smoutput (2#LF) (] (' :' (i.&1@:e. split ]) S: 0 LF cut {.)~ i.&1@:E.) S: 0 'From [^\n]+\nFrom: [^\n]+\nDate: [^\n]+\nSubject: [^\n]+\nMessage-ID: [^\n]+\n\n' (] <;.1~ rxE) 0 : 0 [ load 'regex' From christopher.f.lett at citigroup.com Thu Dec 1 00:55:28 2005 From: christopher.f.lett at citigroup.com (Lett, Christopher F) Date: Thu Dec 1 00:56:29 2005 Subject: [Jbeta] Behavior of Scrollbars in Grid Control Message-ID: I noticed that in earlier releases of the Grid control, I could continuously scroll the window using the scrollbars by placing the mouse in an empty portion of the bar and pressing and holding the mouse button. (You can try this in the "Grid Basic Examples" lab, frame 7). But with the new Grid control in J601, this behavior doesn't work. (Try the Grid demo, "Countries" example). I don't know if this is really a bug or just an incompatibility introduced with the new control, but it was a little unexpected and doesn't seem to follow standard Windows widget behavior. /Christopher Lett From gosinn at gmail.com Thu Dec 1 01:31:35 2005 From: gosinn at gmail.com (=?ISO-8859-1?Q?Bj=F6rn_Helgason?=) Date: Thu Dec 1 01:30:01 2005 Subject: [Jbeta] gl2 - schizophrenia Message-ID: <4031fef0511300931o73a4ecdam@mail.gmail.com> There are two gl2.ijt files one in system\extras\labs\system other in system\extras\labs\graphics both claim to be: Lab: Graphics - gl2 commands From r.e.boss at planet.nl Thu Dec 1 02:17:39 2005 From: r.e.boss at planet.nl (R.E. Boss) Date: Thu Dec 1 02:15:59 2005 Subject: [Jbeta] Behavior of Scrollbars in Grid Control In-Reply-To: Message-ID: <003401c5f5da$58d3ac30$9900000a@BOS> I reported this before but it is repaired in the bbeta-relaease. R.E. Boss -----Oorspronkelijk bericht----- Van: beta-bounces@jsoftware.com [mailto:beta-bounces@jsoftware.com] Namens Lett, Christopher F Verzonden: woensdag 30 november 2005 17:55 Aan: beta@jsoftware.com Onderwerp: [Jbeta] Behavior of Scrollbars in Grid Control I noticed that in earlier releases of the Grid control, I could continuously scroll the window using the scrollbars by placing the mouse in an empty portion of the bar and pressing and holding the mouse button. (You can try this in the "Grid Basic Examples" lab, frame 7). But with the new Grid control in J601, this behavior doesn't work. (Try the Grid demo, "Countries" example). I don't know if this is really a bug or just an incompatibility introduced with the new control, but it was a little unexpected and doesn't seem to follow standard Windows widget behavior. /Christopher Lett ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm From gosinn at gmail.com Thu Dec 1 03:22:41 2005 From: gosinn at gmail.com (=?ISO-8859-1?Q?Bj=F6rn_Helgason?=) Date: Thu Dec 1 03:21:10 2005 Subject: [Jbeta] Lab: Turtle Geometry - JTurtle User Guide error Message-ID: <4031fef0511301122g5afe9951w@mail.gmail.com> Lab: Turtle Geometry - JTurtle User Guide ?? ((1) 4 of 30) The idea of a turtle ??????????????????????? error in jijs_ctrl_fkey domain error: spos f~y From fraser.jackson at xtra.co.nz Thu Dec 1 03:56:34 2005 From: fraser.jackson at xtra.co.nz (Fraser Jackson) Date: Thu Dec 1 03:55:04 2005 Subject: [Jbeta] Lab: Turtle Geometry - JTurtle User Guide error References: <4031fef0511301122g5afe9951w@mail.gmail.com> Message-ID: <001301c5f5e8$2b442070$0101a8c0@vaio> Bjorn, the revisions in the gl2 handling have impacted on this script and I have updated it but it has not got into the beta yet. Fraser ----- Original Message ----- From: "Bjorn Helgason" To: "Beta forum" Sent: Thursday, December 01, 2005 8:22 AM Subject: [Jbeta] Lab: Turtle Geometry - JTurtle User Guide error > Lab: Turtle Geometry - JTurtle User Guide > > ?? ((1) 4 of 30) The idea of a turtle ??????????????????????? > > error in jijs_ctrl_fkey > domain error: spos > f~y > -------------------------------------------------------------------------------- > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm From gosinn at gmail.com Thu Dec 1 04:12:04 2005 From: gosinn at gmail.com (=?ISO-8859-1?Q?Bj=F6rn_Helgason?=) Date: Thu Dec 1 04:10:34 2005 Subject: [Jbeta] Lab: Turtle Geometry - JTurtle User Guide error In-Reply-To: <001301c5f5e8$2b442070$0101a8c0@vaio> References: <4031fef0511301122g5afe9951w@mail.gmail.com> <001301c5f5e8$2b442070$0101a8c0@vaio> Message-ID: <4031fef0511301212n2551d9bk@mail.gmail.com> fine I am just testing as much as I can so it does not get forgotten seems like this j601 is a big step not the least for graphics 2005/11/30, Fraser Jackson : > Bjorn, > the revisions in the gl2 handling have impacted on this script and I have > updated it but it has not got into the beta yet. > Fraser > ----- Original Message ----- > From: "Bjorn Helgason" > To: "Beta forum" > Sent: Thursday, December 01, 2005 8:22 AM > Subject: [Jbeta] Lab: Turtle Geometry - JTurtle User Guide error > > > > Lab: Turtle Geometry - JTurtle User Guide > > > > ?? ((1) 4 of 30) The idea of a turtle ??????????????????????? > > > > error in jijs_ctrl_fkey > > domain error: spos > > f~y > > > > > -------------------------------------------------------------------------------- > > > > ---------------------------------------------------------------------- > > For information about J forums see http://www.jsoftware.com/forums.htm > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > -- Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 Fugl&Fiskur ehf, ?erneyjarsund 23, 802 Selfoss ,t-p?st: gosinn@gmail.com http://www.linux.is/forritun/data/J/grunnur.pdf T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num /|_ .-----------------------------------. ,' .\ / | Me? l?ttri lund ver?ur | ,--' _,' | Dagurinn ? dag | / / | Enn betri en g?rdagurinn | ( -. | `-----------------------------------' | ) | (`-. '--.) `. )----' From gosinn at gmail.com Fri Dec 2 23:09:48 2005 From: gosinn at gmail.com (=?ISO-8859-1?Q?Bj=F6rn_Helgason?=) Date: Fri Dec 2 23:08:17 2005 Subject: [Jbeta] studio advance hard coded? Message-ID: <4031fef0512020709q5e3ade4bx@mail.gmail.com> I was trying to change the studio advance from ctrl+j into some one char like f11 but it seems to be hardcoded In configure shortcuts it looks like you can change it but it does not work When you want to step through many labs it would be nice to be able to just press just one char instead of the little bit awkward ctrl+j From gkirouac at riq.qc.ca Sun Dec 4 07:40:09 2005 From: gkirouac at riq.qc.ca (Gilles Kirouac) Date: Sun Dec 4 07:38:32 2005 Subject: [Jbeta] Ill-formed error message Message-ID: <20051203230150.M23821@riq.qc.ca> What is the meaning of loa? The complete session is: ]v=:12 23 34 44 59 70 98, 12 15 25 27 32 88 99,:15 35 78 82 90 95 97 12 23 34 44 59 70 98 12 15 25 27 32 88 99 15 35 78 82 90 95 97 (mean; stddev ; var)"1 v │value error: mean │ (mean;stddev;var)"1 v load 'c:\j601\system\packages\stats\statfns.ijs' ]"1v │ill-formed number: loa <====== Gilles -) From rsdonovan at hotmail.com Sun Dec 4 19:44:54 2005 From: rsdonovan at hotmail.com (Richard Donovan) Date: Sun Dec 4 19:43:15 2005 Subject: [Jbeta] Pocket PC In-Reply-To: <20051203230150.M23821@riq.qc.ca> Message-ID: Does the new J6 beta work for Pocket Pc? Thanks in advance From eiverson at sympatico.ca Sun Dec 4 22:08:25 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Sun Dec 4 22:06:45 2005 Subject: [Jbeta] Pocket PC References: Message-ID: <001901c5f8dc$33e20070$0701a8c0@xp> The currently available J601 beta won't work on the PocketPC. We'd hoped to get the PocketPc out by now, but it has slipped and won't be out til the new year. ----- Original Message ----- From: "Richard Donovan" To: Sent: Sunday, December 04, 2005 6:44 AM Subject: [Jbeta] Pocket PC > Does the new J6 beta work for Pocket Pc? > > Thanks in advance > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm From gkirouac at riq.qc.ca Tue Dec 6 10:45:38 2005 From: gkirouac at riq.qc.ca (Gilles Kirouac) Date: Tue Dec 6 10:43:58 2005 Subject: [Jbeta] Enter_info.htm+ needs updating Message-ID: <20051206024351.M50416@riq.qc.ca> The line wd `'''''pshow' in file:///c:/j601/system/extras/help/user/enter_info.htm looks weird and results in a domain error. At the top of the same page, open 'system\examples\demo\name.ijs' should be, I think, load jpath '~system\examples\demo\name.ijs' The page c:/j601/system/extras/help/user/form_locales.htm also needs adjustments around the load verb. Gilles -) From gkirouac at riq.qc.ca Tue Dec 6 10:48:09 2005 From: gkirouac at riq.qc.ca (Gilles Kirouac) Date: Tue Dec 6 10:46:26 2005 Subject: [Jbeta] add folder Addons Message-ID: <20051206024622.M51754@riq.qc.ca> Could you please add folder Addons to the folders accessible in Find in Files of the JADE. Won't hurt anyone. Gilles -) From gkirouac at riq.qc.ca Tue Dec 6 10:50:10 2005 From: gkirouac at riq.qc.ca (Gilles Kirouac) Date: Tue Dec 6 10:48:28 2005 Subject: [Jbeta] The new grid : for Linux? Message-ID: <20051206024833.M67820@riq.qc.ca> The current doc for the new grid says: A grid is drawn in an isigraph control on a Windows form. Does that imply that it won't be available for Linux and Mac? Gilles -) From gkirouac at riq.qc.ca Tue Dec 6 12:31:54 2005 From: gkirouac at riq.qc.ca (Gilles Kirouac) Date: Tue Dec 6 12:30:12 2005 Subject: [Jbeta] New grid object : virtual grid possible? Message-ID: <20051206043023.M62020@riq.qc.ca> Will the new grid object allow for a virtual grid? A virtual grid allows for the display of a small subset of a large/long spreadsheet. The programmer needs to catch scrollbar events and program the size and position of the scrollbar himself. Without this, the maximum size of a displayed spreadsheet is much smaller and takes much more preparation time. I have not seen these event and method mentioned in the doc I read. It would be really nice not to have to redo the new grid class ourselves, even if it is possible. The grid class is not light stuff! Gilles -) From eiverson at sympatico.ca Wed Dec 7 02:58:37 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Wed Dec 7 05:03:44 2005 Subject: [Jbeta] The new grid : for Linux? References: <20051206024833.M67820@riq.qc.ca> Message-ID: <000901c5faa8$c845e240$0701a8c0@xp> The same grid is available on all J601 platforms (linux, mac, pocketpc). I guess the doc should say: A grid is drawn in an isigraph control on a J form. ----- Original Message ----- From: "Gilles Kirouac" To: Sent: Monday, December 05, 2005 9:50 PM Subject: [Jbeta] The new grid : for Linux? > The current doc for the new grid says: > > A grid is drawn in an isigraph control on a Windows form. > > Does that imply that it won't be available for Linux and Mac? > > Gilles -) > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm From gosinn at gmail.com Wed Dec 7 05:34:51 2005 From: gosinn at gmail.com (=?ISO-8859-1?Q?Bj=F6rn_Helgason?=) Date: Wed Dec 7 05:34:05 2005 Subject: [Jbeta] The new grid : for Linux? In-Reply-To: <000901c5faa8$c845e240$0701a8c0@xp> References: <20051206024833.M67820@riq.qc.ca> <000901c5faa8$c845e240$0701a8c0@xp> Message-ID: <4031fef0512061334x3ebb8349y@mail.gmail.com> Windows is not only M$ Windows it is also X Windows 2005/12/6, Eric Iverson : > The same grid is available on all J601 platforms (linux, mac, pocketpc). I > guess the doc should say: > > A grid is drawn in an isigraph control on a J form. > > ----- Original Message ----- > From: "Gilles Kirouac" > To: > Sent: Monday, December 05, 2005 9:50 PM > Subject: [Jbeta] The new grid : for Linux? > > > > The current doc for the new grid says: > > > > A grid is drawn in an isigraph control on a Windows form. > > > > Does that imply that it won't be available for Linux and Mac? > > > > Gilles -) > > ---------------------------------------------------------------------- > > For information about J forums see http://www.jsoftware.com/forums.htm > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > -- Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 Fugl&Fiskur ehf, ?erneyjarsund 23, 802 Selfoss ,t-p?st: gosinn@gmail.com http://www.linux.is/forritun/data/J/grunnur.pdf T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num /|_ .-----------------------------------. ,' .\ / | Me? l?ttri lund ver?ur | ,--' _,' | Dagurinn ? dag | / / | Enn betri en g?rdagurinn | ( -. | `-----------------------------------' | ) | (`-. '--.) `. )----' From bobtherriault at shaw.ca Wed Dec 7 05:36:14 2005 From: bobtherriault at shaw.ca (Bob Therriault) Date: Wed Dec 7 05:34:27 2005 Subject: [Jbeta] Question about rank of verb #: antibase Message-ID: <000001c5faad$16216260$6501a8c0@diane> In the process of learning this remarkable language, every once in a while I stumble upon something that doesn't seem quite right. In the case of #: it looks to me as if it acts and is expected to act like a monadic verb of rank 0. However the basic function b. returns a rank of _ and I have a tough time visualizing how antibase would apply to an entire matrix if not on an atom by atom basis. I eagerly await enlightenment. #: 5 1 0 1 #:i.3 2 0 0 0 0 0 1 0 1 0 0 1 1 1 0 0 1 0 1 #: b. 0 _ 1 0 This also applies to J504, but it seems appropriate to address the current beta. My apologies in advance if I have posted incorrectly. Cheers, bob From RHui000 at shaw.ca Wed Dec 7 05:55:54 2005 From: RHui000 at shaw.ca (Roger Hui) Date: Wed Dec 7 05:55:01 2005 Subject: [Jbeta] Question about rank of verb #: antibase References: <000001c5faad$16216260$6501a8c0@diane> Message-ID: <03a601c5faaf$dca1a8d0$f7c35718@lap> The problem is that the positional number system pads on the left, but rank in J pads on the right. Thus: x=: 1 12 123 1234 #: x 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 1 1 1 1 0 1 1 1 0 0 1 1 0 1 0 0 1 0 #:"0 x 1 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 1 1 1 1 0 1 1 0 0 0 0 1 0 0 1 1 0 1 0 0 1 0 10&#.^:_1 x 0 0 0 1 0 0 1 2 0 1 2 3 1 2 3 4 10&#.^:_1"0 x 1 0 0 0 1 2 0 0 1 2 3 0 1 2 3 4 ----- Original Message ----- From: "Bob Therriault" To: Sent: Tuesday, December 06, 2005 1:36 PM Subject: [Jbeta] Question about rank of verb #: antibase In the process of learning this remarkable language, every once in a while I stumble upon something that doesn't seem quite right. In the case of #: it looks to me as if it acts and is expected to act like a monadic verb of rank 0. However the basic function b. returns a rank of _ and I have a tough time visualizing how antibase would apply to an entire matrix if not on an atom by atom basis. I eagerly await enlightenment. #: 5 1 0 1 #:i.3 2 0 0 0 0 0 1 0 1 0 0 1 1 1 0 0 1 0 1 #: b. 0 _ 1 0 This also applies to J504, but it seems appropriate to address the current beta. My apologies in advance if I have posted incorrectly. Cheers, bob From jf at dbron.net Wed Dec 7 06:09:16 2005 From: jf at dbron.net (Dan Bron) Date: Wed Dec 7 06:07:30 2005 Subject: [Jbeta] Question about rank of verb #: antibase In-Reply-To: <000001c5faad$16216260$6501a8c0@diane> Message-ID: If 0 = {. #: b. 0 then the meaning of the output of #: would be different (and it would not be usefully invertible). Fundamentally, this is because a normal fill (ala #:"0 ) would create trailing zeros, like > 1 ; 1 2 ; 1 2 3 . This is a problem because LEADING zeros are insignificant, but trailing zeros are not. To put it another way: defining monad #: to have infinite rank allows it to fill "backwards". If the verb can see the entire array, then it can find the maximal atom in the input, and thereby know the appropriate number of columns for the final output. Examples: (<@#~ [: #: @:i. 2: ^ #) 'abc' NB. All subsets of 'abc' ++-+-+--+-+--+--+---+ ||c|b|bc|a|ac|ab|abc| ++-+-+--+-+--+--+---+ (<@#~ [: #:"0 @:i. 2: ^ #) 'abc' NB. Not as useful ++-+-+--+-+--+--+---+ ||a|a|ab|a|ac|ab|abc| ++-+-+--+-+--+--+---+ (4&{.)&.:(#:) i. 16 NB. Under binary space works as expected 0 1 2 3 (4&{.)&.:(#:"0) i. 16 NB. ... as long as the rank is correct 0 8 8 12 #. 4 {. #: i. 16 NB. Where does the difference lie? 0 1 2 3 #. 4 {. #:"0 i. 16 0 8 8 12 4 {. #: i. 16 0 0 0 0 0 0 0 1 0 0 1 0 0 0 1 1 4 {. #:"0 i. 16 NB. Comes down to filling. 0 0 0 0 1 0 0 0 1 0 0 0 1 1 0 0 -Dan From gkirouac at riq.qc.ca Wed Dec 7 14:42:44 2005 From: gkirouac at riq.qc.ca (Gilles Kirouac) Date: Wed Dec 7 14:41:03 2005 Subject: [Jbeta] DrWatson dump useful? Message-ID: <20051207064004.M87188@riq.qc.ca> Is a V6.01 DrWatson log and .dmp of any use to you, if I do not recall the steps leading to the crash? Gilles - From cburke at jsoftware.com Wed Dec 7 16:38:33 2005 From: cburke at jsoftware.com (Chris Burke) Date: Wed Dec 7 16:36:49 2005 Subject: [Jbeta] New grid object : virtual grid possible? In-Reply-To: <20051206043023.M62020@riq.qc.ca> References: <20051206043023.M62020@riq.qc.ca> Message-ID: <43969F89.6060502@jsoftware.com> Gilles Kirouac wrote: > Will the new grid object allow for a virtual grid? A virtual grid allows for > the display of a small subset of a large/long spreadsheet. The programmer > needs to catch scrollbar events and program the size and position of the > scrollbar himself. Without this, the maximum size of a displayed spreadsheet > is much smaller and takes much more preparation time. > > I have not seen these event and method mentioned in the doc I read. It would > be really nice not to have to redo the new grid class ourselves, even if it is > possible. The grid class is not light stuff! Yes. The function table demo is a virtual grid, but there will be another virtual grid demo in the next beta that displays a noun, and this should be a better example. The base jzgrid class has more functionality than the old jzgrid, but the derived classes are much lighter. The intention is that it should be fairly easy to create derived classes for specific grids. Thanks for your other problem reports. From eiverson at sympatico.ca Wed Dec 7 23:14:00 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Wed Dec 7 23:12:24 2005 Subject: [Jbeta] DrWatson dump useful? References: <20051207064004.M87188@riq.qc.ca> Message-ID: <000f01c5fb40$de50e790$0701a8c0@xp> Not at this time. Reproducible steps always appreciated. ----- Original Message ----- From: "Gilles Kirouac" To: Sent: Wednesday, December 07, 2005 1:42 AM Subject: [Jbeta] DrWatson dump useful? > > Is a V6.01 DrWatson log and .dmp of any use to you, if I do not recall the > steps leading to the crash? > > Gilles - > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm From bobtherriault at shaw.ca Thu Dec 8 00:52:50 2005 From: bobtherriault at shaw.ca (Bob Therriault) Date: Thu Dec 8 00:52:51 2005 Subject: [Jbeta] Re: Question about rank of verb #: antibase In-Reply-To: <20051207040002.BD40674C674@alice.jsoftware.com> Message-ID: <000001c5fb4e$ab82e3f0$6501a8c0@diane> Thank you to both Dan and Roger for an excellent clarification. I really am enjoying having my mind stretched by J. My thanks to all on the forum. Cheers, bob Message: 6 Date: Tue, 6 Dec 2005 17:09:16 -0500 From: Dan Bron Subject: Re: [Jbeta] Question about rank of verb #: antibase To: Beta forum Message-ID: Content-Type: text/plain; charset=iso-8859-1 If 0 = {. #: b. 0 then the meaning of the output of #: would be different (and it would not be usefully invertible). Fundamentally, this is because a normal fill (ala #:"0 ) would create trailing zeros, like > 1 ; 1 2 ; 1 2 3 . This is a problem because LEADING zeros are insignificant, but trailing zeros are not. To put it another way: defining monad #: to have infinite rank allows it to fill "backwards". If the verb can see the entire array, then it can find the maximal atom in the input, and thereby know the appropriate number of columns for the final output. Examples: (<@#~ [: #: @:i. 2: ^ #) 'abc' NB. All subsets of 'abc' ++-+-+--+-+--+--+---+ ||c|b|bc|a|ac|ab|abc| ++-+-+--+-+--+--+---+ (<@#~ [: #:"0 @:i. 2: ^ #) 'abc' NB. Not as useful ++-+-+--+-+--+--+---+ ||a|a|ab|a|ac|ab|abc| ++-+-+--+-+--+--+---+ (4&{.)&.:(#:) i. 16 NB. Under binary space works as expected 0 1 2 3 (4&{.)&.:(#:"0) i. 16 NB. ... as long as the rank is correct 0 8 8 12 #. 4 {. #: i. 16 NB. Where does the difference lie? 0 1 2 3 #. 4 {. #:"0 i. 16 0 8 8 12 4 {. #: i. 16 0 0 0 0 0 0 0 1 0 0 1 0 0 0 1 1 4 {. #:"0 i. 16 NB. Comes down to filling. 0 0 0 0 1 0 0 0 1 0 0 0 1 1 0 0 -Dan ------------------------------ ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm End of Beta Digest, Vol 4, Issue 6 ********************************** From gkirouac at riq.qc.ca Fri Dec 9 13:44:27 2005 From: gkirouac at riq.qc.ca (Gilles Kirouac) Date: Fri Dec 9 13:42:43 2005 Subject: [Jbeta] Script window behaviour Message-ID: <20051209053536.M73101@riq.qc.ca> 0) * in title bar of a script window In J5.04 and J6.01, when opening an existing script, I expect not to see the *. But it is always there. Why? [An execution window appears to have the same problem, but is not a concern for me.] 1) No cancel option when running/closing a modified script window When I run or close a modified script window, I am offered to save it or not. Is a cancel option too much of a luxury? Without it, I necessarily lose either the new version or the old one. Gilles - From gkirouac at riq.qc.ca Fri Dec 9 23:36:47 2005 From: gkirouac at riq.qc.ca (Gilles Kirouac) Date: Fri Dec 9 23:35:01 2005 Subject: [Jbeta] Crash/freeze of J6.01b (2005-11-21) Message-ID: <20051209153345.M3079@riq.qc.ca> Executing this crashes J6.01b: m=:?.101 36$999 m The crash is right after the display. Executing this freezes J6.01b: m=:?.81 36$999 m ctl-d Gilles - From eiverson at sympatico.ca Sat Dec 10 06:39:04 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Sat Dec 10 06:37:23 2005 Subject: [Jbeta] J 601 c beta available Message-ID: <003e01c5fd11$5ed6e460$0701a8c0@xp> The c beta is available: http://www.jsoftware.com/download/j601cbeta_win.exe User feed back on the previous beta showed we were off on the wrong foot with our unicode support. Web research, experiments, and working with Bill Lam and Bjorn, showed that adopting UTF-8 as a byte encoding of unicode data was the right approach. This new beta reflects that. See Studio|Demos|unicode simple and Studio|Demos|unicode for examples of unicode applications and look at the code for both. User Manual section Unicode has a very rough and preliminary discussion of unicode and UTF-8. Also see the documentation in the dictionary vocabulary on new facilities in primitive u: . Problems with OLE/COM (OCX controls and OLE Automation) have been fixed. This area was quite disturbed by the unicode work and we'd appreciate users testing their use of these facilities. And once the old stuff is working, it would be nice to see examples of using unicode for non-English language applications with OLE/COM. The beta fixes problems with the new grid. Study the demo and the docs. From eiverson at sympatico.ca Sat Dec 10 06:46:44 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Sat Dec 10 06:44:58 2005 Subject: [Jbeta] Crash/freeze of J6.01b (2005-11-21) References: <20051209153345.M3079@riq.qc.ca> Message-ID: <001201c5fd12$6f9aecf0$0701a8c0@xp> These appear to be fixed in the new c beta. ----- Original Message ----- From: "Gilles Kirouac" To: Sent: Friday, December 09, 2005 10:36 AM Subject: [Jbeta] Crash/freeze of J6.01b (2005-11-21) > Executing this crashes J6.01b: > > m=:?.101 36$999 > m > > The crash is right after the display. > > Executing this freezes J6.01b: > > m=:?.81 36$999 > m > ctl-d > > > Gilles - > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm From juneaftn at gmail.com Sat Dec 10 08:40:19 2005 From: juneaftn at gmail.com (June Kim) Date: Sat Dec 10 08:38:32 2005 Subject: [Jbeta] J 601 c beta available In-Reply-To: <003e01c5fd11$5ed6e460$0701a8c0@xp> References: <003e01c5fd11$5ed6e460$0701a8c0@xp> Message-ID: <6f80c1520512091640k1843c528p7ff7c92fa73896e4@mail.gmail.com> Thank you for the new beta. Now I can type in multibyte characters in the ijx window. However, I have found out some problem (or my ignorance?) HAN=.'?' NB. this is a korean character, which sounds as "han". You can make the letter instead with "HAN=.8 u: u: 54620" if you don't have the IME. ]codepoint=.3 u: 7 u: HAN 54620 3!:0 HAN 2 3!:0 (7 u: HAN) 131072 3!:0 (4 u: 54620) 131072 #(7 u: HAN) NB. this is one letter 1 (7 u: HAN) = 7 u: HAN 1 HAN_A=.HAN,'a' NB. append the alphabet "a" #7 u: HAN_A NB. now it's two letters 2 So far so good. Until I see (7 u: HAN) = 7 u: HAN_A |length error | (7 u:HAN) =7 u:HAN_A But somehow I found out a working(but ugly and painful) version (4 u: 0{3 u: 7 u: HAN) = 7 u: HAN_A 1 0 June Kim On 12/10/05, Eric Iverson wrote: > The c beta is available: > http://www.jsoftware.com/download/j601cbeta_win.exe > > User feed back on the previous beta showed we were off on the wrong foot > with our unicode support. Web research, experiments, and working with Bill > Lam and Bjorn, showed that adopting UTF-8 as a byte encoding of unicode data > was the right approach. This new beta reflects that. > > See Studio|Demos|unicode simple and Studio|Demos|unicode for examples of > unicode applications and look at the code for both. > > User Manual section Unicode has a very rough and preliminary discussion of > unicode and UTF-8. Also see the documentation in the dictionary vocabulary > on new facilities in primitive u: . > > Problems with OLE/COM (OCX controls and OLE Automation) have been fixed. > This area was quite disturbed by the unicode work and we'd appreciate users > testing their use of these facilities. And once the old stuff is working, it > would be nice to see examples of using unicode for non-English language > applications with OLE/COM. > > The beta fixes problems with the new grid. Study the demo and the docs. > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > From juneaftn at gmail.com Sat Dec 10 08:50:45 2005 From: juneaftn at gmail.com (June Kim) Date: Sat Dec 10 08:49:06 2005 Subject: [Jbeta] J 601 c beta available In-Reply-To: <6f80c1520512091640k1843c528p7ff7c92fa73896e4@mail.gmail.com> References: <003e01c5fd11$5ed6e460$0701a8c0@xp> <6f80c1520512091640k1843c528p7ff7c92fa73896e4@mail.gmail.com> Message-ID: <6f80c1520512091650n135e2f23x6bda771f839e6a1@mail.gmail.com> Okay, I have discovered a simpler version. (0{7 u: HAN) = 7 u: HAN_A 1 0 And there seems like a difference between one lettered ascii and one lettered unicode $'a' $ 7 u: HAN 1 #'a' 1 # 7 u: HAN 1 On 12/10/05, June Kim wrote: > Thank you for the new beta. > > Now I can type in multibyte characters in the ijx window. However, I > have found out some problem (or my ignorance?) > > HAN=.'?' NB. this is a korean character, which sounds as "han". > You can make the letter instead with "HAN=.8 u: u: 54620" if you don't > have the IME. > > ]codepoint=.3 u: 7 u: HAN > 54620 > 3!:0 HAN > 2 > 3!:0 (7 u: HAN) > 131072 > 3!:0 (4 u: 54620) > 131072 > #(7 u: HAN) NB. this is one letter > 1 > (7 u: HAN) = 7 u: HAN > 1 > HAN_A=.HAN,'a' NB. append the alphabet "a" > #7 u: HAN_A NB. now it's two letters > 2 > > So far so good. Until I see > > (7 u: HAN) = 7 u: HAN_A > |length error > | (7 u:HAN) =7 u:HAN_A > > But somehow I found out a working(but ugly and painful) version > > (4 u: 0{3 u: 7 u: HAN) = 7 u: HAN_A > 1 0 > > > > June Kim > > On 12/10/05, Eric Iverson wrote: > > The c beta is available: > > http://www.jsoftware.com/download/j601cbeta_win.exe > > > > User feed back on the previous beta showed we were off on the wrong foot > > with our unicode support. Web research, experiments, and working with Bill > > Lam and Bjorn, showed that adopting UTF-8 as a byte encoding of unicode data > > was the right approach. This new beta reflects that. > > > > See Studio|Demos|unicode simple and Studio|Demos|unicode for examples of > > unicode applications and look at the code for both. > > > > User Manual section Unicode has a very rough and preliminary discussion of > > unicode and UTF-8. Also see the documentation in the dictionary vocabulary > > on new facilities in primitive u: . > > > > Problems with OLE/COM (OCX controls and OLE Automation) have been fixed. > > This area was quite disturbed by the unicode work and we'd appreciate users > > testing their use of these facilities. And once the old stuff is working, it > > would be nice to see examples of using unicode for non-English language > > applications with OLE/COM. > > > > The beta fixes problems with the new grid. Study the demo and the docs. > > > > ---------------------------------------------------------------------- > > For information about J forums see http://www.jsoftware.com/forums.htm > > > From juneaftn at gmail.com Sat Dec 10 09:37:44 2005 From: juneaftn at gmail.com (June Kim) Date: Sat Dec 10 09:35:58 2005 Subject: [Jbeta] J 601 c beta available In-Reply-To: <003e01c5fd11$5ed6e460$0701a8c0@xp> References: <003e01c5fd11$5ed6e460$0701a8c0@xp> Message-ID: <6f80c1520512091737l6b0d99c8k51f8695b37fb8e8d@mail.gmail.com> In some fonts (like "Bitstream Vera Sans Mono"), the box-drawing looks broken. I had to Toggle Ascii Box Drawing from the Tools menu everytime I run J. Is there anyway I can toggle it programmatically, like putting some sentence on the startup file? On 12/10/05, Eric Iverson wrote: > The c beta is available: > http://www.jsoftware.com/download/j601cbeta_win.exe > > User feed back on the previous beta showed we were off on the wrong foot > with our unicode support. Web research, experiments, and working with Bill > Lam and Bjorn, showed that adopting UTF-8 as a byte encoding of unicode data > was the right approach. This new beta reflects that. > > See Studio|Demos|unicode simple and Studio|Demos|unicode for examples of > unicode applications and look at the code for both. > > User Manual section Unicode has a very rough and preliminary discussion of > unicode and UTF-8. Also see the documentation in the dictionary vocabulary > on new facilities in primitive u: . > > Problems with OLE/COM (OCX controls and OLE Automation) have been fixed. > This area was quite disturbed by the unicode work and we'd appreciate users > testing their use of these facilities. And once the old stuff is working, it > would be nice to see examples of using unicode for non-English language > applications with OLE/COM. > > The beta fixes problems with the new grid. Study the demo and the docs. > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > From eiverson at sympatico.ca Sat Dec 10 11:50:24 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Sat Dec 10 11:48:46 2005 Subject: [Jbeta] J 601 c beta available References: <003e01c5fd11$5ed6e460$0701a8c0@xp><6f80c1520512091640k1843c528p7ff7c92fa73896e4@mail.gmail.com> <6f80c1520512091650n135e2f23x6bda771f839e6a1@mail.gmail.com> Message-ID: <003c01c5fd3c$dfc59460$0701a8c0@xp> The problem is that the result of 7 u: for a 2 or 3 character sequence that returns a single wchar is a list. As the argument is a list there is no way to determine if the result should be a scalar or a list. The ASCII arg to 7 u: is currently returned as unchanged. It might be more consistent to make 7 u: always return a list even in the ASCII case. But this wouldn't help your situation. A possible change would be to make 7 u: return a scalar rather than a list if there was only 1 wchar in the result, but then when you wanted a list you'd be upset and I suspect this would just move the problem around. The problem is that the these transformation don't have the nice shape and rank conformation of typical J primitives. That is, 1, 2, or 3 char sequences in the arg make 1 wchar in the result. Feedback and suggestions welcome. I could redo all your HAN experiments, but the character doesn't display properly for me in ijs or ijx. It does display properly for me in your email and in notepad. What font are you using in ijs? Do you know why it won't display in my ijs window? I can display Chinese, Icelandic, etc ok. ----- Original Message ----- From: "June Kim" To: "Beta forum" Sent: Friday, December 09, 2005 7:50 PM Subject: Re: [Jbeta] J 601 c beta available > Okay, I have discovered a simpler version. > > (0{7 u: HAN) = 7 u: HAN_A > 1 0 > > And there seems like a difference between one lettered ascii and one > lettered unicode > > $'a' > > $ 7 u: HAN > 1 > #'a' > 1 > # 7 u: HAN > 1 > > > On 12/10/05, June Kim wrote: >> Thank you for the new beta. >> >> Now I can type in multibyte characters in the ijx window. However, I >> have found out some problem (or my ignorance?) >> >> HAN=.'?' NB. this is a korean character, which sounds as "han". >> You can make the letter instead with "HAN=.8 u: u: 54620" if you don't >> have the IME. >> >> ]codepoint=.3 u: 7 u: HAN >> 54620 >> 3!:0 HAN >> 2 >> 3!:0 (7 u: HAN) >> 131072 >> 3!:0 (4 u: 54620) >> 131072 >> #(7 u: HAN) NB. this is one letter >> 1 >> (7 u: HAN) = 7 u: HAN >> 1 >> HAN_A=.HAN,'a' NB. append the alphabet "a" >> #7 u: HAN_A NB. now it's two letters >> 2 >> >> So far so good. Until I see >> >> (7 u: HAN) = 7 u: HAN_A >> |length error >> | (7 u:HAN) =7 u:HAN_A >> >> But somehow I found out a working(but ugly and painful) version >> >> (4 u: 0{3 u: 7 u: HAN) = 7 u: HAN_A >> 1 0 >> >> >> >> June Kim >> >> On 12/10/05, Eric Iverson wrote: >> > The c beta is available: >> > http://www.jsoftware.com/download/j601cbeta_win.exe >> > >> > User feed back on the previous beta showed we were off on the wrong >> > foot >> > with our unicode support. Web research, experiments, and working with >> > Bill >> > Lam and Bjorn, showed that adopting UTF-8 as a byte encoding of unicode >> > data >> > was the right approach. This new beta reflects that. >> > >> > See Studio|Demos|unicode simple and Studio|Demos|unicode for examples >> > of >> > unicode applications and look at the code for both. >> > >> > User Manual section Unicode has a very rough and preliminary discussion >> > of >> > unicode and UTF-8. Also see the documentation in the dictionary >> > vocabulary >> > on new facilities in primitive u: . >> > >> > Problems with OLE/COM (OCX controls and OLE Automation) have been >> > fixed. >> > This area was quite disturbed by the unicode work and we'd appreciate >> > users >> > testing their use of these facilities. And once the old stuff is >> > working, it >> > would be nice to see examples of using unicode for non-English language >> > applications with OLE/COM. >> > >> > The beta fixes problems with the new grid. Study the demo and the docs. >> > >> > ---------------------------------------------------------------------- >> > For information about J forums see http://www.jsoftware.com/forums.htm >> > >> > -------------------------------------------------------------------------------- > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm From cburke at jsoftware.com Sat Dec 10 13:06:33 2005 From: cburke at jsoftware.com (Chris Burke) Date: Sat Dec 10 13:04:48 2005 Subject: [Jbeta] J 601 c beta available In-Reply-To: <6f80c1520512091737l6b0d99c8k51f8695b37fb8e8d@mail.gmail.com> References: <003e01c5fd11$5ed6e460$0701a8c0@xp> <6f80c1520512091737l6b0d99c8k51f8695b37fb8e8d@mail.gmail.com> Message-ID: <439A6259.6080709@jsoftware.com> June Kim wrote: > In some fonts (like "Bitstream Vera Sans Mono"), the box-drawing looks > broken. I had to Toggle Ascii Box Drawing from the Tools menu > everytime I run J. Is there anyway I can toggle it programmatically, > like putting some sentence on the startup file? 9!:7 '+++++++++|-' From rdmiller at usatoday.com Sat Dec 10 16:01:01 2005 From: rdmiller at usatoday.com (Miller, Raul D) Date: Sat Dec 10 15:59:15 2005 Subject: [Jbeta] J 601 c beta available References: <003e01c5fd11$5ed6e460$0701a8c0@xp><6f80c1520512091640k1843c528p7ff7c92fa73896e4@mail.gmail.com><6f80c1520512091650n135e2f23x6bda771f839e6a1@mail.gmail.com> <003c01c5fd3c$dfc59460$0701a8c0@xp> Message-ID: Eric Iverson wrote: > The problem is that the result of 7 u: for a 2 or 3 character sequence that > returns a single wchar is a list. As the argument is a list there is no way > to determine if the result should be a scalar or a list. The ASCII arg to 7 > u: is currently returned as unchanged. It might be more consistent to make 7 > u: always return a list even in the ASCII case. But this wouldn't help your > situation. I don't understand that last comment. I'd think that having 7 u: always return a list would be the right answer, and I don't know enough about the interpreter internals to understand why that would be a bad solution. Could you elaborate a bit? Thanks, -- Raul From bill_lam at myrealbox.com Sat Dec 10 16:53:49 2005 From: bill_lam at myrealbox.com (bill lam) Date: Sat Dec 10 16:52:05 2005 Subject: [Jbeta] missing gl2 commands Message-ID: <1134204829.c8296adcbill_lam@myrealbox.com> I found many gl2 commands are missing in J601, especially glmap and printer support. Will they be add back later, or is there any new equivalent commands? From bill_lam at myrealbox.com Sat Dec 10 17:19:14 2005 From: bill_lam at myrealbox.com (bill lam) Date: Sat Dec 10 17:17:26 2005 Subject: [Jbeta] 1!:x and unicode Message-ID: <1134206354.c8296adcbill_lam@myrealbox.com> It appears 1!:x family is not unicode aware. e.g. file name supplied to 1!:0 must be multibyte character. unicode and utf8 is not supported. The file names returned is also in multibyte (code page dependent). From bill_lam at myrealbox.com Sat Dec 10 17:29:52 2005 From: bill_lam at myrealbox.com (bill lam) Date: Sat Dec 10 17:28:06 2005 Subject: [Jbeta] tolower and toupper Message-ID: <1134206992.c8296adcbill_lam@myrealbox.com> tolower and toupper will give index error for character above 255. Chinese (and Japanese?) does not have this but Many European languages (Islandic?) do have lower/upper case alphabets. It will be nice to make tolower/toupper also work for them. From gosinn at gmail.com Sat Dec 10 17:46:22 2005 From: gosinn at gmail.com (=?ISO-8859-1?Q?Bj=F6rn_Helgason?=) Date: Sat Dec 10 17:44:34 2005 Subject: [Jbeta] tolower and toupper In-Reply-To: <1134206992.c8296adcbill_lam@myrealbox.com> References: <1134206992.c8296adcbill_lam@myrealbox.com> Message-ID: <4031fef0512100146j168c9700g@mail.gmail.com> a=. 'a?bc?de?fghi?jklmno?pqrstu?vwxy?z???A?BC?DE?FGHI?JKLMNO?PQRSTU?VWXY?Z???' +/255: > tolower and toupper will give index error for character above 255. > Chinese (and Japanese?) does not have this but Many European languages > (Islandic?) do have lower/upper case alphabets. It will be nice to > make tolower/toupper also work for them. > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > -- Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 Fugl&Fiskur ehf, ?erneyjarsund 23, 802 Selfoss ,t-p?st: gosinn@gmail.com http://www.linux.is/forritun/data/J/grunnur.pdf T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num /|_ .-----------------------------------. ,' .\ / | Me? l?ttri lund ver?ur | ,--' _,' | Dagurinn ? dag | / / | Enn betri en g?rdagurinn | ( -. | `-----------------------------------' | ) | (`-. '--.) `. )----' From cburke at jsoftware.com Sat Dec 10 18:36:34 2005 From: cburke at jsoftware.com (Chris Burke) Date: Sat Dec 10 18:34:49 2005 Subject: [Jbeta] tolower and toupper In-Reply-To: <4031fef0512100146j168c9700g@mail.gmail.com> References: <1134206992.c8296adcbill_lam@myrealbox.com> <4031fef0512100146j168c9700g@mail.gmail.com> Message-ID: <439AAFB2.1060709@jsoftware.com> Bj?rn Helgason wrote: > cpcount=: #@(9&u:) > > cpcount is not working It won't, because 9 u: is not defined in the cbeta. See the docs for u: and the definition for ucpcount (renamed from cpcount) in the cbeta. From bill_lam at myrealbox.com Sat Dec 10 18:36:50 2005 From: bill_lam at myrealbox.com (bill lam) Date: Sat Dec 10 18:35:03 2005 Subject: [Jbeta] tolower and toupper Message-ID: <1134211010.c7ea4ebcbill_lam@myrealbox.com> now it uses 7&u: and 8&u: as stated in Vocabulary. In J504, I use something like toupperw=: (toupper :: ])"0 but is doesn't work with accent characters either. In your example a=. 'a?bc?de?fghi?jklmno?pqrstu?vwxy?z???A?BC?DE?FGHI?JKLMNO?PQRSTU?VWXY?Z???' #a 92 #@(7&u:) a 72 toupper a A?BC?DE?FGHI?JKLMNO?PQRSTU?VWXY?Z???A?BC?DE?FGHI?JKLMNO?PQRSTU?VWXY?Z??? toupper&.(7&u:) a ABCDEFGHIJKLMNOPQRSTUVWXYZABCDEFGHIJKLMNOPQRSTUVWXYZ I suppose both toupper and toupper&.(7&u:) give incorrect result for ?slandish. From cburke at jsoftware.com Sat Dec 10 18:58:24 2005 From: cburke at jsoftware.com (Chris Burke) Date: Sat Dec 10 18:56:39 2005 Subject: [Jbeta] missing gl2 commands In-Reply-To: <1134204829.c8296adcbill_lam@myrealbox.com> References: <1134204829.c8296adcbill_lam@myrealbox.com> Message-ID: <439AB4D0.7030108@jsoftware.com> bill lam wrote: > I found many gl2 commands are missing in J601, > especially glmap and printer support. Will they be add back later, or is there any new equivalent commands? The gl2 family has been simplified in J601, but the new set should cover all requirements, unless we missed something. glmap is no longer required, because only pixels mode is supported (the old raw mapping mode), see the gl2 commands reference. Printing is now handled by drawing on the printer, just as you would draw on an isigraph, except only for page handling. There is a lab on this, but I notice that the lab has not been updated since it was first written a few months back, and is unfinished. But that and the print scripts should give you some idea of how it works. From r.e.boss at planet.nl Sat Dec 10 21:03:54 2005 From: r.e.boss at planet.nl (R.E. Boss) Date: Sat Dec 10 21:02:05 2005 Subject: [Jbeta] J 601 c beta available In-Reply-To: <003e01c5fd11$5ed6e460$0701a8c0@xp> Message-ID: <000001c5fd8a$2c8f7340$9900000a@BOS> http://www.jsoftware.com/beta.htm still mentions b beta R.E. Boss From gosinn at gmail.com Sat Dec 10 21:30:36 2005 From: gosinn at gmail.com (=?ISO-8859-1?Q?Bj=F6rn_Helgason?=) Date: Sat Dec 10 21:28:49 2005 Subject: [Jbeta] missing gl2 commands In-Reply-To: <439AB4D0.7030108@jsoftware.com> References: <1134204829.c8296adcbill_lam@myrealbox.com> <439AB4D0.7030108@jsoftware.com> Message-ID: <4031fef0512100530y57d5457do@mail.gmail.com> Will there be a new grid lab? 2005/12/10, Chris Burke : > bill lam wrote: > > I found many gl2 commands are missing in J601, > > especially glmap and printer support. Will they be add back later, or is there any new equivalent commands? > > The gl2 family has been simplified in J601, but the new set should cover > all requirements, unless we missed something. > > glmap is no longer required, because only pixels mode is supported (the > old raw mapping mode), see the gl2 commands reference. > > Printing is now handled by drawing on the printer, just as you would > draw on an isigraph, except only for page handling. There is a lab on > this, but I notice that the lab has not been updated since it was first > written a few months back, and is unfinished. But that and the print > scripts should give you some idea of how it works. > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > -- Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 Fugl&Fiskur ehf, ?erneyjarsund 23, 802 Selfoss ,t-p?st: gosinn@gmail.com http://www.linux.is/forritun/data/J/grunnur.pdf T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num /|_ .-----------------------------------. ,' .\ / | Me? l?ttri lund ver?ur | ,--' _,' | Dagurinn ? dag | / / | Enn betri en g?rdagurinn | ( -. | `-----------------------------------' | ) | (`-. '--.) `. )----' From bill_lam at myrealbox.com Sat Dec 10 21:53:35 2005 From: bill_lam at myrealbox.com (bill lam) Date: Sat Dec 10 21:51:48 2005 Subject: [Jbeta] missing gl2 commands Message-ID: <1134222815.c82b7ddcbill_lam@myrealbox.com> glqdevmode,glzdevmode and glzorientation are essential for printing under win32, without them it is difficult to do professional printing. Are they still available under J601 or docs showing how to get it done? From gosinn at gmail.com Sat Dec 10 22:25:19 2005 From: gosinn at gmail.com (=?ISO-8859-1?Q?Bj=F6rn_Helgason?=) Date: Sat Dec 10 22:23:32 2005 Subject: [Jbeta] tolower and toupper In-Reply-To: <1134211010.c7ea4ebcbill_lam@myrealbox.com> References: <1134211010.c7ea4ebcbill_lam@myrealbox.com> Message-ID: <4031fef0512100625xf371ba5o@mail.gmail.com> (a.&i. { (a. ((a. i. '??????????'),97+i.26)}~'??????????ABCDEFGHIJKLMNOPQRSTUVWXYZ')"_) a A?BC?DE?FGHI?JKLMNO?PQRSTU?VWXY?Z???A?BC?DE?FGHI?JKLMNO?PQRSTU?VWXY?Z??? 2005/12/10, bill lam : > now it uses 7&u: and 8&u: as stated in Vocabulary. > In J504, I use something like > toupperw=: (toupper :: ])"0 > but is doesn't work with accent characters either. > In your example > a=. 'a?bc?de?fghi?jklmno?pqrstu?vwxy?z???A?BC?DE?FGHI?JKLMNO?PQRSTU?VWXY?Z???' > #a > 92 > #@(7&u:) a > 72 > toupper a > A?BC?DE?FGHI?JKLMNO?PQRSTU?VWXY?Z???A?BC?DE?FGHI?JKLMNO?PQRSTU?VWXY?Z??? > toupper&.(7&u:) a > ABCDEFGHIJKLMNOPQRSTUVWXYZABCDEFGHIJKLMNOPQRSTUVWXYZ > > I suppose both toupper and toupper&.(7&u:) give incorrect result for ?slandish. > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > -- Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 Fugl&Fiskur ehf, ?erneyjarsund 23, 802 Selfoss ,t-p?st: gosinn@gmail.com http://www.linux.is/forritun/data/J/grunnur.pdf T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num /|_ .-----------------------------------. ,' .\ / | Me? l?ttri lund ver?ur | ,--' _,' | Dagurinn ? dag | / / | Enn betri en g?rdagurinn | ( -. | `-----------------------------------' | ) | (`-. '--.) `. )----' From ramacd at nbnet.nb.ca Sat Dec 10 22:46:38 2005 From: ramacd at nbnet.nb.ca (Randy MacDonald) Date: Sat Dec 10 22:44:42 2005 Subject: [Jbeta] tolower and toupper References: <1134206992.c8296adcbill_lam@myrealbox.com> Message-ID: <00c101c5fd98$86842270$0302a8c0@LONDON4Y78ST8> toupper and tolower working on Unicode is certainly nice-to-have. In which script are toupper and tolower defined? I wonder if there's a upper/lower codepoint mapping defined by Unicode.org. ----------------------------------------------------------------------- |\/| Randy A MacDonald | you can't pay for it, |/\| ramacd@nbnet.nb.ca | even if you want to. BSc(Math) UNBF'83 Sapere Aude | APL: If you can say it, it's done.. Natural Born APL'er | Demo website: http://142.166.105.166/ ----------------------------------------------------(INTP)----{ gnat }- ----- Original Message ----- From: "bill lam" To: Sent: Saturday, December 10, 2005 5:29 AM Subject: [Jbeta] tolower and toupper tolower and toupper will give index error for character above 255. Chinese (and Japanese?) does not have this but Many European languages (Islandic?) do have lower/upper case alphabets. It will be nice to make tolower/toupper also work for them. ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm From gosinn at gmail.com Sat Dec 10 22:51:26 2005 From: gosinn at gmail.com (=?ISO-8859-1?Q?Bj=F6rn_Helgason?=) Date: Sat Dec 10 22:49:39 2005 Subject: [Jbeta] tolower and toupper In-Reply-To: <00c101c5fd98$86842270$0302a8c0@LONDON4Y78ST8> References: <1134206992.c8296adcbill_lam@myrealbox.com> <00c101c5fd98$86842270$0302a8c0@LONDON4Y78ST8> Message-ID: <4031fef0512100651x4cb4a740j@mail.gmail.com> system/main/stdlib.ijs NB.*tolower v convert text to lower case NB.*toupper v convert text to upper case 'l u'=. (a.i.'aA') +each : > toupper and tolower working on Unicode is certainly nice-to-have. > > In which script are toupper and tolower defined? > > I wonder if there's a upper/lower codepoint mapping defined by Unicode.org. > > ----------------------------------------------------------------------- > |\/| Randy A MacDonald | you can't pay for it, > |/\| ramacd@nbnet.nb.ca | even if you want to. > BSc(Math) UNBF'83 Sapere Aude | APL: If you can say it, it's done.. > Natural Born APL'er | Demo website: http://142.166.105.166/ > ----------------------------------------------------(INTP)----{ gnat }- > > ----- Original Message ----- > From: "bill lam" > To: > Sent: Saturday, December 10, 2005 5:29 AM > Subject: [Jbeta] tolower and toupper > > > tolower and toupper will give index error for character above 255. > Chinese (and Japanese?) does not have this but Many European languages > (Islandic?) do have lower/upper case alphabets. It will be nice to > make tolower/toupper also work for them. > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > -- Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 Fugl&Fiskur ehf, ?erneyjarsund 23, 802 Selfoss ,t-p?st: gosinn@gmail.com http://www.linux.is/forritun/data/J/grunnur.pdf T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num /|_ .-----------------------------------. ,' .\ / | Me? l?ttri lund ver?ur | ,--' _,' | Dagurinn ? dag | / / | Enn betri en g?rdagurinn | ( -. | `-----------------------------------' | ) | (`-. '--.) `. )----' From ramacd at nbnet.nb.ca Sat Dec 10 23:04:05 2005 From: ramacd at nbnet.nb.ca (Randy MacDonald) Date: Sat Dec 10 23:02:23 2005 Subject: [Jbeta] tolower and toupper References: <1134206992.c8296adcbill_lam@myrealbox.com><00c101c5fd98$86842270$0302a8c0@LONDON4Y78ST8> <4031fef0512100651x4cb4a740j@mail.gmail.com> Message-ID: <014801c5fd9a$f6f30600$0302a8c0@LONDON4Y78ST8> Thanks for the pointer (and the actual script... who needs bandwidth when one has J)! It looks like extending toupper and tolower is an exercise for the reader, something to share perhaps on the jwiki. The following link looks like a workable standard: http://www.unicode.org/reports/tr21/tr21-5.html ----------------------------------------------------------------------- |\/| Randy A MacDonald | you can't pay for it, |/\| ramacd@nbnet.nb.ca | even if you want to. BSc(Math) UNBF'83 Sapere Aude | APL: If you can say it, it's done.. Natural Born APL'er | Demo website: http://142.166.105.166/ ----------------------------------------------------(INTP)----{ gnat }- ----- Original Message ----- From: "Bj?rn Helgason" To: "Beta forum" Sent: Saturday, December 10, 2005 10:51 AM Subject: Re: [Jbeta] tolower and toupper system/main/stdlib.ijs NB.*tolower v convert text to lower case NB.*toupper v convert text to upper case 'l u'=. (a.i.'aA') +each : > toupper and tolower working on Unicode is certainly nice-to-have. > > In which script are toupper and tolower defined? > > I wonder if there's a upper/lower codepoint mapping defined by Unicode.org. > > ----------------------------------------------------------------------- > |\/| Randy A MacDonald | you can't pay for it, > |/\| ramacd@nbnet.nb.ca | even if you want to. > BSc(Math) UNBF'83 Sapere Aude | APL: If you can say it, it's done.. > Natural Born APL'er | Demo website: http://142.166.105.166/ > ----------------------------------------------------(INTP)----{ gnat }- > > ----- Original Message ----- > From: "bill lam" > To: > Sent: Saturday, December 10, 2005 5:29 AM > Subject: [Jbeta] tolower and toupper > > > tolower and toupper will give index error for character above 255. > Chinese (and Japanese?) does not have this but Many European languages > (Islandic?) do have lower/upper case alphabets. It will be nice to > make tolower/toupper also work for them. > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > -- Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 Fugl&Fiskur ehf, ?erneyjarsund 23, 802 Selfoss ,t-p?st: gosinn@gmail.com http://www.linux.is/forritun/data/J/grunnur.pdf T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num /|_ .-----------------------------------. ,' .\ / | Me? l?ttri lund ver?ur | ,--' _,' | Dagurinn ? dag | / / | Enn betri en g?rdagurinn | ( -. | `-----------------------------------' | ) | (`-. '--.) `. )----' ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm From eiverson at sympatico.ca Sat Dec 10 23:05:10 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Sat Dec 10 23:15:03 2005 Subject: [Jbeta] J 601 c beta available References: <003e01c5fd11$5ed6e460$0701a8c0@xp><6f80c1520512091640k1843c528p7ff7c92fa73896e4@mail.gmail.com><6f80c1520512091650n135e2f23x6bda771f839e6a1@mail.gmail.com><003c01c5fd3c$dfc59460$0701a8c0@xp> Message-ID: <003f01c5fd9c$bffc84d0$0701a8c0@xp> The users complaint was that what they thought of as a scalar single char ended up as a list after 7 u: and required something like 0{ or {. or ''$ to get the scalar they wanted. Making 7 u: always return a list (that is even for ASCII data which currently is returned as is) would make 7 u: more consistent and is probably what we will do, but it doesn't remove the need for {. in the stated problem. ----- Original Message ----- From: "Miller, Raul D" To: "Beta forum" ; "Beta forum" Sent: Saturday, December 10, 2005 3:01 AM Subject: RE: [Jbeta] J 601 c beta available Eric Iverson wrote: > The problem is that the result of 7 u: for a 2 or 3 character sequence > that > returns a single wchar is a list. As the argument is a list there is no > way > to determine if the result should be a scalar or a list. The ASCII arg to > 7 > u: is currently returned as unchanged. It might be more consistent to make > 7 > u: always return a list even in the ASCII case. But this wouldn't help > your > situation. I don't understand that last comment. I'd think that having 7 u: always return a list would be the right answer, and I don't know enough about the interpreter internals to understand why that would be a bad solution. Could you elaborate a bit? Thanks, -- Raul ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm From eiverson at sympatico.ca Sat Dec 10 23:16:47 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Sat Dec 10 23:15:04 2005 Subject: [Jbeta] missing gl2 commands References: <1134204829.c8296adcbill_lam@myrealbox.com> Message-ID: <004001c5fd9c$c1243bf0$0701a8c0@xp> The notes in Release Highlights should be made stronger. There are MAJOR incompatible changes in gl2 with this release. Mapping modes are decommitted as well as many commands. The gl2 commands and implementation are streamlined to improve performance, provide a more basic core implemented in C++ and JAVA, more functionality provided in J code, and better portability across platforms. Any old gl2 code will need careful thought and conversion. The problems are much greater if you used mapping modes. Probably the place to start is to work through the new gl2 commands lab and read the new User Manual docs. We know this is going to be a big nuisance for some users and will help as much as we can through the beta forum. My main confidence that this was the right thing to do was that Chris drove many of the changes in his requirements for grid and was happy to make the changes for plot. No pain, no gain. ----- Original Message ----- From: "bill lam" To: Sent: Saturday, December 10, 2005 3:53 AM Subject: [Jbeta] missing gl2 commands I found many gl2 commands are missing in J601, especially glmap and printer support. Will they be add back later, or is there any new equivalent commands? ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm From gosinn at gmail.com Sat Dec 10 23:18:19 2005 From: gosinn at gmail.com (=?ISO-8859-1?Q?Bj=F6rn_Helgason?=) Date: Sat Dec 10 23:16:33 2005 Subject: [Jbeta] tolower and toupper In-Reply-To: <014801c5fd9a$f6f30600$0302a8c0@LONDON4Y78ST8> References: <1134206992.c8296adcbill_lam@myrealbox.com> <00c101c5fd98$86842270$0302a8c0@LONDON4Y78ST8> <4031fef0512100651x4cb4a740j@mail.gmail.com> <014801c5fd9a$f6f30600$0302a8c0@LONDON4Y78ST8> Message-ID: <4031fef0512100718s12a4c925k@mail.gmail.com> I used to need these kind of verbs a lot earlier (maybe in the future??!!??) There are several codepages for ?slandish chars So there are hell of a lot of files created in one operating system with one codepage and when you receive such a file using another codepage you need to these kind of exercises to move the chars around codepage 850, 861, EBCDIC, Roman8, 1262, some special codepages as well >From now on I hope to only need to move the chars up to Unicode and not bother with them ever again after that (wishful thinking!!) That is one of the reasons I am very very happy about get such a strong support for Unicode in J There are similar considerations with the APL chars between different implementations of Quad AV - they will also get easier in Unicode You can actually display APL chars very nicely with this new Unicode support in J As usual J is way ahead of the other APL dialects 2005/12/10, Randy MacDonald : > Thanks for the pointer (and the actual script... who needs bandwidth when > one has J)! > > It looks like extending toupper and tolower is an exercise for the reader, > something to share perhaps on the jwiki. > > > The following link looks like a workable standard: > > http://www.unicode.org/reports/tr21/tr21-5.html > > ----------------------------------------------------------------------- > |\/| Randy A MacDonald | you can't pay for it, > |/\| ramacd@nbnet.nb.ca | even if you want to. > BSc(Math) UNBF'83 Sapere Aude | APL: If you can say it, it's done.. > Natural Born APL'er | Demo website: http://142.166.105.166/ > ----------------------------------------------------(INTP)----{ gnat }- > > ----- Original Message ----- > From: "Bj?rn Helgason" > To: "Beta forum" > Sent: Saturday, December 10, 2005 10:51 AM > Subject: Re: [Jbeta] tolower and toupper > > > system/main/stdlib.ijs > NB.*tolower v convert text to lower case > NB.*toupper v convert text to upper case > 'l u'=. (a.i.'aA') +each tolower=: a.&i. { ((l{a.) u} a.)"_ > toupper=: a.&i. { ((u{a.) l} a.)"_ > > > > > 2005/12/10, Randy MacDonald : > > toupper and tolower working on Unicode is certainly nice-to-have. > > > > In which script are toupper and tolower defined? > > > > I wonder if there's a upper/lower codepoint mapping defined by > Unicode.org. > > > > ----------------------------------------------------------------------- > > |\/| Randy A MacDonald | you can't pay for it, > > |/\| ramacd@nbnet.nb.ca | even if you want to. > > BSc(Math) UNBF'83 Sapere Aude | APL: If you can say it, it's done.. > > Natural Born APL'er | Demo website: http://142.166.105.166/ > > ----------------------------------------------------(INTP)----{ gnat }- > > > > ----- Original Message ----- > > From: "bill lam" > > To: > > Sent: Saturday, December 10, 2005 5:29 AM > > Subject: [Jbeta] tolower and toupper > > > > > > tolower and toupper will give index error for character above 255. > > Chinese (and Japanese?) does not have this but Many European languages > > (Islandic?) do have lower/upper case alphabets. It will be nice to > > make tolower/toupper also work for them. > > > > ---------------------------------------------------------------------- > > For information about J forums see http://www.jsoftware.com/forums.htm > > > > > > ---------------------------------------------------------------------- > > For information about J forums see http://www.jsoftware.com/forums.htm > > > > > -- > Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 > Fugl&Fiskur ehf, ?erneyjarsund 23, > 802 Selfoss ,t-p?st: gosinn@gmail.com > http://www.linux.is/forritun/data/J/grunnur.pdf > > > T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans > > g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num > /|_ .-----------------------------------. > ,' .\ / | Me? l?ttri lund ver?ur | > ,--' _,' | Dagurinn ? dag | > / / | Enn betri en g?rdagurinn | > ( -. | `-----------------------------------' > | ) | > (`-. '--.) > `. )----' > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > -- Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 Fugl&Fiskur ehf, ?erneyjarsund 23, 802 Selfoss ,t-p?st: gosinn@gmail.com http://www.linux.is/forritun/data/J/grunnur.pdf T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num /|_ .-----------------------------------. ,' .\ / | Me? l?ttri lund ver?ur | ,--' _,' | Dagurinn ? dag | / / | Enn betri en g?rdagurinn | ( -. | `-----------------------------------' | ) | (`-. '--.) `. )----' From eiverson at sympatico.ca Sat Dec 10 23:22:42 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Sat Dec 10 23:20:57 2005 Subject: [Jbeta] missing gl2 commands References: <1134222815.c82b7ddcbill_lam@myrealbox.com> Message-ID: <005b01c5fd9d$93a11f30$0701a8c0@xp> Please work your way through the gl2 printing lab. Some capabilities were dropped in favor of making J printing identical across platforms. I don't recall the details right now. Let us know if you feel critical stuff is missing and we can see about adding it back in as appropriate. ----- Original Message ----- From: "bill lam" To: Sent: Saturday, December 10, 2005 8:53 AM Subject: Re: [Jbeta] missing gl2 commands glqdevmode,glzdevmode and glzorientation are essential for printing under win32, without them it is difficult to do professional printing. Are they still available under J601 or docs showing how to get it done? ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm From eiverson at sympatico.ca Sat Dec 10 23:32:40 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Sat Dec 10 23:30:55 2005 Subject: [Jbeta] 1!:x and unicode References: <1134206354.c8296adcbill_lam@myrealbox.com> Message-ID: <006901c5fd9e$f6e46330$0701a8c0@xp> There are some nasty issues here that we have not faced up to yet. I'll dig into the problem and see what can be done. I assume you'd like the 1!:x filename interface to be based on utf8. I'm not sure how this works out in windows and java. ----- Original Message ----- From: "bill lam" To: Sent: Saturday, December 10, 2005 4:19 AM Subject: [Jbeta] 1!:x and unicode It appears 1!:x family is not unicode aware. e.g. file name supplied to 1!:0 must be multibyte character. unicode and utf8 is not supported. The file names returned is also in multibyte (code page dependent). ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm From ramacd at nbnet.nb.ca Sat Dec 10 23:33:50 2005 From: ramacd at nbnet.nb.ca (Randy MacDonald) Date: Sat Dec 10 23:31:57 2005 Subject: [Jbeta] tolower and toupper References: <1134206992.c8296adcbill_lam@myrealbox.com><00c101c5fd98$86842270$0302a8c0@LONDON4Y78ST8><4031fef0512100651x4cb4a740j@mail.gmail.com><014801c5fd9a$f6f30600$0302a8c0@LONDON4Y78ST8> <4031fef0512100718s12a4c925k@mail.gmail.com> Message-ID: <015b01c5fd9f$1ecc9ac0$0302a8c0@LONDON4Y78ST8> As far as I can tell, a codepage is a way to map i.256 onto glyphs. This suggests a verb that takes a code page and an element of i.256, or a character noun, and returns a Unicode character: 850 unicode_of_cp_n 'a' a 850 unicode_of_cp_n 97 a This demo, of course, would need to be extended to character values >127. ----------------------------------------------------------------------- |\/| Randy A MacDonald | you can't pay for it, |/\| ramacd@nbnet.nb.ca | even if you want to. BSc(Math) UNBF'83 Sapere Aude | APL: If you can say it, it's done.. Natural Born APL'er | Demo website: http://142.166.105.166/ ----------------------------------------------------(INTP)----{ gnat }- ----- Original Message ----- From: "Bj?rn Helgason" To: "Beta forum" Sent: Saturday, December 10, 2005 11:18 AM Subject: Re: [Jbeta] tolower and toupper I used to need these kind of verbs a lot earlier (maybe in the future??!!??) There are several codepages for ?slandish chars So there are hell of a lot of files created in one operating system with one codepage and when you receive such a file using another codepage you need to these kind of exercises to move the chars around codepage 850, 861, EBCDIC, Roman8, 1262, some special codepages as well >From now on I hope to only need to move the chars up to Unicode and not bother with them ever again after that (wishful thinking!!) That is one of the reasons I am very very happy about get such a strong support for Unicode in J There are similar considerations with the APL chars between different implementations of Quad AV - they will also get easier in Unicode You can actually display APL chars very nicely with this new Unicode support in J As usual J is way ahead of the other APL dialects 2005/12/10, Randy MacDonald : > Thanks for the pointer (and the actual script... who needs bandwidth when > one has J)! > > It looks like extending toupper and tolower is an exercise for the reader, > something to share perhaps on the jwiki. > > > The following link looks like a workable standard: > > http://www.unicode.org/reports/tr21/tr21-5.html > > ----------------------------------------------------------------------- > |\/| Randy A MacDonald | you can't pay for it, > |/\| ramacd@nbnet.nb.ca | even if you want to. > BSc(Math) UNBF'83 Sapere Aude | APL: If you can say it, it's done.. > Natural Born APL'er | Demo website: http://142.166.105.166/ > ----------------------------------------------------(INTP)----{ gnat }- > > ----- Original Message ----- > From: "Bj?rn Helgason" > To: "Beta forum" > Sent: Saturday, December 10, 2005 10:51 AM > Subject: Re: [Jbeta] tolower and toupper > > > system/main/stdlib.ijs > NB.*tolower v convert text to lower case > NB.*toupper v convert text to upper case > 'l u'=. (a.i.'aA') +each tolower=: a.&i. { ((l{a.) u} a.)"_ > toupper=: a.&i. { ((u{a.) l} a.)"_ > > > > > 2005/12/10, Randy MacDonald : > > toupper and tolower working on Unicode is certainly nice-to-have. > > > > In which script are toupper and tolower defined? > > > > I wonder if there's a upper/lower codepoint mapping defined by > Unicode.org. > > > > ----------------------------------------------------------------------- > > |\/| Randy A MacDonald | you can't pay for it, > > |/\| ramacd@nbnet.nb.ca | even if you want to. > > BSc(Math) UNBF'83 Sapere Aude | APL: If you can say it, it's done.. > > Natural Born APL'er | Demo website: http://142.166.105.166/ > > ----------------------------------------------------(INTP)----{ gnat }- > > > > ----- Original Message ----- > > From: "bill lam" > > To: > > Sent: Saturday, December 10, 2005 5:29 AM > > Subject: [Jbeta] tolower and toupper > > > > > > tolower and toupper will give index error for character above 255. > > Chinese (and Japanese?) does not have this but Many European languages > > (Islandic?) do have lower/upper case alphabets. It will be nice to > > make tolower/toupper also work for them. > > > > ---------------------------------------------------------------------- > > For information about J forums see http://www.jsoftware.com/forums.htm > > > > > > ---------------------------------------------------------------------- > > For information about J forums see http://www.jsoftware.com/forums.htm > > > > > -- > Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 > Fugl&Fiskur ehf, ?erneyjarsund 23, > 802 Selfoss ,t-p?st: gosinn@gmail.com > http://www.linux.is/forritun/data/J/grunnur.pdf > > > T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans > > g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num > /|_ .-----------------------------------. > ,' .\ / | Me? l?ttri lund ver?ur | > ,--' _,' | Dagurinn ? dag | > / / | Enn betri en g?rdagurinn | > ( -. | `-----------------------------------' > | ) | > (`-. '--.) > `. )----' > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > -- Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 Fugl&Fiskur ehf, ?erneyjarsund 23, 802 Selfoss ,t-p?st: gosinn@gmail.com http://www.linux.is/forritun/data/J/grunnur.pdf T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num /|_ .-----------------------------------. ,' .\ / | Me? l?ttri lund ver?ur | ,--' _,' | Dagurinn ? dag | / / | Enn betri en g?rdagurinn | ( -. | `-----------------------------------' | ) | (`-. '--.) `. )----' ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm From gosinn at gmail.com Sat Dec 10 23:37:42 2005 From: gosinn at gmail.com (=?ISO-8859-1?Q?Bj=F6rn_Helgason?=) Date: Sat Dec 10 23:35:55 2005 Subject: [Jbeta] tolower and toupper In-Reply-To: <015b01c5fd9f$1ecc9ac0$0302a8c0@LONDON4Y78ST8> References: <1134206992.c8296adcbill_lam@myrealbox.com> <00c101c5fd98$86842270$0302a8c0@LONDON4Y78ST8> <4031fef0512100651x4cb4a740j@mail.gmail.com> <014801c5fd9a$f6f30600$0302a8c0@LONDON4Y78ST8> <4031fef0512100718s12a4c925k@mail.gmail.com> <015b01c5fd9f$1ecc9ac0$0302a8c0@LONDON4Y78ST8> Message-ID: <4031fef0512100737k5b0b4fdn@mail.gmail.com> a=.'????' ucpcount a 4 #a 8 a. i. a 195 190 195 166 195 182 195 176 2005/12/10, Randy MacDonald : > As far as I can tell, a codepage is a way to map i.256 onto glyphs. This > suggests a verb that takes a code page and an element of i.256, or a > character noun, and returns a Unicode character: > > 850 unicode_of_cp_n 'a' > a > 850 unicode_of_cp_n 97 > a > > This demo, of course, would need to be extended to character values >127. > > ----------------------------------------------------------------------- > |\/| Randy A MacDonald | you can't pay for it, > |/\| ramacd@nbnet.nb.ca | even if you want to. > BSc(Math) UNBF'83 Sapere Aude | APL: If you can say it, it's done.. > Natural Born APL'er | Demo website: http://142.166.105.166/ > ----------------------------------------------------(INTP)----{ gnat }- > > ----- Original Message ----- > From: "Bj?rn Helgason" > To: "Beta forum" > Sent: Saturday, December 10, 2005 11:18 AM > Subject: Re: [Jbeta] tolower and toupper > > > I used to need these kind of verbs a lot earlier > (maybe in the future??!!??) > There are several codepages for ?slandish chars > So there are hell of a lot of files created in one operating system > with one codepage and when you receive such a file using another > codepage you need to these kind of exercises to move the chars around > codepage 850, 861, EBCDIC, Roman8, 1262, some special codepages as well > > >From now on I hope to only need to move the chars up to Unicode and > not bother with them ever again after that (wishful thinking!!) > > That is one of the reasons I am very very happy about get such a > strong support for Unicode in J > > There are similar considerations with the APL chars between different > implementations of Quad AV - they will also get easier in Unicode > You can actually display APL chars very nicely with this new Unicode > support in J > > As usual J is way ahead of the other APL dialects > > 2005/12/10, Randy MacDonald : > > Thanks for the pointer (and the actual script... who needs bandwidth when > > one has J)! > > > > It looks like extending toupper and tolower is an exercise for the reader, > > something to share perhaps on the jwiki. > > > > > > The following link looks like a workable standard: > > > > http://www.unicode.org/reports/tr21/tr21-5.html > > > > ----------------------------------------------------------------------- > > |\/| Randy A MacDonald | you can't pay for it, > > |/\| ramacd@nbnet.nb.ca | even if you want to. > > BSc(Math) UNBF'83 Sapere Aude | APL: If you can say it, it's done.. > > Natural Born APL'er | Demo website: http://142.166.105.166/ > > ----------------------------------------------------(INTP)----{ gnat }- > > > > ----- Original Message ----- > > From: "Bj?rn Helgason" > > To: "Beta forum" > > Sent: Saturday, December 10, 2005 10:51 AM > > Subject: Re: [Jbeta] tolower and toupper > > > > > > system/main/stdlib.ijs > > NB.*tolower v convert text to lower case > > NB.*toupper v convert text to upper case > > 'l u'=. (a.i.'aA') +each > tolower=: a.&i. { ((l{a.) u} a.)"_ > > toupper=: a.&i. { ((u{a.) l} a.)"_ > > > > > > > > > > 2005/12/10, Randy MacDonald : > > > toupper and tolower working on Unicode is certainly nice-to-have. > > > > > > In which script are toupper and tolower defined? > > > > > > I wonder if there's a upper/lower codepoint mapping defined by > > Unicode.org. > > > > > > ----------------------------------------------------------------------- > > > |\/| Randy A MacDonald | you can't pay for it, > > > |/\| ramacd@nbnet.nb.ca | even if you want to. > > > BSc(Math) UNBF'83 Sapere Aude | APL: If you can say it, it's done.. > > > Natural Born APL'er | Demo website: > http://142.166.105.166/ > > > ----------------------------------------------------(INTP)----{ gnat }- > > > > > > ----- Original Message ----- > > > From: "bill lam" > > > To: > > > Sent: Saturday, December 10, 2005 5:29 AM > > > Subject: [Jbeta] tolower and toupper > > > > > > > > > tolower and toupper will give index error for character above 255. > > > Chinese (and Japanese?) does not have this but Many European languages > > > (Islandic?) do have lower/upper case alphabets. It will be nice to > > > make tolower/toupper also work for them. > > > > > > ---------------------------------------------------------------------- > > > For information about J forums see http://www.jsoftware.com/forums.htm > > > > > > > > > ---------------------------------------------------------------------- > > > For information about J forums see http://www.jsoftware.com/forums.htm > > > > > > > > > -- > > Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 > > Fugl&Fiskur ehf, ?erneyjarsund 23, > > 802 Selfoss ,t-p?st: gosinn@gmail.com > > http://www.linux.is/forritun/data/J/grunnur.pdf > > > > > > T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans > > > > g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num > > /|_ .-----------------------------------. > > ,' .\ / | Me? l?ttri lund ver?ur | > > ,--' _,' | Dagurinn ? dag | > > / / | Enn betri en g?rdagurinn | > > ( -. | `-----------------------------------' > > | ) | > > (`-. '--.) > > `. )----' > > ---------------------------------------------------------------------- > > For information about J forums see http://www.jsoftware.com/forums.htm > > > > > > ---------------------------------------------------------------------- > > For information about J forums see http://www.jsoftware.com/forums.htm > > > > > -- > Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 > Fugl&Fiskur ehf, ?erneyjarsund 23, > 802 Selfoss ,t-p?st: gosinn@gmail.com > http://www.linux.is/forritun/data/J/grunnur.pdf > > > T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans > > g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num > /|_ .-----------------------------------. > ,' .\ / | Me? l?ttri lund ver?ur | > ,--' _,' | Dagurinn ? dag | > / / | Enn betri en g?rdagurinn | > ( -. | `-----------------------------------' > | ) | > (`-. '--.) > `. )----' > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > -- Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 Fugl&Fiskur ehf, ?erneyjarsund 23, 802 Selfoss ,t-p?st: gosinn@gmail.com http://www.linux.is/forritun/data/J/grunnur.pdf T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num /|_ .-----------------------------------. ,' .\ / | Me? l?ttri lund ver?ur | ,--' _,' | Dagurinn ? dag | / / | Enn betri en g?rdagurinn | ( -. | `-----------------------------------' | ) | (`-. '--.) `. )----' From bill_lam at myrealbox.com Sun Dec 11 00:46:29 2005 From: bill_lam at myrealbox.com (bill lam) Date: Sun Dec 11 00:44:43 2005 Subject: [Jbeta] missing gl2 commands Message-ID: <1134233189.c822e55cbill_lam@myrealbox.com> I tried gl2 printing lab and realised J601 try to make printing on printer the same as painting on screen. I suppose this is a mistake. Physically, screen is in a state of constant change, therefore an event handler is called to repaint, but printing on paper is immutable once paper is ejected. In programming logic, if I need to print a report of 1000 records, I will write a loop to print each record, test if the page is full, eject one page and start a new page and continue to write next record. The paint event handler is suitable for printing plots but not for printing text reports. glqdevmode,glzdevmode are critical because they can be used to select paper size, set custom form size, number of copy. Also it can get the printable paper size before actually starting print job. IMHO, it is unreasonable to cut functionality that was previously available in win32 just because they are absent in Linux/Mac. From gosinn at gmail.com Sun Dec 11 00:55:39 2005 From: gosinn at gmail.com (=?ISO-8859-1?Q?Bj=F6rn_Helgason?=) Date: Sun Dec 11 00:53:54 2005 Subject: [Jbeta] missing gl2 commands In-Reply-To: <1134233189.c822e55cbill_lam@myrealbox.com> References: <1134233189.c822e55cbill_lam@myrealbox.com> Message-ID: <4031fef0512100855r6bc2d99bh@mail.gmail.com> Can you not paint out to a printfile and send it later to a printer? 2005/12/10, bill lam : > I tried gl2 printing lab and realised J601 try to make printing on printer > the same as painting on screen. I suppose this is a mistake. > Physically, screen is in a state of constant change, therefore an > event handler is called to repaint, but printing on paper is immutable > once paper is ejected. In programming logic, if I need to print a report > of 1000 records, I will write a loop to print each record, test if the > page is full, eject one page and start a new page and continue to > write next record. The paint event handler is suitable for printing > plots but not for printing text reports. > > glqdevmode,glzdevmode are critical because they can be used to select paper > size, set custom form size, number of copy. Also it can get the printable > paper size before actually starting print job. > > IMHO, it is unreasonable to cut functionality that was previously available in win32 > just because they are absent in Linux/Mac. > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > -- Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 Fugl&Fiskur ehf, ?erneyjarsund 23, 802 Selfoss ,t-p?st: gosinn@gmail.com http://www.linux.is/forritun/data/J/grunnur.pdf T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num /|_ .-----------------------------------. ,' .\ / | Me? l?ttri lund ver?ur | ,--' _,' | Dagurinn ? dag | / / | Enn betri en g?rdagurinn | ( -. | `-----------------------------------' | ) | (`-. '--.) `. )----' From gosinn at gmail.com Sun Dec 11 01:32:29 2005 From: gosinn at gmail.com (=?ISO-8859-1?Q?Bj=F6rn_Helgason?=) Date: Sun Dec 11 01:30:42 2005 Subject: [Jbeta] missing gl2 commands In-Reply-To: <005b01c5fd9d$93a11f30$0701a8c0@xp> References: <1134222815.c82b7ddcbill_lam@myrealbox.com> <005b01c5fd9d$93a11f30$0701a8c0@xp> Message-ID: <4031fef0512100932i70abc47eo@mail.gmail.com> These two system\extras\labs\graphics\gl2.ijt system\extras\labs\system\gl2.ijt Are both LABTITLE=: 'Graphics - gl2 commands' I guess I need to delete one of them Probably system\extras\labs\system\gl2.ijt Right? 2005/12/10, Eric Iverson : > Please work your way through the gl2 printing lab. Some capabilities were > dropped in favor of making J printing identical across platforms. I don't > recall the details right now. Let us know if you feel critical stuff is > missing and we can see about adding it back in as appropriate. > > ----- Original Message ----- > From: "bill lam" > To: > Sent: Saturday, December 10, 2005 8:53 AM > Subject: Re: [Jbeta] missing gl2 commands > > > glqdevmode,glzdevmode and glzorientation are essential for printing > under win32, without them it is difficult to do professional printing. > Are they still available under J601 or docs showing how to get it done? > > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > -- Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 Fugl&Fiskur ehf, ?erneyjarsund 23, 802 Selfoss ,t-p?st: gosinn@gmail.com http://www.linux.is/forritun/data/J/grunnur.pdf T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num /|_ .-----------------------------------. ,' .\ / | Me? l?ttri lund ver?ur | ,--' _,' | Dagurinn ? dag | / / | Enn betri en g?rdagurinn | ( -. | `-----------------------------------' | ) | (`-. '--.) `. )----' From davidvj at verizon.net Sun Dec 11 04:11:36 2005 From: davidvj at verizon.net (David Vincent-Jones) Date: Sun Dec 11 04:10:05 2005 Subject: [Jbeta] Wdstatus display Message-ID: <000f01c5fdc5$efcbe970$2c01a8c0@DAVID> There appears to be something odd regarding wdstatus display on the beta. The information box initially appears without a background colour but changes appearance (with another odd appearance) if clicked during a long run. This may be a more general wd display problem. stattest=:verb define pos=:0 fin=:1000000 while. pos<005b01c5fd9d$93a11f30$0701a8c0@xp> <4031fef0512100932i70abc47eo@mail.gmail.com> Message-ID: <002401c5fdca$eb74f920$0701a8c0@xp> You need to download the latest c beta. The duplicate gl2 labs is fixed there. Note that this c version is newer than the test c beta I gave you earlier when working on the uft-8 problems. ----- Original Message ----- From: "Bj?rn Helgason" To: "Beta forum" Sent: Saturday, December 10, 2005 12:32 PM Subject: Re: [Jbeta] missing gl2 commands These two system\extras\labs\graphics\gl2.ijt system\extras\labs\system\gl2.ijt Are both LABTITLE=: 'Graphics - gl2 commands' I guess I need to delete one of them Probably system\extras\labs\system\gl2.ijt Right? 2005/12/10, Eric Iverson : > Please work your way through the gl2 printing lab. Some capabilities were > dropped in favor of making J printing identical across platforms. I don't > recall the details right now. Let us know if you feel critical stuff is > missing and we can see about adding it back in as appropriate. > > ----- Original Message ----- > From: "bill lam" > To: > Sent: Saturday, December 10, 2005 8:53 AM > Subject: Re: [Jbeta] missing gl2 commands > > > glqdevmode,glzdevmode and glzorientation are essential for printing > under win32, without them it is difficult to do professional printing. > Are they still available under J601 or docs showing how to get it done? > > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > -- Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 Fugl&Fiskur ehf, ?erneyjarsund 23, 802 Selfoss ,t-p?st: gosinn@gmail.com http://www.linux.is/forritun/data/J/grunnur.pdf T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num /|_ .-----------------------------------. ,' .\ / | Me? l?ttri lund ver?ur | ,--' _,' | Dagurinn ? dag | / / | Enn betri en g?rdagurinn | ( -. | `-----------------------------------' | ) | (`-. '--.) `. )----' ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm From gosinn at gmail.com Sun Dec 11 04:53:36 2005 From: gosinn at gmail.com (=?ISO-8859-1?Q?Bj=F6rn_Helgason?=) Date: Sun Dec 11 04:51:49 2005 Subject: [Jbeta] missing gl2 commands In-Reply-To: <002401c5fdca$eb74f920$0701a8c0@xp> References: <1134222815.c82b7ddcbill_lam@myrealbox.com> <005b01c5fd9d$93a11f30$0701a8c0@xp> <4031fef0512100932i70abc47eo@mail.gmail.com> <002401c5fdca$eb74f920$0701a8c0@xp> Message-ID: <4031fef0512101253q4972f283o@mail.gmail.com> looks fine now I renamed the old j601b and reinstalled 2005/12/10, Eric Iverson : > You need to download the latest c beta. The duplicate gl2 labs is fixed > there. Note that this c version is newer than the test c beta I gave you > earlier when working on the uft-8 problems. > > ----- Original Message ----- > From: "Bj?rn Helgason" > To: "Beta forum" > Sent: Saturday, December 10, 2005 12:32 PM > Subject: Re: [Jbeta] missing gl2 commands > > > These two > > system\extras\labs\graphics\gl2.ijt > > system\extras\labs\system\gl2.ijt > > Are both > > LABTITLE=: 'Graphics - gl2 commands' > > I guess I need to delete one of them > > Probably > system\extras\labs\system\gl2.ijt > > Right? > > > 2005/12/10, Eric Iverson : > > Please work your way through the gl2 printing lab. Some capabilities were > > dropped in favor of making J printing identical across platforms. I don't > > recall the details right now. Let us know if you feel critical stuff is > > missing and we can see about adding it back in as appropriate. > > > > ----- Original Message ----- > > From: "bill lam" > > To: > > Sent: Saturday, December 10, 2005 8:53 AM > > Subject: Re: [Jbeta] missing gl2 commands > > > > > > glqdevmode,glzdevmode and glzorientation are essential for printing > > under win32, without them it is difficult to do professional printing. > > Are they still available under J601 or docs showing how to get it done? > > > > > > > > ---------------------------------------------------------------------- > > For information about J forums see http://www.jsoftware.com/forums.htm > > > > ---------------------------------------------------------------------- > > For information about J forums see http://www.jsoftware.com/forums.htm > > > > > -- > Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 > Fugl&Fiskur ehf, ?erneyjarsund 23, > 802 Selfoss ,t-p?st: gosinn@gmail.com > http://www.linux.is/forritun/data/J/grunnur.pdf > > > T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans > > g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num > /|_ .-----------------------------------. > ,' .\ / | Me? l?ttri lund ver?ur | > ,--' _,' | Dagurinn ? dag | > / / | Enn betri en g?rdagurinn | > ( -. | `-----------------------------------' > | ) | > (`-. '--.) > `. )----' > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > -- Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 Fugl&Fiskur ehf, ?erneyjarsund 23, 802 Selfoss ,t-p?st: gosinn@gmail.com http://www.linux.is/forritun/data/J/grunnur.pdf T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num /|_ .-----------------------------------. ,' .\ / | Me? l?ttri lund ver?ur | ,--' _,' | Dagurinn ? dag | / / | Enn betri en g?rdagurinn | ( -. | `-----------------------------------' | ) | (`-. '--.) `. )----' From eiverson at sympatico.ca Sun Dec 11 05:04:49 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Sun Dec 11 05:48:04 2005 Subject: [Jbeta] Wdstatus display References: <000f01c5fdc5$efcbe970$2c01a8c0@DAVID> Message-ID: <000c01c5fdd3$a794c9c0$0701a8c0@xp> The problem seems to be the same in J504 so is not new to the J601 beta. If you do a wdstatus first and give the system a chance to paint its background then stattest works as you would like. The trouble is in your example compute loop the system never gets to paint the background whereas the text area is being updated immediattely. Some minor changes to how wdstatus is coded could fix this, but it won't be a high priority as it has been this way for quite some time. ----- Original Message ----- From: "David Vincent-Jones" To: "JSoftware Beta" Sent: Saturday, December 10, 2005 3:11 PM Subject: [Jbeta] Wdstatus display > There appears to be something odd regarding wdstatus display on the beta. > > The information box initially appears without a background colour but > changes appearance (with another odd appearance) if clicked during a long > run. > > This may be a more general wd display problem. > > stattest=:verb define > pos=:0 > fin=:1000000 > while. pos if. 0=1000|pos > do.' position' wdstatus 'position ',(":pos),' Completion > ',(0j1":100*pos%$fin),'%' > end. > pos=:pos+1 > end. > wdstatus '' > ) > > David > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm From eiverson at sympatico.ca Sun Dec 11 05:49:50 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Sun Dec 11 05:48:06 2005 Subject: [Jbeta] missing gl2 commands References: <1134233189.c822e55cbill_lam@myrealbox.com> Message-ID: <000d01c5fdd3$a8a24220$0701a8c0@xp> There are two different issues here. One is that printing of a page is done in an event handler and the other is the functionality of devmode etc. If there is missing functionality, it can be added back in to meet critical requirements. The more important question is the page print handler. Although it is a slightly different view of things than existed in J504, it is in fact more typical of modern printing interfaces and I think with adjustments to your worldview works well in the mode you describe. In both cases there is a loop that runs printing one page and then the next. I suggest that it isn't a big difference if the master loop is your J while. or whether the master loop is in the host as the code called inside the loop is essentially the same. I may be wrong on this, but ask you to please take a careful look. ----- Original Message ----- From: "bill lam" To: Sent: Saturday, December 10, 2005 11:46 AM Subject: Re: [Jbeta] missing gl2 commands I tried gl2 printing lab and realised J601 try to make printing on printer the same as painting on screen. I suppose this is a mistake. Physically, screen is in a state of constant change, therefore an event handler is called to repaint, but printing on paper is immutable once paper is ejected. In programming logic, if I need to print a report of 1000 records, I will write a loop to print each record, test if the page is full, eject one page and start a new page and continue to write next record. The paint event handler is suitable for printing plots but not for printing text reports. glqdevmode,glzdevmode are critical because they can be used to select paper size, set custom form size, number of copy. Also it can get the printable paper size before actually starting print job. IMHO, it is unreasonable to cut functionality that was previously available in win32 just because they are absent in Linux/Mac. ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm From eiverson at sympatico.ca Sun Dec 11 05:51:34 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Sun Dec 11 05:49:48 2005 Subject: [Jbeta] 1!:x and unicode References: <1134206354.c8296adcbill_lam@myrealbox.com> Message-ID: <001b01c5fdd3$e54cce70$0701a8c0@xp> I have a test version running that has utf-8 support for 1!:x. It looks good and I'll probably get it out tomorrow if tests don't show up problems. ----- Original Message ----- From: "bill lam" To: Sent: Saturday, December 10, 2005 4:19 AM Subject: [Jbeta] 1!:x and unicode It appears 1!:x family is not unicode aware. e.g. file name supplied to 1!:0 must be multibyte character. unicode and utf8 is not supported. The file names returned is also in multibyte (code page dependent). ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm From juneaftn at gmail.com Sun Dec 11 08:17:18 2005 From: juneaftn at gmail.com (June Kim) Date: Sun Dec 11 08:15:30 2005 Subject: [Jbeta] J 601 c beta available In-Reply-To: <003c01c5fd3c$dfc59460$0701a8c0@xp> References: <003e01c5fd11$5ed6e460$0701a8c0@xp> <6f80c1520512091640k1843c528p7ff7c92fa73896e4@mail.gmail.com> <6f80c1520512091650n135e2f23x6bda771f839e6a1@mail.gmail.com> <003c01c5fd3c$dfc59460$0701a8c0@xp> Message-ID: <6f80c1520512101617k1321733av98e211768925aeea@mail.gmail.com> Thank you for the explanation. On 12/10/05, Eric Iverson wrote: > The problem is that the result of 7 u: for a 2 or 3 character sequence that > returns a single wchar is a list. As the argument is a list there is no way > to determine if the result should be a scalar or a list. The ASCII arg to 7 > u: is currently returned as unchanged. It might be more consistent to make 7 > u: always return a list even in the ASCII case. But this wouldn't help your > situation. > > A possible change would be to make 7 u: return a scalar rather than a list > if there was only 1 wchar in the result, but then when you wanted a list > you'd be upset and I suspect this would just move the problem around. > > The problem is that the these transformation don't have the nice shape and > rank conformation of typical J primitives. That is, 1, 2, or 3 char > sequences in the arg make 1 wchar in the result. > > Feedback and suggestions welcome. I think, when any character(in any encoding) is decoded into unicode, all of them should be treated equally, whether it be korean or not, and that's one reason of why unicode was born. > > I could redo all your HAN experiments, but the character doesn't display > properly for me in ijs or ijx. It does display properly for me in your email > and in notepad. What font are you using in ijs? Do you know why it won't > display in my ijs window? I can display Chinese, Icelandic, etc ok. > I use Bitstream Vera Sans Mono usually(you can download it free from the internet) but the font doesn't have korean characters and when I try to use korean characters using the font, the system automatically fall back to other korean default font instead. So I explicitly changed the J system font into Arial Unicode MS and it worked. If you use Arial Unicode MS, it should show the korean character. Also note that the HAN character as in UTF8(that is, no conversion or no explicit encoding/decoding) is shown correctly but the unicode form (7 u: HAN) is not shown properly. From gosinn at gmail.com Sun Dec 11 08:31:39 2005 From: gosinn at gmail.com (=?ISO-8859-1?Q?Bj=F6rn_Helgason?=) Date: Sun Dec 11 08:29:51 2005 Subject: [Jbeta] J 601 c beta available In-Reply-To: <6f80c1520512101617k1321733av98e211768925aeea@mail.gmail.com> References: <003e01c5fd11$5ed6e460$0701a8c0@xp> <6f80c1520512091640k1843c528p7ff7c92fa73896e4@mail.gmail.com> <6f80c1520512091650n135e2f23x6bda771f839e6a1@mail.gmail.com> <003c01c5fd3c$dfc59460$0701a8c0@xp> <6f80c1520512101617k1321733av98e211768925aeea@mail.gmail.com> Message-ID: <4031fef0512101631v7e0422cx@mail.gmail.com> It displays fine with "@Arial Unicode MS" 12 ucpcount HAN 1 #HAN 3 a. i. HAN 237 149 156 2005/12/11, June Kim : > Thank you for the explanation. > > On 12/10/05, Eric Iverson wrote: > > The problem is that the result of 7 u: for a 2 or 3 character sequence that > > returns a single wchar is a list. As the argument is a list there is no way > > to determine if the result should be a scalar or a list. The ASCII arg to 7 > > u: is currently returned as unchanged. It might be more consistent to make 7 > > u: always return a list even in the ASCII case. But this wouldn't help your > > situation. > > > > A possible change would be to make 7 u: return a scalar rather than a list > > if there was only 1 wchar in the result, but then when you wanted a list > > you'd be upset and I suspect this would just move the problem around. > > > > The problem is that the these transformation don't have the nice shape and > > rank conformation of typical J primitives. That is, 1, 2, or 3 char > > sequences in the arg make 1 wchar in the result. > > > > Feedback and suggestions welcome. > > I think, when any character(in any encoding) is decoded into unicode, > all of them should be treated equally, whether it be korean or not, > and that's one reason of why unicode was born. > > > > > I could redo all your HAN experiments, but the character doesn't display > > properly for me in ijs or ijx. It does display properly for me in your email > > and in notepad. What font are you using in ijs? Do you know why it won't > > display in my ijs window? I can display Chinese, Icelandic, etc ok. > > > > I use Bitstream Vera Sans Mono usually(you can download it free from > the internet) but the font doesn't have korean characters and when I > try to use korean characters using the font, the system automatically > fall back to other korean default font instead. So I explicitly > changed the J system font into Arial Unicode MS and it worked. If you > use Arial Unicode MS, it should show the korean character. > > Also note that the HAN character as in UTF8(that is, no conversion or > no explicit encoding/decoding) is shown correctly but the unicode form > (7 u: HAN) is not shown properly. > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > -- Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 Fugl&Fiskur ehf, ?erneyjarsund 23, 802 Selfoss ,t-p?st: gosinn@gmail.com http://www.linux.is/forritun/data/J/grunnur.pdf T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num /|_ .-----------------------------------. ,' .\ / | Me? l?ttri lund ver?ur | ,--' _,' | Dagurinn ? dag | / / | Enn betri en g?rdagurinn | ( -. | `-----------------------------------' | ) | (`-. '--.) `. )----' From bill_lam at myrealbox.com Sun Dec 11 14:47:34 2005 From: bill_lam at myrealbox.com (bill lam) Date: Sun Dec 11 14:45:47 2005 Subject: [Jbeta] missing gl2 commands Message-ID: <1134283654.c82f09bcbill_lam@myrealbox.com> >it is in fact more typical of modern printing interfaces and I think with >adjustments to your worldview works well in the mode you describe. Thank you for pointing out the world has changed this way. I still consider it counter productive. I did web search and found out in Net framework, gui boys' primary concern is how to paint graphics on printer. This is unfortunate but I have to live with it. May be I can use workaround like wrapper or direct window api call. From bill_lam at myrealbox.com Sun Dec 11 14:56:51 2005 From: bill_lam at myrealbox.com (bill lam) Date: Sun Dec 11 14:55:04 2005 Subject: [Jbeta] utf8 and EAV Message-ID: <1134284211.c82f09bcbill_lam@myrealbox.com> EAV is used extensively as delimiter in wd commands, but I found there are some problems in J601. 1. fe and ff should be illegal characters inside utf-8 encoding. 7&u: should give domain error if string contains them. 2. EAV either as single char (255{a.) or as utf-8 (195 191{a.) does not function as delimiter. For example, replacing LF in unisimple demo by EAV. From bill_lam at myrealbox.com Sun Dec 11 18:12:34 2005 From: bill_lam at myrealbox.com (bill lam) Date: Sun Dec 11 18:10:48 2005 Subject: [Jbeta] olesetlic Message-ID: <1134295954.c826347cbill_lam@myrealbox.com> I believe olesetlic is not working. although olegetlic can still get the license key (is it cached?),but it will cause domain error when try to create a form with ocx. From bill_lam at myrealbox.com Sun Dec 11 18:25:23 2005 From: bill_lam at myrealbox.com (bill lam) Date: Sun Dec 11 18:23:35 2005 Subject: [Jbeta] 0.3 ": Message-ID: <1134296723.c7f8399cbill_lam@myrealbox.com> there is a 0.3 ": inside C:\j601\system\main\dates.ijs it should be changed to 0j3 From eiverson at sympatico.ca Sun Dec 11 22:58:58 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Sun Dec 11 22:57:16 2005 Subject: [Jbeta] J 601 c beta available References: <003e01c5fd11$5ed6e460$0701a8c0@xp><6f80c1520512091640k1843c528p7ff7c92fa73896e4@mail.gmail.com><6f80c1520512091650n135e2f23x6bda771f839e6a1@mail.gmail.com><003c01c5fd3c$dfc59460$0701a8c0@xp><6f80c1520512101617k1321733av98e211768925aeea@mail.gmail.com> <4031fef0512101631v7e0422cx@mail.gmail.com> Message-ID: <001a01c5fe63$6ca5f4c0$0701a8c0@xp> Your suggestions don't work for me. It would seem (even though I've installed the MS east asian language stuff) that I don't have a font on my machine with Korean. Note the following: k=: 237 149 156{a. NB.Korean utf8 character c=: 229 137 172{a. NB. Chinese utf8 character k utf16 value is 16bd55c and c is 16b526c When I look in Character map for the courier new font I normally use I see that neither of these characters is in that font. But when I display k the system manages to find the character (presumably in some other font) and do the display. But it can't find the k character so the display is a box. My conclusion is that I don't have a font on my machine with the Korean characters. And font stuff is always more complicated and mysterious than one imagines. That said, I do think utf-8 is going to work well for the J community. ----- Original Message ----- From: "Bj?rn Helgason" To: "Beta forum" Sent: Saturday, December 10, 2005 7:31 PM Subject: Re: [Jbeta] J 601 c beta available It displays fine with "@Arial Unicode MS" 12 ucpcount HAN 1 #HAN 3 a. i. HAN 237 149 156 2005/12/11, June Kim : > Thank you for the explanation. > > On 12/10/05, Eric Iverson wrote: > > The problem is that the result of 7 u: for a 2 or 3 character sequence > > that > > returns a single wchar is a list. As the argument is a list there is no > > way > > to determine if the result should be a scalar or a list. The ASCII arg > > to 7 > > u: is currently returned as unchanged. It might be more consistent to > > make 7 > > u: always return a list even in the ASCII case. But this wouldn't help > > your > > situation. > > > > A possible change would be to make 7 u: return a scalar rather than a > > list > > if there was only 1 wchar in the result, but then when you wanted a list > > you'd be upset and I suspect this would just move the problem around. > > > > The problem is that the these transformation don't have the nice shape > > and > > rank conformation of typical J primitives. That is, 1, 2, or 3 char > > sequences in the arg make 1 wchar in the result. > > > > Feedback and suggestions welcome. > > I think, when any character(in any encoding) is decoded into unicode, > all of them should be treated equally, whether it be korean or not, > and that's one reason of why unicode was born. > > > > > I could redo all your HAN experiments, but the character doesn't display > > properly for me in ijs or ijx. It does display properly for me in your > > email > > and in notepad. What font are you using in ijs? Do you know why it won't > > display in my ijs window? I can display Chinese, Icelandic, etc ok. > > > > I use Bitstream Vera Sans Mono usually(you can download it free from > the internet) but the font doesn't have korean characters and when I > try to use korean characters using the font, the system automatically > fall back to other korean default font instead. So I explicitly > changed the J system font into Arial Unicode MS and it worked. If you > use Arial Unicode MS, it should show the korean character. > > Also note that the HAN character as in UTF8(that is, no conversion or > no explicit encoding/decoding) is shown correctly but the unicode form > (7 u: HAN) is not shown properly. > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > -- Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 Fugl&Fiskur ehf, ?erneyjarsund 23, 802 Selfoss ,t-p?st: gosinn@gmail.com http://www.linux.is/forritun/data/J/grunnur.pdf T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num /|_ .-----------------------------------. ,' .\ / | Me? l?ttri lund ver?ur | ,--' _,' | Dagurinn ? dag | / / | Enn betri en g?rdagurinn | ( -. | `-----------------------------------' | ) | (`-. '--.) `. )----' ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm From eiverson at sympatico.ca Sun Dec 11 23:06:45 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Sun Dec 11 23:05:00 2005 Subject: [Jbeta] utf8 and EAV References: <1134284211.c82f09bcbill_lam@myrealbox.com> Message-ID: <002501c5fe64$84179590$0701a8c0@xp> Yes, there is a problem with the 'old' wd use of EAV (255{a.) as a delimiter. I need to point this out clearly in the release highlights. As you point out the use of utf8 means we can't use EAV as the delimiter. With hindsight this was a bad decision made many years ago. To move to utf-8 we need to abandon EAV as a wd delimiter. Chris has now defined DEL as 127{a. and this is now the delimiter recognized by wd. Chris and I weren't sure what the easiest migration path was. By default we picked: 1. define DEL as 127{a. and do careful search/replace of EAV by DEL If this doesn't work fairly easily for you, perhaps you could suggest a better approach. ----- Original Message ----- From: "bill lam" To: Sent: Sunday, December 11, 2005 1:56 AM Subject: [Jbeta] utf8 and EAV EAV is used extensively as delimiter in wd commands, but I found there are some problems in J601. 1. fe and ff should be illegal characters inside utf-8 encoding. 7&u: should give domain error if string contains them. 2. EAV either as single char (255{a.) or as utf-8 (195 191{a.) does not function as delimiter. For example, replacing LF in unisimple demo by EAV. ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm From donguinn at pvtnetworks.net Sun Dec 11 23:20:16 2005 From: donguinn at pvtnetworks.net (Don Guinn) Date: Sun Dec 11 23:18:35 2005 Subject: [Jbeta] gl2 labs In-Reply-To: <006901c5fd9e$f6e46330$0701a8c0@xp> References: <1134206354.c8296adcbill_lam@myrealbox.com> <006901c5fd9e$f6e46330$0701a8c0@xp> Message-ID: <439C43B0.1080104@pvtn.net> Just playing with the gl2 labs. Looks good. Might I suggest that you add gl2lab_jctrl_fkey=: jijs_labadvance_button_jijs_ to the lab. Then one doesn't have to switch to some ij? window before advancing. I find that I am in the lab window a lot when I press ctrl-j and nothing happens. From bill_lam at myrealbox.com Sun Dec 11 23:23:03 2005 From: bill_lam at myrealbox.com (bill lam) Date: Sun Dec 11 23:21:15 2005 Subject: [Jbeta] gl2 docs Message-ID: <1134314583.c80201dcbill_lam@myrealbox.com> In gl2 docs gltext text ; write text in the glfont font. Where and how the text is displayed is affected by the gltextalign and gltextxy commands. but gltextalign has been decommited, how to do gltextalign in J601? fontspec The size is in logical units. For wd commands this is generally point size. In gl2, size is in mapping mode units. In gl3, size is in pixels. gl print lab said font size is in point for printer, but for screen, is font size also always in point, or the fontspec docs is correct? isigraph events An isigraph control supports character and mouse events. The Code Dialog in the form editor lists all events for an isigraph control. size resize event (OpenGL, gl2 must request with glwantresize) but glwantresize is decommitted. For mouse events the wdhandler variable sysdata contains: x y width height leftbutton rightbutton ctrl shift In J504 lower left is 0 0 but in J601 upper left is 0 0, J601 seems more reasonable, is this an intended change? in Window Driver Command Reference Overview It still point to a mapping mode page, but glmap is already decommitted. now screen always use MM_RAW and printer always use MM_TWIPS. Also in win32, origin of MM_TWIPS is at lower left corner, is J601 use upper left corner as origin for MM_TWIPS. From eiverson at sympatico.ca Sun Dec 11 23:28:55 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Sun Dec 11 23:27:07 2005 Subject: [Jbeta] gl2 docs References: <1134314583.c80201dcbill_lam@myrealbox.com> Message-ID: <00a101c5fe67$9b2384d0$0701a8c0@xp> Thank you for pointing out these many documentation errors. I'll go through them and get to you in detail as required. ----- Original Message ----- From: "bill lam" To: Sent: Sunday, December 11, 2005 10:23 AM Subject: [Jbeta] gl2 docs In gl2 docs gltext text ; write text in the glfont font. Where and how the text is displayed is affected by the gltextalign and gltextxy commands. but gltextalign has been decommited, how to do gltextalign in J601? fontspec The size is in logical units. For wd commands this is generally point size. In gl2, size is in mapping mode units. In gl3, size is in pixels. gl print lab said font size is in point for printer, but for screen, is font size also always in point, or the fontspec docs is correct? isigraph events An isigraph control supports character and mouse events. The Code Dialog in the form editor lists all events for an isigraph control. size resize event (OpenGL, gl2 must request with glwantresize) but glwantresize is decommitted. For mouse events the wdhandler variable sysdata contains: x y width height leftbutton rightbutton ctrl shift In J504 lower left is 0 0 but in J601 upper left is 0 0, J601 seems more reasonable, is this an intended change? in Window Driver Command Reference Overview It still point to a mapping mode page, but glmap is already decommitted. now screen always use MM_RAW and printer always use MM_TWIPS. Also in win32, origin of MM_TWIPS is at lower left corner, is J601 use upper left corner as origin for MM_TWIPS. ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm From bill_lam at myrealbox.com Sun Dec 11 23:55:39 2005 From: bill_lam at myrealbox.com (bill lam) Date: Sun Dec 11 23:53:51 2005 Subject: [Jbeta] utf8 and EAV Message-ID: <1134316539.c809ee3cbill_lam@myrealbox.com> >Chris and I weren't sure what the easiest migration path was. By default we picked: >1. define DEL as 127{a. and do careful search/replace of EAV by DEL > >If this doesn't work fairly easily for you, perhaps you could suggest a better approach. 127{a. is a good choice. I also use it as delimiter is DOS many years ago. It was displayed as a triangle. From gosinn at gmail.com Mon Dec 12 00:14:05 2005 From: gosinn at gmail.com (=?ISO-8859-1?Q?Bj=F6rn_Helgason?=) Date: Mon Dec 12 00:12:21 2005 Subject: [Jbeta] utf8 and EAV In-Reply-To: <1134316539.c809ee3cbill_lam@myrealbox.com> References: <1134316539.c809ee3cbill_lam@myrealbox.com> Message-ID: <4031fef0512110814r58d6892fk@mail.gmail.com> Sounds like a good choice Looks like a house in some fonts like --- one char below - del d e l ----------- del ends in others 127 7F Del In EBCDIC EBCDIC Codepage 1026 (Turkish) 7F 127 ? 220 LATIN CAPITAL LETTER U WITH DIAERESIS 2005/12/11, bill lam : > >Chris and I weren't sure what the easiest migration path was. By default we picked: > >1. define DEL as 127{a. and do careful search/replace of EAV by DEL > > > >If this doesn't work fairly easily for you, perhaps you could suggest a better approach. > > 127{a. is a good choice. I also use it as delimiter is DOS many years ago. It was displayed as a triangle. > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > -- Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 Fugl&Fiskur ehf, ?erneyjarsund 23, 802 Selfoss ,t-p?st: gosinn@gmail.com http://www.linux.is/forritun/data/J/grunnur.pdf T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num /|_ .-----------------------------------. ,' .\ / | Me? l?ttri lund ver?ur | ,--' _,' | Dagurinn ? dag | / / | Enn betri en g?rdagurinn | ( -. | `-----------------------------------' | ) | (`-. '--.) `. )----' From bill_lam at myrealbox.com Mon Dec 12 00:50:10 2005 From: bill_lam at myrealbox.com (bill lam) Date: Mon Dec 12 00:48:24 2005 Subject: [Jbeta] verb verb undefname Message-ID: <1134319810.c82c593cbill_lam@myrealbox.com> Has 'verb verb undefname' as originally purposed by Henry Rich been implemented in J601? From RHui000 at shaw.ca Mon Dec 12 05:52:25 2005 From: RHui000 at shaw.ca (Roger Hui) Date: Mon Dec 12 05:50:59 2005 Subject: [Jbeta] verb verb undefname References: <1134319810.c82c593cbill_lam@myrealbox.com> Message-ID: <001101c5fe9d$36e31720$f7c35718@lap> If it is not in the release notes it has not been implemented yet. What is the original proposal for verb verb undefname? And how is it to be distinguished from verb verb verb since an undefined name is assumed to be a verb? ----- Original Message ----- From: "bill lam" To: Sent: Sunday, December 11, 2005 8:50 AM Subject: [Jbeta] verb verb undefname Has 'verb verb undefname' as originally purposed by Henry Rich been implemented in J601? From HenryHRich at nc.rr.com Mon Dec 12 10:10:01 2005 From: HenryHRich at nc.rr.com (Henry Rich) Date: Mon Dec 12 10:08:14 2005 Subject: [Jbeta] verb verb undefname In-Reply-To: <001101c5fe9d$36e31720$f7c35718@lap> Message-ID: <200512120210.jBC2A01u028168@ms-smtp-04-eri0.southeast.rr.com> If I recall, the idea was to have a sentence that produced a non-noun and didn't start with a copula, produce the 'otiose sentence' error. Refinements were proposed. Henry Rich > -----Original Message----- > From: beta-bounces@jsoftware.com > [mailto:beta-bounces@jsoftware.com] On Behalf Of Roger Hui > Sent: Sunday, December 11, 2005 4:52 PM > To: Beta forum > Subject: Re: [Jbeta] verb verb undefname > > If it is not in the release notes it has not been implemented yet. > > What is the original proposal for verb verb undefname? > And how is it to be distinguished from verb verb verb > since an undefined name is assumed to be a verb? > > > > ----- Original Message ----- > From: "bill lam" > To: > Sent: Sunday, December 11, 2005 8:50 AM > Subject: [Jbeta] verb verb undefname > > Has 'verb verb undefname' as originally purposed by Henry > Rich been implemented in J601? > > > ---------------------------------------------------------------------- > For information about J forums see > http://www.jsoftware.com/forums.htm From bill_lam at myrealbox.com Mon Dec 12 18:41:33 2005 From: bill_lam at myrealbox.com (bill lam) Date: Mon Dec 12 18:39:45 2005 Subject: [Jbeta] passing ole/ocx parameter Message-ID: <1134384093.c7e176dcbill_lam@myrealbox.com> It seems ole/ocx has problem in passing parameter, only the first character is read. I try using delimiter and using *, same result. xlopen=: 3 : 0 wd 'pc xlauto' wd 'cc xl oleautomation:excel.application' wd 'oleget xl base workbooks;oleid xl wb' wd 'olemethod xl wb open *C:\j601\system\examples\data\expense.xls' ) xlopen'' |domain error: wd | wd'olemethod xl wb open *C:\j601\system\examples\data\expense.xls' wd'qer' ole - 'C.xls' could not be found. Check the spelling of the file name, and verify that the file location is correct. If you are trying to open the file from your list of most recently used files on the File menu, make sure that the file has not been renamed, moved, or deleted. : 0 From HenryHRich at nc.rr.com Tue Dec 13 00:53:35 2005 From: HenryHRich at nc.rr.com (Henry Rich) Date: Tue Dec 13 00:51:47 2005 Subject: [Jbeta] J6 grid: requests In-Reply-To: <200512120210.jBC2A01u028168@ms-smtp-04-eri0.southeast.rr.com> Message-ID: <200512121653.jBCGrYWe022159@ms-smtp-01-eri0.southeast.rr.com> Trying to port my application that uses glgridxxx calls to the new grid, I find I need clarifications/extensions. 0. I use CTRL+leftclick, CTRL+leftdoubleclick, CTRL+richtclick in addition to those same clicks without the CTRL key. a. Am I right that left-clicking a cell will cause grid_handler 'click' to be called? And that I should look at the code for click to see what globals are set for this call? Can I find out what the button and shift-key state was? b. I don't see rclick and rdblclick events. Can they be added? 1. I use boldface in some of my cells. I see that support for boldface is listed as a not-too-hard feature. Any chance of getting that soon? Henry Rich From gkirouac at riq.qc.ca Tue Dec 13 01:03:09 2005 From: gkirouac at riq.qc.ca (Gilles Kirouac) Date: Tue Dec 13 01:01:20 2005 Subject: [Jbeta] 0.3 ": In-Reply-To: <1134296723.c7f8399cbill_lam@myrealbox.com> References: <1134296723.c7f8399cbill_lam@myrealbox.com> Message-ID: <20051212165539.M74561@riq.qc.ca> Same problem also on page j601/system/extras/help/user/grid_class.htm P.S. The insert button on the Find in files page is a little marvel of usefulness, helping to build regex expressions and look for things like above. Thanks Chris! Gilles ---------- Original Message ----------- From: "bill lam" To: beta@jsoftware.com Sent: Sun, 11 Dec 2005 10:25:23 +0000 Subject: [Jbeta] 0.3 ": > there is a 0.3 ": inside C:\j601\system\main\dates.ijs > it should be changed to 0j3 > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm ------- End of Original Message ------- From HenryHRich at nc.rr.com Tue Dec 13 01:13:55 2005 From: HenryHRich at nc.rr.com (Henry Rich) Date: Tue Dec 13 01:12:06 2005 Subject: [Jbeta] J6 grid: requests In-Reply-To: <200512121653.jBCGrYWe022159@ms-smtp-01-eri0.southeast.rr.com> Message-ID: <200512121713.jBCHDs1u013285@ms-smtp-04-eri0.southeast.rr.com> I also need to connect a scrollbar to my grid. There must be some way to indicate the index of the top-left displayed cell, but I can't find it. What is it? And do I have to support the scrollbar in code of my own? Henry Rich > -----Original Message----- > From: beta-bounces@jsoftware.com > [mailto:beta-bounces@jsoftware.com] On Behalf Of Henry Rich > Sent: Monday, December 12, 2005 11:54 AM > To: 'Beta forum' > Subject: [Jbeta] J6 grid: requests > > Trying to port my application that uses glgridxxx calls to the new > grid, I find I need clarifications/extensions. > > 0. I use CTRL+leftclick, CTRL+leftdoubleclick, CTRL+richtclick > in addition to those same clicks without the CTRL key. > > a. Am I right that left-clicking a cell will cause > > grid_handler 'click' > > to be called? And that I should look at the code for > click to see what globals are set for this call? > Can I find out what the button and shift-key state was? > > b. I don't see rclick and rdblclick events. Can they be added? > > 1. I use boldface in some of my cells. I see that support > for boldface > is listed as a not-too-hard feature. Any chance of getting that > soon? > > Henry Rich > > > > ---------------------------------------------------------------------- > For information about J forums see > http://www.jsoftware.com/forums.htm From cburke at jsoftware.com Tue Dec 13 05:58:57 2005 From: cburke at jsoftware.com (Chris Burke) Date: Tue Dec 13 05:57:13 2005 Subject: [Jbeta] J6 grid: requests In-Reply-To: <200512121713.jBCHDs1u013285@ms-smtp-04-eri0.southeast.rr.com> References: <200512121713.jBCHDs1u013285@ms-smtp-04-eri0.southeast.rr.com> Message-ID: <439DF2A1.8030802@jsoftware.com> Henry Rich wrote: > I also need to connect a scrollbar to my grid. There must be some > way to indicate the index of the top-left displayed cell, but > I can't find it. What is it? And do I have to support the scrollbar > in code of my own? The grid displays scrollbars as needed. You don't need code for this. Take a look at the grid demos, and try resizing the windows. If you need to find the position after an event, then: CELLMARK = current marked cell or region Scrollr,Scrollc = amount the grid has been scrolled (so top left). From cburke at jsoftware.com Tue Dec 13 06:15:07 2005 From: cburke at jsoftware.com (Chris Burke) Date: Tue Dec 13 06:13:24 2005 Subject: [Jbeta] J6 grid: requests In-Reply-To: <200512121653.jBCGrYWe022159@ms-smtp-01-eri0.southeast.rr.com> References: <200512121653.jBCGrYWe022159@ms-smtp-01-eri0.southeast.rr.com> Message-ID: <439DF66B.9040103@jsoftware.com> Henry Rich wrote: > Trying to port my application that uses glgridxxx calls to the new > grid, I find I need clarifications/extensions. > > 0. I use CTRL+leftclick, CTRL+leftdoubleclick, CTRL+richtclick > in addition to those same clicks without the CTRL key. > > a. Am I right that left-clicking a cell will cause > > grid_handler 'click' > > to be called? And that I should look at the code for > click to see what globals are set for this call? > Can I find out what the button and shift-key state was? Yes. The click code is below, and the globals should give you what is needed. NB. click, after mbldown event NB. NB. Px, Py mouse pixel position (0,0)=top left NB. Row, Col index into visible cells NB. (0,0) = top left data area NB. negative values for headers NB. Ctrl,Shift flags click=: 3 : 0 'Px Py Row Col Ctrl Shift'=: y. if. gridhandler 'click' do. clickx'' end. ) > b. I don't see rclick and rdblclick events. Can they be added? Yes, I'll do that for the next beta (this week). > 1. I use boldface in some of my cells. I see that support for boldface > is listed as a not-too-hard feature. Any chance of getting that > soon? Probably for the next but one beta. From eiverson at sympatico.ca Tue Dec 13 06:35:14 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Tue Dec 13 06:33:30 2005 Subject: [Jbeta] J601 d beta available Message-ID: <006001c5ff6c$5486fa10$0701a8c0@xp> The d beta is available: http://www.jsoftware.com/download/j601dbeta_win.exe This beta fixes some, but not all, problem reports for the previous version. 1!:x family (file function) now support utf8 file and directory names 7 u: always returns list display of wchar data now works olesetlic - a bug has been fixed, but I currently don't have a licensed ocx for testing. Please confirm if it works or if there is still a problem xlopen (olemethod call fails) - fixed From HenryHRich at nc.rr.com Tue Dec 13 07:24:54 2005 From: HenryHRich at nc.rr.com (Henry Rich) Date: Tue Dec 13 07:23:04 2005 Subject: [Jbeta] Function Plots In-Reply-To: <439DF66B.9040103@jsoftware.com> Message-ID: <200512122324.jBCNOr1u008362@ms-smtp-04-eri0.southeast.rr.com> I would like to call the attention of members of this list to a new feature in J6 called Function Plots (you can see a pointer to the description on the main plot page). This feature grew out of some suggestions Kip Murray made. I had a part in the coding, and I would like to know if users need anything added or changed. Function Plot is a quick way to call for the plot of a function. Examples: plot _10 10 ; 'sin`cos' NB. sin & cos from _10 to +10 plot 0.01 0.1 ; 'sin % y.' NB. 1/sin(x) Henry Rich From gosinn at gmail.com Tue Dec 13 16:56:07 2005 From: gosinn at gmail.com (=?ISO-8859-1?Q?Bj=F6rn_Helgason?=) Date: Tue Dec 13 16:54:16 2005 Subject: [Jbeta] J601 d beta available In-Reply-To: <006001c5ff6c$5486fa10$0701a8c0@xp> References: <006001c5ff6c$5486fa10$0701a8c0@xp> Message-ID: <4031fef0512130056j2bd0937ar@mail.gmail.com> I downloaded and installed the new beta Trying to start J601 I got ?file name error ? ({.a.) 1!:12 fn;0 I restarted the pc renamed J601 to J601c Downloaded again Trying to start J601 i get ?file name error ? ({.a.) 1!:12 fn;0 again Trying Jconsole gives ?file name error ? ({.a.) 1!:12 fn;0 Anything I should try ? 2005/12/12, Eric Iverson : > The d beta is available: > http://www.jsoftware.com/download/j601dbeta_win.exe > > This beta fixes some, but not all, problem reports for the previous version. > > 1!:x family (file function) now support utf8 file and directory names > > 7 u: always returns list > > display of wchar data now works > > olesetlic - a bug has been fixed, but I currently don't have a licensed ocx > for testing. Please confirm if it works or if there is still a problem > > xlopen (olemethod call fails) - fixed > > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > -- Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 Fugl&Fiskur ehf, ?erneyjarsund 23, 802 Selfoss ,t-p?st: gosinn@gmail.com http://www.linux.is/forritun/data/J/grunnur.pdf T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num /|_ .-----------------------------------. ,' .\ / | Me? l?ttri lund ver?ur | ,--' _,' | Dagurinn ? dag | / / | Enn betri en g?rdagurinn | ( -. | `-----------------------------------' | ) | (`-. '--.) `. )----' From eiverson at sympatico.ca Tue Dec 13 22:23:41 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Tue Dec 13 22:22:17 2005 Subject: [Jbeta] J601 d beta available References: <006001c5ff6c$5486fa10$0701a8c0@xp> <4031fef0512130056j2bd0937ar@mail.gmail.com> Message-ID: <002401c5fff0$de04fd30$0701a8c0@xp> Thanks for the problem report. It is an interesting one. I haven't dug into it completely yet so my analysis may be somewhat off or incomplete. The line with the error ({.a.) 1!:12 fn;0 is from break.ijs and it is setting up the break file that shares the byte with the external break program (the new yellow J icon). The break file name comes from host API GetTempPathA and GetLongPathNameA. The A suffix indicates these are the char versions, not the unicode versions. This means the path returned is an ANSI path with the code page required for the proper meaning. But the 1!:x family in the d beta require utf8 format of the unicode file name and this won't exist as it is trying to use an ANSI filename as a utf8 filename. When you get the failure, please report the following results to me: 9!:46'' NB. break file name in temp directory a. i. 9!:46'' NB. values in the file name This bug will affect anyone with non-ascii (7 bit) values in the temp path results. I'll provide a version of break.ijs shortly that uses the W flavours of the temp path api and expect this will fix the problem. ----- Original Message ----- From: "Bj?rn Helgason" To: "Beta forum" Sent: Tuesday, December 13, 2005 3:56 AM Subject: Re: [Jbeta] J601 d beta available >I downloaded and installed the new beta > Trying to start J601 I got > > ?file name error > ? ({.a.) 1!:12 fn;0 > > I restarted the pc > renamed J601 to J601c > > Downloaded again > Trying to start J601 i get > > ?file name error > ? ({.a.) 1!:12 fn;0 > > again > > Trying Jconsole gives > ?file name error > ? ({.a.) 1!:12 fn;0 > > Anything I should try ? > > 2005/12/12, Eric Iverson : >> The d beta is available: >> http://www.jsoftware.com/download/j601dbeta_win.exe >> >> This beta fixes some, but not all, problem reports for the previous >> version. >> >> 1!:x family (file function) now support utf8 file and directory names >> >> 7 u: always returns list >> >> display of wchar data now works >> >> olesetlic - a bug has been fixed, but I currently don't have a licensed >> ocx >> for testing. Please confirm if it works or if there is still a problem >> >> xlopen (olemethod call fails) - fixed >> >> >> >> ---------------------------------------------------------------------- >> For information about J forums see http://www.jsoftware.com/forums.htm >> > > > -- > Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 > Fugl&Fiskur ehf, ?erneyjarsund 23, > 802 Selfoss ,t-p?st: gosinn@gmail.com > http://www.linux.is/forritun/data/J/grunnur.pdf > > > T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans > > g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num > /|_ .-----------------------------------. > ,' .\ / | Me? l?ttri lund ver?ur | > ,--' _,' | Dagurinn ? dag | > / / | Enn betri en g?rdagurinn | > ( -. | `-----------------------------------' > | ) | > (`-. '--.) > `. )----' > -------------------------------------------------------------------------------- > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm From gosinn at gmail.com Tue Dec 13 22:32:57 2005 From: gosinn at gmail.com (=?ISO-8859-1?Q?Bj=F6rn_Helgason?=) Date: Tue Dec 13 22:31:07 2005 Subject: [Jbeta] J601 d beta available In-Reply-To: <002401c5fff0$de04fd30$0701a8c0@xp> References: <006001c5ff6c$5486fa10$0701a8c0@xp> <4031fef0512130056j2bd0937ar@mail.gmail.com> <002401c5fff0$de04fd30$0701a8c0@xp> Message-ID: <4031fef0512130632p40abffb1j@mail.gmail.com> This happens as soon as i start J and have not started anything ?file name error ? ({.a.) 1!:12 fn;0 9!:46'' NB. break file name in temp directory a. i. 9!:46'' NB. values in the file name 9!:14'' j601/beta/2005-12-08/17:30 Results just empty - nothing Same for jconsole 2005/12/13, Eric Iverson : > Thanks for the problem report. It is an interesting one. I haven't dug into > it completely yet so my analysis may be somewhat off or incomplete. > > The line with the error ({.a.) 1!:12 fn;0 is from break.ijs and it is > setting up the break file that shares the byte with the external break > program (the new yellow J icon). > > The break file name comes from host API GetTempPathA and GetLongPathNameA. > The A suffix indicates these are the char versions, not the unicode > versions. This means the path returned is an ANSI path with the code page > required for the proper meaning. But the 1!:x family in the d beta require > utf8 format of the unicode file name and this won't exist as it is trying to > use an ANSI filename as a utf8 filename. > > When you get the failure, please report the following results to me: > 9!:46'' NB. break file name in temp directory > a. i. 9!:46'' NB. values in the file name > > This bug will affect anyone with non-ascii (7 bit) values in the temp path > results. > > I'll provide a version of break.ijs shortly that uses the W flavours of the > temp path api and expect this will fix the problem. > > ----- Original Message ----- > From: "Bj?rn Helgason" > To: "Beta forum" > Sent: Tuesday, December 13, 2005 3:56 AM > Subject: Re: [Jbeta] J601 d beta available > > > >I downloaded and installed the new beta > > Trying to start J601 I got > > > > ?file name error > > ? ({.a.) 1!:12 fn;0 > > > > I restarted the pc > > renamed J601 to J601c > > > > Downloaded again > > Trying to start J601 i get > > > > ?file name error > > ? ({.a.) 1!:12 fn;0 > > > > again > > > > Trying Jconsole gives > > ?file name error > > ? ({.a.) 1!:12 fn;0 > > > > Anything I should try ? > > > > 2005/12/12, Eric Iverson : > >> The d beta is available: > >> http://www.jsoftware.com/download/j601dbeta_win.exe > >> > >> This beta fixes some, but not all, problem reports for the previous > >> version. > >> > >> 1!:x family (file function) now support utf8 file and directory names > >> > >> 7 u: always returns list > >> > >> display of wchar data now works > >> > >> olesetlic - a bug has been fixed, but I currently don't have a licensed > >> ocx > >> for testing. Please confirm if it works or if there is still a problem > >> > >> xlopen (olemethod call fails) - fixed > >> > >> > >> > >> ---------------------------------------------------------------------- > >> For information about J forums see http://www.jsoftware.com/forums.htm > >> > > > > > > -- > > Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 > > Fugl&Fiskur ehf, ?erneyjarsund 23, > > 802 Selfoss ,t-p?st: gosinn@gmail.com > > http://www.linux.is/forritun/data/J/grunnur.pdf > > > > > > T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans > > > > g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num > > /|_ .-----------------------------------. > > ,' .\ / | Me? l?ttri lund ver?ur | > > ,--' _,' | Dagurinn ? dag | > > / / | Enn betri en g?rdagurinn | > > ( -. | `-----------------------------------' > > | ) | > > (`-. '--.) > > `. )----' > > > > > -------------------------------------------------------------------------------- > > > > ---------------------------------------------------------------------- > > For information about J forums see http://www.jsoftware.com/forums.htm > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > -- Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 Fugl&Fiskur ehf, ?erneyjarsund 23, 802 Selfoss ,t-p?st: gosinn@gmail.com http://www.linux.is/forritun/data/J/grunnur.pdf T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num /|_ .-----------------------------------. ,' .\ / | Me? l?ttri lund ver?ur | ,--' _,' | Dagurinn ? dag | / / | Enn betri en g?rdagurinn | ( -. | `-----------------------------------' | ) | (`-. '--.) `. )----' From bill_lam at myrealbox.com Tue Dec 13 23:01:06 2005 From: bill_lam at myrealbox.com (bill lam) Date: Tue Dec 13 22:59:16 2005 Subject: [Jbeta] J601 d beta available Message-ID: <1134486066.c7dcc2dcbill_lam@myrealbox.com> I tested and confirm olesetlic working properly. After some modification, my old application almost run except the 3!:1/3!:2 problem that I reported early. 3!:1 of J601 being not compatible with J504 is documented and reasonable, however 3!:2 of J601 can not decode results from J504 is a nasty problem, it means data can not migrate from J504 to J601 because these data may inside component file or sql database. Is there any workaround such as J script that can decode 3!:1 of J504? Also, has the "bug in cd with *w parameter" (missing trailing 0 0 when passing *w in cd) that reported in J504 been fixed in J601? From HenryHRich at nc.rr.com Tue Dec 13 23:16:55 2005 From: HenryHRich at nc.rr.com (Henry Rich) Date: Tue Dec 13 23:15:04 2005 Subject: [Jbeta] J601 d beta available In-Reply-To: <1134486066.c7dcc2dcbill_lam@myrealbox.com> Message-ID: <200512131516.jBDFGr1u027160@ms-smtp-04-eri0.southeast.rr.com> I must have missed this earlier. All my data is stored as the output of 3!:1 and I read it in with 3!:2. I will need to know how to read these files into J6. Henry Rich > -----Original Message----- > From: beta-bounces@jsoftware.com > [mailto:beta-bounces@jsoftware.com] On Behalf Of bill lam > Sent: Tuesday, December 13, 2005 10:01 AM > To: beta@jsoftware.com > Subject: Re: [Jbeta] J601 d beta available > > I tested and confirm olesetlic working properly. > After some modification, my old application almost run except > the 3!:1/3!:2 problem that I reported early. > 3!:1 of J601 being not compatible with J504 is documented and > reasonable, however 3!:2 of J601 can not decode results from J504 > is a nasty problem, it means data can not migrate from J504 to > J601 because these data may inside component file or sql database. > Is there any workaround such as J script that can decode 3!:1 of J504? > > Also, has the "bug in cd with *w parameter" (missing trailing > 0 0 when passing *w in cd) > that reported in J504 been fixed in J601? > > > > ---------------------------------------------------------------------- > For information about J forums see > http://www.jsoftware.com/forums.htm From RHui000 at shaw.ca Tue Dec 13 23:19:26 2005 From: RHui000 at shaw.ca (Roger Hui) Date: Tue Dec 13 23:18:01 2005 Subject: [Jbeta] verb verb undefname References: <200512120210.jBC2A01u028168@ms-smtp-04-eri0.southeast.rr.com> Message-ID: <010501c5fff8$a363dbd0$f7c35718@lap> I found the thread from 2005-09-13. At that time I said: It probably makes sense to change the syntax to signal error in an explicit definition on a sentence which: is not the executed last sentence, and does not have assignment as the last operation, and does not result in a noun This will necessarily be a dynamic rather than static analysis, as the conditions will only be known at the time the explicit definition is executed. Looking at it again today, I believe the rule should apply only to an explicit verb, or to an explicit adverb or conjunction which has both the x./y. and u./v./m./n. arguments. Moreover, the "executed last sentence" is unnecessary, as that is already an error. Thus: A "redundant sentence" error is signalled in an explicit verb, or in an explicit adverb or conjunction which has both the x./y. and u./v./m./n. arguments, if a sentence is executed which - does not have assignment as the last operation, and - does not result in a noun The idea is that under such conditions it is likely that the program has a typographical mistake, and signalling error at the point of the mistake facilitates debugging. ----- Original Message ----- From: "Henry Rich" To: "'Beta forum'" Sent: Sunday, December 11, 2005 6:10 PM Subject: RE: [Jbeta] verb verb undefname If I recall, the idea was to have a sentence that produced a non-noun and didn't start with a copula, produce the 'otiose sentence' error. Refinements were proposed. Henry Rich > -----Original Message----- > From: beta-bounces@jsoftware.com > [mailto:beta-bounces@jsoftware.com] On Behalf Of Roger Hui > Sent: Sunday, December 11, 2005 4:52 PM > To: Beta forum > Subject: Re: [Jbeta] verb verb undefname > > If it is not in the release notes it has not been implemented yet. > > What is the original proposal for verb verb undefname? > And how is it to be distinguished from verb verb verb > since an undefined name is assumed to be a verb? > > > > ----- Original Message ----- > From: "bill lam" > To: > Sent: Sunday, December 11, 2005 8:50 AM > Subject: [Jbeta] verb verb undefname > > Has 'verb verb undefname' as originally purposed by Henry > Rich been implemented in J601? From HenryHRich at nc.rr.com Tue Dec 13 23:45:44 2005 From: HenryHRich at nc.rr.com (Henry Rich) Date: Tue Dec 13 23:43:54 2005 Subject: [Jbeta] verb verb undefname In-Reply-To: <010501c5fff8$a363dbd0$f7c35718@lap> Message-ID: <200512131545.jBDFjhWe022693@ms-smtp-01-eri0.southeast.rr.com> That sounds good to me. I expect to find several such redundant sentences in heavily-maintained apps. Henry Rich > -----Original Message----- > From: beta-bounces@jsoftware.com > [mailto:beta-bounces@jsoftware.com] On Behalf Of Roger Hui > Sent: Tuesday, December 13, 2005 10:19 AM > To: Beta forum > Subject: Re: [Jbeta] verb verb undefname > > I found the thread from 2005-09-13. At that time I said: > > It probably makes sense to change the syntax to signal > error in an explicit definition on a sentence which: > is not the executed last sentence, and > does not have assignment as the last operation, and > does not result in a noun > This will necessarily be a dynamic rather than static > analysis, as the conditions will only be known at > the time the explicit definition is executed. > > Looking at it again today, I believe the rule should > apply only to an explicit verb, or to an explicit adverb > or conjunction which has both the x./y. and u./v./m./n. > arguments. Moreover, the "executed last sentence" is > unnecessary, as that is already an error. Thus: > > A "redundant sentence" error is signalled in an explicit verb, > or in an explicit adverb or conjunction which has both the x./y. > and u./v./m./n. arguments, if a sentence is executed which > - does not have assignment as the last operation, and > - does not result in a noun > > The idea is that under such conditions it is likely that > the program has a typographical mistake, and signalling > error at the point of the mistake facilitates debugging. > > > > ----- Original Message ----- > From: "Henry Rich" > To: "'Beta forum'" > Sent: Sunday, December 11, 2005 6:10 PM > Subject: RE: [Jbeta] verb verb undefname > > > If I recall, the idea was to have a sentence that produced > a non-noun and didn't start with a copula, produce the > 'otiose sentence' error. Refinements were proposed. > > Henry Rich > > > > > -----Original Message----- > > From: beta-bounces@jsoftware.com > > [mailto:beta-bounces@jsoftware.com] On Behalf Of Roger Hui > > Sent: Sunday, December 11, 2005 4:52 PM > > To: Beta forum > > Subject: Re: [Jbeta] verb verb undefname > > > > If it is not in the release notes it has not been implemented yet. > > > > What is the original proposal for verb verb undefname? > > And how is it to be distinguished from verb verb verb > > since an undefined name is assumed to be a verb? > > > > > > > > ----- Original Message ----- > > From: "bill lam" > > To: > > Sent: Sunday, December 11, 2005 8:50 AM > > Subject: [Jbeta] verb verb undefname > > > > Has 'verb verb undefname' as originally purposed by Henry > > Rich been implemented in J601? > > > ---------------------------------------------------------------------- > For information about J forums see > http://www.jsoftware.com/forums.htm From eiverson at sympatico.ca Wed Dec 14 00:04:47 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Wed Dec 14 00:11:05 2005 Subject: [Jbeta] J601 d beta available References: <1134486066.c7dcc2dcbill_lam@myrealbox.com> Message-ID: <000d01c60000$14fcce30$0701a8c0@xp> I now see that the bug your report with 3!:1/2 between 504 and 601 is only with wchar data (u:). This is not related to any change in the flag field as mentioned in Release Highlights. The bug/difference I strongly suspect is related to the change (I think requested by you) that wchar strings have a terminating null for use with cd etc. We will dig into this and get back to you. ----- Original Message ----- From: "bill lam" To: Sent: Tuesday, December 13, 2005 10:01 AM Subject: Re: [Jbeta] J601 d beta available I tested and confirm olesetlic working properly. After some modification, my old application almost run except the 3!:1/3!:2 problem that I reported early. 3!:1 of J601 being not compatible with J504 is documented and reasonable, however 3!:2 of J601 can not decode results from J504 is a nasty problem, it means data can not migrate from J504 to J601 because these data may inside component file or sql database. Is there any workaround such as J script that can decode 3!:1 of J504? Also, has the "bug in cd with *w parameter" (missing trailing 0 0 when passing *w in cd) that reported in J504 been fixed in J601? ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm From eiverson at sympatico.ca Tue Dec 13 23:52:14 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Wed Dec 14 00:11:07 2005 Subject: [Jbeta] J601 d beta available References: <1134486066.c7dcc2dcbill_lam@myrealbox.com> Message-ID: <000b01c60000$1099c230$0701a8c0@xp> The intended incompatibility in 3!:1/3!:2 is very minor. It will not affect the reading of 504 data by 601. And with luck won't affect the reading of 601 data by 504. I don't believe these changes have been made yet. I missed your earlier problem report and will dig into it now. The intended incompatibility only affects the flag bits. The flag bits are currently 32 0 bits to indicate 32 bit data and 64 1 bits to indicate 64 bit data. The endiannes of the data is currently a bit tricky to figure out and we'd like to add that info to the flag bits. Bottom line: You can definitely depend on 601 being able to read 504 3!:1 data. You can probably depend on 504 being able to read 601 data, but we need to check on that. If the the trouble report is a bug I don't think it is related to this change and it will be fixed. ----- Original Message ----- From: "bill lam" To: Sent: Tuesday, December 13, 2005 10:01 AM Subject: Re: [Jbeta] J601 d beta available I tested and confirm olesetlic working properly. After some modification, my old application almost run except the 3!:1/3!:2 problem that I reported early. 3!:1 of J601 being not compatible with J504 is documented and reasonable, however 3!:2 of J601 can not decode results from J504 is a nasty problem, it means data can not migrate from J504 to J601 because these data may inside component file or sql database. Is there any workaround such as J script that can decode 3!:1 of J504? Also, has the "bug in cd with *w parameter" (missing trailing 0 0 when passing *w in cd) that reported in J504 been fixed in J601? ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm From eiverson at sympatico.ca Wed Dec 14 00:15:37 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Wed Dec 14 00:14:02 2005 Subject: [Jbeta] J601 d beta available References: <1134486066.c7dcc2dcbill_lam@myrealbox.com> Message-ID: <002401c60000$766c0af0$0701a8c0@xp> I think you'll find that wchar data (*w in cd) now has the trailing null. And that this is probably the cause of the 3!:1/2 troubles. ----- Original Message ----- From: "bill lam" To: Sent: Tuesday, December 13, 2005 10:01 AM Subject: Re: [Jbeta] J601 d beta available I tested and confirm olesetlic working properly. After some modification, my old application almost run except the 3!:1/3!:2 problem that I reported early. 3!:1 of J601 being not compatible with J504 is documented and reasonable, however 3!:2 of J601 can not decode results from J504 is a nasty problem, it means data can not migrate from J504 to J601 because these data may inside component file or sql database. Is there any workaround such as J script that can decode 3!:1 of J504? Also, has the "bug in cd with *w parameter" (missing trailing 0 0 when passing *w in cd) that reported in J504 been fixed in J601? ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm From RHui000 at shaw.ca Wed Dec 14 00:38:04 2005 From: RHui000 at shaw.ca (Roger Hui) Date: Wed Dec 14 00:36:32 2005 Subject: [Jbeta] J601 d beta available References: <200512131516.jBDFGr1u027160@ms-smtp-04-eri0.southeast.rr.com> Message-ID: <018901c60003$a1cbd5b0$f7c35718@lap> I believe the incompatibility only occurs on the unicode datatype (131072). The change was that like 1-byte character data, unicode data now also has a trailing 0-byte. I'll see if the 3!:2 code can be relaxed to tolerate data that do not have the trailing 0 byte. ----- Original Message ----- From: "Henry Rich" To: "'Beta forum'" Sent: Tuesday, December 13, 2005 7:16 AM Subject: RE: [Jbeta] J601 d beta available I must have missed this earlier. All my data is stored as the output of 3!:1 and I read it in with 3!:2. I will need to know how to read these files into J6. Henry Rich From donguinn at pvtnetworks.net Wed Dec 14 00:39:15 2005 From: donguinn at pvtnetworks.net (Don Guinn) Date: Wed Dec 14 00:37:37 2005 Subject: [Jbeta] J601 d beta available In-Reply-To: <000b01c60000$1099c230$0701a8c0@xp> References: <1134486066.c7dcc2dcbill_lam@myrealbox.com> <000b01c60000$1099c230$0701a8c0@xp> Message-ID: <439EF933.9040304@pvtn.net> So does 1!:x support file sizes over 4GB in J6? Eric Iverson wrote: > The intended incompatibility in 3!:1/3!:2 is very minor. It will not > affect the reading of 504 data by 601. And with luck won't affect the > reading of 601 data by 504. I don't believe these changes have been > made yet. > > I missed your earlier problem report and will dig into it now. > > The intended incompatibility only affects the flag bits. The flag bits > are currently 32 0 bits to indicate 32 bit data and 64 1 bits to > indicate 64 bit data. The endiannes of the data is currently a bit > tricky to figure out and we'd like to add that info to the flag bits. > > Bottom line: You can definitely depend on 601 being able to read 504 > 3!:1 data. You can probably depend on 504 being able to read 601 data, > but we need to check on that. > > If the the trouble report is a bug I don't think it is related to > this change and it will be fixed. > > . . . From eiverson at sympatico.ca Wed Dec 14 01:23:06 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Wed Dec 14 01:21:33 2005 Subject: [Jbeta] J601 d beta available References: <1134486066.c7dcc2dcbill_lam@myrealbox.com><000b01c60000$1099c230$0701a8c0@xp> <439EF933.9040304@pvtn.net> Message-ID: <002301c60009$e41f25b0$0701a8c0@xp> The limit on file size support is unrelated to any of the current unicode driven changes. I'm pretty sure J32 has the same restrictions it has always had and there is no intent to change them. J64 (all releases and platforms) does not have the restriction. ----- Original Message ----- From: "Don Guinn" To: "Beta forum" Sent: Tuesday, December 13, 2005 11:39 AM Subject: Re: [Jbeta] J601 d beta available > So does 1!:x support file sizes over 4GB in J6? > > Eric Iverson wrote: > >> The intended incompatibility in 3!:1/3!:2 is very minor. It will not >> affect the reading of 504 data by 601. And with luck won't affect the >> reading of 601 data by 504. I don't believe these changes have been made >> yet. >> >> I missed your earlier problem report and will dig into it now. >> >> The intended incompatibility only affects the flag bits. The flag bits >> are currently 32 0 bits to indicate 32 bit data and 64 1 bits to indicate >> 64 bit data. The endiannes of the data is currently a bit tricky to >> figure out and we'd like to add that info to the flag bits. >> >> Bottom line: You can definitely depend on 601 being able to read 504 3!:1 >> data. You can probably depend on 504 being able to read 601 data, but we >> need to check on that. >> >> If the the trouble report is a bug I don't think it is related to this >> change and it will be fixed. >> >> . . . > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm From eiverson at sympatico.ca Wed Dec 14 02:03:52 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Wed Dec 14 02:02:11 2005 Subject: [Jbeta] new version of d beta available Message-ID: <004201c6000f$9970e520$0701a8c0@xp> A new version of the d beta is available http://www.jsoftware.com/download/j601dbeta_win.exe This version has a fix forthe problem Bjorn reported. That is, | file name error | ({.a.) 1!:12 fn:0 Bjorn, please let me know if the problem isn't fixed. From gosinn at gmail.com Wed Dec 14 06:04:33 2005 From: gosinn at gmail.com (=?ISO-8859-1?Q?Bj=F6rn_Helgason?=) Date: Wed Dec 14 06:02:45 2005 Subject: [Jbeta] new version of d beta available In-Reply-To: <004201c6000f$9970e520$0701a8c0@xp> References: <004201c6000f$9970e520$0701a8c0@xp> Message-ID: <4031fef0512131404y39267bd8q@mail.gmail.com> Well.... Now it is like this ?domain error ? 9!:47 fn this here above comes after starting I then tried this as before 9!:46'' a. i. 9!:46'' 9!:14'' j601/beta/2005-12-08/17:30 jconsole is same as this ------------------ another matter I tried to run through the files lab and got this ?? ((7) 2 of 8) fview ????? fview jpath '~system\main\files.ijs' ?value error: wdview ? empty(>{.y) wdview dat 2005/12/13, Eric Iverson : > A new version of the d beta is available > http://www.jsoftware.com/download/j601dbeta_win.exe > > This version has a fix forthe problem Bjorn reported. That is, > | file name error > | ({.a.) 1!:12 fn:0 > > Bjorn, please let me know if the problem isn't fixed. > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > -- Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 Fugl&Fiskur ehf, ?erneyjarsund 23, 802 Selfoss ,t-p?st: gosinn@gmail.com http://www.linux.is/forritun/data/J/grunnur.pdf T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num /|_ .-----------------------------------. ,' .\ / | Me? l?ttri lund ver?ur | ,--' _,' | Dagurinn ? dag | / / | Enn betri en g?rdagurinn | ( -. | `-----------------------------------' | ) | (`-. '--.) `. )----' From eiverson at sympatico.ca Thu Dec 15 01:07:48 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Thu Dec 15 01:05:59 2005 Subject: [Jbeta] e beta available Message-ID: <000f01c600d0$eb5fa7b0$0701a8c0@xp> e beta available http://www.jsoftware.com/download/j601ebeta_win.exe This version has some (not all) of the library and documentation problems fixed. It also fixes (again) Bjorn's problem with 9!:47 setting the break file name with utf8 characters. Bjorn, please let me know if the problem isn't fixed. From eiverson at sympatico.ca Thu Dec 15 01:29:43 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Thu Dec 15 01:30:01 2005 Subject: [Jbeta] glbkmode Message-ID: <000a01c600d4$4620a890$0701a8c0@xp> I've looked into glbkmode a bit more and now see why it was decommitted for 601 and why this is OK. 504 default for text background is opaque with color white. Color can be changed with glbkcolor and opaque can be changed to transparent with glbkmode. Transparent is much more useful and you need to do glbkmode 1 to get that. 601 text is always drawn transparent as this is most useful. If you really want a text background then you can draw a rectangle of the size of the text in the color of your choice. This allows us to kill of glbkmode and glcolor in 601. These are not supported directly in Java and can be easily supported by simple J code. The push on gl2 is to provide a core portable set and to let J do the computation where reasonable. So glbkmode and glcolor stay decomitted in j601. On the original question about printing text tightly together: 1. in 501 use glbkmode 1 as Bill suggests 2. in 601 the default is transparent and you'll get what you want From gosinn at gmail.com Thu Dec 15 02:41:42 2005 From: gosinn at gmail.com (=?ISO-8859-1?Q?Bj=F6rn_Helgason?=) Date: Thu Dec 15 02:39:50 2005 Subject: [Jbeta] e beta available In-Reply-To: <000f01c600d0$eb5fa7b0$0701a8c0@xp> References: <000f01c600d0$eb5fa7b0$0701a8c0@xp> Message-ID: <4031fef0512141041j6defa159k@mail.gmail.com> I am not at home I did download and try this out where I am now I did not get the problem now 2005/12/14, Eric Iverson : > e beta available > http://www.jsoftware.com/download/j601ebeta_win.exe > > This version has some (not all) of the library and documentation problems > fixed. It also fixes (again) Bjorn's problem with 9!:47 setting the break > file name with utf8 characters. > > Bjorn, please let me know if the problem isn't fixed. > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > -- Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 Fugl&Fiskur ehf, ?erneyjarsund 23, 802 Selfoss ,t-p?st: gosinn@gmail.com http://www.linux.is/forritun/data/J/grunnur.pdf T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num /|_ .-----------------------------------. ,' .\ / | Me? l?ttri lund ver?ur | ,--' _,' | Dagurinn ? dag | / / | Enn betri en g?rdagurinn | ( -. | `-----------------------------------' | ) | (`-. '--.) `. )----' From gkirouac at riq.qc.ca Thu Dec 15 06:49:37 2005 From: gkirouac at riq.qc.ca (Gilles Kirouac) Date: Thu Dec 15 06:47:45 2005 Subject: [Jbeta] Problem tracing execution Message-ID: <20051214224819.M77367@riq.qc.ca> J601ebeta, trace.ijs of 1 d?cembre, 2005, 12:37:04. plus=: + load'trace' trace'2 plus 3' --------------- 2 Dyad ------- 2 plus 3 │value error: plus │ t_z=.(2) (plus)(3) Gilles - From RHui000 at shaw.ca Thu Dec 15 07:36:22 2005 From: RHui000 at shaw.ca (Roger Hui) Date: Thu Dec 15 07:35:14 2005 Subject: [Jbeta] Problem tracing execution References: <20051214224819.M77367@riq.qc.ca> Message-ID: <00a101c60107$36114c60$f7c35718@lap> Instead of trace '2 plus 3' , use trace '2 plus__ 3' ----- Original Message ----- From: "Gilles Kirouac" To: Sent: Wednesday, December 14, 2005 2:49 PM Subject: [Jbeta] Problem tracing execution J601ebeta, trace.ijs of 1 d?cembre, 2005, 12:37:04. plus=: + load'trace' trace'2 plus 3' --------------- 2 Dyad ------- 2 plus 3 │value error: plus │ t_z=.(2) (plus)(3) From steve at ni.aist.go.jp Thu Dec 15 10:20:26 2005 From: steve at ni.aist.go.jp (Steven Phillips) Date: Thu Dec 15 10:18:39 2005 Subject: [Jbeta] e beta available In-Reply-To: <000f01c600d0$eb5fa7b0$0701a8c0@xp> References: <000f01c600d0$eb5fa7b0$0701a8c0@xp> Message-ID: <43A0D2EA.7030905@ni.aist.go.jp> The unicode verb u: has 1&u: and 2&u: , 3&u: and 4&u: , and 7&u: and 8&u: as inverse pairs. This departs from the "convention" of having (-x)&f as the inverse of x&f for verbs o. , 3!:5 , and s: . Why not also define the inverse of 1&u: as _1&u: , etc? Steve ---------------------------------------------------------- Steven Phillips, PhD Cognitive and Behavioral Sciences Group, Neuroscience Research Institute, AIST 1-1-1 Umezono, Tsukuba Central 2, Tsukuba, Ibaraki 305-8568 Japan Tel: +81 298 61 5165 (PHS: 45165) Fax: +81 298 61 5857 WWW: ---------------------------------------------------------- Eric Iverson wrote: > e beta available > http://www.jsoftware.com/download/j601ebeta_win.exe > > This version has some (not all) of the library and documentation > problems fixed. It also fixes (again) Bjorn's problem with 9!:47 > setting the break file name with utf8 characters. > > Bjorn, please let me know if the problem isn't fixed. > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > > From bill_lam at myrealbox.com Thu Dec 15 21:37:18 2005 From: bill_lam at myrealbox.com (bill lam) Date: Thu Dec 15 21:35:25 2005 Subject: [Jbeta] glbkmode Message-ID: <1134653838.c7d16e9cbill_lam@myrealbox.com> in isigraph events size resize event (OpenGL, gl2 must request with glwantresize) since glwantresize is decommitted so will the size event. I suppose gltextalign is useful, will it be added back, or some workaround suggested? From donguinn at pvtnetworks.net Thu Dec 15 21:48:15 2005 From: donguinn at pvtnetworks.net (Don Guinn) Date: Thu Dec 15 21:46:32 2005 Subject: [Jbeta] glbkmode In-Reply-To: <1134653838.c7d16e9cbill_lam@myrealbox.com> References: <1134653838.c7d16e9cbill_lam@myrealbox.com> Message-ID: <43A1741F.4080200@pvtn.net> It looks like the resize event is always enabled so glwantresize is not necessary. bill lam wrote: >in isigraph events >size resize event (OpenGL, gl2 must request with glwantresize) >since glwantresize is decommitted so will the size event. > >I suppose gltextalign is useful, will it be added back, or some >workaround suggested? > >---------------------------------------------------------------------- >For information about J forums see http://www.jsoftware.com/forums.htm > > > From bill_lam at myrealbox.com Thu Dec 15 22:20:15 2005 From: bill_lam at myrealbox.com (bill lam) Date: Thu Dec 15 22:18:22 2005 Subject: [Jbeta] unicode odbc ? Message-ID: <1134656415.c7d16e9cbill_lam@myrealbox.com> as odbc api already has a unicode interface, is there any plan to update odbc.ijs to make it unicode aware? From bill_lam at myrealbox.com Thu Dec 15 22:26:31 2005 From: bill_lam at myrealbox.com (bill lam) Date: Thu Dec 15 22:24:37 2005 Subject: [Jbeta] glbkmode Message-ID: <1134656791.c7d16e9cbill_lam@myrealbox.com> >It looks like the resize event is always enabled so glwantresize is not necessary. No, it looks like there is no resize event anymore. It now uses paint event to redraw. From donguinn at pvtnetworks.net Thu Dec 15 22:30:30 2005 From: donguinn at pvtnetworks.net (Don Guinn) Date: Thu Dec 15 22:28:45 2005 Subject: [Jbeta] glbkmode In-Reply-To: <1134656791.c7d16e9cbill_lam@myrealbox.com> References: <1134656791.c7d16e9cbill_lam@myrealbox.com> Message-ID: <43A17E06.1020701@pvtn.net> Well, somehow it knows when the isi object is resized. Maybe it's called something else now. I am anxious to see the new lab for grid. bill lam wrote: >>It looks like the resize event is always enabled so glwantresize is not necessary. >> >> > >No, it looks like there is no resize event anymore. It now uses paint event to redraw. > > >---------------------------------------------------------------------- >For information about J forums see http://www.jsoftware.com/forums.htm > > > From bill_lam at myrealbox.com Thu Dec 15 22:55:59 2005 From: bill_lam at myrealbox.com (bill lam) Date: Thu Dec 15 22:54:06 2005 Subject: [Jbeta] 4!:3 Message-ID: <1134658559.c7d16e9cbill_lam@myrealbox.com> Besides 1!:x family, I suppose 0!:x also use utf8 now. But it seems file names returned by 4!:3 are still not yet utf8. PS. I am also anxious to see the new lab for grid. (^o^) From eiverson at sympatico.ca Thu Dec 15 23:07:35 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Thu Dec 15 23:06:05 2005 Subject: [Jbeta] glbkmode References: <1134653838.c7d16e9cbill_lam@myrealbox.com> Message-ID: <001d01c60189$4a68fe60$0701a8c0@xp> In 601 glwantresize is decomitted. In 601 an opengl only gets paint events (there is no longer a size event). The documentation on opengl need to be updated. Wd do not plan to add gltextalign back in. With glqextent and the new glqextentw you can determine the size of the text you display and can calculate and maintain whatever poisition information you want directly in J. Again, this is part of the move to let J do all the calculations and state management that is reasonable possible as this is more open and general. Doing glqextent and gltextxy is a complete (and more general) replacement for gltextalign. But it is very incompatible and if you need to manually do conversions of old systems that used it. If your use of gltext and gltextalign was consistent and simple it can probably be replaced with a utility that does the required calculations and steps. ----- Original Message ----- From: "bill lam" To: Sent: Thursday, December 15, 2005 8:37 AM Subject: Re: [Jbeta] glbkmode in isigraph events size resize event (OpenGL, gl2 must request with glwantresize) since glwantresize is decommitted so will the size event. I suppose gltextalign is useful, will it be added back, or some workaround suggested? ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm From eiverson at sympatico.ca Thu Dec 15 23:11:47 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Thu Dec 15 23:10:29 2005 Subject: [Jbeta] unicode odbc ? References: <1134656415.c7d16e9cbill_lam@myrealbox.com> Message-ID: <002801c60189$e0534a20$0701a8c0@xp> This (unicode version odbc.ijs) is not a high priority for Jsoftware. It would be an excellent project for someone to contribute back to the community. The original work was done by John Baker and there have been minor changes since then. A unicode version would be nice. Do I see a raised hand out there in netland? ----- Original Message ----- From: "bill lam" To: Sent: Thursday, December 15, 2005 9:20 AM Subject: [Jbeta] unicode odbc ? as odbc api already has a unicode interface, is there any plan to update odbc.ijs to make it unicode aware? ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm From eiverson at sympatico.ca Thu Dec 15 23:16:39 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Thu Dec 15 23:21:29 2005 Subject: [Jbeta] 4!:3 References: <1134658559.c7d16e9cbill_lam@myrealbox.com> Message-ID: <004501c6018b$75ea9e70$0701a8c0@xp> Thanks for pointing out that 4!:3 (and probably a few related places) also need utf8 changes. ----- Original Message ----- From: "bill lam" To: Sent: Thursday, December 15, 2005 9:55 AM Subject: [Jbeta] 4!:3 Besides 1!:x family, I suppose 0!:x also use utf8 now. But it seems file names returned by 4!:3 are still not yet utf8. PS. I am also anxious to see the new lab for grid. (^o^) ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm From bakerjd99 at gmail.com Thu Dec 15 23:29:45 2005 From: bakerjd99 at gmail.com (John Baker) Date: Thu Dec 15 23:27:52 2005 Subject: [Jbeta] VB jsvr1 example broken in e beta Message-ID: <377611440512150729u16643d9ahf05a580b606d8772@mail.gmail.com> I just compiled the J VB example ..\system\examples\ole\vb for the latest e beta and it is still broken. It would be useful if these examples ( C++ and VB ) were reworked and made to run in the new unicode'fied J. From eiverson at sympatico.ca Fri Dec 16 00:48:10 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Fri Dec 16 00:46:49 2005 Subject: [Jbeta] VB jsvr1 example broken in e beta References: <377611440512150729u16643d9ahf05a580b606d8772@mail.gmail.com> Message-ID: <001101c60197$59c1f250$0701a8c0@xp> I don't have VB immediately available. Could you provide a little more info about the failure? If it is a simple problem in a particular area I could probably find and fix without having to go through the reinstall of VB right now. ----- Original Message ----- From: "John Baker" To: Sent: Thursday, December 15, 2005 10:29 AM Subject: [Jbeta] VB jsvr1 example broken in e beta I just compiled the J VB example ..\system\examples\ole\vb for the latest e beta and it is still broken. It would be useful if these examples ( C++ and VB ) were reworked and made to run in the new unicode'fied J. ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm From christopher.f.lett at citigroup.com Fri Dec 16 00:55:47 2005 From: christopher.f.lett at citigroup.com (Lett, Christopher F) Date: Fri Dec 16 00:55:45 2005 Subject: [Jbeta] Stop Manager in Debugger Message-ID: Using the e-beta, in a new J session: (9!:12 '');9!:14 '' +-+--------------------------+ |6|j601/beta/2005-12-08/17:30| +-+--------------------------+ If I type: scrout=: [: wd 'smappend "'"_ , ,&'" 0' At this point I bring up the debugger by pressing Ctrl+K, then click on the Stop Manager button. I get: ?domain error: jdb_isexplicit ?jdb_isexplicit[0] I realize this is an old wd command (in fact, none of my code uses it any more), but the Debugger/Stop Manager in 504 copes with this assignment without complaint. /Chris Lett From bakerjd99 at gmail.com Fri Dec 16 01:33:44 2005 From: bakerjd99 at gmail.com (John Baker) Date: Fri Dec 16 01:31:51 2005 Subject: [Jbeta] VB jsvr1 example broken in e beta In-Reply-To: <001101c60197$59c1f250$0701a8c0@xp> References: <377611440512150729u16643d9ahf05a580b606d8772@mail.gmail.com> <001101c60197$59c1f250$0701a8c0@xp> Message-ID: <377611440512150933m7f9739d9o8abfc2c00ce025fa@mail.gmail.com> I've done a bit of poking around with the sample. The error occurs on the first call of js.do. The argument is a string that causes a spelling error in J. Usually a symptom of improper VB string to J string conversion. The js.setb and js.getb verbs appear to work. I would suggest adding a variant of the js.do verb js.dob so there is no confunsion about how the vb string will be converted. js.show and js.log do not seem to work. I cannot make the J exe server visible. The testing was done with VB6. I haven't tried the sample with VB.NET Hope this helps. On 12/15/05, Eric Iverson wrote: > > I don't have VB immediately available. Could you provide a little more > info > about the failure? If it is a simple problem in a particular area I could > probably find and fix without having to go through the reinstall of VB > right > now. > > ----- Original Message ----- > From: "John Baker" > To: > Sent: Thursday, December 15, 2005 10:29 AM > Subject: [Jbeta] VB jsvr1 example broken in e beta > > > I just compiled the J VB example ..\system\examples\ole\vb for the > latest > e beta > and it is still broken. > > It would be useful if these examples ( C++ and VB ) were reworked and made > to > run in the new unicode'fied J. > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > From gosinn at gmail.com Fri Dec 16 02:10:30 2005 From: gosinn at gmail.com (=?ISO-8859-1?Q?Bj=F6rn_Helgason?=) Date: Fri Dec 16 02:08:36 2005 Subject: [Jbeta] Re: junklab not created In-Reply-To: <4031fef0511300218y748bb9c5i@mail.gmail.com> References: <4031fef0511300218y748bb9c5i@mail.gmail.com> Message-ID: <4031fef0512151010n40610bc6r@mail.gmail.com> Lab: DLL: Using System DLLs (file examples) 9!:14'' j601/beta/2005-12-08/17:30 ?? (29 of 35) CreateDirectory ???????????????????? createdirectory=: 3 : 'CreateDirectory y.;<<0' createdirectory 'user\junklab' NB. 0 is failure ?domain error: CreateDirectory ? CreateDirectory y.;<<0 2005/11/30, Bj?rn Helgason : > > Lab: DLL: Using System DLLs (file examples) > > First the text seems to me to be wrong > > ?? (11 of 35) CreateFile (ctd) ?????????????????????????????? > > In our examples we use access GENERIC_READ+GENERIC_READ, > share 0, and attributes 0. > ) > > should be Write > > ?? (11 of 35) CreateFile (ctd) ?????????????????????????????? > > In our examples we use access GENERIC_READ+GENERIC_WRITE, > share 0, and attributes 0. > ) > > junklab dir does not get created > > --------------- > createdirectory 'user\junklab' NB. 0 is failure > ?domain error: CreateDirectory > ? CreateDirectory y.;<<0 > -- Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 Fugl&Fiskur ehf, ?erneyjarsund 23, 802 Selfoss ,t-p?st: gosinn@gmail.com http://www.linux.is/forritun/data/J/grunnur.pdf T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num /|_ .-----------------------------------. ,' .\ / | Me? l?ttri lund ver?ur | ,--' _,' | Dagurinn ? dag | / / | Enn betri en g?rdagurinn | ( -. | `-----------------------------------' | ) | (`-. '--.) `. )----' From eiverson at sympatico.ca Fri Dec 16 04:54:34 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Fri Dec 16 04:55:41 2005 Subject: [Jbeta] Re: junklab not created References: <4031fef0511300218y748bb9c5i@mail.gmail.com> <4031fef0512151010n40610bc6r@mail.gmail.com> Message-ID: <004f01c601ba$2a2b8d80$0701a8c0@xp> Thanks for the bug report. The problem is the win32api gives an incorrect prototype for CreateDirectoryA. After the domain error a cder'' shows the 2nd parameter is bad. The prototype from win32api is: kernel32 CreateDirectoryA i *c i This requires the second argument to be an integer and instead a null pointer <<0 is passed. The proper prototype is: kernel32 CreateDirectoryA i *c * The second parameter is a pointer to a security structure. This problem will probably be fixed by our changing the prototype. If you are at the failure point in the lab you can fix it by doing: CreateDirectory=: 'kernel32 CreateDirectoryA i *c *'&(15!:0) createdirectory 'user\morejunk' NB. should work now with new prototype There is a more interesting problem with this lab. It uses the ansi version of all host api (that is, CreateFileA, DeleteFileA, CreateDirectoryA) where the filename arguments are ansi character strings. For unicode support the lab needs to be reworked (or have an alternate version) that uses CreateFileW etc and provides unicode arguments (from 7 u: of utf8 arguments. For now we'll probably just add a caution at the front of the lab that it uses ansi filenames. After all, it is a demo of dll calls not file operations. ----- Original Message ----- From: "Bj?rn Helgason" To: "Beta forum" Sent: Thursday, December 15, 2005 1:10 PM Subject: [Jbeta] Re: junklab not created > Lab: DLL: Using System DLLs (file examples) > > 9!:14'' > j601/beta/2005-12-08/17:30 > > ?? (29 of 35) CreateDirectory ???????????????????? createdirectory=: 3 : > 'CreateDirectory y.;<<0' > createdirectory 'user\junklab' NB. 0 is failure > ?domain error: CreateDirectory > ? CreateDirectory y.;<<0 > > > 2005/11/30, Bj?rn Helgason : >> >> Lab: DLL: Using System DLLs (file examples) >> >> First the text seems to me to be wrong >> >> ?? (11 of 35) CreateFile (ctd) ?????????????????????????????? >> >> In our examples we use access GENERIC_READ+GENERIC_READ, >> share 0, and attributes 0. >> ) >> >> should be Write >> >> ?? (11 of 35) CreateFile (ctd) ?????????????????????????????? >> >> In our examples we use access GENERIC_READ+GENERIC_WRITE, >> share 0, and attributes 0. >> ) >> >> junklab dir does not get created >> >> --------------- >> createdirectory 'user\junklab' NB. 0 is failure >> ?domain error: CreateDirectory >> ? CreateDirectory y.;<<0 >> > > > > -- > Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 > Fugl&Fiskur ehf, ?erneyjarsund 23, > 802 Selfoss ,t-p?st: gosinn@gmail.com > http://www.linux.is/forritun/data/J/grunnur.pdf > > > T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans > > g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num > /|_ .-----------------------------------. > > ,' .\ / | Me? l?ttri lund ver?ur | > ,--' _,' | Dagurinn ? dag | > / / | Enn betri en g?rdagurinn | > ( -. | `-----------------------------------' > | ) | > (`-. '--.) > `. )----' > -------------------------------------------------------------------------------- > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm From gosinn at gmail.com Fri Dec 16 05:03:47 2005 From: gosinn at gmail.com (=?ISO-8859-1?Q?Bj=F6rn_Helgason?=) Date: Fri Dec 16 05:01:56 2005 Subject: [Jbeta] Re: junklab not created In-Reply-To: <004f01c601ba$2a2b8d80$0701a8c0@xp> References: <4031fef0511300218y748bb9c5i@mail.gmail.com> <4031fef0512151010n40610bc6r@mail.gmail.com> <004f01c601ba$2a2b8d80$0701a8c0@xp> Message-ID: <4031fef0512151303m71faecaeh@mail.gmail.com> Thx works great! 2005/12/15, Eric Iverson : This problem will probably be fixed by our changing the prototype. If you > are at the failure point in the lab you can fix it by doing: > CreateDirectory=: 'kernel32 CreateDirectoryA i *c *'&(15!:0) > createdirectory 'user\morejunk' NB. should work now with new prototype > > From eiverson at sympatico.ca Fri Dec 16 06:07:53 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Fri Dec 16 06:06:05 2005 Subject: [Jbeta] VB jsvr1 example broken in e beta References: <377611440512150729u16643d9ahf05a580b606d8772@mail.gmail.com><001101c60197$59c1f250$0701a8c0@xp> <377611440512150933m7f9739d9o8abfc2c00ce025fa@mail.gmail.com> Message-ID: <004d01c601c4$033b8d10$0701a8c0@xp> John, There is an obvious coding error with the change to unicode for js.do for jexeserver. jdllserver probably works. You can get a j.exe with this fixed at: http://www.jsoftware.com/download/jexe.zip js.do may work with this j.exe or it may just uncover the next problem. The failure of show and log may be because no js.do was done to create an ijx window. If js.do works, please try them again. I haven't run vb in at least two years and haven't missed it at all. I don't have it installed on my current dev system and will have to dig out some old cds if this doesn't sort out fairly easily. Your time and patience on this is much appreciated. ----- Original Message ----- From: "John Baker" To: "Beta forum" Sent: Thursday, December 15, 2005 12:33 PM Subject: Re: [Jbeta] VB jsvr1 example broken in e beta I've done a bit of poking around with the sample. The error occurs on the first call of js.do. The argument is a string that causes a spelling error in J. Usually a symptom of improper VB string to J string conversion. The js.setb and js.getb verbs appear to work. I would suggest adding a variant of the js.do verb js.dob so there is no confunsion about how the vb string will be converted. js.show and js.log do not seem to work. I cannot make the J exe server visible. The testing was done with VB6. I haven't tried the sample with VB.NET Hope this helps. On 12/15/05, Eric Iverson wrote: > > I don't have VB immediately available. Could you provide a little more > info > about the failure? If it is a simple problem in a particular area I could > probably find and fix without having to go through the reinstall of VB > right > now. > > ----- Original Message ----- > From: "John Baker" > To: > Sent: Thursday, December 15, 2005 10:29 AM > Subject: [Jbeta] VB jsvr1 example broken in e beta > > > I just compiled the J VB example ..\system\examples\ole\vb for the > latest > e beta > and it is still broken. > > It would be useful if these examples ( C++ and VB ) were reworked and made > to > run in the new unicode'fied J. > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm From gkirouac at riq.qc.ca Fri Dec 16 07:51:58 2005 From: gkirouac at riq.qc.ca (Gilles Kirouac) Date: Fri Dec 16 07:50:04 2005 Subject: [Jbeta] verb verb undefname In-Reply-To: <010501c5fff8$a363dbd0$f7c35718@lap> References: <200512120210.jBC2A01u028168@ms-smtp-04-eri0.southeast.rr.com> <010501c5fff8$a363dbd0$f7c35718@lap> Message-ID: <20051215234348.M4429@riq.qc.ca> ---------- Original Message ----------- From: Roger Hui Sent: Tue, 13 Dec 2005 07:19:26 -0800 Subject: Re: [Jbeta] verb verb undefname ... > A "redundant sentence" error is signalled in an explicit verb, > or in an explicit adverb or conjunction which has both the x./y. > and u./v./m./n. arguments, if a sentence is executed which > - does not have assignment as the last operation, and > - does not result in a noun > ------- End of Original Message ------- If I understand correctly, the error can happen only if a verb is called dyadically, so that both x. and y. are defined. Why is the requirement of having both the x. and y. defined? Gilles -) From HenryHRich at nc.rr.com Fri Dec 16 08:11:07 2005 From: HenryHRich at nc.rr.com (Henry Rich) Date: Fri Dec 16 08:09:14 2005 Subject: [Jbeta] verb verb undefname In-Reply-To: <20051215234348.M4429@riq.qc.ca> Message-ID: <200512160011.jBG0B5We014536@ms-smtp-01-eri0.southeast.rr.com> He means one of more of x./y. and one or more of u./v./m./n. defined. In early J, x. and y. meant what u. and v. do now. When u. and v. were introduced, for compatibility x. and y. kept the old interpretation if u./v. etc. were not present, otherwise they had their current meanings. Henry Rich > -----Original Message----- > From: beta-bounces@jsoftware.com > [mailto:beta-bounces@jsoftware.com] On Behalf Of Gilles Kirouac > Sent: Thursday, December 15, 2005 6:52 PM > To: Beta forum > Subject: Re: [Jbeta] verb verb undefname > > > ---------- Original Message ----------- > From: Roger Hui > Sent: Tue, 13 Dec 2005 07:19:26 -0800 > Subject: Re: [Jbeta] verb verb undefname > > ... > > A "redundant sentence" error is signalled in an explicit verb, > > or in an explicit adverb or conjunction which has both the x./y. > > and u./v./m./n. arguments, if a sentence is executed which > > - does not have assignment as the last operation, and > > - does not result in a noun > > > ------- End of Original Message ------- > > If I understand correctly, the error can happen only if a > verb is called > dyadically, so that both x. and y. are defined. > > Why is the requirement of having both the x. and y. defined? > > Gilles -) > ---------------------------------------------------------------------- > For information about J forums see > http://www.jsoftware.com/forums.htm From RHui000 at shaw.ca Fri Dec 16 08:27:14 2005 From: RHui000 at shaw.ca (Roger Hui) Date: Fri Dec 16 08:25:35 2005 Subject: [Jbeta] verb verb undefname References: <200512120210.jBC2A01u028168@ms-smtp-04-eri0.southeast.rr.com> <010501c5fff8$a363dbd0$f7c35718@lap> <20051215234348.M4429@riq.qc.ca> Message-ID: <01d101c601d7$7a9d1870$f7c35718@lap> The x./y. and u./v./m./n. comment applies to "an explicit adverb or conjunction". Thus a redundant error can not be signalled for "key" but can be signalled for "slashdot". key=: 1 : '=@[ u.@# ]' slashdot=: 1 : 0 (oind y.) u.@osub"0 1 ,@(<"_2) y. : (=x.) u.@# y. ) osub=: >@]`(>@[ >@:{ ]) @. (*@#@]) oind=: (+/&i./ To: "Beta forum" Sent: Thursday, December 15, 2005 3:51 PM Subject: Re: [Jbeta] verb verb undefname ---------- Original Message ----------- From: Roger Hui Sent: Tue, 13 Dec 2005 07:19:26 -0800 Subject: Re: [Jbeta] verb verb undefname ... > A "redundant sentence" error is signalled in an explicit verb, > or in an explicit adverb or conjunction which has both the x./y. > and u./v./m./n. arguments, if a sentence is executed which > - does not have assignment as the last operation, and > - does not result in a noun > ------- End of Original Message ------- If I understand correctly, the error can happen only if a verb is called dyadically, so that both x. and y. are defined. Why is the requirement of having both the x. and y. defined? Gilles -) From bakerjd99 at gmail.com Fri Dec 16 23:16:25 2005 From: bakerjd99 at gmail.com (John Baker) Date: Fri Dec 16 23:14:30 2005 Subject: [Jbeta] VB jsvr1 example broken in e beta In-Reply-To: <004d01c601c4$033b8d10$0701a8c0@xp> References: <377611440512150729u16643d9ahf05a580b606d8772@mail.gmail.com> <001101c60197$59c1f250$0701a8c0@xp> <377611440512150933m7f9739d9o8abfc2c00ce025fa@mail.gmail.com> <004d01c601c4$033b8d10$0701a8c0@xp> Message-ID: <377611440512160716j13a024f8rb259919150bc68b5@mail.gmail.com> I'll take a look - we are buried in snow up in Ottawa today so I am home. I'll try your exe. On 12/15/05, Eric Iverson wrote: > > John, > > There is an obvious coding error with the change to unicode for js.do for > jexeserver. jdllserver probably works. You can get a j.exe with this fixed > at: > http://www.jsoftware.com/download/jexe.zip > js.do may work with this j.exe or it may just uncover the next problem. > > The failure of show and log may be because no js.do was done to create an > ijx window. If js.do works, please try them again. > > I haven't run vb in at least two years and haven't missed it at all. I > don't > have it installed on my current dev system and will have to dig out some > old > cds if this doesn't sort out fairly easily. Your time and patience on this > is much appreciated. > > ----- Original Message ----- > From: "John Baker" > To: "Beta forum" > Sent: Thursday, December 15, 2005 12:33 PM > Subject: Re: [Jbeta] VB jsvr1 example broken in e beta > > > I've done a bit of poking around with the sample. > > The error occurs on the first call of js.do. The argument is a string > that causes a spelling error in J. Usually a symptom of improper > VB string to J string conversion. > > The js.setb and js.getb verbs appear to work. I would suggest adding > a variant of the js.do verb js.dob so there is no confunsion about how > the vb string will be converted. > > js.show and js.log do not seem to work. I cannot make the J exe server > visible. > > The testing was done with VB6. I haven't tried the sample with VB.NET > > Hope this helps. > > > On 12/15/05, Eric Iverson wrote: > > > > I don't have VB immediately available. Could you provide a little more > > info > > about the failure? If it is a simple problem in a particular area I > could > > probably find and fix without having to go through the reinstall of VB > > right > > now. > > > > ----- Original Message ----- > > From: "John Baker" > > To: > > Sent: Thursday, December 15, 2005 10:29 AM > > Subject: [Jbeta] VB jsvr1 example broken in e beta > > > > > > I just compiled the J VB example ..\system\examples\ole\vb for the > > latest > > e beta > > and it is still broken. > > > > It would be useful if these examples ( C++ and VB ) were reworked and > made > > to > > run in the new unicode'fied J. > > ---------------------------------------------------------------------- > > For information about J forums see http://www.jsoftware.com/forums.htm > > > > ---------------------------------------------------------------------- > > For information about J forums see http://www.jsoftware.com/forums.htm > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > From eiverson at sympatico.ca Fri Dec 16 23:27:48 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Fri Dec 16 23:27:44 2005 Subject: [Jbeta] f beta available Message-ID: <006001c60255$80716a70$0701a8c0@xp> f beta available http://www.jsoftware.com/download/j601fbeta_win.exe This version fixes problems reported with J as an OLE automation client. From eiverson at sympatico.ca Sat Dec 17 08:23:22 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Sat Dec 17 08:21:29 2005 Subject: [Jbeta] holidays Message-ID: <000b01c602a0$19875400$0701a8c0@xp> There probably won't be a new beta until early in the new year. From donguinn at pvtnetworks.net Sat Dec 17 22:29:18 2005 From: donguinn at pvtnetworks.net (Don Guinn) Date: Sat Dec 17 22:30:29 2005 Subject: [Jbeta] Scripts rank error Message-ID: <43A420BE.7010304@pvtn.net> Why is it a rank error when I try to use the 3-digit form for scripts? I tried this in J5 and it fails there too. (0!:1) 'xxx=:3+4' xxx=:3+4 (0!:1 0 1) 'xxx=:3+4' |rank error | ( 0!:1 0 1)'xxx=:3+4' 9!:14'' j601/beta/2005-12-08/17:30 I had always wondered how the form 0!:0 could determine whether the input was from a noun or a file. I played with it a little and found that if y was boxed it assumed that it named a file but if unboxed it executed y as a noun. Shouldn't this information be added to the Dictionary? From jf at dbron.net Sat Dec 17 23:28:50 2005 From: jf at dbron.net (Dan Bron) Date: Sat Dec 17 23:26:49 2005 Subject: [Jbeta] Scripts rank error In-Reply-To: <43A420BE.7010304@pvtn.net> Message-ID: <20051217152647.B301A74C64D@alice.jsoftware.com> The 3 digit form refers to a single number with 3 decimal digits. That is, in place of: (0!:1 0 1) 'xxx=:3+4' use (0!:101 ) 'xxx=:3+4' xxx=:3+4 xxx 7 -Dan From donguinn at pvtnetworks.net Sat Dec 17 23:33:44 2005 From: donguinn at pvtnetworks.net (Don Guinn) Date: Sat Dec 17 23:31:57 2005 Subject: [Jbeta] Scripts rank error In-Reply-To: <20051217152647.B301A74C64D@alice.jsoftware.com> References: <20051217152647.B301A74C64D@alice.jsoftware.com> Message-ID: <43A42FD8.2040505@pvtn.net> Of course, how stupid of me! Thanks. Dan Bron wrote: >The 3 digit form refers to a single number with 3 decimal digits. > >That is, in place of: > > (0!:1 0 1) 'xxx=:3+4' >use > > (0!:101 ) 'xxx=:3+4' > xxx=:3+4 > > xxx > 7 > >-Dan > >---------------------------------------------------------------------- >For information about J forums see http://www.jsoftware.com/forums.htm > > > From juneaftn at gmail.com Mon Dec 19 02:44:09 2005 From: juneaftn at gmail.com (June Kim) Date: Mon Dec 19 02:42:10 2005 Subject: [Jbeta] line number not shown in the window title Message-ID: <6f80c1520512181044x6b81f95fje8c002883d1a521e@mail.gmail.com> Hi As far as I know, the only way to know the current line number where the curson is on is to see the window title (supposing the Window Title Style set properly beforehand). In the j601 betas, the line number does not change after a file is first loaded, or sometimes not shown at all( displayed as [] instead). j504 shows the line number properly. (not the column though). From gosinn at gmail.com Mon Dec 19 04:02:37 2005 From: gosinn at gmail.com (=?ISO-8859-1?Q?Bj=F6rn_Helgason?=) Date: Mon Dec 19 04:00:39 2005 Subject: [Jbeta] line number not shown in the window title In-Reply-To: <6f80c1520512181044x6b81f95fje8c002883d1a521e@mail.gmail.com> References: <6f80c1520512181044x6b81f95fje8c002883d1a521e@mail.gmail.com> Message-ID: <4031fef0512181202v3e77160ah@mail.gmail.com> In my J601 beta the line numbers list fine within the brackets[] with no prior settings When I create a new ijs it starts [] but changes to a linenumber as soon as I type anything When I open a script it displays properly 9!:14'' j601/beta/2005-12-08/17:30 2005/12/18, June Kim : > > Hi > > As far as I know, the only way to know the current line number where > the curson is on is to see the window title (supposing the Window > Title Style set properly beforehand). > > In the j601 betas, the line number does not change after a file is > first loaded, or sometimes not shown at all( displayed as [] instead). > > j504 shows the line number properly. (not the column though). > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > -- Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 Fugl&Fiskur ehf, ?erneyjarsund 23, 802 Selfoss ,t-p?st: gosinn@gmail.com http://www.linux.is/forritun/data/J/grunnur.pdf T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num /|_ .-----------------------------------. ,' .\ / | Me? l?ttri lund ver?ur | ,--' _,' | Dagurinn ? dag | / / | Enn betri en g?rdagurinn | ( -. | `-----------------------------------' | ) | (`-. '--.) `. )----' From juneaftn at gmail.com Mon Dec 19 04:07:10 2005 From: juneaftn at gmail.com (June Kim) Date: Mon Dec 19 04:05:13 2005 Subject: [Jbeta] line number not shown in the window title In-Reply-To: <4031fef0512181202v3e77160ah@mail.gmail.com> References: <6f80c1520512181044x6b81f95fje8c002883d1a521e@mail.gmail.com> <4031fef0512181202v3e77160ah@mail.gmail.com> Message-ID: <6f80c1520512181207l51fa8b54sef32cd4c26f3d3ee@mail.gmail.com> Could you please test moving your cursor around in an ijs window after you type in a couple of lines, and see if the linenumber changes? In j504 it changed but not in the lastest beta. On 12/19/05, Bj?rn Helgason wrote: > In my J601 beta the line numbers list fine within the brackets[] with no > prior settings > When I create a new ijs it starts [] but changes to a linenumber as soon as > I type anything > When I open a script it displays properly > > 9!:14'' > j601/beta/2005-12-08/17:30 > > > 2005/12/18, June Kim : > > > > Hi > > > > As far as I know, the only way to know the current line number where > > the curson is on is to see the window title (supposing the Window > > Title Style set properly beforehand). > > > > In the j601 betas, the line number does not change after a file is > > first loaded, or sometimes not shown at all( displayed as [] instead). > > > > j504 shows the line number properly. (not the column though). > > ---------------------------------------------------------------------- > > For information about J forums see http://www.jsoftware.com/forums.htm > > > > > > -- > Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 > Fugl&Fiskur ehf, ?erneyjarsund 23, > 802 Selfoss ,t-p?st: gosinn@gmail.com > http://www.linux.is/forritun/data/J/grunnur.pdf > > > T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans > > g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num > /|_ .-----------------------------------. > > ,' .\ / | Me? l?ttri lund ver?ur | > ,--' _,' | Dagurinn ? dag | > / / | Enn betri en g?rdagurinn | > ( -. | `-----------------------------------' > | ) | > (`-. '--.) > `. )----' > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > From thomasarneson at comcast.net Mon Dec 19 05:17:03 2005 From: thomasarneson at comcast.net (tom arneson) Date: Mon Dec 19 05:15:31 2005 Subject: [Jbeta] line number not shown in the window title In-Reply-To: <6f80c1520512181207l51fa8b54sef32cd4c26f3d3ee@mail.gmail.com> Message-ID: <20051218211529.CD00C74C648@alice.jsoftware.com> It is working for me in beta: JVERSION Binary: j601binfbeta_win Library: j601libfbeta Help: j601hlpfbeta Engine: j601/beta/2005-12-08/17:30 -----Original Message----- From: beta-bounces@jsoftware.com [mailto:beta-bounces@jsoftware.com] On Behalf Of June Kim Sent: Sunday, December 18, 2005 14:07 To: Beta forum Subject: Re: [Jbeta] line number not shown in the window title Could you please test moving your cursor around in an ijs window after you type in a couple of lines, and see if the linenumber changes? In j504 it changed but not in the lastest beta. On 12/19/05, Bj?rn Helgason wrote: > In my J601 beta the line numbers list fine within the brackets[] with no > prior settings > When I create a new ijs it starts [] but changes to a linenumber as soon as > I type anything > When I open a script it displays properly > > 9!:14'' > j601/beta/2005-12-08/17:30 > > > 2005/12/18, June Kim : > > > > Hi > > > > As far as I know, the only way to know the current line number where > > the curson is on is to see the window title (supposing the Window > > Title Style set properly beforehand). > > > > In the j601 betas, the line number does not change after a file is > > first loaded, or sometimes not shown at all( displayed as [] instead). > > > > j504 shows the line number properly. (not the column though). > > ---------------------------------------------------------------------- > > For information about J forums see http://www.jsoftware.com/forums.htm > > > > > > -- > Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 > Fugl&Fiskur ehf, ?erneyjarsund 23, > 802 Selfoss ,t-p?st: gosinn@gmail.com > http://www.linux.is/forritun/data/J/grunnur.pdf > > > T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans > > g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num > /|_ .-----------------------------------. > > ,' .\ / | Me? l?ttri lund ver?ur | > ,--' _,' | Dagurinn ? dag | > / / | Enn betri en g?rdagurinn | > ( -. | `-----------------------------------' > | ) | > (`-. '--.) > `. )----' > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm From bill_lam at myrealbox.com Mon Dec 19 18:07:17 2005 From: bill_lam at myrealbox.com (bill lam) Date: Mon Dec 19 18:05:18 2005 Subject: [Jbeta] CELLDATA/CELLDRAW in jzgrid Message-ID: <1134986837.c7f52e7cbill_lam@myrealbox.com> I try jzgrid and can not tell the difference between CELLDATA and CELLDRAW. It seems CELLDATA is more useful, but what's CELLDRAW for? From bill_lam at myrealbox.com Mon Dec 19 18:32:35 2005 From: bill_lam at myrealbox.com (bill lam) Date: Mon Dec 19 18:30:36 2005 Subject: [Jbeta] line number not shown in the window title Message-ID: <1134988355.c80da1fcbill_lam@myrealbox.com> I suppose it is broken or changed in J601(fbeta) in J504 there is 2 numbers within [], the first is line number inside script and the second is line number within explicit definition. in J601, the second line number is missing. From bakerjd99 at gmail.com Mon Dec 19 22:43:20 2005 From: bakerjd99 at gmail.com (John Baker) Date: Mon Dec 19 22:41:22 2005 Subject: [Jbeta] VB jsvr1 example broken in e beta In-Reply-To: <377611440512160716j13a024f8rb259919150bc68b5@mail.gmail.com> References: <377611440512150729u16643d9ahf05a580b606d8772@mail.gmail.com> <001101c60197$59c1f250$0701a8c0@xp> <377611440512150933m7f9739d9o8abfc2c00ce025fa@mail.gmail.com> <004d01c601c4$033b8d10$0701a8c0@xp> <377611440512160716j13a024f8rb259919150bc68b5@mail.gmail.com> Message-ID: <377611440512190643t7948dcd2t592a6c9693d05a7c@mail.gmail.com> The jsrv1 demo now works. I have also tested this with my psm.exe application that uses the vb ole/com interface and so far things look good. I will be testing more throughly over the holidays. On 12/16/05, John Baker wrote: > > I'll take a look - we are buried in snow up in Ottawa today so I am home. > I'll try > your exe. > > > On 12/15/05, Eric Iverson wrote: > > > > John, > > > > There is an obvious coding error with the change to unicode for js.dofor > > jexeserver. jdllserver probably works. You can get a j.exe with this > > fixed > > at: > > http://www.jsoftware.com/download/jexe.zip > > js.do may work with this j.exe or it may just uncover the next problem. > > > > The failure of show and log may be because no js.do was done to create > > an > > ijx window. If js.do works, please try them again. > > > > I haven't run vb in at least two years and haven't missed it at all. I > > don't > > have it installed on my current dev system and will have to dig out some > > old > > cds if this doesn't sort out fairly easily. Your time and patience on > > this > > is much appreciated. > > > > ----- Original Message ----- > > From: "John Baker" > > To: "Beta forum" > > Sent: Thursday, December 15, 2005 12:33 PM > > Subject: Re: [Jbeta] VB jsvr1 example broken in e beta > > > > > > I've done a bit of poking around with the sample. > > > > The error occurs on the first call of js.do. The argument is a string > > that causes a spelling error in J. Usually a symptom of improper > > VB string to J string conversion. > > > > The js.setb and js.getb verbs appear to work. I would suggest adding > > a variant of the js.do verb js.dob so there is no confunsion about > > how > > the vb string will be converted. > > > > js.show and js.log do not seem to work. I cannot make the J exe server > > visible. > > > > The testing was done with VB6. I haven't tried the sample with VB.NET > > > > Hope this helps. > > > > > > On 12/15/05, Eric Iverson < eiverson@sympatico.ca> wrote: > > > > > > I don't have VB immediately available. Could you provide a little more > > > info > > > about the failure? If it is a simple problem in a particular area I > > could > > > probably find and fix without having to go through the reinstall of VB > > > right > > > now. > > > > > > ----- Original Message ----- > > > From: "John Baker" < bakerjd99@gmail.com> > > > To: > > > Sent: Thursday, December 15, 2005 10:29 AM > > > Subject: [Jbeta] VB jsvr1 example broken in e beta > > > > > > > > > I just compiled the J VB example ..\system\examples\ole\vb for the > > > latest > > > e beta > > > and it is still broken. > > > > > > It would be useful if these examples ( C++ and VB ) were reworked and > > made > > > to > > > run in the new unicode'fied J. > > > ---------------------------------------------------------------------- > > > For information about J forums see http://www.jsoftware.com/forums.htm > > > > > > > > ---------------------------------------------------------------------- > > > For information about J forums see http://www.jsoftware.com/forums.htm > > > > > ---------------------------------------------------------------------- > > For information about J forums see http://www.jsoftware.com/forums.htm > > > > ---------------------------------------------------------------------- > > For information about J forums see http://www.jsoftware.com/forums.htm > > > > From gosinn at gmail.com Mon Dec 19 23:36:14 2005 From: gosinn at gmail.com (=?ISO-8859-1?Q?Bj=F6rn_Helgason?=) Date: Mon Dec 19 23:34:17 2005 Subject: [Jbeta] f beta available In-Reply-To: <006001c60255$80716a70$0701a8c0@xp> References: <006001c60255$80716a70$0701a8c0@xp> Message-ID: <4031fef0512190736h2babfc42v@mail.gmail.com> There was a question recently about background colour I was then looking at background and changed the background of the J session - ijx It works fine I then open an ijs window and it has the same colour If I change the background colour it changes both ijx and all ijs windows I have no idea if it would be difficult to change so that it would be possible to have the background in the ijs windows different from the ijx window It would be nice to have such an option Not to mention that if there is some background colour when you open an ijs that it would stay the same even if you change in config and next opened would get the new colour from config It would thus be easier to distinguish between open windows by the colours 2005/12/16, Eric Iverson : > > f beta available > http://www.jsoftware.com/download/j601fbeta_win.exe > > This version fixes problems reported with J as an OLE automation client. > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > -- Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 Fugl&Fiskur ehf, ?erneyjarsund 23, 802 Selfoss ,t-p?st: gosinn@gmail.com http://www.linux.is/forritun/data/J/grunnur.pdf T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num /|_ .-----------------------------------. ,' .\ / | Me? l?ttri lund ver?ur | ,--' _,' | Dagurinn ? dag | / / | Enn betri en g?rdagurinn | ( -. | `-----------------------------------' | ) | (`-. '--.) `. )----' From rdmiller at usatoday.com Tue Dec 20 07:35:50 2005 From: rdmiller at usatoday.com (Miller, Raul D) Date: Tue Dec 20 07:33:54 2005 Subject: [Jbeta] Bug? (not really specific to beta) Message-ID: According to the dictionary, u b. 1 is the identity function for u. However, experimentation shows that currently it's a list of characters. Likewise for the obverse (that is: assigning rank to one of these produces a constant function which returns the indicated display form). Checking back, j504 had the same kind of behavior. -- Raul From RHui000 at shaw.ca Tue Dec 20 07:49:39 2005 From: RHui000 at shaw.ca (Roger Hui) Date: Tue Dec 20 07:48:45 2005 Subject: [Jbeta] Bug? (not really specific to beta) References: Message-ID: <013f01c604f6$e90913f0$f7c35718@lap> The description in the dictionary needs to be corrected. ----- Original Message ----- From: "Miller, Raul D" To: Sent: Monday, December 19, 2005 3:35 PM Subject: [Jbeta] Bug? (not really specific to beta) According to the dictionary, u b. 1 is the identity function for u. However, experimentation shows that currently it's a list of characters. Likewise for the obverse (that is: assigning rank to one of these produces a constant function which returns the indicated display form). Checking back, j504 had the same kind of behavior. From cburke at jsoftware.com Tue Dec 20 09:35:00 2005 From: cburke at jsoftware.com (Chris Burke) Date: Tue Dec 20 09:33:03 2005 Subject: [Jbeta] CELLDATA/CELLDRAW in jzgrid In-Reply-To: <1134986837.c7f52e7cbill_lam@myrealbox.com> References: <1134986837.c7f52e7cbill_lam@myrealbox.com> Message-ID: <43A75FC4.3010000@jsoftware.com> bill lam wrote: > I try jzgrid and can not tell the difference between > CELLDATA and CELLDRAW. It seems CELLDATA is more useful, > but what's CELLDRAW for? CELLDRAW (raw data) is for the built-in sort only. If present, then the data is sorted on it, instead of on CELLDATA. From bakerjd99 at gmail.com Fri Dec 23 03:34:35 2005 From: bakerjd99 at gmail.com (John Baker) Date: Fri Dec 23 03:32:34 2005 Subject: [Jbeta] image3 first lab fails on step 7 of 30 Message-ID: <377611440512221134r177fb49eg4a4951c1954f4e41@mail.gmail.com> I realize that jsoftware is not responsible for addons but the image3 addon lab fails on step 7 of 30 for the f beta. The error is caused by a value error glwindowext. From reiterc at mail.lafayette.edu Fri Dec 23 03:39:30 2005 From: reiterc at mail.lafayette.edu (Cliff Reiter) Date: Fri Dec 23 03:37:27 2005 Subject: [Jbeta] image3 first lab fails on step 7 of 30 In-Reply-To: <377611440512221134r177fb49eg4a4951c1954f4e41@mail.gmail.com> References: <377611440512221134r177fb49eg4a4951c1954f4e41@mail.gmail.com> Message-ID: <43AB00F2.2020103@mail.lafayette.edu> I have noticed that image3 and many other of my utilities are broken under J6.01. Since the default mapping mode is now raw (did I say that correctly?) I suspect that the fix will amount to removing or commentling out the offending lines :), this is on my to do list for January. Best, Cliff John Baker wrote: > I realize that jsoftware is not responsible for addons but the image3 addon > lab fails on step 7 of 30 for the f beta. The error is caused by a value > error > glwindowext. > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > -- Clifford A. Reiter Mathematics Department, Lafayette College Easton, PA 18042 USA, 610-330-5277 http://www.lafayette.edu/~reiterc From donguinn at pvtnetworks.net Fri Dec 23 04:57:57 2005 From: donguinn at pvtnetworks.net (Don Guinn) Date: Fri Dec 23 04:56:21 2005 Subject: [Jbeta] Form Edit Message-ID: <43AB1355.40304@pvtn.net> Just noticed two new controls - editijs and editijx. What are they? How do they work? I stuck an editijx in a form and I couldn't get any events from it. Also, when I wanted to delete it to do something else I couldn't get the del button to remove it. It was selected. I got it to go away by pressing Undo. 9!:14'' j601/beta/2005-12-08/17:30 From cburke at jsoftware.com Fri Dec 23 05:15:16 2005 From: cburke at jsoftware.com (Chris Burke) Date: Fri Dec 23 05:13:13 2005 Subject: [Jbeta] Form Edit In-Reply-To: <43AB1355.40304@pvtn.net> References: <43AB1355.40304@pvtn.net> Message-ID: <43AB1764.4060509@jsoftware.com> Don Guinn wrote: > Just noticed two new controls - editijs and editijx. What are they? > How do they work? I stuck an editijx in a form and I couldn't get any > events from it. > > Also, when I wanted to delete it to do something else I couldn't get the > del button to remove it. It was selected. I got it to go away by > pressing Undo. These controls are the J session windows, and were introduced in J5. They probably should not be shown in the form editor. I suspect that originally we had the idea that the controls could be put on any form. At least, the editijx cannot be put on a form, since only one ijx window is supported at the moment. From lorin.lund at gmail.com Fri Dec 23 09:54:30 2005 From: lorin.lund at gmail.com (Lorin Lund) Date: Fri Dec 23 09:52:28 2005 Subject: [Jbeta] no error message for trailing undefined Message-ID: <123898aa0512221754kcaa689aye3315ef81ab7fbdd@mail.gmail.com> I just joined the beta list recently. Maybe this is a known and accepted difference. Sorry If I'm covery old ground. With j504: asdf |value error: asdf echo 'asdf: ',asdf asdf: |syntax error | echo'asdf: ',asdf With J6 asdf ?value error: asdf echo 'asdf: ',asdf ?????????????????????? ?echo????????????????? ? ??asdf: ?,?asdf?? ? ????????????????? ?????????????????????? when something like echo 'asdf: ',asdf occurs in an explicit verb there is no error message. From donguinn at pvtnetworks.net Fri Dec 23 10:30:08 2005 From: donguinn at pvtnetworks.net (Don Guinn) Date: Fri Dec 23 10:28:23 2005 Subject: [Jbeta] Form Edit In-Reply-To: <43AB1764.4060509@jsoftware.com> References: <43AB1355.40304@pvtn.net> <43AB1764.4060509@jsoftware.com> Message-ID: <43AB6130.1020609@pvtn.net> Another Form Edit problem. Pressing escape on the form to be edited cancels the edit session without saving and changes. In version 5 it canceled a command in progress on the edit window. From RHui000 at shaw.ca Fri Dec 23 11:25:50 2005 From: RHui000 at shaw.ca (Roger Hui) Date: Fri Dec 23 11:24:45 2005 Subject: [Jbeta] no error message for trailing undefined Message-ID: <048a01c60770$975094d0$f7c35718@lap> The difference is that n v1 v2 is a verb in J6.01 but an error in J5.04. 'asdf:',asdf is the pattern n v1 v2. (An undefined name is assumed to be a verb.) Yestday I came across two more occurrences of the n v1 v2 train: ctf=: 3 #. #: Cantor ternary function. ctf i is the i-th number whose base-3 representation consists of only 0 and 1 primerace=: [: +/\ _2 + 4 | p: The surplus or deficit of primes that are 3 mod 4 (equiv. to _1 mod 4) vs. those that are 1 mod 4. ----- Original Message ----- From: "Lorin Lund" To: Sent: Thursday, December 22, 2005 5:54 PM Subject: [Jbeta] no error message for trailing undefined I just joined the beta list recently. Maybe this is a known and accepted difference. Sorry If I'm covery old ground. With j504: asdf |value error: asdf echo 'asdf: ',asdf asdf: |syntax error | echo'asdf: ',asdf With J6 asdf ?value error: asdf echo 'asdf: ',asdf ?????????????????????? ?echo????????????????? ? ??asdf: ?,?asdf?? ? ????????????????? ?????????????????????? when something like echo 'asdf: ',asdf occurs in an explicit verb there is no error message. From bakerjd99 at gmail.com Fri Dec 23 21:57:35 2005 From: bakerjd99 at gmail.com (John Baker) Date: Fri Dec 23 21:55:32 2005 Subject: [Jbeta] image3 first lab fails on step 7 of 30 In-Reply-To: <43AB00F2.2020103@mail.lafayette.edu> References: <377611440512221134r177fb49eg4a4951c1954f4e41@mail.gmail.com> <43AB00F2.2020103@mail.lafayette.edu> Message-ID: <377611440512230557i490369e6xe139382c729fbc10@mail.gmail.com> Sounds good. My addon also broken in j6.01. I've decided to rename it and do some things I've been thinking about for awhile. But not over the holidays. Cheers. On 12/22/05, Cliff Reiter wrote: > > I have noticed that image3 and many other of my utilities are broken > under J6.01. Since the default mapping mode is now raw (did I say > that correctly?) I suspect that the fix will amount to removing or > commentling out the offending lines :), this is on my to do list for > January. > Best, > Cliff > > John Baker wrote: > > > I realize that jsoftware is not responsible for addons but the image3 > addon > > lab fails on step 7 of 30 for the f beta. The error is caused by a > value > > error > > glwindowext. > > ---------------------------------------------------------------------- > > For information about J forums see http://www.jsoftware.com/forums.htm > > > > -- > Clifford A. Reiter > Mathematics Department, Lafayette College > Easton, PA 18042 USA, 610-330-5277 > http://www.lafayette.edu/~reiterc > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > From eiverson at sympatico.ca Sat Dec 24 04:37:39 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Sat Dec 24 04:35:41 2005 Subject: [Jbeta] line number not shown in the window title References: <6f80c1520512181044x6b81f95fje8c002883d1a521e@mail.gmail.com> Message-ID: <005a01c60800$b97b3700$77dad1d8@xp> As far as I can tell the line numbering in the title bar works properly in the latest J f beta. Please verify that you are using the f beta and give the simplest sequence of steps to show the problem. ----- Original Message ----- From: "June Kim" To: "Beta forum" Sent: Sunday, December 18, 2005 1:44 PM Subject: [Jbeta] line number not shown in the window title Hi As far as I know, the only way to know the current line number where the curson is on is to see the window title (supposing the Window Title Style set properly beforehand). In the j601 betas, the line number does not change after a file is first loaded, or sometimes not shown at all( displayed as [] instead). j504 shows the line number properly. (not the column though). ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm From eiverson at sympatico.ca Sat Dec 24 05:53:31 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Sat Dec 24 05:51:40 2005 Subject: [Jbeta] line number not shown in the window title References: <1134988355.c80da1fcbill_lam@myrealbox.com> Message-ID: <01c001c6080b$572edfb0$77dad1d8@xp> Bill, thanks, the 2nd number (line number in explicit) is broken in 601 and will be fixed. June, unless I hear otherwise I'll assume that fixing the above will fix your problem as well. ----- Original Message ----- From: "bill lam" To: Sent: Monday, December 19, 2005 5:32 AM Subject: RE: [Jbeta] line number not shown in the window title I suppose it is broken or changed in J601(fbeta) in J504 there is 2 numbers within [], the first is line number inside script and the second is line number within explicit definition. in J601, the second line number is missing. ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm From juneaftn at gmail.com Sat Dec 24 09:48:34 2005 From: juneaftn at gmail.com (June Kim) Date: Sat Dec 24 09:46:29 2005 Subject: [Jbeta] line number not shown in the window title In-Reply-To: <01c001c6080b$572edfb0$77dad1d8@xp> References: <1134988355.c80da1fcbill_lam@myrealbox.com> <01c001c6080b$572edfb0$77dad1d8@xp> Message-ID: <6f80c1520512231748q66238cc5j6ab6bb4eba48ca3a@mail.gmail.com> I am using the 601 f beta and following the steps shown below I can demonstrate the problem in my windows machine(windows xp sp1 with korean language): Run the J window and see the ijx window shown with the title : 3.ijx* [0] c:\j601\temp\3.ijx Now I type in three lines: 1 1 2 2 3 3 I can see the line no change in the title from 0 to 2,4, and 6. Now I move the cursor up to the first line and the line no does not change. Similar thing happens on ijs windows as well. On 12/24/05, Eric Iverson wrote: > Bill, thanks, the 2nd number (line number in explicit) is broken in 601 and > will be fixed. > > June, unless I hear otherwise I'll assume that fixing the above will fix > your problem as well. > > ----- Original Message ----- > From: "bill lam" > To: > Sent: Monday, December 19, 2005 5:32 AM > Subject: RE: [Jbeta] line number not shown in the window title > > > I suppose it is broken or changed in J601(fbeta) > in J504 there is 2 numbers within [], the first is line > number inside script and the second is line number > within explicit definition. > in J601, the second line number is missing. > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > From r.e.boss at planet.nl Sat Dec 24 20:13:02 2005 From: r.e.boss at planet.nl (R.E. Boss) Date: Sat Dec 24 20:10:55 2005 Subject: [Jbeta] line number not shown in the window title In-Reply-To: <6f80c1520512231748q66238cc5j6ab6bb4eba48ca3a@mail.gmail.com> Message-ID: <001d01c60883$63334440$9900000a@BOS> I tried the same in 601 a beta (j601/beta/2005-11-17/08:10) and the line no just worked fine. R.E. Boss -----Oorspronkelijk bericht----- Van: beta-bounces@jsoftware.com [mailto:beta-bounces@jsoftware.com] Namens June Kim Verzonden: zaterdag 24 december 2005 2:49 Aan: Beta forum Onderwerp: Re: [Jbeta] line number not shown in the window title I am using the 601 f beta and following the steps shown below I can demonstrate the problem in my windows machine(windows xp sp1 with korean language): Run the J window and see the ijx window shown with the title : 3.ijx* [0] c:\j601\temp\3.ijx Now I type in three lines: 1 1 2 2 3 3 I can see the line no change in the title from 0 to 2,4, and 6. Now I move the cursor up to the first line and the line no does not change. Similar thing happens on ijs windows as well. From eiverson at sympatico.ca Sun Dec 25 03:40:03 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Sun Dec 25 03:38:05 2005 Subject: [Jbeta] line number not shown in the window title References: <1134988355.c80da1fcbill_lam@myrealbox.com><01c001c6080b$572edfb0$77dad1d8@xp> <6f80c1520512231748q66238cc5j6ab6bb4eba48ca3a@mail.gmail.com> Message-ID: <001301c608c1$d8388190$3151d0d8@xp> Strange. It works properly for me here, and by various reports works for most other users. The mechanism is pretty simple and curious that it fails. What font are you using? Could you please revert back to "courier new" and see if it works any differently. (can't see what font has to do with it, but can't think of anything else). Is this the f beta out of the box, or do you have startup scripts that you run? ----- Original Message ----- From: "June Kim" To: "Beta forum" Sent: Friday, December 23, 2005 8:48 PM Subject: Re: [Jbeta] line number not shown in the window title I am using the 601 f beta and following the steps shown below I can demonstrate the problem in my windows machine(windows xp sp1 with korean language): Run the J window and see the ijx window shown with the title : 3.ijx* [0] c:\j601\temp\3.ijx Now I type in three lines: 1 1 2 2 3 3 I can see the line no change in the title from 0 to 2,4, and 6. Now I move the cursor up to the first line and the line no does not change. Similar thing happens on ijs windows as well. On 12/24/05, Eric Iverson wrote: > Bill, thanks, the 2nd number (line number in explicit) is broken in 601 > and > will be fixed. > > June, unless I hear otherwise I'll assume that fixing the above will fix > your problem as well. > > ----- Original Message ----- > From: "bill lam" > To: > Sent: Monday, December 19, 2005 5:32 AM > Subject: RE: [Jbeta] line number not shown in the window title > > > I suppose it is broken or changed in J601(fbeta) > in J504 there is 2 numbers within [], the first is line > number inside script and the second is line number > within explicit definition. > in J601, the second line number is missing. > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm From eiverson at sympatico.ca Sun Dec 25 03:48:46 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Sun Dec 25 03:46:46 2005 Subject: [Jbeta] line number not shown in the window title References: <1134988355.c80da1fcbill_lam@myrealbox.com><01c001c6080b$572edfb0$77dad1d8@xp> <6f80c1520512231748q66238cc5j6ab6bb4eba48ca3a@mail.gmail.com> Message-ID: <004301c608c3$0fa563e0$3151d0d8@xp> Does the 2nd number behave properly? That is, line position in explicit defn ending in ) . ----- Original Message ----- From: "June Kim" To: "Beta forum" Sent: Friday, December 23, 2005 8:48 PM Subject: Re: [Jbeta] line number not shown in the window title I am using the 601 f beta and following the steps shown below I can demonstrate the problem in my windows machine(windows xp sp1 with korean language): Run the J window and see the ijx window shown with the title : 3.ijx* [0] c:\j601\temp\3.ijx Now I type in three lines: 1 1 2 2 3 3 I can see the line no change in the title from 0 to 2,4, and 6. Now I move the cursor up to the first line and the line no does not change. Similar thing happens on ijs windows as well. From gosinn at gmail.com Sun Dec 25 03:59:36 2005 From: gosinn at gmail.com (=?ISO-8859-1?Q?Bj=F6rn_Helgason?=) Date: Sun Dec 25 03:57:34 2005 Subject: [Jbeta] line number not shown in the window title In-Reply-To: <6f80c1520512231748q66238cc5j6ab6bb4eba48ca3a@mail.gmail.com> References: <1134988355.c80da1fcbill_lam@myrealbox.com> <01c001c6080b$572edfb0$77dad1d8@xp> <6f80c1520512231748q66238cc5j6ab6bb4eba48ca3a@mail.gmail.com> Message-ID: <4031fef0512241159r144851aey@mail.gmail.com> Does not behave like this at my end 9!:14'' j601/beta/2005-12-08/17:30 9!:12'' 6 I am using win xp Are you using old windows? 2005/12/24, June Kim : > > I am using the 601 f beta and following the steps shown below I can > demonstrate the problem in my windows machine(windows xp sp1 with > korean language): > > Run the J window and see the ijx window shown with the title : 3.ijx* > [0] c:\j601\temp\3.ijx > > Now I type in three lines: > 1 > 1 > 2 > 2 > 3 > 3 > > I can see the line no change in the title from 0 to 2,4, and 6. Now I > move the cursor up to the first line and the line no does not change. > > Similar thing happens on ijs windows as well. > > On 12/24/05, Eric Iverson wrote: > > Bill, thanks, the 2nd number (line number in explicit) is broken in 601 > and > > will be fixed. > > > > June, unless I hear otherwise I'll assume that fixing the above will fix > > your problem as well. > > > > ----- Original Message ----- > > From: "bill lam" > > To: > > Sent: Monday, December 19, 2005 5:32 AM > > Subject: RE: [Jbeta] line number not shown in the window title > > > > > > I suppose it is broken or changed in J601(fbeta) > > in J504 there is 2 numbers within [], the first is line > > number inside script and the second is line number > > within explicit definition. > > in J601, the second line number is missing. > > > > > > ---------------------------------------------------------------------- > > For information about J forums see http://www.jsoftware.com/forums.htm > > > > ---------------------------------------------------------------------- > > For information about J forums see http://www.jsoftware.com/forums.htm > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > -- Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 Fugl&Fiskur ehf, ?erneyjarsund 23, 802 Selfoss ,t-p?st: gosinn@gmail.com http://www.linux.is/forritun/data/J/grunnur.pdf T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num /|_ .-----------------------------------. ,' .\ / | Me? l?ttri lund ver?ur | ,--' _,' | Dagurinn ? dag | / / | Enn betri en g?rdagurinn | ( -. | `-----------------------------------' | ) | (`-. '--.) `. )----' From juneaftn at gmail.com Sun Dec 25 09:53:05 2005 From: juneaftn at gmail.com (June Kim) Date: Sun Dec 25 09:51:00 2005 Subject: [Jbeta] line number not shown in the window title In-Reply-To: <004301c608c3$0fa563e0$3151d0d8@xp> References: <1134988355.c80da1fcbill_lam@myrealbox.com> <01c001c6080b$572edfb0$77dad1d8@xp> <6f80c1520512231748q66238cc5j6ab6bb4eba48ca3a@mail.gmail.com> <004301c608c3$0fa563e0$3151d0d8@xp> Message-ID: <6f80c1520512241753y773376ffgaccd2fe1b3b53a80@mail.gmail.com> 9!:14'' j601/beta/2005-12-08/17:30 9!:12'' 6 I am using windows xp sp1. 504 works right but j601 doesn't work as I mentioned in the original post. I once copied the j.exe file from the j601 folder to the j504 folder and ran the j601 exe file in the j504 folder to test with the same environment and the line numbers didn't change as I moved the cursor. On 12/25/05, Eric Iverson wrote: > Does the 2nd number behave properly? That is, line position in explicit defn > ending in ) . Interestingly, the current line number changes as I move the cursor back and forth while in explicit define is on. x=:monad define but the right behavior ends as I close the definition with a paranthesis. June > > ----- Original Message ----- > From: "June Kim" > To: "Beta forum" > Sent: Friday, December 23, 2005 8:48 PM > Subject: Re: [Jbeta] line number not shown in the window title > > > I am using the 601 f beta and following the steps shown below I can > demonstrate the problem in my windows machine(windows xp sp1 with > korean language): > > Run the J window and see the ijx window shown with the title : 3.ijx* > [0] c:\j601\temp\3.ijx > > Now I type in three lines: > 1 > 1 > 2 > 2 > 3 > 3 > > I can see the line no change in the title from 0 to 2,4, and 6. Now I > move the cursor up to the first line and the line no does not change. > > Similar thing happens on ijs windows as well. > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > From juneaftn at gmail.com Sun Dec 25 09:58:05 2005 From: juneaftn at gmail.com (June Kim) Date: Sun Dec 25 09:55:59 2005 Subject: [Jbeta] line number not shown in the window title In-Reply-To: <6f80c1520512241753y773376ffgaccd2fe1b3b53a80@mail.gmail.com> References: <1134988355.c80da1fcbill_lam@myrealbox.com> <01c001c6080b$572edfb0$77dad1d8@xp> <6f80c1520512231748q66238cc5j6ab6bb4eba48ca3a@mail.gmail.com> <004301c608c3$0fa563e0$3151d0d8@xp> <6f80c1520512241753y773376ffgaccd2fe1b3b53a80@mail.gmail.com> Message-ID: <6f80c1520512241758w34529f91q746ae6d1f536ca95@mail.gmail.com> Oh, I missed one thing. In j504, if I open up the explicit define mode, there are two numbers shown in the title bar and as I move the cursor both the numbers change. However in j601, in explicit define mode, there are only one number shown(the line number of the whole window), and it changes as I move the cursor. On 12/25/05, June Kim wrote: > 9!:14'' > j601/beta/2005-12-08/17:30 > 9!:12'' > 6 > > I am using windows xp sp1. > > 504 works right but j601 doesn't work as I mentioned in the original > post. I once copied the j.exe file from the j601 folder to the j504 > folder and ran the j601 exe file in the j504 folder to test with the > same environment and the line numbers didn't change as I moved the > cursor. > > > On 12/25/05, Eric Iverson wrote: > > Does the 2nd number behave properly? That is, line position in explicit defn > > ending in ) . > > Interestingly, the current line number changes as I move the cursor > back and forth while in explicit define is on. > > x=:monad define > > > but the right behavior ends as I close the definition with a paranthesis. > > June > > > > ----- Original Message ----- > > From: "June Kim" > > To: "Beta forum" > > Sent: Friday, December 23, 2005 8:48 PM > > Subject: Re: [Jbeta] line number not shown in the window title > > > > > > I am using the 601 f beta and following the steps shown below I can > > demonstrate the problem in my windows machine(windows xp sp1 with > > korean language): > > > > Run the J window and see the ijx window shown with the title : 3.ijx* > > [0] c:\j601\temp\3.ijx > > > > Now I type in three lines: > > 1 > > 1 > > 2 > > 2 > > 3 > > 3 > > > > I can see the line no change in the title from 0 to 2,4, and 6. Now I > > move the cursor up to the first line and the line no does not change. > > > > Similar thing happens on ijs windows as well. > > > > > > ---------------------------------------------------------------------- > > For information about J forums see http://www.jsoftware.com/forums.htm > > > From juneaftn at gmail.com Sun Dec 25 10:25:55 2005 From: juneaftn at gmail.com (June Kim) Date: Sun Dec 25 10:23:49 2005 Subject: [Jbeta] line number not shown in the window title In-Reply-To: <6f80c1520512241758w34529f91q746ae6d1f536ca95@mail.gmail.com> References: <1134988355.c80da1fcbill_lam@myrealbox.com> <01c001c6080b$572edfb0$77dad1d8@xp> <6f80c1520512231748q66238cc5j6ab6bb4eba48ca3a@mail.gmail.com> <004301c608c3$0fa563e0$3151d0d8@xp> <6f80c1520512241753y773376ffgaccd2fe1b3b53a80@mail.gmail.com> <6f80c1520512241758w34529f91q746ae6d1f536ca95@mail.gmail.com> Message-ID: <6f80c1520512241825q552a0feag544600c86373f7dd@mail.gmail.com> I think I have identified the cause of this malfunction. I opened up my task manager(ctrl-alt-esc) and kill the Tablet.exe process. Now magically the line numbering works right in j601fbeta as well. Tablet.exe is wacom tablet driver loader: http://www.windowsstartup.com/wso/detail.php?id=2824 I didn't connect the tablet to my computer's usb port even. I think you can reproduce the same malfunction if you download and run the Tablet.exe (maybe from the wacom site?). Somehow, there got in a conflict between Tablet.exe and J, since J601. There was no conflict with J504. On 12/25/05, June Kim wrote: > Oh, I missed one thing. In j504, if I open up the explicit define > mode, there are two numbers shown in the title bar and as I move the > cursor both the numbers change. > > However in j601, in explicit define mode, there are only one number > shown(the line number of the whole window), and it changes as I move > the cursor. > > On 12/25/05, June Kim wrote: > > 9!:14'' > > j601/beta/2005-12-08/17:30 > > 9!:12'' > > 6 > > > > I am using windows xp sp1. > > > > 504 works right but j601 doesn't work as I mentioned in the original > > post. I once copied the j.exe file from the j601 folder to the j504 > > folder and ran the j601 exe file in the j504 folder to test with the > > same environment and the line numbers didn't change as I moved the > > cursor. > > > > > > On 12/25/05, Eric Iverson wrote: > > > Does the 2nd number behave properly? That is, line position in explicit defn > > > ending in ) . > > > > Interestingly, the current line number changes as I move the cursor > > back and forth while in explicit define is on. > > > > x=:monad define > > > > > > but the right behavior ends as I close the definition with a paranthesis. > > > > June > > > > > > ----- Original Message ----- > > > From: "June Kim" > > > To: "Beta forum" > > > Sent: Friday, December 23, 2005 8:48 PM > > > Subject: Re: [Jbeta] line number not shown in the window title > > > > > > > > > I am using the 601 f beta and following the steps shown below I can > > > demonstrate the problem in my windows machine(windows xp sp1 with > > > korean language): > > > > > > Run the J window and see the ijx window shown with the title : 3.ijx* > > > [0] c:\j601\temp\3.ijx > > > > > > Now I type in three lines: > > > 1 > > > 1 > > > 2 > > > 2 > > > 3 > > > 3 > > > > > > I can see the line no change in the title from 0 to 2,4, and 6. Now I > > > move the cursor up to the first line and the line no does not change. > > > > > > Similar thing happens on ijs windows as well. > > > > > > > > > ---------------------------------------------------------------------- > > > For information about J forums see http://www.jsoftware.com/forums.htm > > > > > > From juneaftn at gmail.com Sun Dec 25 11:06:54 2005 From: juneaftn at gmail.com (June Kim) Date: Sun Dec 25 11:04:48 2005 Subject: [Jbeta] line number not shown in the window title In-Reply-To: <6f80c1520512241825q552a0feag544600c86373f7dd@mail.gmail.com> References: <1134988355.c80da1fcbill_lam@myrealbox.com> <01c001c6080b$572edfb0$77dad1d8@xp> <6f80c1520512231748q66238cc5j6ab6bb4eba48ca3a@mail.gmail.com> <004301c608c3$0fa563e0$3151d0d8@xp> <6f80c1520512241753y773376ffgaccd2fe1b3b53a80@mail.gmail.com> <6f80c1520512241758w34529f91q746ae6d1f536ca95@mail.gmail.com> <6f80c1520512241825q552a0feag544600c86373f7dd@mail.gmail.com> Message-ID: <6f80c1520512241906g190eb74fu50809d18ed646842@mail.gmail.com> I have solved the problem by updating the tablet driver with a newer version. My driver's version was 4.7.6.8, which caused the problem with J. Those versions above 4.9.x.x doesn't interfere with J's line number display in the title bar. I think a caution in the README file or somewhere in the distribution would help users who might be using wacom's previous drivers. However, the line number thing is not perfect yet. In j504, I see two numbers displayed when in explicit define mode but I see only one number in j601f. Is it already a known problem here? On 12/25/05, June Kim wrote: > I think I have identified the cause of this malfunction. > > I opened up my task manager(ctrl-alt-esc) and kill the Tablet.exe > process. Now magically the line numbering works right in j601fbeta as > well. > > Tablet.exe is wacom tablet driver loader: > http://www.windowsstartup.com/wso/detail.php?id=2824 > > I didn't connect the tablet to my computer's usb port even. I think > you can reproduce the same malfunction if you download and run the > Tablet.exe (maybe from the wacom site?). > > Somehow, there got in a conflict between Tablet.exe and J, since J601. > There was no conflict with J504. > > > On 12/25/05, June Kim wrote: > > Oh, I missed one thing. In j504, if I open up the explicit define > > mode, there are two numbers shown in the title bar and as I move the > > cursor both the numbers change. > > > > However in j601, in explicit define mode, there are only one number > > shown(the line number of the whole window), and it changes as I move > > the cursor. > > > > On 12/25/05, June Kim wrote: > > > 9!:14'' > > > j601/beta/2005-12-08/17:30 > > > 9!:12'' > > > 6 > > > > > > I am using windows xp sp1. > > > > > > 504 works right but j601 doesn't work as I mentioned in the original > > > post. I once copied the j.exe file from the j601 folder to the j504 > > > folder and ran the j601 exe file in the j504 folder to test with the > > > same environment and the line numbers didn't change as I moved the > > > cursor. > > > > > > > > > On 12/25/05, Eric Iverson wrote: > > > > Does the 2nd number behave properly? That is, line position in explicit defn > > > > ending in ) . > > > > > > Interestingly, the current line number changes as I move the cursor > > > back and forth while in explicit define is on. > > > > > > x=:monad define > > > > > > > > > but the right behavior ends as I close the definition with a paranthesis. > > > > > > June > > > > > > > > ----- Original Message ----- > > > > From: "June Kim" > > > > To: "Beta forum" > > > > Sent: Friday, December 23, 2005 8:48 PM > > > > Subject: Re: [Jbeta] line number not shown in the window title > > > > > > > > > > > > I am using the 601 f beta and following the steps shown below I can > > > > demonstrate the problem in my windows machine(windows xp sp1 with > > > > korean language): > > > > > > > > Run the J window and see the ijx window shown with the title : 3.ijx* > > > > [0] c:\j601\temp\3.ijx > > > > > > > > Now I type in three lines: > > > > 1 > > > > 1 > > > > 2 > > > > 2 > > > > 3 > > > > 3 > > > > > > > > I can see the line no change in the title from 0 to 2,4, and 6. Now I > > > > move the cursor up to the first line and the line no does not change. > > > > > > > > Similar thing happens on ijs windows as well. > > > > > > > > > > > > ---------------------------------------------------------------------- > > > > For information about J forums see http://www.jsoftware.com/forums.htm > > > > > > > > > > From gosinn at gmail.com Sun Dec 25 18:24:36 2005 From: gosinn at gmail.com (=?ISO-8859-1?Q?Bj=F6rn_Helgason?=) Date: Sun Dec 25 18:22:31 2005 Subject: [Jbeta] line number not shown in the window title In-Reply-To: <6f80c1520512241906g190eb74fu50809d18ed646842@mail.gmail.com> References: <1134988355.c80da1fcbill_lam@myrealbox.com> <01c001c6080b$572edfb0$77dad1d8@xp> <6f80c1520512231748q66238cc5j6ab6bb4eba48ca3a@mail.gmail.com> <004301c608c3$0fa563e0$3151d0d8@xp> <6f80c1520512241753y773376ffgaccd2fe1b3b53a80@mail.gmail.com> <6f80c1520512241758w34529f91q746ae6d1f536ca95@mail.gmail.com> <6f80c1520512241825q552a0feag544600c86373f7dd@mail.gmail.com> <6f80c1520512241906g190eb74fu50809d18ed646842@mail.gmail.com> Message-ID: <4031fef0512250224i32ffac3dq@mail.gmail.com> ---------- Forwarded message ---------- From: Eric Iverson Date: 23.12.2005 21:53 Subject: Re: [Jbeta] line number not shown in the window title To: Beta forum Bill, thanks, the 2nd number (line number in explicit) is broken in 601 and will be fixed. 2005/12/25, June Kim : > > However, the line number thing is not perfect yet. In j504, I see two > numbers displayed when in explicit define mode but I see only one > number in j601f. Is it already a known problem here? > From bill_lam at myrealbox.com Mon Dec 26 23:35:40 2005 From: bill_lam at myrealbox.com (bill lam) Date: Mon Dec 26 23:33:33 2005 Subject: [Jbeta] J601 gui wish list Message-ID: <1135611340.c7d3d83cbill_lam@myrealbox.com> 1. treeview list 2. textbox for integer 3. textbox for decimal number 4. textbox for date/time although there are ocx available, it will be nice if J have them. From davidmitchell at att.net Tue Dec 27 00:50:40 2005 From: davidmitchell at att.net (David Mitchell) Date: Tue Dec 27 00:47:52 2005 Subject: [Jbeta] SFL Library lab section 9 Message-ID: <43B01F60.5060300@att.net> The SFL Library lab dies in section 9 after: expandblock_jsfl_ compressblock_jsfl_ 20#'a' J has encountered a problem and needs to close. We are sorry for the inconvenience. AppName: j.exe AppVer: 5.0.5.0 ModName: ntdll.dll ModVer: 5.1.2600.1217 Offset: 00033905 9!:14'' j601/beta/2005-12-08/17:30 If I run the sentence outside of the lab, it works fine. Dave Mitchell From gkirouac at riq.qc.ca Tue Dec 27 13:39:46 2005 From: gkirouac at riq.qc.ca (Gilles Kirouac) Date: Tue Dec 27 13:37:41 2005 Subject: [Jbeta] ocx support needs updating Message-ID: <20051227053547.M93902@riq.qc.ca> What follows applies to V5.04 and V6 ebeta. 0)ocxutils.ijs After changing the last few lines of ocxutils.ijs as below, ocxinfo runs ok: ocxchar=: 2 : 'ocxset (@ocx_char)' NB. character ocxnum =: 2 : 'ocxset (@ocx_num)' NB. numeric ocxindex=: 2 : 'ocxset (&.>) (@ocx_index) (;@)' NB. index ocxclist=: 2 : 'ocxset (&.>) (@ocx_clist) (;@)' NB. character list ocxnlist=: 2 : 'ocxset (&.>) (@ocx_nlist) (;@)' NB. numeric list ocxilist=: 2 : 'ocxset (&.>) (@ocx_ilist) (;@)' NB. numeric index ocxdata =: 2 : 'ocxset (@ocxfmt)' NB. data This follows from the bident (C A) deprecated with V5. Surprisingly this was not updated with any 5.0x. However ocxdialog still has a problem. ocxdialog'Shell.Explorer.2' │domain error: wd │ wd'oledlg x0' wd'qer' command failed : 0 1)These pages refer to ".js" scripts which are long outdated. file:///c:/j601/system/extras/help/user/isigraph_events.htm file:///c:/j601/system/extras/help/user/j_ole_client.htm file:///c:/j601/system/extras/help/user/ole_examples.htm Gilles ~ From viton.1 at osu.edu Wed Dec 28 01:15:47 2005 From: viton.1 at osu.edu (Philip A. Viton) Date: Wed Dec 28 02:13:42 2005 Subject: [Jbeta] z fmt Message-ID: <7.0.0.16.2.20051227131115.0222fd30@osu.edu> Is there any way to get the effect of the 'z' modifier in the old fmt verb in the format package, using the new format foreign? That is, under j5, 'z2' fmt 2 -> '02'. If this isn't possible, mightn't it be reasonable to make the old fmt verb available in j6's format package, but under a different name, for compatibility's sake? Or are there now some cases where the old fmt just doesn't work on j6 data? ------------------------ Philip A. Viton City Planning, Ohio State University 275 West Woodruff Avenue, Columbus OH 43210 viton.1@osu.edu From gosinn at gmail.com Wed Dec 28 02:38:52 2005 From: gosinn at gmail.com (=?ISO-8859-1?Q?Bj=F6rn_Helgason?=) Date: Wed Dec 28 02:36:43 2005 Subject: [Jbeta] z fmt In-Reply-To: <7.0.0.16.2.20051227131115.0222fd30@osu.edu> References: <7.0.0.16.2.20051227131115.0222fd30@osu.edu> Message-ID: <4031fef0512271038o4458b621q@mail.gmail.com> If you take the fmt verb from c:\j504\system\main\format.ijs and run it in j601 'z2' fmt 2 02 2005/12/27, Philip A. Viton : > > > Is there any way to get the effect of the 'z' modifier in the old fmt > verb in the format package, using the new format foreign? That is, > under j5, 'z2' fmt 2 -> '02'. > > If this isn't possible, mightn't it be reasonable to make the old fmt > verb available in j6's format package, but under a different name, > for compatibility's sake? Or are there now some cases where the old > fmt just doesn't work on j6 data? > > > > > > ------------------------ > Philip A. Viton > City Planning, Ohio State University > 275 West Woodruff Avenue, Columbus OH 43210 > viton.1@osu.edu > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > -- Bj?rn Helgason, Verkfr??ingur, gsm: +354698 5532 Fugl&Fiskur ehf, ?erneyjarsund 23, 802 Selfoss ,t-p?st: gosinn@gmail.com http://www.linux.is/forritun/data/J/grunnur.pdf T?knikunn?tta h?ndlar hi? fl?kna, sk?punarg?fa er meistari einfaldleikans g??ur kennari getur stigi? ? t?r ?n ?ess a? glansinn fari af sk?num /|_ .-----------------------------------. ,' .\ / | Me? l?ttri lund ver?ur | ,--' _,' | Dagurinn ? dag | / / | Enn betri en g?rdagurinn | ( -. | `-----------------------------------' | ) | (`-. '--.) `. )----' From donguinn at pvtnetworks.net Wed Dec 28 03:19:48 2005 From: donguinn at pvtnetworks.net (Don Guinn) Date: Wed Dec 28 03:17:42 2005 Subject: [Jbeta] z fmt In-Reply-To: <7.0.0.16.2.20051227131115.0222fd30@osu.edu> References: <7.0.0.16.2.20051227131115.0222fd30@osu.edu> Message-ID: <43B193D4.6040306@pvtn.net> A trick I pulled in the past for this was x=.2 }.":100+x 02 If you only need this part of fmt this might be easier than running fmt. Philip A. Viton wrote: > > Is there any way to get the effect of the 'z' modifier in the old fmt > verb in the format package, using the new format foreign? That is, > under j5, 'z2' fmt 2 -> '02'. > > If this isn't possible, mightn't it be reasonable to make the old fmt > verb available in j6's format package, but under a different name, for > compatibility's sake? Or are there now some cases where the old fmt > just doesn't work on j6 data? > > > > > > ------------------------ > Philip A. Viton > City Planning, Ohio State University > 275 West Woodruff Avenue, Columbus OH 43210 > viton.1@osu.edu > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > From eiverson at sympatico.ca Wed Dec 28 06:14:15 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Wed Dec 28 06:12:18 2005 Subject: [Jbeta] ocx support needs updating References: <20051227053547.M93902@riq.qc.ca> Message-ID: <006401c60b32$e7771810$0701a8c0@xp> Chris will fix the script problems you point out. ocxdialog'Shell.Explorer.2' fails because this ocx doesn't provide a dialog and this failure is the same in previous versions. Instead try: ocxdialog'MSCAL.Calendar.7' ----- Original Message ----- From: "Gilles Kirouac" To: Sent: Tuesday, December 27, 2005 12:39 AM Subject: [Jbeta] ocx support needs updating > > What follows applies to V5.04 and V6 ebeta. > > 0)ocxutils.ijs > > After changing the last few lines of ocxutils.ijs as below, ocxinfo runs > ok: > > ocxchar=: 2 : 'ocxset (@ocx_char)' NB. character > ocxnum =: 2 : 'ocxset (@ocx_num)' NB. numeric > ocxindex=: 2 : 'ocxset (&.>) (@ocx_index) (;@)' NB. index > ocxclist=: 2 : 'ocxset (&.>) (@ocx_clist) (;@)' NB. character list > ocxnlist=: 2 : 'ocxset (&.>) (@ocx_nlist) (;@)' NB. numeric list > ocxilist=: 2 : 'ocxset (&.>) (@ocx_ilist) (;@)' NB. numeric index > ocxdata =: 2 : 'ocxset (@ocxfmt)' NB. data > > This follows from the bident (C A) deprecated with V5. > Surprisingly this was not updated with any 5.0x. > > However ocxdialog still has a problem. > ocxdialog'Shell.Explorer.2' > │domain error: wd > │ wd'oledlg x0' > wd'qer' > command failed : 0 > > > 1)These pages refer to ".js" scripts which are long outdated. > > file:///c:/j601/system/extras/help/user/isigraph_events.htm > file:///c:/j601/system/extras/help/user/j_ole_client.htm > file:///c:/j601/system/extras/help/user/ole_examples.htm > > Gilles ~ > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm From eiverson at sympatico.ca Wed Dec 28 06:32:22 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Wed Dec 28 06:30:15 2005 Subject: [Jbeta] SFL Library lab section 9 References: <43B01F60.5060300@att.net> Message-ID: <008701c60b35$69f00e30$0701a8c0@xp> I haven't dug into this but the cd code in expandblock looks suspicious. My quick guess is that it is in error and the problem is unrelated to the beta. The verb expandblock does 2 cd calls as is typical for this kind of routine. The first call is to find out how large the result will be and then allocates the space and the second call fill it with the result. The 1st call in expand block looks wrong. It allocates 10 bytes and calls the routine. The routine has no way to know this is just a call for the size and will overwrites memory with any result greater than 10 bytes. The 1st call should probably be with a null pointer (<<0). If you track this down, let me know what you find and suggest a fix for the lab. Otherwise I'll eventually find the spec for the routine and fix the lab. ----- Original Message ----- From: "David Mitchell" To: "Beta forum" Sent: Monday, December 26, 2005 11:50 AM Subject: [Jbeta] SFL Library lab section 9 > The SFL Library lab dies in section 9 after: > > expandblock_jsfl_ compressblock_jsfl_ 20#'a' > > J has encountered a problem and needs to close. We are sorry for the > inconvenience. > > AppName: j.exe AppVer: 5.0.5.0 ModName: ntdll.dll > ModVer: 5.1.2600.1217 Offset: 00033905 > > 9!:14'' > j601/beta/2005-12-08/17:30 > > If I run the sentence outside of the lab, it works fine. > > Dave Mitchell > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm From davidmitchell at att.net Wed Dec 28 11:13:50 2005 From: davidmitchell at att.net (David Mitchell) Date: Wed Dec 28 11:11:00 2005 Subject: [Jbeta] SFL Library lab section 9 In-Reply-To: <008701c60b35$69f00e30$0701a8c0@xp> References: <43B01F60.5060300@att.net> <008701c60b35$69f00e30$0701a8c0@xp> Message-ID: <43B202EE.2080108@att.net> You were correct to suspect the cd code. As far as I can tell, the code fails with a null pointer for the result buffer also. The code published at http://www.imatix.com/html/sfl/sfl19.htm#TOC32 has no provision for sizing the result. The doc for expand_block says (about the result buffer size): "dst must be large enough to accommodate the largest possible decompressed block." This is in general not easily calculated. I suspect the simplest work around is to capture the size of the uncompressed data at compress time and use the saved size to build the buffer: compress_block=: 'compress_block + s *c *c s'cdm expand_block=: 'expand_block + s *c *c s'cdm compressblock=: verb define source=. y. osize=. #source dest=. osize#' ' 'size source dest sourcesize'=. compress_block source;dest;#source osize;size{.dest ) expandblock=: verb define 'osize source'=. y. source=. ,source dest=. osize#{.a. 'size source dest sourcesize'=. expand_block source;dest;#source size{.dest ) This could be done for all of the expand_ verbs. Eric Iverson wrote: > I haven't dug into this but the cd code in expandblock looks suspicious. > My quick guess is that it is in error and the problem is unrelated to > the beta. The verb expandblock does 2 cd calls as is typical for this > kind of routine. The first call is to find out how large the result will > be and then allocates the space and the second call fill it with the > result. The 1st call in expand block looks wrong. It allocates 10 bytes > and calls the routine. The routine has no way to know this is just a > call for the size and will overwrites memory with any result greater > than 10 bytes. The 1st call should probably be with a null pointer > (<<0). If you track this down, let me know what you find and suggest a > fix for the lab. Otherwise I'll eventually find the spec for the routine > and fix the lab. > > ----- Original Message ----- From: "David Mitchell" > To: "Beta forum" > Sent: Monday, December 26, 2005 11:50 AM > Subject: [Jbeta] SFL Library lab section 9 > > >> The SFL Library lab dies in section 9 after: >> >> expandblock_jsfl_ compressblock_jsfl_ 20#'a' >> >> J has encountered a problem and needs to close. We are sorry for the >> inconvenience. >> >> AppName: j.exe AppVer: 5.0.5.0 ModName: ntdll.dll >> ModVer: 5.1.2600.1217 Offset: 00033905 >> >> 9!:14'' >> j601/beta/2005-12-08/17:30 >> >> If I run the sentence outside of the lab, it works fine. >> >> Dave Mitchell >> ---------------------------------------------------------------------- >> For information about J forums see http://www.jsoftware.com/forums.htm > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > From eiverson at sympatico.ca Thu Dec 29 02:54:01 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Thu Dec 29 02:51:54 2005 Subject: [Jbeta] SFL Library lab section 9 References: <43B01F60.5060300@att.net> <008701c60b35$69f00e30$0701a8c0@xp> <43B202EE.2080108@att.net> Message-ID: <001e01c60be0$139eec80$0701a8c0@xp> I have fixed sflcompress.ijs and the lab and there is a new version of the sfl addon at the web site. Thanks for pointing out this problem. ----- Original Message ----- From: "David Mitchell" To: "Beta forum" Sent: Tuesday, December 27, 2005 10:13 PM Subject: Re: [Jbeta] SFL Library lab section 9 > You were correct to suspect the cd code. As far as I can tell, the code > fails with a null pointer for the result buffer also. The code published > at http://www.imatix.com/html/sfl/sfl19.htm#TOC32 > has no provision for sizing the result. > > The doc for expand_block says (about the result buffer size): > "dst must be large enough to accommodate the largest possible decompressed > block." > > This is in general not easily calculated. I suspect the simplest work > around is to capture the size of the uncompressed data at compress time > and use the saved size to build the buffer: > > compress_block=: 'compress_block + s *c *c s'cdm > expand_block=: 'expand_block + s *c *c s'cdm > > compressblock=: verb define > source=. y. > osize=. #source > dest=. osize#' ' > 'size source dest sourcesize'=. compress_block source;dest;#source > osize;size{.dest > ) > > expandblock=: verb define > 'osize source'=. y. > source=. ,source > dest=. osize#{.a. > 'size source dest sourcesize'=. expand_block source;dest;#source > size{.dest > ) > > This could be done for all of the expand_ verbs. > > Eric Iverson wrote: >> I haven't dug into this but the cd code in expandblock looks suspicious. >> My quick guess is that it is in error and the problem is unrelated to the >> beta. The verb expandblock does 2 cd calls as is typical for this kind of >> routine. The first call is to find out how large the result will be and >> then allocates the space and the second call fill it with the result. The >> 1st call in expand block looks wrong. It allocates 10 bytes and calls the >> routine. The routine has no way to know this is just a call for the size >> and will overwrites memory with any result greater than 10 bytes. The 1st >> call should probably be with a null pointer (<<0). If you track this >> down, let me know what you find and suggest a fix for the lab. Otherwise >> I'll eventually find the spec for the routine and fix the lab. >> >> ----- Original Message ----- From: "David Mitchell" >> >> To: "Beta forum" >> Sent: Monday, December 26, 2005 11:50 AM >> Subject: [Jbeta] SFL Library lab section 9 >> >> >>> The SFL Library lab dies in section 9 after: >>> >>> expandblock_jsfl_ compressblock_jsfl_ 20#'a' >>> >>> J has encountered a problem and needs to close. We are sorry for the >>> inconvenience. >>> >>> AppName: j.exe AppVer: 5.0.5.0 ModName: ntdll.dll >>> ModVer: 5.1.2600.1217 Offset: 00033905 >>> >>> 9!:14'' >>> j601/beta/2005-12-08/17:30 >>> >>> If I run the sentence outside of the lab, it works fine. >>> >>> Dave Mitchell >>> ---------------------------------------------------------------------- >>> For information about J forums see http://www.jsoftware.com/forums.htm >> >> >> ---------------------------------------------------------------------- >> For information about J forums see http://www.jsoftware.com/forums.htm >> > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm From PGauthier at checkfree.com Thu Dec 29 22:45:34 2005 From: PGauthier at checkfree.com (PGauthier@checkfree.com) Date: Thu Dec 29 22:42:36 2005 Subject: [Jbeta] The turtle lab generates an error In-Reply-To: <006401c60b32$e7771810$0701a8c0@xp> Message-ID: The following error appears when I try to {Ctrl+J} through the Turtle Geometry Lab. wdhandler error in :jijs_jctrl_fkey domain error: spos f~v Paul Gauthier APL Software Developer - Senior pgauthier@checkfree.com Phone: 312-739-3467 Fax: 312-739-3496 CheckFree. The Company that Powers Payment on the WebSM. http://www.checkfree.com/paybillsonline From geomap at freesurf.ch Thu Dec 29 22:42:34 2005 From: geomap at freesurf.ch (David Vincent-Jones) Date: Thu Dec 29 23:59:28 2005 Subject: [Jbeta] J in the Background Message-ID: <000001c60c86$1e411310$0201a8c0@DAVID> Not sure what is going on .. this may be my system and .. I have recently switched to a new computer and from W2K to XPpro with 1 gb. memory. I have always been able to run J in the background but find that I cannot even watch TaskManager running at the same time as J I am able to run multiple other programs and do background work but with J the program just stops responding. If I start a second J session the first stops responding but the second continues just fine. Priority is set for the foreground application but not to exclude the background. I have never been aware of this problem in the past. Should I be using a special start-up profile? David From eiverson at sympatico.ca Fri Dec 30 06:47:19 2005 From: eiverson at sympatico.ca (Eric Iverson) Date: Fri Dec 30 06:45:07 2005 Subject: [Jbeta] J in the Background References: <000001c60c86$1e411310$0201a8c0@DAVID> Message-ID: <002b01c60cc9$d4f3b8e0$0701a8c0@xp> This sounds very strange. The first thing is to confirm whether it has anything to do with the new beta. That is, try your experiments on your new platform and see if J504 behaves properly. Let's determine if the problem is J vs your new platform or J6 vs your new platform. ----- Original Message ----- From: "David Vincent-Jones" To: "JSoftware Beta" Sent: Thursday, December 29, 2005 9:42 AM Subject: [Jbeta] J in the Background > Not sure what is going on .. this may be my system and .. > I have recently switched to a new computer and from W2K to XPpro with 1 > gb. > memory. > > I have always been able to run J in the background but find that I cannot > even watch TaskManager running at the same time as J > > I am able to run multiple other programs and do background work but with J > the program just stops responding. > > If I start a second J session the first stops responding but the second > continues just fine. > > Priority is set for the foreground application but not to exclude the > background. > > I have never been aware of this problem in the past. Should I be using a > special start-up profile? > > David > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm From reiterc at mail.lafayette.edu Fri Dec 30 21:13:11 2005 From: reiterc at mail.lafayette.edu (Cliff Reiter) Date: Fri Dec 30 21:11:02 2005 Subject: [Jbeta] J in the Background In-Reply-To: <000001c60c86$1e411310$0201a8c0@DAVID> References: <000001c60c86$1e411310$0201a8c0@DAVID> Message-ID: <43B53267.5020307@mail.lafayette.edu> David, When you changed machines, did you switch from a machine that had hyperthreading (or dual processing) to one without? Or did hyperthreading get turned off on the new machine? While what you describe sounds pretty extreme, I know when J is running, the status of hyperthreading makes a huge difference on our lab machines. Best, Cliff David Vincent-Jones wrote: > Not sure what is going on .. this may be my system and .. > I have recently switched to a new computer and from W2K to XPpro with 1 gb. > memory. > > I have always been able to run J in the background but find that I cannot > even watch TaskManager running at the same time as J > > I am able to run multiple other programs and do background work but with J > the program just stops responding. > > If I start a second J session the first stops responding but the second > continues just fine. > > Priority is set for the foreground application but not to exclude the > background. > > I have never been aware of this problem in the past. Should I be using a > special start-up profile? > > David > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > -- Clifford A. Reiter Mathematics Department, Lafayette College Easton, PA 18042 USA, 610-330-5277 http://www.lafayette.edu/~reiterc From fraser.jackson at xtra.co.nz Fri Dec 30 16:43:19 2005 From: fraser.jackson at xtra.co.nz (Fraser Jackson) Date: Sat Dec 31 14:24:22 2005 Subject: [Jbeta] The turtle lab generates an error References: Message-ID: <000001c60dd3$262afd90$502558db@vaio> This lab is broken with J601 and some reworking will be done in the New Year to get it working again. )