From owner-jsoftware@lists.interlog.com Sun Aug 1 19:28:13 1999 Date: Sun, 1 Aug 1999 16:26:50 -0700 From: gordon@island.net (Ken Ian Gordon) Subject: Re: Jforum: Viewing large arrays. User help Brian Bambrough wrote >A few days ago I posted some questions on the J Forum. Two of them >were to do with wdview and "what's this?" kind of help. Nobody had Sorry about that. ........ > >If anyone has suggestions for improvements I would be greatly >interested. I used an alternative approach to supply help on buttons. As you know it is available for menu entries, but I too could find nothing for buttons in J. I used the F1 key instead. When pressed it ascertained which button had the focus and displayed the appropriate short help message in the status area as taken from a list. Using TAB to move focus and pressing F! then gave all the help messages for the buttons. It also gave short help messages for anything data entry fields, to explain the required entry. These were taken from the meta data. Not as good as right button, or the automatic display for menus, but it appeared to suffice. regards kig -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Mon Aug 2 00:16:50 1999 Date: Mon, 02 Aug 1999 00:15:29 -0400 From: David Ness Subject: Jforum: A (surely) stupid idea ... I find that I have (slightly) less trouble reading J than K. I should say at the outset, that I am not bitching, here, as I find that reading both is enormously profitable for me. But I wonder if one might help me with the other. For example, in the recently discussed `Combinations' problem, Roger produced an answer that I understood after a couple of hours of study, while over on the K-forum, Arthur's equally elegant answer took me an extra couple of days. All of the hours and days were well spent in both endavours, so I am not complaining about any of this. However, in the process I did discover that, at least for me, J's ability to `expose' the parse of the input and give me sort of a `diagrammatic' representation of how it `sees the world' as it executes the input is a step up on understanding why the answers that it produces are right, and my thoughts about what it `should do' are invariably naive. This leads me to wonder if it would be an interesting and instructive exercise to try to produce a `K simulator' in J. I certainly understand that producing a full K simulator would be both hard and pointless. However, short of that, I find I have enough trouble with understanding some of the elemental aspects of K, that `exposing' their opertaion in J would be some contribution to my understanding of both languages. I understand that this is probably all silly, but I have the feeling that if the `better heads' here are able to tell me why, I'll learn something from raising the question. I recognize also that I could equally well raise this question in the K forum. However, given that I have a harder time, given present experience, understanding K, for me this would be an exercise in trying to use the more obscure to un-obfuscate. And thus a more difficult task... -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Mon Aug 2 15:25:30 1999 Delivered-To: fixup-forum@JSoftware.Com@fixme Date: Mon, 02 Aug 1999 12:22:56 -0700 From: greg heil Subject: Re: Jforum: Cross Post: A (surely) stupid idea ... References: <37A51B61.46C29812@Home.Com> David Ness wrote: > ... in the recently discussed `Combinations' problem, Roger produced an answer that I understood after a couple of hours of study, while over on the K-forum, Arthur's equally elegant answer took me an extra couple of days. The combinations problem can be a nice "Rosetta Stone" which was one of the reasons i brought it up on the K-list. Other than that i actually needed the functionality;} Rogers and Arthurs solutions are quite parallel, too. NB. Roger Hui: rc=: 4 : 0 NB. All size x. combinations of i.y. z=.1 0$k=.i.#c=.1,~(y.-x.)$0 for. i.x. do. z=.;k,.&.>(-c=.+/\.c){.&.><1+z end. ) /Arthur Whitney: ac:{,/x{k,''(+\#:'x)#\:,/1+x}/k:|!y-x-:1} K derives much of its brevity in this example from the fact that list operands are the native type, whereas in J the under operator, &., can be combined with boxing, <, to achieve the same end. The k's in both implementations serve the same purpose, to be the addend for each growing combination and are formed analogously through index set generators. Being as {k,''(+\#:'x)#\:,/1+x} is a monad the first x in ac serves as a loop count, in parallel with rc's for. i.x. do. rc's c is implicit in the ac (+\#:'x). Which is to say a scan of the counts of the previous iteration. Putting these together in the iterative step of the rc case: k,.&.>(-c=.+/\.c){.&.><1+z the addends k are laminated (,.&.>) to count (-c=.+/\.c) copies ({.&.>)) of the incremented (<1+z) results of the previous iteration (for. i.x. do.). Much like in ac where k's are laminated (,'') to count (+\#:'x) copies (#\:) of the ravelled (,/) incremented (1+x) results of the previous iteration (x{}/). A strictly kosher way to figure out a phrase's meaning is to use the languages debug facility to dissect it. In K's case this is the stop/trace/null facility. Setting it to trace \b t / and instrumenting ac: ac:{+,/x{\ k,''(\ +\#:'x)#\:,/1+x}/k:|!y-x-:1} ac[3;6] / and running... 1 2 3 4 / 1st iteration counts; and results: (,3 4 (2 4 2 3) (1 4 1 3 1 2) (0 4 0 3 0 2 0 1)) 1 3 6 10 / 2nd iteration (,3 4 5 (2 4 5 2 3 5 2 3 4) (1 4 5 1 3 5 1 3 4 1 2 5 1 2 4 1 2 3) (0 4 5 0 3 5 0 3 4 0 2 5 0 2 4 0 2 3 0 1 5 0 1 4 0 1 3 0 1 2)) / and final result (transposed for brevity) (3 2 2 2 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 4 4 3 3 4 3 3 2 2 2 4 3 3 2 2 2 1 1 1 1 5 5 5 4 5 5 4 5 4 3 5 5 4 5 4 3 5 4 3 2) > This leads me to wonder if it would be an interesting and instructive exercise to try to produce a `K simulator'... Stevan Apter has put out an interpreter for "pure K" a couple times (eg DejaNews for comp.lang.apl). But that is still a long way from capturing all the syntactic sweeteners. greg heil mailto:gheil@acm.org http://www.scn.org/tl/anvil -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Wed Aug 4 01:51:59 1999 From: k-list@iname.com Date: Wed, 4 Aug 1999 01:47:24 -0400 (EDT) Content-Type: Text/Plain Subject: Jforum: implementation of "record lock" "bill lam" I'm trying to implement an address book application using J4.03, it uses ddsel(select *) to fetch a record for view/change, and then uses ddsql(update) for updating. It works fine, until concurrent access occurs. The obvious solution is of course using "record lock", but how? I can think of 2 possible ways, 1) use ODBC "cursor library" 2) use a centralised file for holding record lock information. but I haven't tried these yet, anyone has experience in this subject matter or other suggestions? --------------------------------------------------- Get free personalized email at http://www.iname.com -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Wed Aug 4 08:40:04 1999 From: gosi@centrum.is Subject: Re: Jforum: implementation of record lock Date: Wed, 4 Aug 99 12:38:57 +0000 k-list@iname.com writes: > "bill lam" > I'm trying to implement an address book application using J4.03, > it uses ddsel(select *) to fetch a record for view/change, > and then uses ddsql(update) for updating. It works fine, until > concurrent access occurs. The obvious solution is of course using > "record lock", but how? I can think of 2 possible ways, > 1) use ODBC "cursor library" > 2) use a centralised file for holding record lock information. It all depends on the database you are using and what options that database allows you to do. if your application is the only application using this database then of course you can use your own methods. Otherwise it is up to the database. /Gosi -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Wed Aug 4 17:17:56 1999 From: "Nichols, Peter" Subject: Jforum: TreeView OCX Date: Wed, 4 Aug 1999 15:46:12 -0500 Could someone give me an example of using the TreeView OCX? And also what other tree controls are available and any recommendations would be great. Thanks, Pete Nichols -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Thu Aug 5 06:23:57 1999 From: Subject: Jforum: A chanllenge... Date: Thu, 5 Aug 1999 12:21:25 +0200 I have a future doctor in Operational Research who beleives that no one in the world can achieve better results than the CPLEX corporation solver on integer's linear programming. Anyone interested to give hints on a better solution please contact me. Regards/Paul -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Thu Aug 5 11:37:39 1999 From: "Oleg Kobchenko" References: <01074BD4EEC4D1118E3200805F6542A7177FF9@SPRDALLAS2> Subject: Re: Jforum: TreeView OCX Date: Thu, 5 Aug 1999 18:38:39 +0300 charset="koi8-r" You can check http://members.xoom.com/olegyk/jpage.html Also J distribution should have such sample. TreeView should saticefy most needs. A single example of a different and original type of tree is Xerox Hyperbolic Tree control. ----- Original Message ----- From: Nichols, Peter Sent: Wednesday, August 04, 1999 23:46 Subject: Jforum: TreeView OCX > Could someone give me an example of using the TreeView OCX? And also what > other tree controls are available and any recommendations would be great. > > Thanks, > Pete Nichols -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Thu Aug 5 14:30:50 1999 From: btanksley@hifn.com Subject: RE: Jforum: A chanllenge... Date: Thu, 5 Aug 1999 11:29:15 -0700 charset="iso-8859-1" Could you clarify? I've never heard of that solver, but my professor specializes in integer programming. -Billy >-----Original Message----- >From: paul_gauthier@dynasys.tm.fr [mailto:paul_gauthier@dynasys.tm.fr] >Sent: Thursday, August 05, 1999 3:21 AM >To: forum@jsoftware.com >Subject: Jforum: A chanllenge... > > >I have a future doctor in Operational Research who beleives that no one >in the world can achieve better results than the CPLEX corporation >solver on integer's linear programming. > >Anyone interested to give hints on a better solution please contact me. > >Regards/Paul > >--------------------------------------------------------------- >----------------- >J Forum: for information about this list, see >http://www.jsoftware.com/forum.htm > -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Sat Aug 7 01:25:16 1999 Date: Fri, 06 Aug 1999 23:40:48 -0400 From: Brian Bambrough Subject: Re: Jforum: Viewing large arrays. User help References: <199908012326.QAA25635@mimas.island.net> Ken Ian Gordon wrote: > I used an alternative approach to supply help on buttons. As you > know it is available for menu entries, but I too could find nothing > for buttons in J. I used the F1 key instead. When pressed it > ascertained which button had the focus and displayed the appropriate > short help message in the status area as taken from a list. Using > TAB to move focus and pressing F1 then gave all the help messages > for the buttons. It also gave short help messages for anything data > entry fields, to explain the required entry. These were taken from > the meta data. > Not as good as right button, or the automatic display for menus, but > it appeared to suffice. The big advantage of your approach is that there are no extra controls in the form. In my approach I have to add a "Button help" check box. As I understand it, in your method the only way to set the focus on a button is to tab to it. You can't set the focus with a mouse click as this would invoke the action of the button. Is this right? Martin Neitzel wrote: > re your enquiry about context sensitive help: In the J Workbench, I > use the following approach to jump around between verb definitions: > If the "goto verb" function is triggered (by using Ctrl-G or > clicking a button), I make a check whether there is some text > selection in one of several text fields. If so, I extend the > partial selection to the whole word near to the left selection > border and look that up in my list of identifiers. If I can find > it, I jump to it. Otherwise (no match or no selection) I pop up a > menu. > This works quite well, in particular because ANY text output > benefits from this jump feature (unlike html-text which needs all > links to be prepared). For example, users are able to jump quickly > to a verb listed in a boxed representation. > While this is a very specific context for "jumps", and your question > for Context Sensitive Help solutions appears to be much more > general, I thought I might drop short note. I must confess that I don't grasp what you wrote, although it sounds very interesting. I don't know what a "J Workbench" is. Is it a tool for developing in J? If so, I would like to know more about it. I will be at APL99 next week. Perhaps we could talk about it then. -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Sat Aug 7 13:43:43 1999 Date: Sat, 7 Aug 1999 10:38:40 -0700 From: gordon@island.net (Ken Ian Gordon) Subject: Re: Jforum: Viewing large arrays. User help Brian Bambrough wrote >Ken Ian Gordon wrote: > >> I used an alternative approach to supply help on buttons. As you >> know it is available for menu entries, but I too could find nothing >> for buttons in J. I used the F1 key instead. When pressed it > >The big advantage of your approach is that there are no extra controls >in the form. In my approach I have to add a "Button help" check box. > >As I understand it, in your method the only way to set the focus on a >button is to tab to it. You can't set the focus with a mouse click as >this would invoke the action of the button. Is this right? Unfortunatley yes. -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Sat Aug 7 22:29:45 1999 From: Ajith Prasad Subject: Jforum: KEYED FILES & DICTIONARIES Date: Sun, 8 Aug 1999 10:24:17 +0800 A recent discussion forum resulted in a script ('jzdict')for handling of "Dictionaries" being included in the "system\classes\dict" directory. I have just noticed that J also includes a 'kfiles' script in the "system\packages\files" directory which allows data to be accessed using keywords. The functionality seems similar to that in the dictionary script. Has anyone tried both scripts and prefers one to the other? -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Sun Aug 8 04:37:39 1999 From: "Stefano Lanzavecchia" References: Subject: Re: Jforum: KEYED FILES & DICTIONARIES Date: Sun, 8 Aug 1999 10:27:32 +0200 Organization: APL Italiana charset="iso-8859-1" ----- Original Message ----- From: Ajith Prasad Sent: Sunday, August 08, 1999 4:24 AM Subject: Jforum: KEYED FILES & DICTIONARIES : A recent discussion forum resulted in a script ('jzdict')for handling of : "Dictionaries" being included in the "system\classes\dict" directory. I have : just noticed that J also includes a 'kfiles' script in the : "system\packages\files" directory which allows data to be accessed using In two words: "kfiles" is a component file system (like APL's traditional) where the components are accessed not by number but by name. "jzdict" keeps the dictionary in memory, dictionary which can be exported to file and re-imported from file if the user wants it. -- WildHeart'99 Homepage: http://www.insight.dk/stf/ -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Sun Aug 8 12:42:24 1999 From: "Anne Faust" Subject: Jforum: Press Release - Iverson Software Inc. Date: Sun, 8 Aug 1999 12:45:30 -0400 charset="iso-8859-1" Eric Iverson has been appointed Chief Technical Officer with Adaytum Software. See http://www.adaytum.com for information about Adaytum. In addition to being attracted to Adaytum's business, rapid growth, and dynamic company culture, Eric says that a primary reason for joining is Adaytum's commitment to J as a key technology in building the future for Adaytum and its clients. Adaytum's teams of J developers working on building new products is a significant step forward for J and will help ISI build and maintain J momentum. In addition, the ISI development team are planning the following upcoming releases. Later this year there will be a release that supports sparse arrays. Roger Hui will present a paper on this at the upcoming APL99 conference, see http://www.lingo.com/apl99. Also, early next year there will be a release that includes a Session Manager and Window driver for J on Unix platforms with an integrated set of libraries and utilities allowing many complete applications to move seamlessly between Windows and Unix. -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Sun Aug 8 14:38:45 1999 From: "Gordon Sutcliffe" Subject: Jforum: Use of new Plot option types: Linefit and Step Date: Thu, 5 Aug 1999 15:18:29 +0100 boundary="----=_NextPart_000_0018_01BEDF55.C5D2E6E0" This is a multi-part message in MIME format. ------=_NextPart_000_0018_01BEDF55.C5D2E6E0 charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable The following J 4.03 script (Windows 95), displays exception messages on = the linefit and step options of the plot command: =20 dsp =3D: 1!:2&2 NB. function to copy left argument to display errmsg =3D: 13!:12 =20 require 'plot trig' tp =3D: 3 : 0 NB. Test Plot for all plot types listed in User Manual t =3D. 'area bar density dot errorbar hilo line linefit fbar sbar ' t =3D. <;._2 t, 'pie point radar stick step surface wire ' =20 for_i. t do. dsp i=3D.>i NB. dislay plot option type try. ('type ', i, ';') plot 3 6$i.6 catch. dsp (i, ': fails with error message:'),: errmsg '' end. 6!:3 y. end. 0 0$'' NB. return empty result ) =20 tp 3 NB. call tp, right argument is delay between plots=20 =20 I would be glad of any advice on getting these two options to work. ------=_NextPart_000_0018_01BEDF55.C5D2E6E0 charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
The following J = 4.03 script=20 (Windows 95), displays exception messages on the linefit and step = options of the=20 plot command:
 
dsp  =3D:=20 1!:2&2      NB. function to copy left = argument to=20 display
errmsg =3D:=20 13!:12
 
require 'plot=20 trig'

 
tp =3D: 3 : = 0  NB. Test Plot=20 for all plot types listed in User Manual
 t =3D. 'area bar = density dot=20 errorbar hilo line linefit fbar sbar '
 t =3D. <;._2 t, 'pie = point=20 radar stick step surface wire '
 
 for_i. t = do.
 =20 dsp i=3D.>i         NB. = dislay plot=20 option type
  try. ('type ', i, ';') plot 3 6$i.6
  = catch. dsp=20 (i, ': fails with error message:'),: errmsg ''
  end.
  = 6!:3=20 y.
 end.
0 0$''   NB. return empty=20 result
)
 
tp 3   = NB. call tp,=20 right argument is delay between plots
 
I would be glad of any = advice on=20 getting these two options to = work.
------=_NextPart_000_0018_01BEDF55.C5D2E6E0-- -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Sun Aug 8 15:26:37 1999 Date: Sun, 08 Aug 1999 21:29:48 +0300 From: Anssi Seppala Subject: Re: Jforum: Press Release - Iverson Software Inc. In-Reply-To: <001e01bee1bd$8d0c1380$8e29fea9@t500> Good luck Eric! Would it be too much to ask some more details how Adaytum uses J? Anssi At 12:45 8.8.1999 -0400, you wrote: >Eric Iverson has been appointed Chief Technical Officer with Adaytum >Software. See http://www.adaytum.com for information about Adaytum. > >In addition to being attracted to Adaytum's business, rapid growth, and >dynamic company culture, Eric says that a primary reason for joining is >Adaytum's commitment to J as a key technology in building the future for >Adaytum and its clients. -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Sun Aug 8 19:38:40 1999 From: Ajith Prasad Subject: FW: Jforum: Press Release - Iverson Software Inc. Date: Mon, 9 Aug 1999 07:38:47 +0800 charset="windows-1252" Best Wishes to Eric. I hope his departure will not hold back the pace of development of J but rather result in J being the core technology behind major mainstream decision support software such as Adaytum's products. I am encouraged more by the reference to Adaytum's "teams of J developers" working on products rather than by Adaytum's "committment to J as a key teachnology". I understand that "Adaytum Planning", Adaytum's main product, is a budgeting application that was originally developed at IBM UK using APL but was bought out by its development team and marketed separately. Adaytum has a development team in Denmark presumably using APL. Their US corporate office is in Blooomington, Minnesota. Is that close to where Strand Software is? If Adaytum is also using J, then this would be the second company, after SAP, known to have used J in a mainstream financial application. Would be interesting to know what other companies are using J for mainstream applications. -----Original Message----- At 12:45 8.8.1999 -0400, you wrote: >Eric Iverson has been appointed Chief Technical Officer with Adaytum >Software. See http://www.adaytum.com for information about Adaytum. > >In addition to being attracted to Adaytum's business, rapid growth, and >dynamic company culture, Eric says that a primary reason for joining is >Adaytum's commitment to J as a key technology in building the future for >Adaytum and its clients. ------------------------------------------------------------------------ -------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Sun Aug 8 21:01:13 1999 Date: Mon, 09 Aug 1999 02:00:37 +0100 From: Joachim Hoffmann Subject: Re: FW: Jforum: Press Release - Iverson Software Inc. In-Reply-To: Please also note the Adaytum UK Development Group in York. JoHo At 07:38 AM 8/9/1999 +0800, Ajith Prasad wrote: >... >I understand that "Adaytum Planning", Adaytum's main product, is a budgeting >application that was originally developed at IBM UK using APL but was bought >out by its development team and marketed separately. Adaytum has a >development team in Denmark presumably using APL. Their US corporate office >is in Blooomington, Minnesota. Is that close to where Strand Software is? >... -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Sun Aug 8 21:32:55 1999 Date: Sun, 08 Aug 1999 21:32:45 -0400 From: David Ness Subject: Re: FW: Jforum: Press Release - Iverson Software Inc. References: Ajith Prasad wrote: > > .... Adaytum has a > development team in Denmark presumably using APL. Their US corporate office > is in Blooomington, Minnesota. Is that close to where Strand Software is? > > Shorewood is about 15 mi from Bloomington. That's about the distance from the southern tip of Manhattan up to the northern Bronx... -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Mon Aug 9 08:28:59 1999 From: "Chris Burke" Subject: Jforum: J403c Date: Mon, 9 Aug 1999 08:27:48 -0400 charset="iso-8859-1" J4.03c for Windows 9x/NT is now available at the web site. This has a few minor fixes and breaks the regular expression part of J out from the j.dll into file jregexp.dll. This makes the j.dll a bit smaller and the jregexp.dll is required only if regular expressions are used. -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Mon Aug 9 09:57:58 1999 From: "Chris Burke" References: <000201bee1cd$065f3280$e278063e@default> Subject: Re: Jforum: Use of new Plot option types: Linefit and Step Date: Mon, 9 Aug 1999 10:00:15 -0400 charset="iso-8859-1" Gordon Sorry, the plot documentation incorrectly lists linefit and step as plot types, but these are not supported. Neither should be difficult to add, so I suspect that they were originally listed with the plot types with the intention of adding them before some release. The "step" option has never worked. The "linefit" option simply plots a 2 element list as the original data, plus fitted lines, as in the example below, though I do not think this is a good definition for linefit. Chris require 'plot numeric stats trig' dat=: 0.1 0.4 0.5 0.7 0.7 0.9 ,: 0.61 0.92 0.99 1.52 1.47 2.03 NB. least-squares fit coefficients for polynomials of order 0-4 fits=: (i.5) lsfit"0 _ dat NB. dat compared with fitted dat: dat,fitdat=: fits p."1 {.dat NB. sum of squared deviations from y values: +/"1 *: fitdat-"1 {:dat NB. plot original points, and polys of order 1, 2 and 4: x=: steps 0 1 50 y=: 1 2 4{fits p."1 x 'linefit' plot ( Sent: Thursday, August 05, 1999 10:18 AM Subject: Jforum: Use of new Plot option types: Linefit and Step The following J 4.03 script (Windows 95), displays exception messages on the linefit and step options of the plot command: dsp =: 1!:2&2 NB. function to copy left argument to display errmsg =: 13!:12 require 'plot trig' tp =: 3 : 0 NB. Test Plot for all plot types listed in User Manual t =. 'area bar density dot errorbar hilo line linefit fbar sbar ' t =. <;._2 t, 'pie point radar stick step surface wire ' for_i. t do. dsp i=.>i NB. dislay plot option type try. ('type ', i, ';') plot 3 6$i.6 catch. dsp (i, ': fails with error message:'),: errmsg '' end. 6!:3 y. end. 0 0$'' NB. return empty result ) tp 3 NB. call tp, right argument is delay between plots I would be glad of any advice on getting these two options to work. -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Mon Aug 9 11:21:28 1999 From: "Oleg Kobchenko" Subject: Jforum: New on Oleg's J Page Date: Mon, 9 Aug 1999 18:22:45 +0300 charset="koi8-r" New on Oleg's J Page ~~~~~~~~~~~~~~~~~~~~ regexs Regular expressions extended for Perl-like substitution console Windows Console functionality -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Mon Aug 9 11:45:34 1999 Date: Mon, 09 Aug 1999 18:47:04 +0300 From: Oleg Kobchenko Subject: Jforum: New on Oleg's J Page. Service Pack 1 The URL is http://members.xoom.com/olegyk/jpage.html -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Mon Aug 9 12:03:59 1999 From: "Donald Pittenger" Subject: Jforum: who will mind the store? Date: Mon, 9 Aug 1999 09:03:09 -0700 boundary="----=_NextPart_000_0022_01BEE246.00A7D620" This is a multi-part message in MIME format. ------=_NextPart_000_0022_01BEE246.00A7D620 charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable I don't always enjoy posing the blunt question, but as of the time I'm = drafting this, no one else has--so here goes. =20 With Eric leaving JSoftware, who will be the day-to-day head of the = company? What can users expect for the longer-term future? Who will we = be dealing with, and for what aspects of J? =20 I ask this because I've been moving most of my firm's systems to J from = APL, and Eric's announcement raised an Adrian Smith-like level of = paranoia about being left high & dry. =20 Granted, the 'creative phase' of J language development is probably = pretty much over anyway; the need is to keep up with general = computational environmental change (the Web, Linux, etc), and this might = not require quite as much work as getting J to Windows. =20 Therefore, I hope the J folk will take time to spell out JSoftware's new = organization and the responsibilities of various people involved = full/part time in maintaining & enhancing J. From a public-relations = standpoint, the annouuncement about Eric raises as many questions as = answers. Lacking anything solid about ISA, I for one, might have to = reconsider my shift away from APL. My livelihood could be at stake, so = my attitude necessarily requires sang-froid. =20 Don Pittenger The Demographics Laboratory ------=_NextPart_000_0022_01BEE246.00A7D620 charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
I don't always enjoy posing the = blunt question,=20 but as of the time I'm drafting this, no one else has--so here=20 goes.
 
With Eric leaving JSoftware, who = will be the=20 day-to-day head of the company?  What can users expect for the = longer-term=20 future?  Who will we be dealing with, and for what aspects of=20 J?
 
I ask this because I've been moving = most of my=20 firm's systems to J from APL, and Eric's announcement raised an Adrian=20 Smith-like level of paranoia about being left high & = dry.
 
Granted, the 'creative phase' of J = language=20 development is probably pretty much over anyway; the need is to keep up = with=20 general computational environmental change (the Web, Linux, etc), and = this might=20 not require quite as much work as getting J to Windows.
 
Therefore, I hope the J folk will = take time to=20 spell out JSoftware's new organization and the responsibilities of = various=20 people involved full/part time in maintaining & enhancing J.  = >From a=20 public-relations standpoint, the annouuncement about Eric raises as many = questions as answers.  Lacking anything solid about ISA, I for one, = might=20 have to reconsider my shift away from APL.  My livelihood could be = at=20 stake, so my attitude necessarily requires sang-froid.
 
Don Pittenger
The Demographics=20 Laboratory
------=_NextPart_000_0022_01BEE246.00A7D620-- -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Mon Aug 9 13:32:53 1999 From: "Stefano Lanzavecchia" References: <002501bee280$ada8ba80$67fe02c7@fwh71> Subject: Re: Jforum: who will mind the store? Date: Mon, 9 Aug 1999 19:29:27 +0200 Organization: APL Italiana charset="iso-8859-1" ====== With Eric leaving JSoftware, who will be the day-to-day head of the company? What can users expect for the longer-term future? Who will we be dealing with, and for what aspects of J? ====== One of the reasons why Eric was taken on board here in Adaytum Software, was to tie our relationship with JSoftware. So, if, from a certain point of view, Eric himself is going to have less time in the future to deal directly with the development of J (Adaytum is a large organization and his CTO does not have plenty of spare time), it is also true that this move should guarantee a safer ground for the prosperity of J. The need to keep up with what you call "environmental changes" is also one of our needs, and J can only benefit from this. The announced native support in J for sparse arrays, support, which is going to be showed in Scranton in a few days time, is an example of what I am saying. Also don't forget that Kirk Iverson, Roger Hui and Chris Burke have not left JSoftware... By the way, these are my opinions, I don't speak as an Adaytum official, so please don't quote me on this as if mine was an official statement. -- WildHeart'99 - Senior developer in Adaytum Software Homepage: http://www.insight.dk/stf/ -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Mon Aug 9 15:35:45 1999 Date: Mon, 09 Aug 1999 21:43:29 +0300 From: Anssi Seppala Subject: Re: Jforum: who will mind the store? In-Reply-To: <011001bee28c$bf219320$57a4b9c2@apl.it> References: <002501bee280$ada8ba80$67fe02c7@fwh71> That was something I wanted to hear! For a while I was wondering what is the idea behind sparse matrixes. I have already heard that sparse matrix is a very basic in Adaytum applications. Thanks! Anssi At 19:29 9.8.1999 +0200, you wrote: >of our needs, and J can only benefit from this. The announced native support >in J for sparse arrays, support, which is going to be showed in Scranton in >a few days time, is an example of what I am saying. -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Mon Aug 9 18:03:17 1999 From: "Gordon Sutcliffe" Subject: Jforum: Use of Plot option types: Linefit and Step Date: Mon, 9 Aug 1999 22:43:44 +0100 charset="iso-8859-1" Chris Thank you for the explanation of plot types linefit and step, and also for the interesting example re linefit. Gordon -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Mon Aug 9 22:16:41 1999 Date: Mon, 09 Aug 1999 22:16:13 -0400 From: David Ness Subject: Jforum: DBrowse.IJS problem? I seem to encounter a problem with Oleg's DBrowse.IJS when in encounters a directory that is completely empty. Is it me, or is it a (tiny) bug? -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Tue Aug 10 00:52:06 1999 Date: Tue, 10 Aug 1999 06:49:02 +0200 From: "david alis" Subject: Re: Jforum: who will mind the sparse store? References: <002501bee280$ada8ba80$67fe02c7@fwh71> <4.2.0.58.19990809213748.00bb45c0@pop.kolumbus.fi> WIll Roger be publishing any design notes on his implementation of sparse arrays? Anssi Seppala wrote: > That was something I wanted to hear! For a while I was wondering what is > the idea behind sparse matrixes. I have already heard that sparse matrix is > a very basic in Adaytum applications. Thanks! > > Anssi > > At 19:29 9.8.1999 +0200, you wrote: > >of our needs, and J can only benefit from this. The announced native support > >in J for sparse arrays, support, which is going to be showed in Scranton in > >a few days time, is an example of what I am saying. > > -------------------------------------------------------------------------------- > J Forum: for information about this list, see http://www.jsoftware.com/forum.htm -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Tue Aug 10 06:10:05 1999 From: M.Day@fscii.maff.gov.uk Tue, 10 Aug 1999 10:59:23 +0100 Tue, 10 Aug 1999 10:55:38 +0100 Tue, 10 Aug 1999 10:58:25 +0100 Date: Tue, 10 Aug 1999 10:58:25 +0100 Content-Identifier: m1210810105757aa Alternate-Recipient: Allowed In-Reply-To: <4.2.0.58.19990717192951.0094bde0@pop.kolumbus.fi> Subject: Re: Jforum: J Plot key with linepatterns Anssi wrote on July 17 >Can someone tell if it is possible to have linepatterns in plot key? I try >to make plots for b/w presentation and colour printing is not possible. > >Trying this > style=:'type line;color black;penstyle 0 1 2 3 4;key as1 as2 as3 as4' > style plot i.3 4 I've just realised I can help on this. Apologies for a late response and/or if this has already been dealt with. I suggested some small modifications to the verbs drawkey getkeysize and getkeybox which were incorporated in system/classes/jzplot.ijs, but the effects were not documented. (I've removed tabs here - hope it displays ok) Options Type Default Description Old usage (as documented): keystyle n 0 2 0=horizontal 1=vertical style 0-3 positions from top left New usage: keystyle n 0 2 0=horizontal 1=vertical style rectangle 2=horizontal 3=vertical style line 0-3 positions from top left So in your case you need (for example): style=:'type line;color black;penstyle 0 1 2 3 4;key as1 as2 as3 as4' style=. style, ';keystyle 2 2' style plot i.3 4 Mike 10 Aug 99 -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Tue Aug 10 22:04:07 1999 Date: Tue, 10 Aug 1999 22:03:02 -0400 From: David Ness Subject: Jforum: Debugging `wd'? I must admit to being at a rather complete loss about how to approach `debugging' wd errors. I tend to only get `domain error: wd' without any further help. This doesn't give me much of a clue about what is wrong, so my debugging is reduced to random trial and error, and with J's rich instruction set, every `trial' seems to result in `error' that is no more descriptive. Clearly I am missing something about how to debug `wd' commands. Is there something I should read? -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Tue Aug 10 23:10:52 1999 Date: Tue, 10 Aug 1999 23:20:40 -0400 From: Eric Iverson Organization: Iverson Software Inc. Subject: Re: Jforum: Debugging `wd'? References: <37B0D9D6.4C849A21@Home.Com> David Ness wrote: > I must admit to being at a rather complete loss about how to approach > `debugging' wd errors. wd'qer' NB. info on last error (integer is index of error) -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Tue Aug 10 23:17:58 1999 Date: Tue, 10 Aug 1999 23:17:50 -0400 From: David Ness Subject: Re: Jforum: Debugging `wd'? References: <37B0D9D6.4C849A21@Home.Com> <37B0EC07.17F0@interlog.com> Maybe I'm expecting (way) too much, but, for example in my case typing wd`qer' as you suggested resulted in `bad parameter : 0' which is not a whole lot more informative. I still don't have much of a clue as to what might be wrong. I feel like I am back debugging in octal. Eric Iverson wrote: > > David Ness wrote: > > I must admit to being at a rather complete loss about how to approach > > `debugging' wd errors. > > wd'qer' NB. info on last error (integer is index of error) > > -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Wed Aug 11 05:03:50 1999 Date: Wed, 11 Aug 1999 05:01:38 -0400 From: Fraser Jackson Subject: Jforum: A chanllenge... charset=ISO-8859-1 Content-Disposition: inline Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by plus.interlog.com id FAA14773 This is not the forum for this question on integer linear program solvers. The CPlex solver is very widely used by companies which market interfaces to linear programming codes and clearly has some excellent code for integer programming problems. However it is frequently the case that integer programming problems have some special structure that can be exploited, and there are suppliers who claim that for particular classes of problems their code performs better. However it is not a J problem and should not be explored further here. -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Wed Aug 11 08:02:18 1999 From: "J Tibollo" Subject: Jforum: J Windows Interface Date: Wed, 11 Aug 1999 08:00:07 -0400 boundary="----=_NextPart_000_000C_01BEE3CF.87152D40" This is a multi-part message in MIME format. ------=_NextPart_000_000C_01BEE3CF.87152D40 charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable As near as I can tell, child controls in a form can respond to only a = subset of events that windows generates. Is this true? I have not been = able to find documentation that lays out for each control the events = generated for each control. How, for example, would you catch a form = resize event? Or a lost focus event? Or a double click vs a single = click in a listbox? In most other Windowing environments discussion about windows elements = consists of the properties, events, and methods supported by the = control. Would it be possible to do the same with J? It seems that even some simple things are difficult. If I wish to read = the contents of an edit control (so that I can monitor how full it is = getting and delete some data at the beginning to allow for appending = data to the end) how would I do this? It seems that wd 'q' returns the = contents of all objects, but this is available only following some = event. What If I am writing to an edit control on an ongoing basis, and = every once in awhile want to know how much data is in the edit control. = Without an intervening user event, I can't figure out how I could query = the control. (Sure I can keep a shadow copy of the data that I am = writing and obviate the need to read the edit control, or perhaps use = the Windows API to access the edit - but how kludgey - and how = unnecessary). I think J could benefit a bit from some kind of structure in its windows = interface, otherwise, I suspect that some time down the road, the = interface is going to seem more and more kludgey (hap-hazard, = unorganized, arbitrary, etc.). My two cents worth anyway. In the meantime, like everyone else, I am = trying to figure out what is and is not possible so that I can use the = Windows interface too. Regards, Joe Tibollo ------=_NextPart_000_000C_01BEE3CF.87152D40 charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
As near as I can tell, child controls = in a form=20 can respond to only a subset of events that windows generates.  Is = this=20 true?  I have not been able to find documentation that lays out for = each=20 control the events generated for each control.  How, for example, = would you=20 catch a form resize event?  Or a lost focus event?  Or a = double click=20 vs a single click in a listbox?
 
In most other Windowing environments = discussion=20 about windows elements consists of the properties, events, and methods = supported=20 by the control.  Would it be possible to do the same with=20 J?
 
It seems that even some simple things = are=20 difficult.  If I wish to read the contents of an edit control (so = that I=20 can monitor how full it is getting and delete some data at the beginning = to=20 allow for appending data to the end) how would I do this?  It seems = that wd=20 'q' returns the contents of all objects, but this is available only = following=20 some event.  What If I am writing to an edit control on an ongoing = basis,=20 and every once in awhile want to know how much data is in the edit=20 control.  Without an intervening user event, I can't figure out how = I could=20 query the control.  (Sure I can keep a shadow copy of the data that = I am=20 writing and obviate the need to read the edit control, or perhaps use = the=20 Windows API to access the edit - but how kludgey - and how=20 unnecessary).
 
I think J could benefit a bit from some = kind of=20 structure in its windows interface, otherwise, I suspect that some time = down the=20 road, the interface is going to seem more and more kludgey (hap-hazard,=20 unorganized, arbitrary, etc.).
 
My two cents worth anyway.  In the = meantime,=20 like everyone else, I am trying to figure out what is and is not = possible so=20 that I can use the Windows interface too.
 
Regards,
Joe = Tibollo
------=_NextPart_000_000C_01BEE3CF.87152D40-- -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Wed Aug 11 08:33:44 1999 From: "Chris Burke" References: <37B0D9D6.4C849A21@Home.Com> <37B0EC07.17F0@interlog.com> <37B0EB5E.528F7E63@Home.Com> Subject: Re: Jforum: Debugging `wd'? Date: Wed, 11 Aug 1999 08:32:48 -0400 charset="iso-8859-1" The argument to wd can have several commands, so the number lets you know which command is at fault (it actually gives the position in the string of the command). In this case, it is the first command. The 'bad parameter' means just that. What is the wd command in error? ----- Original Message ----- From: David Ness Sent: Tuesday, August 10, 1999 11:17 PM Subject: Re: Jforum: Debugging `wd'? > Maybe I'm expecting (way) too much, but, for example in my case > typing wd`qer' as you suggested > resulted in `bad parameter : 0' which is not a whole lot more > informative. I still don't have much of a clue as to what might > be wrong. > > I feel like I am back debugging in octal. > > > Eric Iverson wrote: > > > > David Ness wrote: > > > I must admit to being at a rather complete loss about how to approach > > > `debugging' wd errors. > > > > wd'qer' NB. info on last error (integer is index of error) > > > > > > -------------------------------------------------------------------------- ------ > J Forum: for information about this list, see http://www.jsoftware.com/forum.htm > -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Wed Aug 11 08:57:11 1999 From: "Oleg Kobchenko" References: <37AF8B6D.C3A2B0A5@Home.Com> Subject: Re: Jforum: DBrowse.IJS problem? Date: Wed, 11 Aug 1999 15:57:49 +0300 charset="iso-8859-1" Thank you for pointing out this "tiny" bug. The definition of lstdirs=: (#~ 'd'&=@(4&{)@(>@:{:&.|:))@:(1!:0) should be changed to lstdirs=: (#~ 'd'&=@(4&{)@(>@:{:&.|:))^:(*@#)@:(1!:0) or as lstdirs=: (#~ 'd'&=@(4&{)@(>@:{:&.|:))^:(*@(*/@$))@:(1!:0) (which had already been done in my working copy of the lib) ----- Original Message ----- From: David Ness Sent: Tuesday, August 10, 1999 5:16 Subject: Jforum: DBrowse.IJS problem? > I seem to encounter a problem with Oleg's DBrowse.IJS when in > encounters a directory that is completely empty. Is it me, or is it > a (tiny) bug? -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Wed Aug 11 09:28:52 1999 From: "Chris Burke" References: <000f01bee3f1$0edf6ee0$81d23dcf@jat> Subject: Re: Jforum: J Windows Interface Date: Wed, 11 Aug 1999 09:30:25 -0400 charset="iso-8859-1" ----- Original Message ----- From: J Tibollo Sent: Wednesday, August 11, 1999 8:00 AM Subject: Jforum: J Windows Interface >As near as I can tell, child controls in a form can respond >to only a subset of events that windows generates. Is this >true? I have not been able to find documentation that lays >out for each control the events generated for each control. >How, for example, would you catch a form resize event? Or a >lost focus event? Or a double click vs a single click in a >listbox? Most child controls have only one event they produce. When coding the form, you get this when you Ctrl-click on a control. For a full list of the events for a form, use the Design Window|Code dialog. There is no form resize event. We will likely add one, in the meantime, the Size rules parameters take care of most resize requirements. There is no lost focus event, again we probably should have one. When you highlight an item in a listbox, you get a select event. When you double-click or press enter, you get a button event. >In most other Windowing environments discussion about >windows elements consists of the properties, events, and >methods supported by the control. Would it be possible to do >the same with J? It would. The original docs were written some time ago, when the facilities were even simpler. But the interface is still very straightforward, and easy to experiment with. For example, create a form, and define the parent handler to be wdqshow, e.g. myform_handler=: wdqshow Now, all events that occur for the form are displayed. The "events" demo has a similar display. >It seems that even some simple things are difficult. If I >wish to read the contents of an edit control (so that I can >monitor how full it is getting and delete some data at the >beginning to allow for appending data to the end) how would >I do this? It seems that wd 'q' returns the contents of all Use wd 'qd'. "Query" wd commands all begin with a "q". >I think J could benefit a bit from some kind of structure in >its windows interface, otherwise, I suspect that some time >down the road, the interface is going to seem more and more >kludgey (hap-hazard, unorganized, arbitrary, etc.). I dont agree that that it is kludgey or hap-hazard, etc. The wd commands map closely to the underlying WinAPI calls. The event handling mechanism is extremely simple and powerful. The form editor is ideal for what we support. Overall, I find wd very easy to use, much easier for example, than VB, VBA or APL (APL+WIN and APLW). Note that while we will likely continue to improve wd, we dont want to spend time catching up with Microsoft. Here is a relevant comment that Eric recently sent to the forum: >The J UI is adequate for applications with modest UI >requirements. It is not adequate for applications that >require the latest UI sizzle. We've put a lot of effort into >providing a 90% GUI solution. The cost of providing the >remaining 10% remains too high. The problem is that VB won't >stand still. We could put in the significant effort to catch >up to where they are today. But by then they'd have moved >and we'd still be behind. I don't like being in a race I am >guaranteed to lose. > >If your applications really require the latest GUI, then you >should use the best GUI tools available (C++, VB, OCX,...). >But you probably don't want to use them to do serious data >processing (DP). If your application needs sizzling GUI and >serious DP, it is fortunate indeed that you can use the J >COM server to get the best of both worlds. Learning to >program the GUI in tools like VB is actually fairly easy and >J and APL programmers should not be so reluctant to take >that bull by the horns. And developing and debugging hybrid >applications that use VB (or whatever) for the GUI and J for >DP works very well. In fact, I would argue that for big >applications it imposes a rigourous line between GUI and DP >that is to the benefit of all, and allows the use of the >best tool for each part of the job. -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Wed Aug 11 09:44:04 1999 From: "Nichols, Peter" Subject: RE: Jforum: J Windows Interface Date: Wed, 11 Aug 1999 08:39:10 -0500 charset="iso-8859-1" I agree with you Joe. We must have access to all the events so we may make apps more responsive and intuitive. Gaining & losing focus is a good example. I think that J would benefit greatly from a more complete windows interface. We are competing against apps written in VB or Delphi that are very slick. Admittedly, we shine if there is any real processing involved. But many times we want to create simple windows that don't require in depth code, i.e. quick and dirty throwaways. I'm working on an app now that is windows intensive, but that doesn't need any great computational power. Yes, we can code front ends in another language, but I think we write in J because we love J. I also know that we don't have the manpower to completely keep up with the latest and greatest windows interface tools. Perhaps there is a middle ground though. What about having a J script that would take in a scripting language and put out VB or C++ or Delphi code for front ends. If it handled the moving of data from the windows interface into J variables or files, it could be fairly transparent. We would gain access to all the windows bells and whistles without feeling like we're having to choose between J and another language. Best Regards, Pete Nichols -----Original Message----- From: J Tibollo [SMTP:jtibollo@backassociates.com] Sent: Wednesday, August 11, 1999 7:00 AM To: J Forum Subject: Jforum: J Windows Interface As near as I can tell, child controls in a form can respond to only a subset of events that windows generates. Is this true? ... -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Wed Aug 11 10:25:48 1999 From: "Anne Faust" Subject: RE: Jforum: who will mind the store? Date: Wed, 11 Aug 1999 09:26:47 -0500 boundary="----=_NextPart_000_000D_01BEE3DB.A3205260" Importance: Normal In-Reply-To: <002501bee280$ada8ba80$67fe02c7@fwh71> This is a multi-part message in MIME format. ------=_NextPart_000_000D_01BEE3DB.A3205260 charset="iso-8859-1" Don, We see this as overall good news for J users. There will be new people brought on board for parts of J development. However, the rest of the development team remains in place and committed to continual development and support of the J product. There is momentum building with J and we intend to keep pushing ahead. Anne -----Original Message----- From: owner-jsoftware@lists.interlog.com [mailto:owner-jsoftware@lists.interlog.com]On Behalf Of Donald Pittenger Sent: Monday, August 09, 1999 11:03 AM To: forum@jsoftware.com Subject: Jforum: who will mind the store? I don't always enjoy posing the blunt question, but as of the time I'm drafting this, no one else has--so here goes. With Eric leaving JSoftware, who will be the day-to-day head of the company? What can users expect for the longer-term future? Who will we be dealing with, and for what aspects of J? I ask this because I've been moving most of my firm's systems to J from APL, and Eric's announcement raised an Adrian Smith-like level of paranoia about being left high & dry. Granted, the 'creative phase' of J language development is probably pretty much over anyway; the need is to keep up with general computational environmental change (the Web, Linux, etc), and this might not require quite as much work as getting J to Windows. Therefore, I hope the J folk will take time to spell out JSoftware's new organization and the responsibilities of various people involved full/part time in maintaining & enhancing J. >From a public-relations standpoint, the annouuncement about Eric raises as many questions as answers. Lacking anything solid about ISA, I for one, might have to reconsider my shift away from APL. My livelihood could be at stake, so my attitude necessarily requires sang-froid. Don Pittenger The Demographics Laboratory ------=_NextPart_000_000D_01BEE3DB.A3205260 charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable

Don,

We see this as overall good news for J users. There will be new = people=20 brought on board for parts of J development. However, the rest of the=20 development team remains in place and committed to continual development = and=20 support of the J product. There is momentum building with J and we = intend to=20 keep pushing ahead.

Anne

-----Original Message-----
From:=20 owner-jsoftware@lists.interlog.com=20 [mailto:owner-jsoftware@lists.interlog.com]On Behalf Of = Donald=20 Pittenger
Sent: Monday, August 09, 1999 11:03 = AM
To:=20 forum@jsoftware.com
Subject: Jforum: who will mind the=20 store?

I don't always enjoy posing the = blunt=20 question, but as of the time I'm drafting this, no one else has--so = here=20 goes.
 
With Eric leaving JSoftware, who = will be the=20 day-to-day head of the company?  What can users expect for the=20 longer-term future?  Who will we be dealing with, and for what = aspects=20 of J?
 
I ask this because I've been = moving most of=20 my firm's systems to J from APL, and Eric's announcement raised an = Adrian=20 Smith-like level of paranoia about being left high & = dry.
 
Granted, the 'creative phase' of = J language=20 development is probably pretty much over anyway; the need is to keep = up with=20 general computational environmental change (the Web, Linux, etc), = and this=20 might not require quite as much work as getting J to = Windows.
 
Therefore, I hope the J folk = will take time=20 to spell out JSoftware's new organization and the responsibilities = of=20 various people involved full/part time in maintaining & = enhancing=20 J.  >From a public-relations standpoint, the annouuncement = about=20 Eric raises as many questions as answers.  Lacking anything = solid about=20 ISA, I for one, might have to reconsider my shift away from = APL.  My=20 livelihood could be at stake, so my attitude necessarily requires=20 sang-froid.
 
Don Pittenger
The Demographics=20 Laboratory
------=_NextPart_000_000D_01BEE3DB.A3205260-- -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Wed Aug 11 11:07:59 1999 From: Andrew Nikitin Subject: RE: Jforum: J Windows Interface Date: Wed, 11 Aug 1999 10:07:24 -0500 > -----Original Message----- > From: Nichols, Peter [SMTP:pnichols@sprinc.com] > Sent: Wednesday, August 11, 1999 08:39 > > apps more responsive and intuitive. Gaining & losing focus is a good > example. The funny thing with J gui is that you have more access on foreign components (activex) rather than on native ones (edits, buttons etc.). For example if you want to process mousemove event for button -- you can use forms.commandbutton activex control that can send mouse events to its host, but you can do nothing if you are using standard button. I still do not know any component that sends gotfocus and lostfocus events, though. > Admittedly, we shine if there is any real processing involved. But > many > times we want to create simple windows that don't require in depth > code, > i.e. quick and dirty throwaways. K has very interesting 'show' facility that allows really quick and really dirty throwaways to be created. Unfortunately J follows another gui developement style (VB style) where it always be in the tail of mainstreamers. > What about having a J script that would take in a scripting > language and put out VB or C++ or Delphi code for front ends. You can do this right now. Actually you can even do this in 2 different ways. 1. Create forms in VB, delphi, etc and use J as computational engine (Eric Iverson said more on this approach). 2. Create forms with sophisticated behaviour in VB, delphi etc and package them into .ocx as controls. Then insert these controls into convenient J form. -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Wed Aug 11 11:46:45 1999 From: Subject: RE: Jforum: A chanllenge... Date: Wed, 11 Aug 1999 17:46:33 +0200 Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by plus.interlog.com id LAA08589 >-----Message d'origine----- >De: Fraser Jackson [SMTP:Fraser_Jackson@compuserve.com] >Date: mercredi 11 ao�t 1999 11:02 >�: INTERNET:forum@jsoftware.com >Objet: Jforum: A chanllenge... > >This is not the forum for this question on integer linear program solvers. > > >The CPlex solver is very widely used by companies which market interfaces >to linear programming codes and clearly has some excellent code for integer >programming problems. However it is frequently the case that integer >programming problems have some special structure that can be exploited, and >there are suppliers who claim that for particular classes of problems their >code performs better. > >However it is not a J problem and should not be explored further here. > > >[ Paul GAUTHIER] >I am new to linear programming and trying to do it in J. > >Maybe I should not talk about linear programming at all because it is not a J >problem but somehow I have my neophytes doubts about linear programming not >being a J problem... > >Does anybody else in the J forum feel that linear programming is not a J >subject ? If so, can someone explain in layman terms why it should not even >be discussed ? Or better come up with a formal proof that J should never be >used for linear programming ? > >Regards/Paul -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Wed Aug 11 11:59:17 1999 From: "J Tibollo" References: <000f01bee3f1$0edf6ee0$81d23dcf@jat> <009c01bee3fe$07bbd060$eb31fea9@t500> Subject: Re: Jforum: J Windows Interface Date: Wed, 11 Aug 1999 12:02:03 -0400 charset="iso-8859-1" ---- Original Message ----- From: Chris Burke Sent: Wednesday, August 11, 1999 9:30 AM Subject: Re: Jforum: J Windows Interface > >In most other Windowing environments discussion about > >windows elements consists of the properties, events, and > >methods supported by the control. Would it be possible to do > >the same with J? > > It would. The original docs were written some time ago, when > the facilities were even simpler. > > But the interface is still very straightforward, and easy to > experiment with. For example, create a form, and define the > parent handler to be wdqshow, e.g. > > myform_handler=: wdqshow > > Now, all events that occur for the form are displayed. The > "events" demo has a similar display. I appreciate that one can experiment with some dummy code and simply note what events are generated. But, why not document this? Is it simply just the time and effort involved? If so, then okay, I accept this - perhaps documentation of this sort is way down on the list of priorities. Developers don't particularly like translating their work to words... > >I think J could benefit a bit from some kind of structure in > >its windows interface, otherwise, I suspect that some time > >down the road, the interface is going to seem more and more > >kludgey (hap-hazard, unorganized, arbitrary, etc.). > > I dont agree that that it is kludgey or hap-hazard, etc. The > wd commands map closely to the underlying WinAPI > calls. The event handling mechanism is extremely simple and > powerful. The form editor is ideal for what we support. > > Overall, I find wd very easy to use, much easier for > example, than VB, VBA or APL (APL+WIN and APLW). Fair enough. I suppose we can agree to disagree. The Dyalog Windows interface, for example, is like a Cadillac. I suppose that putting in the effort to generate an interface similar to that is just too prohibitive. Just looking at a simple Dyalog APL button object I note that it has 39 properties, 17 events, and 2 methods. Their edit object has 46 properties, 20 events, and 2 methods. And so on. Many controls have more than just the simple click event - how about drag and drop, etc.? > Note that while we will likely continue to improve wd, we > dont want to spend time catching up with Microsoft. Here is > a relevant comment that Eric recently sent to the forum: > > >The J UI is adequate for applications with modest UI > >requirements. It is not adequate for applications that > >require the latest UI sizzle. We've put a lot of effort into > >providing a 90% GUI solution. The cost of providing the > >remaining 10% remains too high. The problem is that VB won't > >stand still. We could put in the significant effort to catch > >up to where they are today. But by then they'd have moved > >and we'd still be behind. I don't like being in a race I am > >guaranteed to lose. and so on. I can't dispute the thinking here. The issue of how much of the Windows Interface to implement in J is not a technical issue - its one of philosophy. Anyone interested in J has nothing but the highest regard for the programming language. Its power and flexibility is unparallelled. Full Stop. However, the usefullness of a language can sometimes ( not always, but sometimes) depend on the interfaces built into the language. Over the years, I have gotten the same answer from vendors of APL - the language is the important thing, the interfaces are secondary. Fair enough. This really can't be debated. Its too much like arguing over religion or politics. I remember how long it has taken to get colour support added to common Windows controls. The answer for not adding colour support was the same then - we can't compete with Microsoft. Okay. But I don't see how adding colour competes with Microsoft. I suppose part of the problem is the definition of 90% implementation. I suppose if what you need is implemented that 90% looks pretty good. If you need something and it is missing, obviously, 90% is small comfort. But, I can't complain. As a J user I have only one vote to cast. That happens when I buy the product. Since I already have done so, and continue to do so, I suppose my vote isn't much good in influencing future directions. Thanks for for reply and your examples. Regards, Joe -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Wed Aug 11 12:11:48 1999 From: Andrew Nikitin Subject: RE: Jforum: J Windows Interface Date: Wed, 11 Aug 1999 11:10:57 -0500 > -----Original Message----- > From: J Tibollo [SMTP:jtibollo@backassociates.com] > Sent: Wednesday, August 11, 1999 11:02 > Just looking at a simple Dyalog APL > button object I note that it has 39 properties, 17 events, and > 2 methods. Their edit object has 46 properties, 20 events, and > 2 methods. And so on. ??? Did I understand you in a correct way -- are you claiming this mess of events, properties and methods for the simplest possible control as an advantage? Are you really sure you want to see some days a 794 page book titled "Button control for advanced J users -- tips, tricks and secrets"? -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Wed Aug 11 12:55:54 1999 From: "Stefano Lanzavecchia" References: Subject: Re: Jforum: J Windows Interface Date: Wed, 11 Aug 1999 18:32:21 +0200 Organization: APL Italiana charset="iso-8859-1" ----- Original Message ----- From: Andrew Nikitin Subject: RE: Jforum: J Windows Interface : > Just looking at a simple Dyalog APL : > button object I note that it has 39 properties, 17 events, and : > 2 methods. Their edit object has 46 properties, 20 events, and : > 2 methods. And so on. : ??? : Did I understand you in a correct way -- are you claiming this mess of : events, properties and methods for the simplest possible control as an : advantage? Only experience can tell that that mess is actually required if anybody is to write a serious application with a fine control on the GUI. -- WildHeart'99 Homepage: http://www.insight.dk/stf/ -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Wed Aug 11 13:16:13 1999 From: "Donald Pittenger" Subject: Re: Jforum: who will mind the store? Date: Wed, 11 Aug 1999 10:10:40 -0700 boundary="----=_NextPart_000_000E_01BEE3E1.C3F80270" This is a multi-part message in MIME format. ------=_NextPart_000_000E_01BEE3E1.C3F80270 charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Please consider this as constructive criticism: From a PR perspective, = if potentially disruptive or unsettling news is presented, extra care = should be taken to give recipients reasonable reassurance that things = will still be okay by spelling out what can be expected in the near = future. In other words, Microsoft can't get away with announcing 'Bill Gates = stepped down as chairman to pursue his long-held dream of forming = virtual-campground website: assured stockholders things will be just = fine', and leaving it at that. There would be all sorts of info about = what Steve Ballmer & all the supporting cast will be doing to fill = Bill's shoes. Sure, MS stock would still probably tank, but at least = the folks at Waggoner-Edstrom would have done the best they could under = the circumstances. So, when one of the key ISI players leaves, there is a similar need to = reassure the J community by offering concrete info on how those shoes = will be filled. Otherwise, people will start to wonder. In my case, I = have an economic stake. If ISI goes south, it is not fatal, but would = mean several months work bringing an old APL system into line with the = new J system I'm completing. Nothing immediate, but I would have to do = something within 2 years, and that 'something' would be a diversion from = doing more productive stuff. So I (and others) need to be reassured = that ISI is taking steps to prevent itself from 'going south'. That = should have been part of the original announcement: the initial = reassurances were oblique at best. I strongly suggest that ISI or Strand put something up on the Forum that = is concrete, detailed, and thereby lets the J community (including folks = like me who are still are not 100% committed to J) what (in general) can = be expected to happen, and who will be doing it. Maybe do it this = coming weekend after the dust settles from APL99. Don Pittenger -----Original Message----- From: Anne Faust To: forum@jsoftware.com Date: Wednesday, August 11, 1999 8:29 AM Subject: RE: Jforum: who will mind the store? =20 =20 Don, We see this as overall good news for J users. There will be new = people brought on board for parts of J development. However, the rest of = the development team remains in place and committed to continual = development and support of the J product. There is momentum building = with J and we intend to keep pushing ahead. Anne -----Original Message----- From: owner-jsoftware@lists.interlog.com = [mailto:owner-jsoftware@lists.interlog.com]On Behalf Of Donald Pittenger Sent: Monday, August 09, 1999 11:03 AM To: forum@jsoftware.com Subject: Jforum: who will mind the store? =20 =20 I don't always enjoy posing the blunt question, but as of the = time I'm drafting this, no one else has--so here goes. =20 With Eric leaving JSoftware, who will be the day-to-day head of = the company? What can users expect for the longer-term future? Who = will we be dealing with, and for what aspects of J? =20 I ask this because I've been moving most of my firm's systems to = J from APL, and Eric's announcement raised an Adrian Smith-like level of = paranoia about being left high & dry. =20 Granted, the 'creative phase' of J language development is = probably pretty much over anyway; the need is to keep up with general = computational environmental change (the Web, Linux, etc), and this might = not require quite as much work as getting J to Windows. =20 Therefore, I hope the J folk will take time to spell out = JSoftware's new organization and the responsibilities of various people = involved full/part time in maintaining & enhancing J. >From a = public-relations standpoint, the annouuncement about Eric raises as many = questions as answers. Lacking anything solid about ISA, I for one, = might have to reconsider my shift away from APL. My livelihood could be = at stake, so my attitude necessarily requires sang-froid. =20 Don Pittenger The Demographics Laboratory ------=_NextPart_000_000E_01BEE3E1.C3F80270 charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Please consider this as constructive = criticism:=20 >From a PR perspective, if potentially disruptive or unsettling news is=20 presented, extra care should be taken to give recipients reasonable = reassurance=20 that things will still be okay by spelling out what can be expected in = the near=20 future.
 
In other words, Microsoft can't get away with = announcing 'Bill=20 Gates stepped down as chairman to pursue his long-held dream of forming=20 virtual-campground website: assured stockholders things will be just = fine', and=20 leaving it at that.  There would be all sorts of info about what = Steve=20 Ballmer & all the supporting cast will be doing to fill Bill's = shoes. =20 Sure, MS stock would still probably tank, but at least the folks at=20 Waggoner-Edstrom would have done the best they could under the=20 circumstances.
 
So, when one of the key ISI players leaves, there is = a similar=20 need to reassure the J community by offering concrete info on how those = shoes=20 will be filled.  Otherwise, people will start to wonder.  In = my case,=20 I have an economic stake.  If ISI goes south, it is not fatal, but = would=20 mean several months work bringing an old APL system into line with the = new J=20 system I'm completing.  Nothing immediate, but I would have to do = something=20 within 2 years, and that 'something' would be a diversion from doing = more=20 productive stuff.  So I (and others) need to be reassured that ISI = is=20 taking steps to prevent itself from 'going south'.  That should = have been=20 part of the original announcement: the initial reassurances were oblique = at=20 best.
 
I strongly suggest that ISI or Strand put something = up on the=20 Forum that is concrete, detailed, and thereby lets the J community = (including=20 folks like me who are still are not 100% committed to J) what (in = general) can=20 be expected to happen, and who will be doing it.  Maybe do it this = coming=20 weekend after the dust settles from APL99.
 
Don Pittenger
-----Original = Message-----
From:=20 Anne Faust <amfaust@mindspring.com>
= To:=20 forum@jsoftware.com = <forum@jsoftware.com>
Dat= e:=20 Wednesday, August 11, 1999 8:29 AM
Subject: RE: = Jforum: who=20 will mind the store?

Don,

We see this as overall good news for J users. There will be new = people=20 brought on board for parts of J development. However, the rest of = the=20 development team remains in place and committed to continual = development and=20 support of the J product. There is momentum building with J and we = intend to=20 keep pushing ahead.

Anne

-----Original Message-----
From:=20 owner-jsoftware@lists.interlog.com=20 [mailto:owner-jsoftware@lists.interlog.com]On Behalf Of = Donald=20 Pittenger
Sent: Monday, August 09, 1999 11:03 = AM
To:=20 forum@jsoftware.com
Subject: Jforum: who will mind the = store?

I don't always enjoy posing = the blunt=20 question, but as of the time I'm drafting this, no one else = has--so here=20 goes.
 
With Eric leaving JSoftware, = who will be=20 the day-to-day head of the company?  What can users expect = for the=20 longer-term future?  Who will we be dealing with, and for = what=20 aspects of J?
 
I ask this because I've been = moving most=20 of my firm's systems to J from APL, and Eric's announcement = raised an=20 Adrian Smith-like level of paranoia about being left high &=20 dry.
 
Granted, the 'creative = phase' of J=20 language development is probably pretty much over anyway; the = need is to=20 keep up with general computational environmental change (the = Web, Linux,=20 etc), and this might not require quite as much work as getting J = to=20 Windows.
 
Therefore, I hope the J folk = will take=20 time to spell out JSoftware's new organization and the = responsibilities=20 of various people involved full/part time in maintaining & = enhancing=20 J.  >From a public-relations standpoint, the = annouuncement about=20 Eric raises as many questions as answers.  Lacking anything = solid=20 about ISA, I for one, might have to reconsider my shift away = from=20 APL.  My livelihood could be at stake, so my attitude = necessarily=20 requires sang-froid.
 
Don Pittenger
The Demographics=20 = Laboratory
------=_NextPart_000_000E_01BEE3E1.C3F80270-- -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Wed Aug 11 13:41:00 1999 From: Andrew Nikitin Subject: RE: Jforum: J Windows Interface Date: Wed, 11 Aug 1999 12:16:24 -0500 > Only experience can tell that that mess is actually required if > anybody is > to write a serious application with a fine control on the GUI. My experience tells me: simple things should be simple. Button control is intended to tip user that there is some action available at the point and trigger this action if he (or she) wants to. All other tricks like changing text/color on mouse over, animation on click, etc. just look silly and nothing more. And they are rather annoying if form is used frequently. I agree, that application full of such silly gadgets can be sold much better, though. -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Wed Aug 11 13:41:08 1999 From: "J Tibollo" References: Subject: Re: Jforum: J Windows Interface Date: Wed, 11 Aug 1999 12:26:31 -0400 charset="iso-8859-1" ----- Original Message ----- From: Andrew Nikitin Sent: Wednesday, August 11, 1999 12:10 PM Subject: RE: Jforum: J Windows Interface > > -----Original Message----- > > From: J Tibollo [SMTP:jtibollo@backassociates.com] > > Sent: Wednesday, August 11, 1999 11:02 > > Just looking at a simple Dyalog APL > > button object I note that it has 39 properties, 17 events, and > > 2 methods. Their edit object has 46 properties, 20 events, and > > 2 methods. And so on. > ??? > Did I understand you in a correct way -- are you claiming this mess of > events, properties and methods for the simplest possible control as an > advantage? > Are you really sure you want to see some days a 794 page book titled > "Button control for advanced J users -- tips, tricks and secrets"? This is not my point at all. I do not think that we need a Cadillac solution. I was trying to point out how far away from the Cadillac solution we now are. I was disputing the idea that we have implemented 90% of what is possible. I am simply saying that we are all free to judge for ourselves whether the current J Windows interface fits the needs of the majority or even a signficant minority of the J community. You judge for yourself. >From my vantage point, I do not feel we are following the proverbial 90-10 or 80 - 20 but rather closer to 60 40 or 50 50 level. But I appreciate that it all depends on your application and how good you personally are at extending the J environment, such as it is, through the windows API or by using 3rd party tools for the front end. Also, don't misunderstand. Just because Dialog APL implements all those properties and all those events it doesn't add to the complexity of the product in the least. Defaults are implemented so that a programmer has the choice but not the burden of using this or that feature. Very nice. Regards, Joe -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Wed Aug 11 14:44:22 1999 From: "Nichols, Peter" Subject: RE: Jforum: J Windows Interface Date: Wed, 11 Aug 1999 13:36:29 -0500 charset="iso-8859-1" The hard fact is that if a person who is signing paychecks sees an app with all the windows features versus a simple J solution and they both do the same thing we are going to lose. He probably has never heard of J but he also knows that VB programmers are a dime a dozen. Also the mouse over/animation/color changing extras are not silly. In the hands of a good developer they can make a tremendous difference in how a user perceives the application. And perception IS reality. Again Joe is right, we're around 60/40 or 50/50 and we need the 90/10. -----Original Message----- From: Andrew Nikitin [SMTP:anikitin@fastenal.com] Sent: Wednesday, August 11, 1999 12:16 PM To: 'forum@jsoftware.com' Subject: RE: Jforum: J Windows Interface > Only experience can tell that that mess is actually required if > anybody is > to write a serious application with a fine control on the GUI. My experience tells me: simple things should be simple. Button control is intended to tip user that there is some action available at the point and trigger this action if he (or she) wants to. All other tricks like changing text/color on mouse over, animation on click, etc. just look silly and nothing more. And they are rather annoying if form is used frequently. I agree, that application full of such silly gadgets can be sold much better, though. ---------------------------------------------------------------------------- ---- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Wed Aug 11 15:02:13 1999 From: Subject: RE: Jforum: J Windows Interface Date: Wed, 11 Aug 1999 20:58:57 +0200 Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by plus.interlog.com id OAA10843 >-----Message d'origine----- >De: Andrew Nikitin [SMTP:anikitin@fastenal.com] >Date: mercredi 11 ao�t 1999 18:11 >�: 'forum@jsoftware.com' >Objet: RE: Jforum: J Windows Interface > >??? >Did I understand you in a correct way -- are you claiming this mess of >events, properties and methods for the simplest possible control as an >advantage? >Are you really sure you want to see some days a 794 page book titled >"Button control for advanced J users -- tips, tricks and secrets"? > >[ Paul GAUTHIER] >As a dreamer, I would personally prefer to get rid of Keyboards, mouses and >screen. I would like to talk to (maybe a hologram grewed from some magical >crystal) and have a voice driven interface builder that would give me better >3D graphics than OpenGL with celestial sound effects. > >In the mean time I have both the frustration of trying to get something done >in a specific way with J and ask a C++ or Java programmer to do the job for a >viable interface. Of course one can use the OCX extensions etc. > >But the frustration expressed by Joe is understandable. Maybe we should just >accept that the burden of including Dyalog type of windowing is just out of >price and concentrate on sparsed arrays etc. > >On several occasions I was told that Dyalog would have a hard time to follow >the GUI development pace of Microsoft and that statement appears reasonable, >but they do keep it, the fact that we can use other tools/languages also seem >reasonable but the claim that Joe makes is also understandable. > >Of course, I wish for J's future that more money is driven their way so that >they can tackle that problem. In the mean time, one can use Dyalog to develop >an interface and call J if they want... As far as I know, they would still be >frustrated not to have this or that and will still have to tackle the OCX or >whatever... > >In brief, I agree with Andrew because I know that in the jungle of GUI tools, >one is never satisfied, no matter what... So I prefer a Roger Hui >implementing sparsed arrays then new GUI features, it took me a while to >accept J's official position on GUI but I finally came to peace with it after >having lived frustrations with Dyalog as well... > >Regards/Paul -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Wed Aug 11 16:46:32 1999 From: Andrew Nikitin Subject: RE: Jforum: J Windows Interface Date: Wed, 11 Aug 1999 15:46:09 -0500 > -----Original Message----- > From: Nichols, Peter [SMTP:pnichols@sprinc.com] > Sent: Wednesday, August 11, 1999 13:36 > > Also the mouse over/animation/color changing extras are not silly. In > the > hands of a good developer they can make a tremendous difference in how > a > user perceives the application. And perception IS reality. I disagree. When somebody uses some application 40 hours a week, he admires all this tricky stuff only for the first 3 days. Then he just don't see it at all. But the fact that animated button insert some noticeable delay in the process can be very boring. So, if the supposed lifetime of an application is 3 days or less -- then it is ok, perception really IS reality that matters. > Again Joe is right, we're around 60/40 or 50/50 and we need the 90/10. I would even say 48.3/51.7 -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Wed Aug 11 20:21:55 1999 From: bszuch@wsa-fincon.com.au (WSA-fincon--bill Szuch) Subject: Jforum: JSVIEW - ERROR Date: Thu, 12 Aug 1999 10:20:12 +1000 Organization: WSA Financial Consulting Pty Ltd JSVIEW.IJS - ERROR I have installed J403c and got an error when I carried out the following sets in the installation. (a) J403c was installed as normal from the "j403c.exe", this part was OK. (b) I then used PROJECT MANAGER to have all the library files included as "initial" scripts when J is first started. This also included the file "jsview.ijs" (c) After installing all the library files and restarting J, I got an error that did not allow J to fully instal. (d) The error was: domain error getconfig_j_ n= . {{.1"1 CONFIG) i. Boxopen y. and was from the the script jsview.ijs. When the script file jsview.ijs in not part of the initial scripts the start up is OK - no error. Is this an error or am I doing something wrong by having jsview.ijs included in the initial scripts?. Bill Szuch -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Thu Aug 12 01:48:17 1999 Date: Thu, 12 Aug 1999 08:52:29 +0300 From: Anssi Seppala Subject: Re: Jforum: JSVIEW - ERROR In-Reply-To: <01BEE4AC.49A5C0E0.bszuch@wsa-fincon.com.au> Try drop + add again the script in PM. I remember i had similar experience with on update. Anssi At 10:20 12.8.1999 +1000, you wrote: >JSVIEW.IJS - ERROR -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Thu Aug 12 04:43:04 1999 From: "Hans Fahlin" Subject: RE: Jforum: J Windows Interface Date: Thu, 12 Aug 1999 10:40:11 +0200 X-Msmail-Priority: Normal X-Mimeole: Produced By Microsoft MimeOLE V4.72.3110.3 Importance: Normal In-Reply-To: <01ff01bee417$18572e60$57a4b9c2@apl.it> charset="iso-8859-1" This is my *one* vote in the J Windows Interface design discussion. The purpose of all the bells and whistles of available to the developer in a modern GUI interface is of course to facilitate the development of a user interface that is friendly, easy to learn, and from a number of perspectives lets the intended user use a computer program as productively as possible. The complexity (from a development perspective) of modern GUI elements makes it easier to create a simpler interface (from a usability perspective). By all means, it also puts in the hands of developers the power to create the most terribly confusing interface. Good UI design still necessitates careful attention to the task. Another aspect of user friendliness is a program's ability to look like and behave like other programs in their environment. If users recognize interface elements from other programs from other vendors, they will learn quicker and find usage easier. As time goes by, this makes it impossible for developers not to follow e.g. Microsoft in how programs look and behave. I stress "follow". If the developers of J want to position their product as a computational engine with limited user interface development facilities, I am disappointed. J is also a great tool for rapid application development, for quick-and-dirty-apps if you will, where the main strength is not only its effectiveness in implementing solutions to complex problems, but its amazing productivity. But to play this role, J needs to continue developing with the state of GUI design. In my opinion, J today lacks some important GUI interface features such as multiple document interface functionality, drag and drop, etc... These features are hardly the latest sizzle. They have been around for a long time. J's ability to interface with other development tools such as VB is a great feature. But using it carries a cost. In my experience, it adds an additional layer of complexity, and consequently another source of bugs, to applications. It can speak against J at the time of deciding what tool to use, particularly in a RAD-context. The clear separation between user interface and program logic should and can be maintained anyway. /Hans -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Thu Aug 12 07:52:19 1999 Date: Thu, 12 Aug 1999 12:51:48 +0100 (BST) From: J E H Shaw Subject: Jforum: Quick Agenda I often need to apply different verbs to different ranges of y., where y. may be a large array. Gerund+agenda with rank 0 can sometimes be improved upon considerably in terms of speed, as in the following example to return the log_gamma function ------------------------------------------------------------------[sect.1] RSQRT2PI=: % %: 2p1 HL2PI=: - ^. RSQRT2PI merge=: /:@/:@[ { ] NB. cf "J Phrases" 3B underravel=: adverb def '$@] $ x.@,' NB. apply x. to (,y.) then reshape NB. define gamma function and log_gamma for small,large arguments gamma=: ! & <: gt100=: | > 100"_ lgsmall=: ^.@:gamma lgbig=: HL2PI"_ - ] - ([: % 12&*) - ([: % 360"_ * ^&3) - -&0.5 * ^. NB. versions of log_gamma using agenda, "vectorised agenda" lgagenda=: lgsmall`lgbig @. gt100 lgquick=: (] (] merge lgsmall@:(#~ -.) , lgbig@:#~) gt100) underravel NB. test above constructions lgagenda"0 ] 95 + i. 2 5 336.261 340.815 345.379 349.954 354.539 359.134 363.739 368.354 372.979 377.614 lgquick 95 + i. 2 5 336.261 340.815 345.379 349.954 354.539 359.134 363.739 368.354 372.979 377.614 10 timex 'lgagenda"0 ] i. 3 4 5 6' 0.786 10 timex 'lgquick i. 3 4 5 6' 0.049 ------------------------------------------------------------------[sect.2] I'd like to define a general quick agenda conjunction, so that v0`v1`...`vk agenda select has the same effect as v0`v1`...`vk @. select "0 for zero rank verbs v0,v1,...,vk and select. This is easy(ish) in cases like log_gamma, where v0..vk are monadic and select returns either 0 or 1, as follows ------------------------------------------------------------------[sect.3] NB. define and use conjunction ifft (if false/true) if0=: @:(#~ -.) if1=: @:(#~) ifft=: 2 : '(] (] merge (x.@.0:)if0 , (x.@.1:)if1) y.) underravel' lggen=: lgsmall`lgbig ifft gt100 NB. check OK 10 timex 'lggen i. 3 4 5 6' 0.05 (lggen -: lgagenda"0) i. 3 4 5 6 1 ------------------------------------------------------------------[sect.4] However, I haven't yet come up with a neat agenda as outlined in [sect.2], so throw out the following questions: 1) Have I missed something obvious that would speed up the existing @. ? 2) Has anyone already written a quick agenda conjunction ? Or can anyone come up with one that doesn't require a long define ? 3) It would help to have a 'boxwise' adverb with the following effect: ; ((+:each)`(*:each))/. 1 ; 2 3 ; 4 5 6 2 4 9 8 10 12 ; +:`*: boxwise 1 ; 2 3 ; 4 5 6 NB. I wish! 2 4 9 8 10 12 I feel sure there's a nice way to do this, but I'm stuck. Can anyone help me out here? 4) Is this of sufficient general interest for me to write up a 'Quick Agenda' note (even if restricted to the ifft case as above) for, say, Vector? Many thanks - Ewart Shaw PS. The above timings use J3 on a 486DX33! I'm eagerly awaiting delivery of a new home computer, which will be able to run J4, and revolutionise my life. -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Thu Aug 12 08:45:52 1999 Date: Thu, 12 Aug 1999 05:45:43 -0700 (PDT) From: Joey K Tuttle Subject: RE: Jforum: J Windows Interface In-Reply-To: <000a01bee49e$4a1a0ee0$628b67c2@absanc9098.sto.alfredberg.se> On Thu, 12 Aug 1999, Hans Fahlin wrote: > The purpose of all the bells and whistles of available to the developer in a > modern GUI interface is of course to facilitate the development of a user > interface that is friendly, easy to learn, and from a number of perspectives > lets the intended user use a computer program as productively as possible. This is a trap. GUI (even good ones) rarely allow an experienced user to reach maximum productivity. They deceive by appearing to be easy to use (but that is frequently not the case) and while it is true that simplistic things can be taught quckly - this is quite removed from real productivity issues. > > Another aspect of user friendliness is a program's ability to look like and > behave like other programs in their environment. If users recognize interface > elements from other programs from other vendors, they will learn quicker and > find usage easier. As time goes by, this makes it impossible for developers not > to follow e.g. Microsoft in how programs look and behave. I stress "follow". > You are asserting that Microsoft innovate their GUI elements? But what you say is true, developers feel constrained to make user interfaces "look like" familiar ones - even if that is clumsy and inappropriate... > The clear separation between user interface and program logic > should and can be maintained anyway. On this, we agree. I am aware of many systems where the majority of the development time (not usually rapid...) was spent on GUI, to no particular benefit. This is the trap I refer to, and a very easy one to fall into. I should add that my personal computer of choice is a Macintosh (and has been for many years) so it is not that I am a "DOS head" who insists on doing everything on command lines. On the other hand, for other than tasks which are "best done" with a GUI (e.g. Photoshop, Illustrator, and GUI development), I prefer to use a Unix/Linux environment (primarily command line). The things that an experienced user can do with simple environments and a tool like J often astound even the most jaded users of the "best systems". I like J to be focused on the leanest meanest way to do do the fundamental things - not being a picture editor. Notice in the above that I didn't mention MS Word (or any "word processor" as being an example). In fact, they illustrate my point. They appear to be simple to use, but they can present big challenges to doing complex word processing - and they aren't very flexible or powerful editors either. What such sytems do frequently lead to are things like Email messages and general correspondance that take orders of magnitude more data to convey around the net than the simple words intended (e.g. this message). On top of that, the idea that "pretty fonts" etc. enhance the value of written words is often very misguided. Enough said - but cast my vote (again) for J to keep good/simple interfaces to the environment (and associated tools) it is running in. Given a choice, I hope that J development time will be spent on fundamental enhancements and efficiencies - not on pretty display. -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Thu Aug 12 09:04:27 1999 From: "J Tibollo" References: <000a01bee49e$4a1a0ee0$628b67c2@absanc9098.sto.alfredberg.se> Subject: Re: Jforum: J Windows Interface Date: Thu, 12 Aug 1999 09:04:02 -0400 charset="iso-8859-1" ----- Original Message ----- From: Hans Fahlin Sent: Thursday, August 12, 1999 4:40 AM Subject: RE: Jforum: J Windows Interface > If the developers of J want to position their product as a computational > engine with limited user interface development facilities, I am disappointed. J > is also a great tool for rapid application development, for quick-and-dirty-apps > if you will, where the main strength is not only its effectiveness in > implementing solutions to complex problems, but its amazing productivity. But to > play this role, J needs to continue developing with the state of GUI design. > > In my opinion, J today lacks some important GUI interface features such as > multiple document interface functionality, drag and drop, etc... These features > are hardly the latest sizzle. They have been around for a long time. I have read other of comments which tried to explain some of the rationale for the decline of APL like languages. I wonder if this isn't a good example of that. Clearly there are many market segments each with their own unique requirements. However, if "purists' insist that J (Apl in general) should focus on features which enhance the DP aspects and skip over the interfaces and other mass market applications - is it any wonder that APL is relegated to a particular niche? And if this niche can't maintain itself and declines over time who is to blame? If you want a language with broad appeal you need a language with broad applicability. Now if you want to say that J already has this broad appeal, then I won't quarrel with your religion. I for one don't know. But just remember being better didn't help beta max (VHS won). Regards, Joe -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Thu Aug 12 11:01:42 1999 Date: Thu, 12 Aug 1999 10:01:38 -0400 From: Alain Miville de =?ISO-8859-1?Q?Ch=EAne?= Subject: Re: Jforum: who will mind the store? charset=ISO-8859-1 Content-Disposition: inline Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by plus.interlog.com id KAA26557 Don, could you change a parameter in your text editor so that it wraps lines. Your paragraphs are one liners which my readre doesn't wrap. -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Thu Aug 12 11:01:58 1999 Date: Thu, 12 Aug 1999 10:01:34 -0400 From: Alain Miville de =?ISO-8859-1?Q?Ch=EAne?= Subject: Re: Jforum: J Windows Interface charset=ISO-8859-1 Content-Disposition: inline Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by plus.interlog.com id KAA26555 Message text written by INTERNET:forum@jsoftware.com >Only experience can tell that that mess is actually required if anybody is to write a serious application with a fine control on the GUI.< The simplified GUI in J forces us to concentrate on the essential. I make perfectly adequate applications without surround-sound video. We might need a few more events to be offered, but not many more. -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Thu Aug 12 11:02:08 1999 Date: Thu, 12 Aug 1999 10:01:46 -0400 From: Alain Miville de =?ISO-8859-1?Q?Ch=EAne?= Subject: Re: Jforum: J Windows Interface charset=ISO-8859-1 Content-Disposition: inline Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by plus.interlog.com id KAA26629 Message text written by INTERNET:forum@jsoftware.com >However, if "purists' insist that J (Apl in general) should focus on features which enhance the DP aspects and skip over the interfaces and other mass market applications - is it any wonder that APL is relegated to a particular niche?< APL has always been a minority language, even since the beginning where it was the best thing around competing with FORTRAN on punched cards. Let's face it, most programmers feel comfortable with the mainstream scalar languages, and it takes particular nuts like us to like a mathematical language initially designed to exchange ideas between human beings. It has nothing to do with GUI. -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Thu Aug 12 11:37:33 1999 From: Charles Fisk Subject: RE: Jforum: J Windows Interface Date: Thu, 12 Aug 1999 10:37:19 -0500 charset="iso-8859-1" Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by plus.interlog.com id LAA25879 I use DHTML, JavaScript and a browser as my interface to APL (I have not tried it with J but I am sure it works). The browser has enough GUI and is available to a wider audience. I do not think J should spend lots of time creating buttons and flashy menus rather than stick to what it really is. So yes, keep J GUI stuff basic and exploit browser technology which is more than enough. I can assure you that when the users see they can enter parameters in a browser form and then get APL generated reports/results displayed in a web page they just love us! We are doing this here at USAA, once again I have done it with APL and JavaScript, not with J but I am sure J can do it and maybe better( sadly I am not allowed to introduce J as a tool in USAA,yet...) -----Original Message----- From: Alain Miville de Ch�ne [SMTP:Infodev@compuserve.com] Sent: Thursday, August 12, 1999 9:02 AM To: INTERNET:forum@jsoftware.com Subject: Re: Jforum: J Windows Interface Message text written by INTERNET:forum@jsoftware.com >However, if "purists' insist that J (Apl in general) should focus on features which enhance the DP aspects and skip over the interfaces and other mass market applications - is it any wonder that APL is relegated to a particular niche?< APL has always been a minority language, even since the beginning where it was the best thing around competing with FORTRAN on punched cards. Let's face it, most programmers feel comfortable with the mainstream scalar languages, and it takes particular nuts like us to like a mathematical language initially designed to exchange ideas between human beings. It has nothing to do with GUI. ---------------------------------------------------------------------------- ---- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Thu Aug 12 11:57:34 1999 Date: Thu, 12 Aug 1999 19:02:54 +0300 From: Anssi Seppala Subject: Re: Jforum: J Windows Interface In-Reply-To: <199908121002_MC2-80C7-F59D@compuserve.com> I like that comment. Lets hope more people become nuts :) Anssi At 10:01 12.8.1999 -0400, you wrote: >APL has always been a minority language, even since the beginning where it >was the best thing around competing with FORTRAN on punched cards. Let's >face it, most programmers feel comfortable with the mainstream scalar >languages, and it takes particular nuts like us to like a mathematical >language initially designed to exchange ideas between human beings. It has >nothing to do with GUI. > >--------------------------------------------------------------------------- >----- >J Forum: for information about this list, see >http://www.jsoftware.com/forum.htm -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Thu Aug 12 13:29:11 1999 From: "David Vincent-Jones" References: <4.2.0.58.19990812190110.00a2be80@pop.kolumbus.fi> Subject: Re: Jforum: J Windows Interface Date: Thu, 12 Aug 1999 09:49:55 -0700 When we developed a commercial package (in C ) we found it convenient and very cost effective to pull a very comprehensive public domaine GUI off the web and make modifications as needed. The clever, highly 'tweeked' and fully tested front end saved us all a lot of time and effort. Although such a model would not solve all of the limitations discusses in the Forum, it might make the whole J interface a whole lot more palatable for all concerned. If the resources of this Forum were put to the common task, who knows what might develop... -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Thu Aug 12 13:51:45 1999 From: "Stefano Lanzavecchia" References: <199908121001_MC2-80C7-F593@compuserve.com> Subject: Re: Jforum: J Windows Interface Date: Thu, 12 Aug 1999 19:48:28 +0200 Organization: APL Italiana charset="iso-8859-1" Content-Transfer-Encoding: 8bit ----- Original Message ----- From: Alain Miville de Ch�ne Sent: Thursday, August 12, 1999 4:01 PM Subject: Re: Jforum: J Windows Interface : Message text written by INTERNET:forum@jsoftware.com : >Only experience can tell that that mess is actually required if anybody is : to write a serious application with a fine control on the GUI.< : : The simplified GUI in J forces us to concentrate on the essential. : I make perfectly adequate applications without surround-sound video. Please, don't misunderstand me. I am a user who enjoys over-engineered GUIs, because I am a freak, a nerd and a geek. Yet it's not what we want the important thing. It's more what THEY want, they being all the mr. Smith's and Jones' out there, and more important our marketing department, not too mention theirs. Let's say your application allows the user to choose from a custom set of objects, in a library structure that is quite like a DOS/Unix filesystem. You could give them a simple object browser which can be written in 1 hour and does the job. But what if they say: it does not look like Windows Explorer a single bit; that's what we really want. Try and write Windows Explorer (with all the details, including drag&drop including timer-controlled expansion of subtrees while hovering a collapsed node or scrolling during a drag&drop, copy/cut/paste, rename in place) with the two more methods suggested. It does not really count the fact that you, I, we don't believe that an explorer-like interface is necessary, or, even, useful. If that's what THEY want, that's what they are willing to pay, you'd better give it to them, or the next VB-guy will eat you for dinner. Frankly I believe, J should not change too much. J is a fantastic language and that's where its true power is. Socket interface and the like are more than welcome. The plot package is so beatiful that it's almost not real. But if people insist on building complete applications trusting the fact that an intelligent user can see through a poor GUI the real application shining, J is going to have a really hard time surviving. That was possible 10, maybe 5 years ago. Today, when even a CD player or an FTP client are so lush with their scripting language, their stunning animations, their colorful buttons and captions, yet only a mere few hundred kilobytes, a teletype-like application, with misplaced and odd-sized buttons, misaligned captions and grey like graveyard, even if it offers antigravity and mind-reading, it's going to disappear in a dusty archive. That's a fact. -- WildHeart'99 Homepage: http://www.insight.dk/stf/ -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Thu Aug 12 14:38:13 1999 From: "Oleg Kobchenko" References: <01ff01bee417$18572e60$57a4b9c2@apl.it> Subject: Re: Jforum: J Windows Interface Date: Thu, 12 Aug 1999 21:39:35 +0300 charset="ISO-8859-1" I have a constructive though tentative proposal to get the best of the two worlds: To use Visual Studio and J as lind of add-in, like InterDev and Script Debugger now work in v6.00. There InterDev and Script Debugger have the same feel and look as VB. It seems that they all use the same environment and only add their own "add-ins". What is important in Visual Studio is - form designer - Object Browser - Auto substitution - Debugger If these could be preserved and J (maybe in Scripting Edition) plugged in that would be the solution. -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Thu Aug 12 14:42:40 1999 From: Andrew Nikitin Subject: RE: Jforum: J Windows Interface Date: Thu, 12 Aug 1999 13:42:23 -0500 > -----Original Message----- > From: Stefano Lanzavecchia [SMTP:stf@apl.it] > Sent: Thursday, August 12, 1999 12:48 > > Yet it's not what we want the > important thing. It's more what THEY want, they being all the mr. > Smith's > and Jones' out there, and more important our marketing department, not > too > mention theirs. ... > Try and write > Windows Explorer ... Have you seen Object Browser by Oleg Kobchenko written in J? It is not as difficult as you can imagine. > If that's what THEY want, that's what they are willing to pay, you'd > better give it to them, or the next VB-guy will eat you for dinner. If saying THEY, you mean "users" then "what they want" does not matter at all. User does not know what he wants in nearly 100% of cases. As for willingness to pay it depends on the salesperson much more than on the quality of the product itself. Of course, there are few exceptions, but I am not aware of them. > their stunning animations, their colorful buttons > and captions, yet only a mere few hundred kilobytes, a teletype-like > application, with misplaced and odd-sized buttons, misaligned captions > and > grey like graveyard, Well, I guess they are even more possible misplaced and odd-sized animated colorful shining blinking and singing buttons with misaligned pictures and grey as graveyard animation. Just because it is much more difficult to produce good animation than to produce short and clear label phrase (that, in turn, also requires some labor an skills). Probably you will not believe me, but the best (fast, clear and convenient) interface known to me consited only of labels and edits (with pop up listboxes). And one global menu bar. The art of creating good user interface consists not in putting as much different controls as possible, but in placing them in right positions. -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Thu Aug 12 15:09:55 1999 From: "Oleg Kobchenko" References: Subject: Jforum: Callback functions Date: Thu, 12 Aug 1999 22:11:33 +0300 charset="koi8-r" There was a discussion earlier in the history of the Forum on that thread. In WinAPI, eps. things that are ported to J: MIDI, Console, etc., there are cases that require CallBack functionality, i.e. for asynchroneous interaction. One solution would be 1. Put in "parstr" 'myverb par1, par2, ...' 2. Put in J var "callstr" assembly code: CALL JDLL.IsReady JE Busy PUSH parstr CALL JDLL.Do POP parstr RET ; code to set success result Busy: ; code to set Busy result RET 3. Set the callback to addr of callstr This is a simplistic scheme, and it needs to know the way to call JDLL for the current process (it might be possible even now). -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Thu Aug 12 15:10:27 1999 From: "John D. Baker" Subject: Re: Jforum: J Windows Interface Date: Thu, 12 Aug 1999 15:10:53 -0400 charset="iso-8859-1" X-Mimeole: Produced By Microsoft MimeOLE V4.72.3110.3 I've been following the GUI debate that has been going on in the J forum with some interest. I agree with Eric Iverson and others that the J interface is adequate for simple applications and if more is required it's easy to use the J automation servers and tools like VB, C++ and Delphi. I've written a number of pure J applications and I've also developed hybrid automation based applications. In my experience the hybrid applications are the easiest to work with because you can exploit the strengths of two environments. Like it or not hybrid programming is the way most systems are built these days and as component based computing becomes more and more pervasive it will eventually dominate. Restricting your programming to one environment is counterproductive. It's also counterproductive to try and satisfy all programming needs with one tool or language. Not only do you end up with overly complicated software tools but you also breed narrow minded programmers as well. I think the future of J is best served by making it one of the best array processing components available to programmers. In this sense sparse array handling is far more valuable than enhancing the GUI. -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Thu Aug 12 15:21:11 1999 From: "J Tibollo" References: Subject: Re: Jforum: J Windows Interface Date: Thu, 12 Aug 1999 15:24:06 -0400 charset="iso-8859-1" I would just like to emphasize some points in my original post: 1. the documentation could benefit from restructuring. It might be nice to adopt the attributes, methods, events paradigm used in other GUI products. 2. it might be very nice to extend our current J GUI interface to include just a few more events, properties, and methods (i.e. it would be nice to be able to read the present contents of an edit control without waiting for an event). 3. keeping up with Microsoft's introduction of additional GUI objects is not nearly as important as insuring that the current objects already supported are sufficiently functional to be of widespread use. For example, the Grid object could benefit by adding a method which automatically resized the columns to fit the contents. In conclusion, I feel it would be a good exercise for everyone to consider the state of our current GUI interface in J and suggest only those improvements, enhancements, and extensions which are considered to be essential for a wide majority of users. Those that don't use a GUI interface or that are happy with the present state of affairs need not be overly concerned. The J development team can assess the tradeoffs in extending the current GUI or adding other additional features to the language. Lets think about it... Regards, Joe Tibollo -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Thu Aug 12 17:20:18 1999 From: Andrew Nikitin Subject: RE: Jforum: Quick Agenda Date: Thu, 12 Aug 1999 16:19:20 -0500 > -----Original Message----- > From: J E H Shaw [SMTP:strgh@csv.warwick.ac.uk] > Sent: Thursday, August 12, 1999 06:52 > > 3) It would help to have a 'boxwise' adverb with the following effect: > > ; ((+:each)`(*:each))/. 1 ; 2 3 ; 4 5 6 > 2 4 9 8 10 12 > ; +:`*: boxwise 1 ; 2 3 ; 4 5 6 NB. I wish! > 2 4 9 8 10 12 > > I feel sure there's a nice way to do this, but I'm stuck. > Can anyone help me out here? > boxwise=: 1 : '(([: , ''&.''"_ ,&< ;&(,''>'')) each m.)/.' ; ((+:each)`(*:each))/. 1 ; 2 3 ; 4 5 6 2 4 9 8 10 12 ; +:`*: boxwise 1 ; 2 3 ; 4 5 6 NB. I wish! 2 4 9 8 10 12 Exactly as you wish. Is it nice enough? Does it help? PS. I liked idea to use oblique to apply each verb from gerund to corresponding items in the list. nsg -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Thu Aug 12 18:04:29 1999 From: "David Vincent-Jones" References: Subject: Re: Jforum: J Windows Interface Date: Thu, 12 Aug 1999 14:43:15 -0700 I could live without many of the clever bells and whistles if only I could get a graphics window with scrolling. Graphical display is too important today, and images are far too large, to tie down to a fixed window device. In the meantime I can only carry on using J as an arithmetic prototyping tool...pity! -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Thu Aug 12 21:01:56 1999 From: k-list@iname.com Date: Thu, 12 Aug 1999 20:58:15 -0400 (EDT) Content-Type: Text/Plain Subject: Re: Jforum: J Windows Interface "bill lam" form editor:- To be fair, J's form editor "is" already very powerful. But the "is" changes to "was" quickly. A shortcut to catch up with the latest GUI is using OCX. J will be MUCH better if it's OCX support adds 1. arguments in event procedure 2. property explorer-like pop-up desginer hybrid application:- Interface between J and scalar programming language is very clumsy, especailly for boxed array. VB call J: J subroutine is often very short, even a one-liner. It looks like most coding spent on passing argments back and fore. This is against productivity. J call VB: same applies here. An additional drawback is events in VB are difficult to handled as callbacks in J. --------------------------------------------------- Get free personalized email at http://www.iname.com -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Thu Aug 12 21:42:03 1999 From: "Donald Pittenger" Subject: Jforum: re: J Windows Interface Date: Thu, 12 Aug 1999 18:40:05 -0700 boundary="----=_NextPart_000_0030_01BEE4F2.1854C5E0" This is a multi-part message in MIME format. ------=_NextPart_000_0030_01BEE4F2.1854C5E0 charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Maybe it's time to think about pluses and minuses of GUI from the ISI = developer standpoint. I've never--and wouldn't dream of--getting into = programming with low-level windows commands, so I have no clear idea = what the J team is up against. But they do. So, perhaps it would be constructive for them to let us = know what features not present under wd are really tough nuts to crack = (presumably, they're already giving us most of the easy stuff). I, for = one, emailed them a year or two ago hoping that MDI could be added, but = the reply was that they couldn't justify the work at that time. Ideally, it would be nice to have a wish-list from J programmers = weighted by number of votes and intensity of need ('it would be cute to = have',..., 'they'll can me if it's not on the screen by December'). And = this would be matched by how much toil & trouble the implementation = would cost. This might highlight what might be added to wd, and when. Then there lurks the problem of GUIs on Linux which might divert ISI = effort in the middle-term future. How much of this might be farmed out = to volunteers or other non-ISI personnel? Finally, I'll add that I tend to agree with Joe Tibollo and Stefano as = opposed to the raw-J set. Aside from experimentation, everything I do = uses a GUI, and this is all in-house work. The J interface (aside from = MDI) handles most of my needs, though I do have a few features I'd like = to see added. Let's vote: the Unix J'ers do. Don Pittenger ------=_NextPart_000_0030_01BEE4F2.1854C5E0 charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Maybe it's time to think about pluses and minuses of GUI from the = ISI=20 developer standpoint.  I've never--and wouldn't dream of--getting = into=20 programming with low-level windows commands, so I have no clear idea = what the J=20 team is up against.
 
But they do.  So, perhaps it would be constructive for them to = let us=20 know what features not present under wd are really tough nuts to crack=20 (presumably, they're already giving us most of the easy stuff).  I, = for=20 one, emailed them a year or two ago hoping that MDI could be added, but = the=20 reply was that they couldn't justify the work at that time.
 
Ideally, it would be nice to have a wish-list from J programmers = weighted=20 by number of votes and intensity of need ('it would be cute to = have',...,=20 'they'll can me if it's not on the screen by December').  And this = would be=20 matched by how much toil & trouble the implementation would = cost.  This=20 might highlight what might be added to wd, and when.
 
Then there lurks the problem of GUIs on Linux which might divert = ISI effort=20 in the middle-term future.  How much of this might be farmed out to = volunteers or other non-ISI personnel?
 
Finally, I'll add that I tend to agree with Joe Tibollo and Stefano = as=20 opposed to the raw-J set.  Aside from experimentation, everything I = do uses=20 a GUI, and this is all in-house work.  The J interface (aside from = MDI)=20 handles most of my needs, though I do have a few features I'd like to = see=20 added.
 
Let's vote: the Unix J'ers do.
 
Don Pittenger
------=_NextPart_000_0030_01BEE4F2.1854C5E0-- -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Fri Aug 13 10:34:18 1999 From: "Seymour Glass" Subject: RE: Jforum: J Windows Interface Date: Fri, 13 Aug 1999 10:30:33 -0400 charset="iso-8859-1" Importance: Normal In-Reply-To: <002601bee50b$b94ec220$3d0881ce@desktop> You could implement scrolling today, except that it would be too slow. I think we need to have a BitBLT command that can use the hardware accelerators; and the ability to specify the RasterOp to be used for brushes. Then we could scroll quickly, and draw XORed boxes and lines. I would like to draw a box that follows the cursor, but I can't figure out how to undraw it (XOR would solve that problem) or how to remove it when the cursor leaves the isigraph control (is that a candidate for a focus event?). Henry Rich > -----Original Message----- > From: owner-jsoftware@lists.interlog.com > [mailto:owner-jsoftware@lists.interlog.com]On Behalf Of David > Vincent-Jones > Sent: Thursday, August 12, 1999 5:43 PM > To: forum@jsoftware.com > Subject: Re: Jforum: J Windows Interface > > > I could live without many of the clever bells and whistles if only I could > get a graphics window with scrolling. > > Graphical display is too important today, and images are far too large, to > tie down to a fixed window device. > > In the meantime I can only carry on using J as an arithmetic prototyping > tool...pity! > > > > ------------------------------------------------------------------ > -------------- > J Forum: for information about this list, see > http://www.jsoftware.com/forum.htm -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Fri Aug 13 12:30:28 1999 Date: Fri, 13 Aug 1999 12:34:55 -0400 From: Tom Chwastyk Organization: US Naval Research Laboratory Subject: Jforum: Enigma demo slower on faster machine? On my old P133 laptop (W95, trid_pci.drv, trid_pci.vxd), the Enigma demo sets up in what looks like about 200 msec. (You can see some unpainted white parts of the form before the color snaps in, but barely). Moves take something under 100 msec. On my newer P350 desktop (NT4.0, Intel740 pci, gfx40.sys, gfx40.dll), setup takes 3500 msec. It's painfully slow to watch the painting. Moves take almost 1500 msec. Any suggestions on how to improve the video performance under J? -- Tom Chwastyk (sounds like Fosdick; Polish: CH='H, W=F (here), A=AH, styk=stick, soften strict 'HFAH-stick) -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Fri Aug 13 12:37:25 1999 From: Andrew Nikitin Subject: RE: Jforum: Enigma demo slower on faster machine? Date: Fri, 13 Aug 1999 11:37:04 -0500 > -----Original Message----- > From: Tom Chwastyk [SMTP:tomC@cms.nrl.navy.mil] > Sent: Friday, August 13, 1999 11:35 > > On my old P133 laptop ... 200 msec. > On my newer P350 desktop ... 3500 msec. > > Any suggestions on how to improve the video performance under J? ? It seems to me that you have already given the answer. Use P133 instead of P350. -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Fri Aug 13 13:11:14 1999 Date: Fri, 13 Aug 1999 13:15:44 -0400 From: Tom Chwastyk Organization: US Naval Research Laboratory Subject: Re: Jforum: Enigma demo slower on faster machine? References: Andrew Nikitin wrote: > > From: Tom Chwastyk > > > > On my old P133 laptop ... 200 msec. > > On my newer P350 desktop ... 3500 msec. > > > > Any suggestions on how to improve the video performance under J? > ? > It seems to me that you have already given the answer. > Use P133 instead of P350. Seriously, NB. P350 intel (1000 time '^.>:i.1000'),10 time 'enigma_b1_button_jigdemo_ ''''' 0.000341 1.407 NB. P133 trident (1000 time '^.>:i.1000'),10 time 'enigma_b1_button_jigdemo_ ''''' 0.00131 0.072 so I'd rather do numerical work on the P350. Any suggestions how to improve the P350 video speed BY SOFTWARE SETTINGS, DRIVER CHANGES, etc.? Is this just inherent in NT or the Intel 740? Anybody else notice this? -- Tom Chwastyk (sounds like Fosdick; Polish: CH='H, W=F (here), A=AH, styk=stick, soften strict 'HFAH-stick) -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Fri Aug 13 13:32:22 1999 Date: Fri, 13 Aug 1999 13:32:02 -0400 From: David Ness Subject: Jforum: Problem with OK's dlistview.ijs When I try to run slv_run'' in Oleg's DLISTVIEW.IJS I get domain error: wd is it just me, or is it a bug? -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Fri Aug 13 15:21:45 1999 From: "Sylvain Hamel" Subject: Jforum: GUI again Date: Fri, 13 Aug 1999 15:16:51 -0400 boundary="----=_NextPart_000_0057_01BEE59E.DEB12F60" This is a multi-part message in MIME format. ------=_NextPart_000_0057_01BEE59E.DEB12F60 charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable I would like to add my opinion to the ongoing discussion about the GUI = side of J. Contrary to the general opinion, I think that the interface part of a = program is very important. I would evaluate that the GUI is at the very = least 50% of the "efficiency" of a program. Lots of research has been = done on this. Denial of the facts never solved anything. And to say that = users dont know what they want is ... real bad to be polite. Are we not = users ourselves ? Are we not discussing the software that we use? That = would mean we do not know what we want ???? Its time we end this myth of = the sheeps-users having to be herded the right way. I think function is = inherently tied to form and it is wrong to consider the two seperatly. = In my experience, just making the application prettier solves a lot of = problems. Welcome to the real world. SO, ... exactly how much is pretty then? In a perfect world you would = have the J language with a GUI designer like VC/VB/DELPHI . = Unfortunatly, since Adam did bite the apple... Right now J has a lot = going for it in the interface dept.. Eric did a great job on OCX = support. We use a few of them and it works great. In short OCXs fill = many of J's GUI shortcomings. I dont think GUI support is a do or die = thing. I think small changes once in a while go a long way. We have = developed some nice commercial applications here and will continue to = use the J GUI designer. We found that using J with VB was a bit of a = hassle. Maybe someday if the communication between J and VB gets better = I will consider the later. But right now, J GUIs WORKED very well for = me. It goes a long way in making J less of a niche market language. It = increases the appeal of J to programmers. The increase in the population = of the "jugglers" specie will greatly improve the J platform. Sylvain Sylvain@webnet.qc.ca P.S. In response to Bill Lam :=20 1. arguments in events procedure : use wdq in event handler. 2. property explorer/designer : use wd'oledlg ocxid' on the ocx during = run-time to pop its property designer. Hope it helped... ------=_NextPart_000_0057_01BEE59E.DEB12F60 charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
I would like to add my opinion to the ongoing = discussion=20 about the GUI side of J.
 
Contrary to the general opinion, I think that = the=20 interface part of a program is very important. I would evaluate = that=20 the GUI is at the very least 50% of the "efficiency" of a=20 program. Lots of research has been done on this. Denial of the = facts never=20 solved anything. And to say that users dont know what they want is ... = real bad=20 to be polite. Are we not users ourselves ? Are we not discussing the = software=20 that we use? That would mean we do not know what we want = ???? Its time=20 we end this myth of the sheeps-users having to be herded the right way. = I think=20 function is inherently tied to form and it is wrong to consider the = two=20 seperatly. In my experience, just making the application = prettier solves a=20 lot of problems. Welcome to the real world.
 
SO, ... exactly how much is pretty then? In a = perfect world=20 you would have the J language with a GUI designer like VC/VB/DELPHI .=20 Unfortunatly, since Adam did bite the apple... Right now J=20 has a lot going for it in the interface dept.. Eric did a great job on=20 OCX support. We use a few of them and it works great. In = short OCXs=20 fill many of J's GUI shortcomings. I dont think GUI support is a do or = die=20 thing. I think small changes once in a while go a long way. We = have=20 developed some nice commercial applications here and will continue = to use=20 the J GUI designer. We found that using J with VB was a bit of a hassle. = Maybe=20 someday if the communication between J and VB gets better I = will consider=20 the later. But right now, J GUIs WORKED very well for me. = It goes=20 a long way in making J less of a niche market language. It increases the = appeal=20 of J to programmers. The increase in the population = of the=20 "jugglers" specie will greatly improve the J platform.
 
Sylvain
Sylvain@webnet.qc.ca
 
P.S. In response to Bill Lam :
 
1. arguments in events procedure : use wdq in event=20 handler.
2. property explorer/designer : use wd'oledlg ocxid' = on the=20 ocx during run-time to pop its property designer.
Hope it helped...
------=_NextPart_000_0057_01BEE59E.DEB12F60-- -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Fri Aug 13 18:03:19 1999 Date: Fri, 13 Aug 1999 18:05:34 -0400 From: Daniel Torres Subject: RE: Jforum: J Windows Interface Content-Type: multipart/mixed; boundary="------------0AC9A706104571D08BCD6C88" This is a multi-part message in MIME format. --------------0AC9A706104571D08BCD6C88 I have developed several applications with J. The way I have structured them is that *all* their input is through files on disk. For example, a user creates a file using a word processor or spreadsheet. So long as the file is saved in the appropriate location, with the appropriate name, and with the appropriate data, when the J application is run, it reads the files, processes them, and saves the results in a new file. The user then looks at the output file with whatever word processor or spreadsheet. This structure has several advantages: A) users who want custom made interfaces can have them developed by another programmer of choice (usually less expensive, usually better at GUI, and in parallel to my development, so long as I have given them the expected file structures and data); B) users can have me, or whatever other programmer, develop the interfaces in any GUI development system that is best for developing the desired final GUI. Given the above, I have until now expected J to aim to always be the ultimate array processing development tool, leaving others' aim to be the best GUI development tools. Maybe what is needed is for ISI to negotiate with a company that develops the GUI tool that ISI thinks shares a similar J philosophy, and agree on a common interface between the two tools. They could then be sold as add-ons to each other. In the meantime, I sleep better knowing that ISI top developers are concentrating their expertise on improving J's array handling, and not its GUI handling. Daniel --------------0AC9A706104571D08BCD6C88 name="torres.vcf" Content-Description: Card for Daniel Torres Content-Disposition: attachment; filename="torres.vcf" begin:vcard n:Torres;Daniel tel;fax:1-978.383-5817 tel;home:1-305.441-0369 tel;work:1-305.461-6829 x-mozilla-html:TRUE adr:;;600 Biltmore Way APT 412;Coral Gables;FL;33134-7529;USA version:2.1 email;internet:torres@SAmerica.com fn:Daniel Torres end:vcard --------------0AC9A706104571D08BCD6C88-- -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Fri Aug 13 20:02:09 1999 Date: Fri, 13 Aug 1999 20:01:47 -0400 From: David Ness Subject: Re: Jforum: J Windows Interface References: <37B496AE.FBAF3DD5@SAmerica.com> I like your approach and use it often. However, it seems to me to apply only to problems where the complete `input set' is defined before any of the results of the calcualtion are available. This is only a small fraction of the situations that I encounter. Most of my tasks are `genuinely interactive' in that they require some input to produce some output which then guides the next stage of input etc. I don't see that your approach applies the this very broad set of problems. How do you handle them? Daniel Torres wrote: > > I have developed several applications with J. The way I have structured them > is that *all* their input is through files on disk. For example, a user > creates a file using a word processor or spreadsheet. So long as the file is > saved in the appropriate location, with the appropriate name, and with the ... -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Sat Aug 14 00:02:29 1999 From: "Anders Ericson" Subject: RE: Jforum: J Windows Interface Date: Fri, 13 Aug 1999 22:55:53 -0500 charset="iso-8859-1" Content-Transfer-Encoding: 8bit In-Reply-To: X-Mimeole: Produced By Microsoft MimeOLE V4.72.3155.0 Importance: Normal Good point! J is really well suited to build the guts of serious DP apps and this edge has to be maintained. Using a WEB browser as the user interface WILL be the way of the future - Most likely, apps will be accessed (for a fee) on the Net. Hence, I would like to see more effort (and documentation) on how J can borrow the GUI from other tools - In other words, rather than spending time on bringing J from 90/10 (or 50/50) to 99/1, concentrate on improving J's core strength and ability to interect with other languages. My observation is that there is a tremendous and growing appetite for number crunching and very few tools to do a good job. By the by, I'm in the business of crunching numbers and have always preferred APL (and now J) to do so... and fancy GUI rarely adds any real value:) > -----Original Message----- > From: owner-jsoftware@lists.interlog.com > [mailto:owner-jsoftware@lists.interlog.com]On Behalf Of Charles Fisk > Sent: Thursday, August 12, 1999 10:37 AM > To: 'forum@jsoftware.com' > Subject: RE: Jforum: J Windows Interface > > > I use DHTML, JavaScript and a browser as my interface to APL (I have not > tried it with J but I am sure it works). The browser has enough GUI and is > available to a wider audience. I do not think J should spend lots of time > creating buttons and flashy menus rather than stick to what it > really is. So > yes, keep J GUI stuff basic and exploit browser technology which is more > than enough. I can assure you that when the users see they can enter > parameters in a browser form and then get APL generated reports/results > displayed in a web page they just love us! We are doing this > here at USAA, > once again I have done it with APL and JavaScript, not with J > but I am sure > J can do it and maybe better( sadly I am not allowed to introduce J as a > tool in USAA,yet...) > -----Original Message----- > From: Alain Miville de Ch�ne [SMTP:Infodev@compuserve.com] > Sent: Thursday, August 12, 1999 9:02 AM > To: INTERNET:forum@jsoftware.com > Subject: Re: Jforum: J Windows Interface > > Message text written by INTERNET:forum@jsoftware.com > >However, if "purists' insist that J (Apl in general) > should focus on features which enhance the DP aspects and skip over > the > interfaces and other mass market applications - is it any wonder > that APL > is > relegated to a particular niche?< > > APL has always been a minority language, even since the beginning > where it > was the best thing around competing with FORTRAN on punched cards. > Let's > face it, most programmers feel comfortable with the mainstream > scalar > languages, and it takes particular nuts like us to like a > mathematical > language initially designed to exchange ideas between human beings. > It has > nothing to do with GUI. > > > ------------------------------------------------------------------ > ---------- > ---- > J Forum: for information about this list, see > http://www.jsoftware.com/forum.htm > > ------------------------------------------------------------------ > -------------- > J Forum: for information about this list, see http://www.jsoftware.com/forum.htm -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Sat Aug 14 03:31:35 1999 From: k-list@iname.com Date: Sat, 14 Aug 1999 03:28:59 -0400 (EDT) Content-Type: Text/Plain Subject: Re: Jforum: J Windows Interface "bill lam" For J4.03c, OCX still can't handling double-byte string properly using oleget/oleset commands under Chinese Win98. Strangely it works properly under English Win98 with Njstar. Apart from ocx, there are no numeric and date input window controls. I am curious what do people use for fool-proof numeric/date input. >Sylvain Thank you, I've never heard of these (undocumented?) ocx variables before. >Stefano >If that's what THEY want, that's what they are willing to pay, you'd >better give it to them, or the next VB-guy will eat you for dinner. I can't agree with you more. No doubt, applications in APL/J with advanced data cracking capability can be sold to all Fortune 100 companies. But fancy-looking naive software in VB/VC/Delphi can be sold to the remaining Fortune 1000000 companies. --------------------------------------------------- Get free personalized email at http://www.iname.com -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Sat Aug 14 09:42:56 1999 From: "J T" Subject: Jforum: Sparse Arrays Date: Sat, 14 Aug 1999 09:42:24 -0400 boundary="----=_NextPart_000_002B_01BEE639.50285F20" This is a multi-part message in MIME format. ------=_NextPart_000_002B_01BEE639.50285F20 charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable I'm just a bit curious about how Sparse Arrays might be implemented. I = think I generally understand the idea behind sparse arrays. But, how = would the interpreter handle the case where information is requested = from an element which has no data? Would that be an error, or would the = interpreter simply return some fill element? If anyone has a clear understanding of how Sparse Arrays would actually = work, I would love to listen. Regards, Joe ------=_NextPart_000_002B_01BEE639.50285F20 charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
I'm just a bit curious about how Sparse = Arrays=20 might be implemented.  I think I generally understand the idea = behind=20 sparse arrays.  But, how would the interpreter handle the case = where=20 information is requested from an element which has no data?  Would = that be=20 an error, or would the interpreter simply return some fill=20 element?
 
If anyone has a clear understanding of = how Sparse=20 Arrays would actually work, I would love to = listen.
 
Regards,
Joe
------=_NextPart_000_002B_01BEE639.50285F20-- -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Sat Aug 14 10:16:54 1999 Date: Sat, 14 Aug 1999 10:13:26 -0400 From: Cliff Reiter Subject: Re: Jforum: Sparse Arrays Organization: Lafayette College Content-type: text/plain; charset=us-ascii References: <002e01bee65a$d7e1d7e0$8fd23dcf@jat.lafayette.edu> J T wrote: > > I'm just a bit curious about how Sparse Arrays might be implemented. > I think I generally understand the idea behind sparse arrays. But, > how would the interpreter handle the case where information is > requested from an element which has no data? Would that be an error, > or would the interpreter simply return some fill element? > > If anyone has a clear understanding of how Sparse Arrays would > actually work, I would love to listen. > > Regards, > Joe Roger Hui spoke on this at APL99 in Scranton yesterday and the work was described as work in progress although it already has been implemented (not available yet) to a significant degree. His paper in the proceedings gives a better discussion than I could give, but the parts that are implemented look very impressive. There is a default fill element and the representation is list of indices and elements for those indices along with various header stuff including the fill element which need not be 0. These arrays are compatible with current arrays in that a=. an array s=. $. a NB. a sparse rep of a s is displayed as indices & elements (some of us encouraged Roger to also display the fill element although admittedly that might look clumsy) s -: a 1 even though they are displayed in differnt forms Since the fill element may be nonzero s1+s2+3 would have (the fill elt of s1) + (fill of s2) + 3 as its fill element. Cool. When I heard this was coming as a feature, I thought to myself that it would probably be useful for PDE type people - serious number crunchers. But as I listened to his talk, I realize I've been forced to use this representation in some of my 3D chaos work and having it built into the language is going to be extremely helpful for my future work. Way to go Roger. Cliff -- Clifford A. Reiter Mathematics Department, Lafayette College Easton, PA 18042 USA, 610-330-5277 http://www.lafayette.edu/~reiterc -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Sat Aug 14 10:26:18 1999 Date: Sat, 14 Aug 1999 10:25:30 -0400 From: Joe Kunkel Subject: Re: Jforum: Sparse Arrays References: <002e01bee65a$d7e1d7e0$8fd23dcf@jat> Joe_T, Here is one approach to sparse arrays from statistics. In factorial analysis sparse arrays or incomplete or unequal sample sizes can be dealt with by solving the general linear equations for individual factors indirectly by subtraction rather than by use of the factors own design matrix columns. If Y = X B where Y is n by k, X is n by p, B is p by k then if B is B1 | B2 | B3 where Bi is pi by k then if the data were not sparse one could solve for the significance of Bi directly from the inversion process in solving Y = Xi Bi but if sparse data in Y results in confounding of factors one can solve for the significance by (if i = 2) then let B13 = B1 | B3 and X13 = X1 | X3 solve for Y = X13 B13 and use the inversion process reduction of Y'Y as the subtraction from the Y = XB solution reduction for all of B to compute the reduction due to B2. I have implemented this in J in a crude translation from my APL version. I will share this script with anyone interested in it at its present crude state and welcome any help in making it more J-like and with less work-arounds due to my inability to find direct translations from my APL. It comes with some sample data sets and some instructions in the script. This is the multivariate extension of the general linear model and is explained in Rao (1965) Linear Statistical Inference and Its Applications. John Wiley & Sons, New York, 522pp. I also have interest in computing contour plots from sparse and non-rectangular arrays of data so I would welcome any J scripts that implement that. Joe_K > J T wrote: > > I'm just a bit curious about how Sparse Arrays might be implemented. > I think I generally understand the idea behind sparse arrays. But, > how would the interpreter handle the case where information is > requested from an element which has no data? Would that be an error, > or would the interpreter simply return some fill element? > > If anyone has a clear understanding of how Sparse Arrays would > actually work, I would love to listen. > > Regards, > Joe -- -------------------- Joseph G. Kunkel, Professor Biology Department joe@bio.umass.edu University of Massachusetts http://www.bio.umass.edu/biology/kunkel Amherst MA 01003 -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Sat Aug 14 10:27:50 1999 Date: Sat, 14 Aug 1999 10:27:43 -0400 From: David Ness Subject: Re: Jforum: Sparse Arrays References: <002e01bee65a$d7e1d7e0$8fd23dcf@jat> Well, here's a model that is unprejudiced by any knowledge (i.e. I don't know much, but am willing to tell you what I know). Think of it this way. Instead of building a matrix where values are all stored in some linear array, I create a function that stores `pairs' that are `(index comb)' and `value'. Here we might have in a non-sparse representation: nonA =: 5 5 $ 1 23 1 # 3 0 5 In a sparse representation we might use sparseA =: 2 2 $ (0 0);3;(4 4);5 both represent the same information, but in quite different ways. In this specific case there isn't much efficiency gained by the sparse representation, but consider a case where we had: nonA =: 10000 10000 $ 1 99999998 1 # 3 0 5 sparseA =: 2 2 $ (0 0);3;(9999 9999);5 The sparse array is the `same size' (in terms of needed `real' storage) as the previous example, but the `non-sparse' array requires 100,000,000 cells instead of the 25 required in the early example. So here, sparse representations really win. The rule for retrieving from the sparseA is clearly to check the indices to see if they match any element. If they do, the value is the corresponding `value' entry. If they don't, return 0. Storing in the sparse array is similar: check the indices and if they are there, change the value to the new one. If they aren't there, append a line with the new indices and new value. Notice that matrix addition, for example, is quite a different implementation in the two cases. I a non-sparse case we add elements in corresponding slots. In the sparse case, we concatenate the lists, carefully `collecting' any situations were there are multiple entries corresponding to some particular index. In a `sparse matrix' implementation, all of these machinations go on without being visible to the user. You say: A + B without caring about whether A and B are sparse or not. The system takes care of `doing the right thing'. > J T wrote: > > I'm just a bit curious about how Sparse Arrays might be implemented. I think I generally > understand the idea behind sparse arrays. But, how would the interpreter handle the case where > information is requested from an element which has no data? Would that be an error, or would the > interpreter simply return some fill element? > > If anyone has a clear understanding of how Sparse Arrays would actually work, I would love to > listen. > > Regards, > Joe -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Sat Aug 14 12:05:26 1999 From: "John D. Baker" Subject: Re: Jforum: J Windows Interface Date: Sat, 14 Aug 1999 12:05:48 -0400 charset="iso-8859-1" >Apart from ocx, there are no numeric and date input window controls. >I am curious what do people use for fool-proof numeric/date input. Whenever I've done this I just define the input fields as character and handle all the numeric validation with J verbs. Works great and allows the entry of J numerics like 23.05j2 23r17 which are typically rejected by most field controls. The downside of such an approach is more coding for moi and input fields that aren't quite as zippy as the native controls. John Baker bakerjd@kos.net -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Sat Aug 14 12:23:02 1999 From: "Oleg Kobchenko" References: <37B452C0.564982BF@cms.nrl.navy.mil> Subject: Re: Jforum: Enigma demo slower on faster machine? Date: Sat, 14 Aug 1999 19:24:44 +0300 charset="iso-8859-1" It might very well be differemce in the speed of the video, which you can't improve. Also it can matter if you switched from Windows 9x with directX 6.0 to NT with directX 2.0, or things like that. ----- Original Message ----- From: Tom Chwastyk Sent: Friday, August 13, 1999 20:15 Subject: Re: Jforum: Enigma demo slower on faster machine? > Andrew Nikitin wrote: > > > > From: Tom Chwastyk > > > > > > On my old P133 laptop ... 200 msec. > > > On my newer P350 desktop ... 3500 msec. > > > > > > Any suggestions on how to improve the video performance under J? > > ? > > It seems to me that you have already given the answer. > > Use P133 instead of P350. > > Seriously, > > NB. P350 intel > (1000 time '^.>:i.1000'),10 time 'enigma_b1_button_jigdemo_ ''''' > 0.000341 1.407 > > NB. P133 trident > (1000 time '^.>:i.1000'),10 time 'enigma_b1_button_jigdemo_ ''''' > 0.00131 0.072 > > so I'd rather do numerical work on the P350. Any suggestions how to improve the P350 > video speed BY SOFTWARE SETTINGS, DRIVER CHANGES, etc.? Is this just inherent in NT > or the Intel 740? Anybody else notice this? > > -- > Tom Chwastyk (sounds like Fosdick; > Polish: CH='H, W=F (here), A=AH, > styk=stick, soften strict 'HFAH-stick) > > > > -------------------------------------------------------------------------------- > J Forum: for information about this list, see http://www.jsoftware.com/forum.htm > -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Sun Aug 15 00:51:53 1999 From: k-list@iname.com Date: Sun, 15 Aug 1999 00:51:25 -0400 (EDT) Content-Type: Text/Plain Subject: Re: Jforum: Enigma demo slower on faster machine? "bill lam" >On my old P133 laptop (W95, trid_pci.drv, trid_pci.vxd), the Enigma demo >On my newer P350 desktop (NT4.0, Intel740 pci, gfx40.sys, gfx40.dll), 1)wd'g* commands have been replaced by wd'gl* commands in newer version of J. 2)GL driver for win95/98 is different from that of winNT. --------------------------------------------------- Get free personalized email at http://www.iname.com -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Sun Aug 15 10:10:16 1999 From: dajoy@hoy.net (Ajoy Victor) Date: Sun, 15 Aug 1999 09:02:30 -0500 Content-type: text/plain; charset=US-ASCII Subject: Jforum: Past Posts? Is there a way to access past posts? Daniel -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Sun Aug 15 10:29:48 1999 Date: Sun, 15 Aug 1999 10:29:11 -0400 From: Alain Miville de =?ISO-8859-1?Q?Ch=EAne?= Subject: Re: Jforum: J Windows Interface charset=ISO-8859-1 Content-Disposition: inline Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by plus.interlog.com id KAA27923 Message text written by INTERNET:forum@jsoftware.com >I am curious what do people use for fool-proof numeric/date input.< We use the controls from Far Point. -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Mon Aug 16 00:57:40 1999 From: k-list@iname.com Date: Mon, 16 Aug 1999 00:56:40 -0400 (EDT) Content-Type: Text/Plain Subject: Jforum: Re: OCX and crash during closing form "bill lam" I've found a workaround to ocx crash problem. Manually destroy all ocx before wd'pclose inside the destroy verb destroy=: 3 : 0 'user32 DestroyWindow i i' 15!:0 <".wd'setenable d1 0; qhwndc' 'user32 DestroyWindow i i' 15!:0 <".wd'setenable d2 0; qhwndc' 'user32 DestroyWindow i i' 15!:0 <".wd'setenable d3 0; qhwndc' 'user32 DestroyWindow i i' 15!:0 <".wd'setenable d4 0; qhwndc' 'user32 DestroyWindow i i' 15!:0 <".wd'setenable d5 0; qhwndc' 'user32 DestroyWindow i i' 15!:0 <".wd'setenable d6 0; qhwndc' 'user32 DestroyWindow i i' 15!:0 <".wd'setenable d7 0; qhwndc' 'user32 DestroyWindow i i' 15!:0 <".wd'setenable d8 0; qhwndc' wd'pclose' codestroy'' ) But J pclose is expected to handle it in long term. By the way, is there a simpler method to select a child control for subsquently wd command? --------------------------------------------------- Get free personalized email at http://www.iname.com -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Mon Aug 16 10:58:47 1999 From: "Sylvain Hamel" Subject: Jforum: Re : Re : GUI again Date: Mon, 16 Aug 1999 10:48:08 -0400 boundary="----=_NextPart_000_004C_01BEE7D4.D3F9D080" This is a multi-part message in MIME format. ------=_NextPart_000_004C_01BEE7D4.D3F9D080 charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable >> Are we not discussing the software that we use? That would mean we do >> not know what we want ???? >Yes that would. >For >example if somebody right now has some very slight keenness, say, in >music he will rather vote for adding musical capabilities to J rather >than for adding sparse arrays support I'd like to believe I have enough judgement to know what's good for me. = In general, people wish what would be usefull to them. Like all things = in life you have to separate the chaff from the wheat, fashion from real = user needs. You seem to think sparse arrays are very important. If I = follow your reasonning, I should not listen to you because you might as = well have asked for an MP3 player instead. Yes there are stupid users = wishing for useless things (wishing for a useless thing is stupid in my = opinion). But overall I think they will wish for usefull improvements. = (can we call it an improvement if its not usefull? ;-) )=20 >> I think function is inherently tied to form and it is wrong to >> consider the two seperatly. In my experience, just making the >> application prettier solves a lot of problems. Welcome to the real >> world. >What particular problems?=20 Well it has been proven that making the user experience more pleasant = (most of the time with minimal effort) prevents a lot of user errors. If = you use the aesthetic aspect to convey meaning then you have = practicability and beauty.Isnt it like marriage? You could have a nice = wife but add beauty and this "until death do us apart" thing will seem = much more bearable. >The subject is that some people consider J's set of GUI tools >appropriate to make powerful, convenient etc. user interface, others >consider it does not contain even half of needed means Finally, about J. With the tremendous work on OCX support (sucking up to = the developpers hey?) there is very little that we cannot do. Still I = think J GUI should never stop to improve. I think efforts should be = directed toward areas where little effort yields great results. The = keyword here is evolution, not revolution. Sylvain Hamel ------=_NextPart_000_004C_01BEE7D4.D3F9D080 charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
>> Are we not discussing the software that we = use? That=20 would mean we do
>> not know what we want ????
>Yes that = would.
>For
>example if somebody right now has = some very=20 slight keenness, say, in
>music he will rather vote for adding = musical=20 capabilities to J rather
>than for adding sparse arrays=20 support
 
I'd like to believe I have enough judgement to know = what's=20 good for me. In general, people wish what would be usefull to them. Like = all=20 things in life you have to separate the chaff from the wheat, = fashion from=20 real user needs. You seem to think sparse arrays are very important. If = I follow=20 your reasonning, I should not listen to you because you might as = well have=20 asked for an MP3 player instead. Yes there are stupid users wishing = for=20 useless things (wishing for a useless thing is stupid in my opinion). = But=20 overall I think they will wish for usefull improvements. (can we = call it an=20 improvement if its not usefull? ;-) )
 
>> I think function is inherently tied to form = and it is=20 wrong to
>> consider the two seperatly. In my experience, just = making=20 the
>> application prettier solves a lot of problems. Welcome = to the=20 real
>> world.
>What particular problems?
 
Well it has been proven that making the user = experience more=20 pleasant (most of the time with minimal effort) prevents a lot of user = errors.=20 If you use the aesthetic aspect to convey meaning then you have=20 practicability and beauty.Isnt it like marriage? You could have a = nice wife=20 but add beauty and this "until death do us apart" thing will seem = much more=20 bearable.
 
>The subject is that some people consider J's set = of GUI=20 tools
>appropriate to make powerful, convenient etc. user = interface,=20 others
>consider it does not contain even half of needed=20 means
 
Finally, about J. With the tremendous work on = OCX=20 support (sucking up to the developpers hey?) there is very little = that we=20 cannot do. Still I think J GUI should never stop to improve. I think = efforts=20 should be directed toward areas where little effort yields great = results. The=20 keyword here is evolution, not revolution.
 
Sylvain = Hamel
------=_NextPart_000_004C_01BEE7D4.D3F9D080-- -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Mon Aug 16 11:15:40 1999 From: Baker Stuart Subject: RE: Jforum: Re : Re : GUI again Date: Mon, 16 Aug 1999 16:15:16 -0000 My two penn'orth - hope it's relevant: > ---------- > <> > Stuart Baker > > Hyder IT - Cardiff, UK > > Stuart.Baker@Hyder.com > Tel +44 (0) 1222 444205 begin 600 j gui.txt M4&5O<&QE(&1I2!D96%L('=I M=&@@=&AE(&UU;F1A;F4@86YD(&]B=FEO=7,N(%1H92!R96%L(&1I9F9E6UB;VQI8R!U2X-"@T*1U5)(&]B:F5C=',@97AP;W-E('1H M96ER("=M96%N:6YG2!B92!S=')E=&-H960[(&)U M='1O;G,@9&5P7-T96TN(%1H M97)E(&ES(&YO(&%P<&%R96YT("=M;V1A;&ET>2<@+2!T:&4@=7-E6]U('1Y<&4@82!S=')I;F<@;V8@6]U(&5X<&QI8VET;'D@"UC;VQO=7)I M;F<@8V%U'1E;F0@=&AE(&-U6UB;VQS(&%R;W5N9"!T;R!D M979E;&]P(&$@=')E92P@2!G:79E(&$@9&%M;B!A8F]U="!T:&]S92!B M96QL6UB;VQS(&AA9"!A(&-E2X-"@T*22!O;F-E(&AA9"!T;R!P2`H1'EA;&]G($%03"`M(&QO=',@;V8@979E;G1S+"!C86QL M8F%C:W,@971C*2X@5&AE;B!)('=E;G0@8F%C:R!T;R!P;&%Y:6YG($%R86-H M;FED("AA(&9R=7-T6%L;V<@1U5)(&EN=&5R9F%C92!W;W5L M9"!L970@;64@)W!L87D@8V%R9',G('=I=&@@=&AE(&9L:6=H="UP86ER2!R97!R97-E;G1I;F<@=&AE(&)A2!S965I;F<@=&AE;2!E>'!L;V1E M(&]N(&-O;G1A8W0L(&]R('-L:61E(&%W87D@9G)O;2!E86-H(&]T:&5R/R!) M)W9E('-E96X@8G5I;&1I;F2!E;&%B;W)A=&5D('=H M96X@=&AE('!R;V=R86T@:7,@97AP;&EC:71L>2!R=6XN(%1H97)E(&ES('-O M(&UU8V@@8V]M<'5T:6YG('!E6)E(&$@)U9I9F]R($HG('=O=6QD(&%L Subject: Jforum: Survey: National Use Characters Date: Mon, 16 Aug 1999 11:00:36 -0400 charset="Windows-1252" 0. Did you know that J has special support for the "national use characters"? (For example, @ can be entered as AT.) 1. If yes, would you be distressed if this support was removed? -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Mon Aug 16 12:43:24 1999 From: Subject: RE: Jforum: Survey: National Use Characters Date: Mon, 16 Aug 1999 18:43:16 +0200 Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by plus.interlog.com id MAA14812 I would not be distressed at all./Paul >-----Message d'origine----- >De: Roger Hui [SMTP:RHui@interlog.com] >Date: lundi 16 ao�t 1999 17:01 >�: J Forum >Objet: Jforum: Survey: National Use Characters >1. If yes, would you be distressed if this support was removed? > -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Mon Aug 16 13:38:41 1999 From: "Chris Burke" References: <002e01bee65a$d7e1d7e0$8fd23dcf@jat> Subject: Re: Jforum: Sparse Arrays Date: Mon, 16 Aug 1999 13:15:39 -0400 Here is a description of the proposed sparse array extension: http://www.jsoftware.com/pubs/sparse.htm (The answer to the question is that the fill element would be returned.) ----- Original Message ----- From: J T Sent: Saturday, August 14, 1999 9:42 AM Subject: Jforum: Sparse Arrays I'm just a bit curious about how Sparse Arrays might be implemented. I think I generally understand the idea behind sparse arrays. But, how would the interpreter handle the case where information is requested from an element which has no data? Would that be an error, or would the interpreter simply return some fill element? If anyone has a clear understanding of how Sparse Arrays would actually work, I would love to listen. Regards, Joe -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Mon Aug 16 15:13:36 1999 Date: Mon, 16 Aug 1999 14:51:48 -0400 From: Brian Bambrough Subject: Re: Jforum: Survey: National Use Characters References: <002101bee803$0a037fe0$5ccb22cf@f3nbp> I did not know such a thing existed and would not be distressed if it was removed. Roger Hui wrote: > 0. Did you know that J has special support for the "national use characters"? > (For example, @ can be entered as AT.) > > 1. If yes, would you be distressed if this support was removed? > > -------------------------------------------------------------------------------- > J Forum: for information about this list, see http://www.jsoftware.com/forum.htm -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Mon Aug 16 15:15:53 1999 From: mellemf@nimo.com X-Lotus-FromDomain: NMPC Date: Mon, 16 Aug 1999 13:37:26 -0400 Subject: Re: Jforum: Enigma demo slower on faster machine? Content-type: text/plain; charset=us-ascii Content-Disposition: inline When my NT desk top at work got pathetically slow (its a 400 Mhz) the support group helped me clean out the Internet cookie dust. Look in your Internet support software to clear your Cookies. -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Mon Aug 16 15:43:23 1999 From: "Seymour Glass" Subject: RE: Jforum: Survey: National Use Characters Date: Mon, 16 Aug 1999 15:22:59 -0400 charset="Windows-1252" Importance: Normal In-Reply-To: <002101bee803$0a037fe0$5ccb22cf@f3nbp> 0. No. 1. No. Henry Rich > -----Original Message----- > From: owner-jsoftware@lists.interlog.com > [mailto:owner-jsoftware@lists.interlog.com]On Behalf Of Roger Hui > Sent: Monday, August 16, 1999 11:01 AM > To: J Forum > Subject: Jforum: Survey: National Use Characters > > > 0. Did you know that J has special support for the "national use > characters"? > (For example, @ can be entered as AT.) > > 1. If yes, would you be distressed if this support was removed? > > > > ------------------------------------------------------------------ > -------------- > J Forum: for information about this list, see > http://www.jsoftware.com/forum.htm -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Mon Aug 16 16:03:39 1999 Subject: Re: Jforum: Survey: National Use Characters In-Reply-To: <002101bee803$0a037fe0$5ccb22cf@f3nbp> from Roger Hui at "Aug 16, 1999 11:00:36 am" From: Keith Smillie Date: Mon, 16 Aug 1999 13:36:23 -0600 (MDT) Roger, I've never used the National Use Characters and can't foresee using them. Keith Smillie -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Mon Aug 16 16:18:55 1999 From: Andrew Nikitin Subject: RE: Jforum: Sparse Arrays Date: Mon, 16 Aug 1999 15:18:26 -0500 > -----Original Message----- > From: Chris Burke [SMTP:cdburke@interlog.com] > Sent: Monday, August 16, 1999 12:16 > > Here is a description of the proposed sparse array extension: > > http://www.jsoftware.com/pubs/sparse.htm > >From the text I understood that monadic , returns dense result on sparse arguments. What are the reasons for this? nsg -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Mon Aug 16 16:42:00 1999 From: "Roger Hui" References: Subject: Re: Jforum: Sparse Arrays Date: Mon, 16 Aug 1999 16:40:58 -0400 charset="iso-8859-1" Andew Nitkitin writes on Monday, August 16: >From the text I understood that monadic , returns dense result on sparse > arguments. What are the reasons for this? The monad , on sparse arguments returns a sparse (vector) result. -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Mon Aug 16 16:52:43 1999 Date: Mon, 16 Aug 1999 21:49:00 +0100 From: Stuart Baker Subject: Re: Jforum: Survey: National Use Characters References: <002101bee803$0a037fe0$5ccb22cf@f3nbp> 0. had forgotten, but remembered with this message. 1. no - if something more interesting were to take its place... do you have something in mind? Roger Hui wrote: > 0. Did you know that J has special support for the "national use characters"? > (For example, @ can be entered as AT.) > > 1. If yes, would you be distressed if this support was removed? > > -------------------------------------------------------------------------------- > J Forum: for information about this list, see http://www.jsoftware.com/forum.htm -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Mon Aug 16 16:59:29 1999 From: Andrew Nikitin Subject: RE: Jforum: Sparse Arrays Date: Mon, 16 Aug 1999 15:59:03 -0500 > -----Original Message----- > From: Roger Hui [SMTP:RHui@interlog.com] > Sent: Monday, August 16, 1999 15:41 > > The monad , on sparse arguments returns a sparse (vector) result. > Ok, this is an answer to my question, but then how would you interpret the following text from "Further examples" part: +/ , s NB. total revenue |limit error NB. the expression failed on ,s because it would | +/ ,s NB. have required a vector of length 2.745e10 (here s is considered to have 1e5 non-fill elements) -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Mon Aug 16 17:38:32 1999 From: "John D. Baker" Subject: Re: Jforum: Survey: National Use Characters Date: Mon, 16 Aug 1999 12:44:03 -0400 charset="Windows-1252" In a word NO. I've never used NUC's and doubt I ever wll. John Baker -----Original Message----- From: Roger Hui Date: Monday, August 16, 1999 12:22 PM Subject: Jforum: Survey: National Use Characters >0. Did you know that J has special support for the "national use characters"? >(For example, @ can be entered as AT.) > >1. If yes, would you be distressed if this support was removed? > > > >------------------------------------------------------------------------------- - >J Forum: for information about this list, see http://www.jsoftware.com/forum.htm > -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Mon Aug 16 17:52:12 1999 From: "Roger Hui" References: Subject: Re: Jforum: Sparse Arrays Date: Mon, 16 Aug 1999 17:48:19 -0400 charset="iso-8859-1" Andrew Nikitin writes on Monday, August 16: > > The monad , on sparse arguments returns a sparse (vector) result. > > Ok, this is an answer to my question, but then how would you interpret > the following text from "Further examples" part: > +/ , s NB. total revenue > |limit error NB. the expression failed on ,s because it would > | +/ ,s NB. have required a vector of length 2.745e10 > > (here s is considered to have 1e5 non-fill elements) The shape of s is 20 50 1000 75 366, and the shape of ,s would be */$s or 2.745e10 . These would be true no matter how many non-sparse elements are in s . Generally speaking, f -: f&.$. for any function f. In particular, any identities on shape that you know and love for ordinary (dense) arrays, will apply to sparse arrays. -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Mon Aug 16 19:40:50 1999 From: "David Vincent-Jones" References: <002e01bee65a$d7e1d7e0$8fd23dcf@jat> <00c501bee80b$0c021560$be99fea9@t500> Subject: Re: Jforum: Sparse Arrays Date: Mon, 16 Aug 1999 16:18:00 -0700 Very interesting descriptive file Chris; just a couple of questions for now. Given data in a sparse form: y1 x1 z1 y5 x5 z5 yn xn zn ..... ie. just a set of random points How do I form the initial Sparse Array? Must I grid the data initially and then convert to sparse format? Secondly; do you see any way of graphically representing the surface of a Sparse Array through the Plot functions? -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Mon Aug 16 19:57:37 1999 Date: Mon, 16 Aug 1999 19:56:55 -0400 From: Joe Kunkel Subject: Re: Jforum: Sparse Arrays References: <002e01bee65a$d7e1d7e0$8fd23dcf@jat> <00c501bee80b$0c021560$be99fea9@t500> <01a401bee83d$c9b84b60$170881ce@desktop> I am beginning to get it. As far as I read the deffinition this is not an array with missing data or with random known data as suggested below. The values at the points not in the sparse form are known but are a constant (= 0). This is more a way of dealing with simplifying redundancies in matrices than dealing with incomplete arrays. I can see practical uses such as in large design matrices in which whole sections of the X'X matrix are independent and thus have zero covariances in those offdiagonal sectors. The sparse array approach might give a useful intermediate form which could help format output more easily. Joe Kunkel David Vincent-Jones wrote: > > Very interesting descriptive file Chris; just a couple of questions for now. > > Given data in a sparse form: > y1 x1 z1 > y5 x5 z5 > yn xn zn ..... ie. just a set of random points > How do I form the initial Sparse Array? > Must I grid the data initially and then convert to sparse format? > > Secondly; do you see any way of graphically representing the surface > of a Sparse Array through the Plot functions? > > -------------------------------------------------------------------------------- > J Forum: for information about this list, see http://www.jsoftware.com/forum.htm -- -------------------- Joseph G. Kunkel, Professor Biology Department joe@bio.umass.edu University of Massachusetts http://www.bio.umass.edu/biology/kunkel Amherst MA 01003 -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Mon Aug 16 20:48:20 1999 From: "Chris Burke" References: <19990815140738.AAA104252@default> Subject: Re: Jforum: Past Posts? Date: Mon, 16 Aug 1999 19:08:23 -0400 Not yet, I'm afraid. ----- Original Message ----- From: Ajoy Victor Sent: Sunday, August 15, 1999 10:02 AM Subject: Jforum: Past Posts? > Is there a way to access past posts? > > Daniel > > > -------------------------------------------------------------------------- ------ > J Forum: for information about this list, see http://www.jsoftware.com/forum.htm > -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Tue Aug 17 02:35:33 1999 From: M.Day@fscii.maff.gov.uk Tue, 17 Aug 1999 07:31:23 +0100 Tue, 17 Aug 1999 07:28:20 +0100 Tue, 17 Aug 1999 07:31:04 +0100 Date: Tue, 17 Aug 1999 07:31:04 +0100 Content-Identifier: m1210817073052aa Alternate-Recipient: Allowed In-Reply-To: Subject: RE: Jforum: Survey: National Use Characters 0 - Yes, but, like others, I'd forgotten. This has reminded me of their existence. 1 - No. I've never used them. While I still find native J difficult to read, APL-like mnemonics were always a pain. Mike Day Roger Hui wrote > Survey: National Use Characters > 0. Did you know that J has special support for the "national use > characters"? > (For example, @ can be entered as AT.) > 1. If yes, would you be distressed if this support was removed? -------------------------------------------------------------------------------- J Forum: for information about this list, see http://www.jsoftware.com/forum.htm From owner-jsoftware@lists.interlog.com Tue Aug 17 05:42:41 1999 From: k-list@iname.com Date: Tue, 17 Aug 1999 05:41:23 -0400 (EDT) Content-Type: Text/Plain Subject: Jforum: #include "bill lam" I tried to write a invoice entry screen with tab control - tab 0 for header information and tab 1 (with a jwgrid) for item lines. There are 3 parent windows but all of them must be inside the same coclass, otherwise event-handling will be difficult. I want to break it into 3 file scripts, and the join them together like this, ---------------------------- coclass 'invoice' #include './header.ijs' NB. #include macro #include './itemlines.ijs' NB. #include macro