💾 Archived View for blitter.com › apl-books › APLSE › www.sigapl.org › Archives › waterloo_archive ›… captured on 2024-08-18 at 19:33:53.
View Raw
More Information
⬅️ Previous capture (2022-07-17)
-=-=-=-=-=-=-
From owner-jsoftware@lists.interlog.com Tue Jun 1 08:31:56 1999
From: mellemf@nimo.com
X-Lotus-FromDomain: NMPC
Date: Tue, 1 Jun 1999 08:13:30 -0400
Subject: Re: Jforum: J 4.03 Beta (8th version) 404 Not Found Error
Content-type: text/plain; charset=us-ascii
Content-Disposition: inline
I am getting the following error: (I hope its just temporary ; I got all other
versions ok)
404 Not Found Error
Sorry! You have reached a page that is unavailable or does not exist. Things you
can do from here:
1.Use any of the menu items on this page to continue working within Interlog
Internet Services.
2.Check your URL for spelling or capitalization mistakes and/or try it again.
Interlog may host the site on our server, however we are not responsible for how
the site functions.
Copyright 1998 - Interlog Internet Services
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Tue Jun 1 09:06:38 1999
Date: Tue, 01 Jun 1999 08:52:19 -0400
From: Eric Iverson <eiverson@interlog.com>
Organization: Iverson Software Inc.
Subject: Re: Jforum: J 4.03 Beta (8th version) 404 Not Found Error
References: <85256783.0043ED53.00@mail.nimo.com>
> 404 Not Found Error (for J 4.03a Beta)
The beta period has expired and the beta is no longer available. The
official release is available at www.jsoftware.com.
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Tue Jun 1 09:07:54 1999
From: "Randy MacDonald" <randy@godin.on.ca>
References: <85256783.0043ED53.00@mail.nimo.com>
Subject: Re: Jforum: J 4.03 Beta (8th version) 404 Not Found Error
Date: Tue, 1 Jun 1999 08:44:21 -0400
charset="iso-8859-1"
If the reason for this page disappearing is the expiration of all beta
versions at
00:00h 1 June, it might be more friendly in the future to replace the beta
with a
"Sorry, Beta Expired, why not look at the full version." page.
----- Original Message -----
From: <mellemf@nimo.com>
Sent: June 1, 1999 8:13 AM
Subject: Re: Jforum: J 4.03 Beta (8th version) 404 Not Found Error
> I am getting the following error: (I hope its just temporary ; I got all
other
> versions ok)
> 404 Not Found Error
> Sorry! You have reached a page that is unavailable or does not exist.
Things you
> can do from here:
> 1.Use any of the menu items on this page to continue working within
Interlog
> Internet Services.
>
> 2.Check your URL for spelling or capitalization mistakes and/or try it
again.
>
> Interlog may host the site on our server, however we are not responsible
for how
> the site functions.
>
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Tue Jun 1 13:56:02 1999
From: "Seymour Glass" <glasss@mindspring.com>
Subject: Jforum: Private to Martin Neitzel
Date: Tue, 1 Jun 1999 13:31:33 -0400
charset="iso-8859-1"
Importance: Normal
Martin,
Your mail server (spog.gaertner.de) is blocking mail from
Mindspring, my ISP (IP address 207.69.200.32), so I can't
reply directly.
But I will try again to upload to CVS, and I will
make one of the changes you suggested about the z locale
(after I figure out which one is best for me).
Henry Rich
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Tue Jun 1 14:16:08 1999
Date: Tue, 1 Jun 1999 14:03:05 -0400 (EDT)
From: Olivier Lefevre <lefevre@panix.com>
Subject: Re: Jforum: DICTIONARIES IN J
In-Reply-To: <3750A609.701A@interlog.com>
> K and Python dictionaries are similar to J locales.
Yes. The suggestions put forth here are of the OO style, where you have
to build dictionaries in userland, so to speak. I use Java collections
all the time and once you have have tried the K approach it just doesn't
cut it. I am not yet very familiar with locales but I believe they offer
a flat namespace, so they wouldn't get you very far (i.e., only 1 level-
deep maps). Is that correct or have J locales been upgraded since last
time I enquired?
Regards,
-- O.L.
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Tue Jun 1 15:15:09 1999
Date: Tue, 1 Jun 1999 15:02:50 -0400 (EDT)
From: Olivier Lefevre <lefevre@panix.com>
Subject: Re: Jforum: DICTIONARIES IN J
In-Reply-To: <3754299E.6285CB14@nassau.cv.net>
> ...once you have have tried the K approach, (Java collections) just (don't) cut it.
Yes.
-- O.L.
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Tue Jun 1 15:20:13 1999
Date: Tue, 1 Jun 1999 15:07:44 -0400 (EDT)
From: Olivier Lefevre <lefevre@panix.com>
Subject: Re: Jforum: DICTIONARIES IN J
In-Reply-To: <3754299E.6285CB14@nassau.cv.net>
I would also like to point out that while "nameless" arrays are well-
suited to numerical tasks, maps are a better match to GUIs, relational
databases and other "business", as opposed to engineering applications.
Doesn't Dyadic APL already offer something very similar to K's 1st-class
namespaces (hence built-in maps)? I think so, and they leverage it in
GUI building. At least that is my recollection of a Dyadic demo a while
back.
-- O.L.
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Tue Jun 1 17:58:21 1999
Date: Tue, 01 Jun 1999 14:42:38 -0400
From: Jake <jm@optonline.net>
Subject: Re: Jforum: DICTIONARIES IN J
References: <Pine.GSU.4.05.9906011358001.10144-100000@panix3.panix.com>
>> I use Java collections all the time and once you have
>> have tried the K approach it just doesn't cut it.
Sorry, for clarification, does "it" refer to the "the K approach" or is the following intended ?
...once you have have tried the K approach, (Java collections) just (don't) cut it.
Sorry for being so dense.
Olivier Lefevre wrote:
>
> > K and Python dictionaries are similar to J locales.
>
> Yes. The suggestions put forth here are of the OO style, where you have
> to build dictionaries in userland, so to speak. I use Java collections
> all the time and once you have have tried the K approach it just doesn't
> cut it. I am not yet very familiar with locales but I believe they offer
> a flat namespace, so they wouldn't get you very far (i.e., only 1 level-
> deep maps). Is that correct or have J locales been upgraded since last
> time I enquired?
>
> Regards,
>
> -- O.L.
>
> --------------------------------------------------------------------------------
> J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
--
- Jake Moskowitz
jm@optonline.net
(jm@nassau.cv.net will still work)
(516) 889-4566
moskowitz_jacob@jpmorgan.com
(212) 235-3787
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Tue Jun 1 18:05:58 1999
From: mellemf@nimo.com
X-Lotus-FromDomain: NMPC
Date: Tue, 1 Jun 1999 09:07:30 -0400
Subject: Re: Jforum: J 4.03 beta test period is over; returning from a
long weekend & reading alot of email.
Content-type: text/plain; charset=us-ascii
Content-Disposition: inline
I should have finished reading the e-mail before crying "The Sky is Falling"
Thanks for the rapid replies.
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Tue Jun 1 18:30:16 1999
From: "Chris Burke" <cdburke@interlog.com>
References: <Pine.GSU.4.05.9906011358001.10144-100000@panix3.panix.com>
Subject: Re: Jforum: DICTIONARIES IN J
Date: Tue, 1 Jun 1999 18:23:58 -0400
charset="iso-8859-1"
> Yes. The suggestions put forth here are of the OO style, where you
have
> to build dictionaries in userland, so to speak. I use Java
collections
> all the time and once you have have tried the K approach it just
doesn't
> cut it. I am not yet very familiar with locales but I believe they
offer
> a flat namespace, so they wouldn't get you very far (i.e., only 1
level-
> deep maps). Is that correct or have J locales been upgraded since
last
> time I enquired?
J401 extended the definition of locales, and you can now specify a
locale path. A good intro to new facilities is the lab Object Oriented
Programming.
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Tue Jun 1 23:51:34 1999
Date: Tue, 01 Jun 1999 23:42:19 -0400
From: Daniel Torres <torres@SAmerica.com>
Subject: Jforum: Re: Typos
References: <19990529231539.24262.rocketmail@attach1.rocketmail.com>
Content-Type: multipart/mixed;
boundary="------------65AE7D68E4BCB5B969CF6B05"
This is a multi-part message in MIME format.
--------------65AE7D68E4BCB5B969CF6B05
Several have requested references regarding my post on devastating purchasing
effects of mispellings.
I have found no research, at all. I cannot even pinpoint how I acquired the
paradigm.
Surprisingly, however, everyone in the field of advertising that I ask,
believes the same way. I have been involved in the field of marketing and
advertising for over 14 years, including as account excutive for Leo Burnett
(I handled all of Heinz products in my region). I guess that everyone simply
takes the statement regarding typos for granted, because no one questions it.
I guess it is analogous to the claim that a typo in a resume can have
devastating effects. Or that dressing adequately at a job interview, including
not wearing socks that might sag, is crucial. I found no hard research on
those statements, either. But I am surprised with the number of people
thinking likewise.
Given the lack of hard research, it is posts like that of Don Guinn's, where he
mentions his wife's reaction to typos, which breath truth to the paradigm.
Maybe the following resume excerpts are appropriate:
1. "I demand a salary commiserate with my extensive experience."
2. "I have lurnt Word Perfect 6.0 computor and spreasheet progroms."
3. "Received a plague for Salesperson of the Year."
4. "Wholly responsible for two (2) failed financial institutions."
5. "Reason for leaving last job: maturity leave."
6. "Failed bar exam with relatively high grades."
7. "It's best for employers that I not work with people."
8. "Let's meet, so you can 'ooh' and 'aah' over my experience."
9. "You will want me to be Head Honcho in no time."
10. "Am a perfectionist and rarely if if ever forget details."
Daniel
> Daniel Torres wrote:
> >
> > The experience in the field of advertising dictates that even if the
> > probability of typos being discovered are low, one should not conclude that the
> > importance of eliminating typos is diminished. The purchasing effect on a
> > potential buyer who discovers a typo in the presentation of a product is
> > devastating.
> >
--------------65AE7D68E4BCB5B969CF6B05
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
--------------65AE7D68E4BCB5B969CF6B05--
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Wed Jun 2 00:56:06 1999
Date: Wed, 02 Jun 1999 00:47:39 -0400
From: David Ness <DNess@home.com>
Subject: Re: Jforum: Re: Typos
References: <19990529231539.24262.rocketmail@attach1.rocketmail.com> <3754A81B.8FD40F97@SAmerica.com>
I guess we have different definitions of `typo'. To me, a `typo' is
a `typographical error'---a very particular type of problem, and only
one very small corner of the field of `Errors in Language'.
To me, the examples you cite are universally errors in language
which are _not_, generally speaking, typos. I won't go thru them
one-by-one, so I'll illustrate with your first example. I doubt
`commiserate' is a typo for `commensurate', I think the (ignorant)
writer did not know the right word and used an incorrect one. He
didn't `mis-type it' he `mis-thought it'.
I would certainly own up to thinking _much more unkindly_ of errors
in language than of typos. It doesn't bother me much to see `dosen't'
that's clearly a typo. It does bother me to see `your' for `you're'
or `there' for `their' or, worse yet, `they're' etc.
Typos are `errors of fingers' and they don't bother me much, and haven't
unduly bothered most of the decision makers I happen to have dealt with.
Errors in language are `errors of the mind and intelligence' and are
in a different class.
Daniel Torres wrote:
>
> Several have requested references regarding my post on devastating purchasing
> effects of mispellings.
>
> I have found no research, at all. I cannot even pinpoint how I acquired the
> paradigm.
> ...
>
> Given the lack of hard research, it is posts like that of Don Guinn's, where he
> mentions his wife's reaction to typos, which breath truth to the paradigm.
>
Please tell me that `... breath ...' is a subtle illustration of
your point and not a telling example...
>
> Maybe the following resume excerpts are appropriate:
>
> 1. "I demand a salary commiserate with my extensive experience."
> 2. "I have lurnt Word Perfect 6.0 computor and spreasheet progroms."
> 3. "Received a plague for Salesperson of the Year."
> 4. "Wholly responsible for two (2) failed financial institutions."
> 5. "Reason for leaving last job: maturity leave."
> 6. "Failed bar exam with relatively high grades."
> 7. "It's best for employers that I not work with people."
> 8. "Let's meet, so you can 'ooh' and 'aah' over my experience."
> 9. "You will want me to be Head Honcho in no time."
> 10. "Am a perfectionist and rarely if if ever forget details."
>
> Daniel
>
> > Daniel Torres wrote:
> > >
> > > The experience in the field of advertising dictates that even if the
> > > probability of typos being discovered are low, one should not conclude that the
> > > importance of eliminating typos is diminished. The purchasing effect on a
> > > potential buyer who discovers a typo in the presentation of a product is
> > > devastating.
> > >
>
>
> Daniel Torres <torres@SAmerica.com>
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Wed Jun 2 03:06:58 1999
From: <paul_gauthier@dynasys.tm.fr>
Subject: RE: Jforum: Re: Typos
Date: Wed, 2 Jun 1999 08:52:41 +0200
Content-Transfer-Encoding: 8bit
X-MIME-Autoconverted: from quoted-printable to 8bit by plus.interlog.com id CAA27904
I am jumping into this discussion of typos for the following reason:
Being french speaking from birth, I could more easily generate typos in
english. So the birth language of a person is also of some importance.
But, in order to turn me off, typos must absolutely be repetitive and in
number. Then it is more likely that negligence is behind the repetitive
aspect of typos. (But it could be too much speed).
I will give an example of the major difference between negligence and a
simple typo. I have seen a lot of abuse of several kind in my short life
already, what always strike me is the regular pattern shown by an abuser
who always deny his abuse by giving examples like: But everybody gives a
blow to a kid once in a while...
Thus, the abuser is using the emotional potential guilt that many normal
people migth have in order to camouflage his repetitive behavior.
Conclusion: One must consider birth language and repetitive aspect of
typos, maybe we could develop a taxonomy of typos and their
implications... Regards/Paul
>-----Message d'origine-----
>De: David Ness [SMTP:DNess@home.com]
>Date: mercredi 2 juin 1999 06:48
>�: forum@jsoftware.com
>Objet: Re: Jforum: Re: Typos
>
>I guess we have different definitions of `typo'. To me, a `typo' is
>a `typographical error'---a very particular type of problem, and only
>one very small corner of the field of `Errors in Language'.
>
>To me, the examples you cite are universally errors in language
>which are _not_, generally speaking, typos. I won't go thru them
>one-by-one, so I'll illustrate with your first example. I doubt
>`commiserate' is a typo for `commensurate', I think the (ignorant)
>writer did not know the right word and used an incorrect one. He
>didn't `mis-type it' he `mis-thought it'.
>
>I would certainly own up to thinking _much more unkindly_ of errors
>in language than of typos. It doesn't bother me much to see `dosen't'
>that's clearly a typo. It does bother me to see `your' for `you're'
>or `there' for `their' or, worse yet, `they're' etc.
>
>Typos are `errors of fingers' and they don't bother me much, and haven't
>unduly bothered most of the decision makers I happen to have dealt with.
>Errors in language are `errors of the mind and intelligence' and are
>in a different class.
>
>Daniel Torres wrote:
>>
>> Several have requested references regarding my post on devastating
>>purchasing
>> effects of mispellings.
>>
>> I have found no research, at all. I cannot even pinpoint how I acquired
>>the
>> paradigm.
>> ...
>>
>> Given the lack of hard research, it is posts like that of Don Guinn's,
>>where he
>> mentions his wife's reaction to typos, which breath truth to the paradigm.
>>
>
>Please tell me that `... breath ...' is a subtle illustration of
>your point and not a telling example...
>
>>
>> Maybe the following resume excerpts are appropriate:
>>
>> 1. "I demand a salary commiserate with my extensive experience."
>> 2. "I have lurnt Word Perfect 6.0 computor and spreasheet progroms."
>> 3. "Received a plague for Salesperson of the Year."
>> 4. "Wholly responsible for two (2) failed financial institutions."
>> 5. "Reason for leaving last job: maturity leave."
>> 6. "Failed bar exam with relatively high grades."
>> 7. "It's best for employers that I not work with people."
>> 8. "Let's meet, so you can 'ooh' and 'aah' over my experience."
>> 9. "You will want me to be Head Honcho in no time."
>> 10. "Am a perfectionist and rarely if if ever forget details."
>>
>> Daniel
>>
>> > Daniel Torres wrote:
>> > >
>> > > The experience in the field of advertising dictates that even if the
>> > > probability of typos being discovered are low, one should not conclude
>>that the
>> > > importance of eliminating typos is diminished. The purchasing effect
>>on a
>> > > potential buyer who discovers a typo in the presentation of a product
>>is
>> > > devastating.
>> > >
>>
>>
>> Daniel Torres <torres@SAmerica.com>
>
>-----------------------------------------------------------------------------
>---
>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 Jun 2 05:17:24 1999
From: Baker Stuart <stuart.baker@hyder.com>
Subject: RE: Jforum: Re: Typos
Date: Wed, 2 Jun 1999 10:06:20 -0000
My two cents' worth...
With typos, errors of linguistic orthodoxy and other noticeable
deviations it occurs to me that - it may be roughly just those occurrences
that we can counter by greater interpretive effort that are even noticed at
all. And if we can say 'that's a mistake' and recover the intended meaning,
then that meaning has exactly not been affected. What has happened is that
the communication has acquired additional 'meaning' or semantic content.
Like swatting vainly at a fly that is buzzing round our head, it is this
extra unwanted attentional burden that is objectionable.
If we think of 'mistakes' as simply part of the broader context of
issues of style, then they are immediately accompanied by other
meta-communications which can be irritating, enchanting, obscure,
inspirational, ... you name it. Personally, I have always found the rather
puritanical terseness of the original set of J documentation 'difficult' -
the labs and the Primer between them have helped me enormously in the past
months.
Looking at the language itself, I find it (personally) much more
difficult in J to make 'typos', in the sense of mistakes which are not
interpretable in the language (at least lexically) and stick out visually.
The symbol-space appears to be so dense compared to conventional languages
that almost everything has a meaning.
This does not mean I am not making mistakes - far from it: in my
current phase of learning I have the curious sensation of walking on eggs
all the time, hoping none of them break. I remember back in the early 70's a
design for a Fortran compiler was published with a front-end that tolerated
spelling and some syntax mistakes. As far as I know nothing came of it - I
wonder why? I reckon it would be really difficult to do the same for J,
apart from dealing with mis-spelled control structures.
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Wed Jun 2 06:22:15 1999
Date: Wed, 02 Jun 1999 13:04:35 +0300
From: Anssi Seppala <anssi.seppala@enease.fi>
Subject: Jforum: HTML publish is not working in J403a
In-Reply-To: <c=FR%a=_%p=DynaSys_-_Schilt%l=DYNASYS01-990602065241Z-9@ma
il.dynasys.tm.fr>
I just loaded J403a and tried to make html publish but it does not work!
Anssi
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Wed Jun 2 07:11:49 1999
From: "Chris Burke" <cdburke@interlog.com>
References: <4.1.19990602125137.00bcd100@pop.kolumbus.fi>
Subject: Re: Jforum: HTML publish is not working in J403a
Date: Wed, 2 Jun 1999 06:49:06 -0400
charset="iso-8859-1"
It works for me. What did you try and with what result? Did it work in
the beta systems?
----- Original Message -----
From: Anssi Seppala <anssi.seppala@enease.fi>
Sent: Wednesday, June 02, 1999 6:04 AM
Subject: Jforum: HTML publish is not working in J403a
> I just loaded J403a and tried to make html publish but it does not
work!
>
> Anssi
>
> --------------------------------------------------------------------
------------
> 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 Jun 2 07:42:04 1999
Date: Wed, 02 Jun 1999 06:29:36 -0500
From: Don Guinn <donguinn@hal-pc.org>
Subject: Re: Jforum: Re: Typos
References: <0488C2686254D111A5C70000F847E8C5030DB32E@LUSIS_EXCH_1>
Typos are easy to make. They are also easy to correct. The question that comes to
my mind when I see a lot of typos is, "If I can find so many mistakes in the product
so quickly, how many more mistakes exist that I haven't found?"
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Wed Jun 2 08:23:41 1999
Date: Wed, 02 Jun 1999 15:09:10 +0300
From: Anssi Seppala <anssi.seppala@enease.fi>
Subject: Re: Jforum: HTML publish is not working in J403a
In-Reply-To: <003601beace7$4556eca0$430114d1@Pcdburke>
References: <4.1.19990602125137.00bcd100@pop.kolumbus.fi>
Content-Transfer-Encoding: 8bit
X-MIME-Autoconverted: from quoted-printable to 8bit by plus.interlog.com id IAA10690
Yes, it did work with beta systems and I am confused. After installation I
select from menu: html publish and nothing happens. Trying the same from PM
it shows error:
Error in pprojectform_pub_button
domain error: wd
msg m =.'publishing:',y.{.~-PATHSEP_j_ i.~|.y.
Anssi
At 06:49 2.6.1999 -0400, you wrote:
>It works for me. What did you try and with what result? Did it work in
>the beta systems?
>
>----- Original Message -----
>From: Anssi Seppala <anssi.seppala@enease.fi>
>To: <forum@jsoftware.com>
>Sent: Wednesday, June 02, 1999 6:04 AM
>Subject: Jforum: HTML publish is not working in J403a
>
>
>> I just loaded J403a and tried to make html publish but it does not
>work!
>>
>> Anssi
>>
>> --------------------------------------------------------------------
>------------
>> 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
-------------------------------------------------------------------
Anssi Sepp�l�, TkT
S�hk�markkinat ja s�hk�njakelu: Sovellukset ja ohjelmistot
e-mail: anssi.seppala@enease.fi tel: +358 9 45400550
Enease Oy fax: +358 9 45400551
Mannerheimintie 66 A mob: +358 400 760077
FIN - 00260 HELSINKI, Finland http://www.kolumbus.fi/enease/
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Wed Jun 2 08:23:44 1999
Date: Wed, 02 Jun 1999 15:09:45 +0300
From: Anssi Seppala <anssi.seppala@enease.fi>
Subject: Re: Jforum: HTML publish is not working in J403a
Yes, it did work with beta systems and I am confused. After installation I
select from menu: html publish and nothing happens. Trying the same from PM
it shows error:
Error in pprojectform_pub_button
domain error: wd
msg m =.'publishing:',y.{.~-PATHSEP_j_ i.~|.y.
Anssi
At 06:49 2.6.1999 -0400, you wrote:
>It works for me. What did you try and with what result? Did it work in
>the beta systems?
>
>----- Original Message -----
>From: Anssi Seppala <anssi.seppala@enease.fi>
>To: <forum@jsoftware.com>
>Sent: Wednesday, June 02, 1999 6:04 AM
>Subject: Jforum: HTML publish is not working in J403a
>
>
>> I just loaded J403a and tried to make html publish but it does not
>work!
>>
>> Anssi
>>
>> --------------------------------------------------------------------
>------------
>> 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 Wed Jun 2 09:25:11 1999
From: k-list@iname.com
Date: Wed, 2 Jun 1999 09:15:05 -0400 (EDT)
Content-Type: Text/Plain
Subject: Jforum: wd & control id
wd commands has no concept of locale. Suppose 2 forms have
text controls of the same id name (say text1), which form's
control will be referenced in the line?
wd 'set text1 abc'
---------------------------------------------------
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 Jun 2 11:48:26 1999
Date: Wed, 02 Jun 1999 11:34:21 -0400
From: Brian Bambrough <b.bambrough@worldnet.att.net>
Subject: Jforum: j403a: minor errors
Here are a couple of minor errors in j403a:
1) When the system is executing a script that takes a long time, the
cursor does not turn into an hour-glass. There does not seem to be
any visual cue as to when the script has finished. Before I started
putting in the sentence, smoutput "finished", in the code, the best I
could do was to hit Ctrl Tab every few seconds.
2) There is a typo in the Dictionary. It is in the printed Dictionary
also:
Example for nub sieve in the Dictionary:
y=: 8 1 8 2 8 1 7 2
~: y
8 1 2 7
Should be ~.y
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Wed Jun 2 18:14:20 1999
Date: Wed, 02 Jun 1999 16:59:56 +0300
From: Anssi Seppala <anssi.seppala@enease.fi>
Subject: Re: Jforum: HTML publish is not working in J403a
In-Reply-To: <003601beace7$4556eca0$430114d1@Pcdburke>
References: <4.1.19990602125137.00bcd100@pop.kolumbus.fi>
Now I am really confused! HTML publish works on my Laptop WIN95 where J is
installed in c:\j403. HTML is NOT working on my destop NT4SP3 where J is
installed in d:\j403.
I am not shure if I did test beta systems' HTML on my desktop.
Anssi
At 06:49 2.6.1999 -0400, you wrote:
>It works for me. What did you try and with what result? Did it work in
>the beta systems?
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Wed Jun 2 18:19:35 1999
From: "Chris Burke" <cdburke@interlog.com>
References: <990602091505CI.21731@weba2.iname.net>
Subject: Re: Jforum: wd & control id
Date: Wed, 2 Jun 1999 09:59:42 -0400
charset="iso-8859-1"
The active form (wd 'psel myform').
Moreover, events in any form are signalled in that form's locale.
See Help|User|Window Driver: Entering Information and Form Locales.
----- Original Message -----
From: <k-list@iname.com>
Sent: Wednesday, June 02, 1999 9:15 AM
Subject: Jforum: wd & control id
> wd commands has no concept of locale. Suppose 2 forms have
> text controls of the same id name (say text1), which form's
> control will be referenced in the line?
> wd 'set text1 abc'
>
>
>
> ---------------------------------------------------
> Get free personalized email at http://www.iname.com
>
> --------------------------------------------------------------------
------------
> 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 Jun 3 02:32:04 1999
From: bszuch@wsa-fincon.com.au (WSA-fincon--bill Szuch)
Subject: Jforum: J and Excel
Date: Thu, 3 Jun 1999 16:04:13 +1000
Organization: WSA Financial Consulting Pty Ltd
APPLICATION ERROR
I have incurred an "Application Error" while running J as client and Excel as server.
After loading J403a I carried out the following:
load 'system\examples\ole\excel\xlutil.ijs'
xlopen ''
xlshow ''
xlcmd 'wb add'
xlquit ''
Everything seemed to work until the final statement.
On executing xlquit '' and still with Excel open I was thrown out of J with the following message:
Application Error
j.exe
Exception access violation: 0xc00000005. Address 0x0045d716
I also get the same problem with using xlexit ''
When I try the same exercise again, however, this time closing Excel and then using xlquit of xlexit I do not have any problems.
This problem was not encountered in J402a which is running OK.
Is one solution at this stage to exit Excel before using xlquit or xlexit ?
I would appreciate some help in running Excel under J as I find it very useful for the type of work I do.
NB.
Using Windows NT4.0
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Thu Jun 3 03:25:37 1999
>Received: from pc1-jwj.oce.nl by smtp02.oce.nl (4.1/SMI-4.1)
id AA10277; Thu, 3 Jun 99 09:13:44 +0200
Date: Thu, 3 Jun 99 09:13:44 +0200
From: jwj@oce.nl (Jan Jacobs)
Subject: Jforum: error with the combination 'extended precision' and 'item ammend'?
ls,
I need to model the storage of 32 bit values in an array (SRAM) which
involves fetching and updating on it.
When running the J code below in J3.03a the system returns on positions 100
and 101 in array SRAM the value: 61967000000070006000500040003787784248761x
and on the 20 subsequent position s also not the expected values of 255x.
Shortly after this the system responds with an exception.
When running the same code in J4.03beta(1) postion 100 contains the value
021473750, also no subsequent new values. The system aborts with illegal
operation.
SRAM=:400$ <: 2x^32
cnt=:22
ptr=:100
p=: 3 : 0
while. cnt>0 do.
SRAM=:255x ptr}SRAM NB. deposit in RAM
ptr=:ptr+1
cnt=:cnt-1
end.
)
Can anyone help me with a nice temporary solution to solve this?
Thanks in advance,
Jan Jacobs.
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Thu Jun 3 07:12:33 1999
From: "Chris Burke" <cdburke@interlog.com>
References: <01BEADDC.8EC9E900.bszuch@wsa-fincon.com.au>
Subject: Re: Jforum: J and Excel
Date: Thu, 3 Jun 1999 06:59:57 -0400
charset="iso-8859-1"
Thanks for the bug report.
The problem is in the last statement of xlquit:
wd 'psel xlauto;pclose'
We'll take a look.
----- Original Message -----
From: WSA-fincon--bill Szuch <bszuch@wsa-fincon.com.au>
Sent: Thursday, June 03, 1999 2:04 AM
Subject: Jforum: J and Excel
> APPLICATION ERROR
>
> I have incurred an "Application Error" while running J as client and
Excel as server.
>
> After loading J403a I carried out the following:
>
> load 'system\examples\ole\excel\xlutil.ijs'
>
> xlopen ''
>
> xlshow ''
>
> xlcmd 'wb add'
>
> xlquit ''
>
>
> Everything seemed to work until the final statement.
>
> On executing xlquit '' and still with Excel open I was thrown out
of J with the following message:
>
> Application Error
> j.exe
> Exception access violation: 0xc00000005. Address 0x0045d716
>
> I also get the same problem with using xlexit ''
>
> When I try the same exercise again, however, this time closing Excel
and then using xlquit of xlexit I do not have any problems.
>
> This problem was not encountered in J402a which is running OK.
>
> Is one solution at this stage to exit Excel before using xlquit or
xlexit ?
>
> I would appreciate some help in running Excel under J as I find it
very useful for the type of work I do.
> NB.
> Using Windows NT4.0
>
>
> --------------------------------------------------------------------
------------
> 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 Jun 3 15:32:00 1999
From: "Oleg Kobchenko" <gccinc@usa.net>
Subject: Jforum: ANSI and OEM national encodings (1)
Date: Thu, 3 Jun 1999 22:07:10 +0300
Content-Type: multipart/mixed;
boundary="----=_NextPart_000_0041_01BEAE0D.6D178B60"
This is a multi-part message in MIME format.
------=_NextPart_000_0041_01BEAE0D.6D178B60
charset="iso-8859-1"
Picture 1
------=_NextPart_000_0041_01BEAE0D.6D178B60
Content-Type: image/bmp;
name="russ.bmp"
Content-Transfer-Encoding: base64
Content-Disposition: attachment;
filename="russ.bmp"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------=_NextPart_000_0041_01BEAE0D.6D178B60--
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Thu Jun 3 18:23:54 1999
Date: Thu, 03 Jun 1999 09:51:51 -0400
From: Cliff Reiter <reiterc@lafvax.lafayette.edu>
Subject: Re: Jforum: HTML publish is not working in J403a
Organization: Lafayette College
Content-type: text/plain; charset=us-ascii
References: <4.1.19990602125137.00bcd100@pop.kolumbus.fi>
<4.1.19990602165549.00bcf3e0@pop.kolumbus.fi>
Anssi Seppala wrote:
>
> Now I am really confused! HTML publish works on my Laptop WIN95 where J is
> installed in c:\j403. HTML is NOT working on my destop NT4SP3 where J is
> installed in d:\j403.
>
> I am not shure if I did test beta systems' HTML on my desktop.
>
> Anssi
>
> At 06:49 2.6.1999 -0400, you wrote:
> >It works for me. What did you try and with what result? Did it work in
> >the beta systems?
I also have been unable to get it to work. Running
"run" "Html Publish" menus seems to access the harddrive,
but nothing happens.
I am running NT4 SP3 too but had similar results on 2 y.o. Win95
machines.
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 Thu Jun 3 18:37:20 1999
From: "Roger Hui" <RHui@interlog.com>
References: <9906030713.AA10277@smtp02.oce.nl>
Subject: Re: Jforum: error with the combination 'extended precision' and 'item ammend'?
Date: Thu, 3 Jun 1999 11:56:52 -0400
charset="iso-8859-1"
Jan Jacob writes on Thursday, June 3:
> I need to model the storage of 32 bit values in an array (SRAM) which
> involves fetching and updating on it.
> When running the J code below in J3.03a the system returns on positions 100
> and 101 in array SRAM the value: 61967000000070006000500040003787784248761x
> and on the 20 subsequent position s also not the expected values of 255x.
> Shortly after this the system responds with an exception.
>
> When running the same code in J4.03beta(1) postion 100 contains the value
> 021473750, also no subsequent new values. The system aborts with illegal
> operation.
>
> SRAM=:400$ <: 2x^32
> cnt=:22
> ptr=:100
>
> p=: 3 : 0
> while. cnt>0 do.
> SRAM=:255x ptr}SRAM NB. deposit in RAM
> ptr=:ptr+1
> cnt=:cnt-1
> end.
> )
>
> Can anyone help me with a nice temporary solution to solve this?
> Thanks in advance,
The problem is that amend in place (x=: blah i}x) is not implemented
for extended precision arrays, but } erroneously neither forces a copy
nor even signal error. To get around this bug for the nonce, you can
force a copy yourself by inserting a ] between the =: and the rest
of the sentence, thus:
SRAM=: ] 255x ptr}SRAM
Thank you for reporting this bug.
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Thu Jun 3 18:48:33 1999
From: "Oleg Kobchenko" <gccinc@usa.net>
Subject: Jforum: ANSI and OEM national encodings
Date: Thu, 3 Jun 1999 22:06:34 +0300
charset="iso-8859-1"
Gosi was right, Russian and Icelandic use just plain 8-bit
encodings in OEM and ANSI. In fact, to the great releaf of
Andrew Nikitin, and possibly Gosi, I managed to make J editor
input and display national in ANSI -- the most useful of the 2 options,
since it may be used in .ijs windows for comments and is used
to build dialog boxes for GUI, which are usually ANSI.
To make it work in Configure/View/Font/Browse select
"Courier" -- plain non-ttf font, not Couier New. The Script
drop-down automatically selects the only available option,
which is current-national-settings-system-default. For Russian
-- it's Cyrillic. Then after closing the Browse dialog, select ANSI.
Also select ASCII for Box Chars. Now when you switch to your national
keyboard layout the editor inputs and displays the national characters
correctly. (see russ.bmp)
For this to work there should be appropriate national support
on Windows 9x/NT. It would be interesting to get feed back
from other national settings than Russian -- how it works.
To my mind, if the Browse dialog 'remembered' the Script settings
for .ttf fonts -- and acted upon them accordingly, the same would
have been possible with "Couirer New" and other multi-charset fonts.
The next step is to enable national characters in OEM -- to be able to
see both the nice boxes and the national chars. This involves
some keyboard mapping at the point of character entering, because
the keyboard layout remains ANSI -- it send the ANSI char codes to
the application, when user presses the keyboard, but the text is in OEM.
Thus, a function, like CharToOem should be called each time when the
char code is received from the keyboard event and is about to be stored
in the text buffer, if current font settings is OEM. (see russOEM.bmp)
This feature of input encoding mapping runs along the lines of
customized key combinations to assign to various actions -- both
these facilities are common in good text editors.
Once, also there was a nice idea spelt about having separate font settings
for .ijs apart from .ijx windows -- the same as we have separate font
settings
for the forms.
-----Original Message-----
From: Eric Iverson <eiverson@interlog.com>
Date: Sunday, May 30, 1999 00:34
Subject: Re: Jforum: new editor features
Oleg Kobchenko wrote:
> Windows has option for fonts to be ANSI or OEM.
> But for keyboards in national settings there is only
> the ANSI option. Prior to beta J403, J 'editor' refused
> to accept keycodes of the upper ANSI set above 127, though
> it could display them. Now it accepts. However the mapping
> is such that it does not correctly correspond to either
> ANSI or OEM.
>
> Moreover, currently the Config/Font selection does not
> preserve the 'Script' and it always remain 'Western'.
> So one cannot even select the proper charset to see the Russian
> letters at all. I suppose the same happens with Icelandic.
> However, French (Canadian) may not suffer such problems,
> since the Western charset includes French letters.
I don't understand. Possibly because my experiments with English windows
and keyboard give different result than on your configuration.
As far as I can tell the support for J402 J403 code editors of ANSI and
OEM character sets is identical (in function, though the code is
completely different). The only difference is that the J403 code editor
insists on a fixed pitch font.
If I select Courier New ANSI, I can enter code points above 127 with the
alt numeric sequence. For example, alt+0200 displays an accented cap E
in the editor. This behaves as expected. Or I can cut/paste from CharMap
or other apps.
Just realized you are probably (Russian and Icelandic) are trying to
work with multi-byte character sets. These might work in the old editor
(as it is a standard edit control), but I think it would get J a bit
confused as I'm not sure how the multibyte codes would get passed
across. There is no multi-byte support in the new code editor.
Eventualy it might be possible to have unicode support both in the
editor and in J and this would be the proper solution in the future. But
the future is not now.
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Thu Jun 3 20:49:41 1999
Date: Thu, 03 Jun 1999 19:37:34 -0500 (CDT)
From: Kip Murray <CMurray@UH.EDU>
Subject: Jforum: Spaces in Windows NT Pathnames
X-VMS-To: IN%"forum@jsoftware.com"
X-VMS-Cc: MATH1IA
Content-type: text/plain; CHARSET=US-ASCII
Can J's facilities for reading Windows NT pathnames deal with a
pathname that has a space in it?
I'm thinking, first of all, of the routine that picks up the user's
profile from the startup icon; and secondly, of names added to
scripts.ijs.
Kip Murray <CMurray@UH.EDU>
Math, University of Houston
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Fri Jun 4 11:12:25 1999
From: Martin Barghoorn <barg@cs.tu-berlin.de>
Date: Fri, 4 Jun 1999 16:57:31 +0200 (MET DST)
Subject: Jforum: J4.03a Download with Solaris-netscape not possible
X-Sun-Charset: US-ASCII
Technical hint for download;
It isn't possible to download
J403a with netscape Solaris browser
Platform type don't show 4.03a
Martin B.
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Fri Jun 4 13:31:56 1999
From: "Martin Barghoorn" <barg@cs.tu-berlin.de>
Date: Fri, 04 Jun 1999 19:19:15 +0200
Subject: Jforum: Excel&J crashed
On Thu, 3 Jun 1999 16:04:13 +1000, WSA-fincon--bill Szuch wrote:
>APPLICATION ERROR
>I have incurred an "Application Error" while running J as client and Excel as server.
>After loading J403a I carried out the following:
load 'system\examples\ole\excel\xlutil.ijs'
> xlopen ''
> xlshow ''
> xlcmd 'wb add'
> xlquit ''
>
The same APPLICATION ERROR I had,
Excel and J crashed after moving the mouse
- **********************************************************************
- Martin BARGHOORN, Computer Science, Technical University of Berlin *
- Sekr. FR 6-9, Franklinstr. 28, D-10587 Berlin, Germany *
- Mail: barg@cs.tu-berlin.de , Web: http://stat.cs.tu-berlin.de/~barg/*
- Telephon(TUB): 030 314 24 392/73571, Faksimile(TUB): 030 314 25 901 *
- **********************************************************************
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Sat Jun 5 17:37:41 1999
Date: Sat, 05 Jun 1999 17:31:50 -0400
From: Daniel Torres <torres@SAmerica.com>
Subject: Jforum: rank of results of Head vs. Behead
Content-Type: multipart/mixed;
boundary="------------B661616AED02E16BB9D8D282"
This is a multi-part message in MIME format.
--------------B661616AED02E16BB9D8D282
Is the following difference in rank of the results intentional?
${.$8 6$2
$}.$8 6$2
1
By the way, I am running J4.02.
Thanks.
Daniel
--------------B661616AED02E16BB9D8D282
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
--------------B661616AED02E16BB9D8D282--
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Sat Jun 5 17:53:17 1999
Date: Sat, 05 Jun 1999 17:47:42 -0400
From: David Ness <DNess@home.com>
Subject: Re: Jforum: rank of results of Head vs. Behead
References: <37599745.F395C0F1@SAmerica.com>
Perhaps naively, but I'd sure think so.
Head selects the leading item of its argument while Behead drops
the leading item. Given that the argument is the vector `8 6'
the head is an atom and the behead is the vector (of one item) (6)...
Daniel Torres wrote:
>
> Is the following difference in rank of the results intentional?
>
> ${.$8 6$2
>
> $}.$8 6$2
> 1
>
> By the way, I am running J4.02.
>
> Thanks.
>
> Daniel
>
> ----------------------------------------------------------------------------------------------------
>
> Daniel Torres <torres@SAmerica.com>
>
> Daniel Torres
> <torres@SAmerica.com>
> HTML Mail
> 600 Biltmore Way APT 412 Fax: 1-978.383-5817
> Coral Gables Home: 1-305.441-0369
> FL Work: 1-305.461-6829
> 33134-7529
> USA
> Additional Information:
> Last Name Torres
> First Name Daniel
> Version 2.1
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Sun Jun 6 01:35:28 1999
From: arneson@pkb.mega.net.id (arneson)
Subject: Jforum: J403a missing system\examples\winapi\winapi.ijs
Date: Sun, 6 Jun 1999 12:15:34 +0700
Content-Type: multipart/mixed; boundary="---- =_NextPart_000_01BEB018.970A7060"
------ =_NextPart_000_01BEB018.970A7060
I tried to run Lab: DLL: Using System DLLs (file examples) from J403a.
It failed on:
require 'system\examples\winapi\winapi.ijs'
because system\examples\winapi\ did not exist. Can I copy it from the J402 directory?
------ =_NextPart_000_01BEB018.970A7060
Content-Type: application/ms-tnef
Content-Transfer-Encoding: base64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------ =_NextPart_000_01BEB018.970A7060--
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Sun Jun 6 06:20:56 1999
From: "Chris Burke" <cdburke@interlog.com>
References: <01BEB018.96F20660@pkb-115.mega.net.id>
Subject: Re: Jforum: J403a missing system\examples\winapi\winapi.ijs
Date: Sun, 6 Jun 1999 06:11:18 -0400
charset="iso-8859-1"
The winapi stuff was moved to system\packages in J403, but this
reference was not updated. Thanks for pointing it out.
You could copy it from J402, or just ignore the error message and
enter:
require 'winapi'
The lab will run from then on.
----- Original Message -----
From: arneson <arneson@pkb.mega.net.id>
Sent: Sunday, June 06, 1999 1:15 AM
Subject: Jforum: J403a missing system\examples\winapi\winapi.ijs
> I tried to run Lab: DLL: Using System DLLs (file examples) from
J403a.
>
> It failed on:
>
> require 'system\examples\winapi\winapi.ijs'
>
> because system\examples\winapi\ did not exist. Can I copy it from
the J402 directory?
>
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Sun Jun 6 08:04:20 1999
From: gosi@centrum.is
Subject: Re: Jforum: J403
Date: Sun, 6 Jun 99 11:57:44 +0000
I have not noticed the new release J403 announced
in c.l.a.
/Gosi
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Sun Jun 6 11:51:47 1999
Date: Sun, 06 Jun 1999 11:42:31 -0400
From: Murray Eisenberg <murray@math.umass.edu>
Organization: Mathematics & Statistics, Univ. of Mass./Amherst
Subject: Re: Jforum: J403a missing system\examples\winapi\winapi.ijs
References: <01BEB018.96F20660@pkb-115.mega.net.id> <000d01beb004$edb9ff20$120114d1@Pcdburke>
Actually, your proposed workaround
require 'winapi'
does NOT work. What does work is typing (or changing the .ijt script to
read):
require 'system\packages\winapi\winapi.ijs'
Chris Burke wrote:
>
> The winapi stuff was moved to system\packages in J403, but this
> reference was not updated. Thanks for pointing it out.
>
> You could copy it from J402, or just ignore the error message and
> enter:
>
> require 'winapi'
>
> The lab will run from then on.
>
> ----- Original Message -----
> From: arneson <arneson@pkb.mega.net.id>
> To: <forum@jsoftware.com>
> Sent: Sunday, June 06, 1999 1:15 AM
> Subject: Jforum: J403a missing system\examples\winapi\winapi.ijs
>
> > I tried to run Lab: DLL: Using System DLLs (file examples) from
> J403a.
> >
> > It failed on:
> >
> > require 'system\examples\winapi\winapi.ijs'
> >
> > because system\examples\winapi\ did not exist. Can I copy it from
> the J402 directory?
> >
>
> --------------------------------------------------------------------------------
> J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
--
Murray Eisenberg murray@math.umass.edu
Mathematics & Statistics Dept. phone 413 549-1020 (H)
Univ. of Massachusetts 413 545-2859 (W)
Amherst, MA 01003-4515
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Sun Jun 6 19:18:36 1999
In-Reply-To: <375184A8.161C@interlog.com>
References: <990530000014B3.09947@weba4.iname.net>
Date: Sat, 5 Jun 1999 00:11:31 -0700
From: Edward Cherlin <cherlin@2cowherd.com>
Subject: Character Overloading Considered Harmful (was Re: Jforum:
double-byte & .js)
At 11:34 -0700 5/30/1999, Eric Iverson wrote:
[snip]
>The
>definition for the mapping of BSTR to J character string is to throw
>away the extra byte of each unicode character. This is fine and
>convenient in 'english' applications where the discarded bytes are
>always 0.
You've been talking to computers too long. It isn't convenient at all in
applications that use the full 'english' character set. The minimum
estimate for professional typesetting is about 500 characters for English,
even without math, which pushes it up to about 1000.
Your method (indeed any method using any of the 8-bit character encodings)
is also heavily platform-specific.
>Since there is no 2 byte character data type in J, such data must be
>represented as J integer lists.
>
>I think you can solve your problem of moving unicode data between VB and
>J if you use VB routines that convert between unicode strings and
>integer arrays and exchange integer arrays with J.
I find it easier to let J think it is dealing with a vector of 8-bit
characters. Then it doesn't mind converting between big- and
little-endian--Reshape to two columns (Mayan writing order!), then reverse
under transpose. I had to give up on reverse rank 1, since it made far too
many temporaries.
--
Edward Cherlin President
Coalition Against Unsolicited Commercial E-mail
Help outlaw Spam. <http://www.cauce.org/>
Talk to us at <news:comp.org.cauce>
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Mon Jun 7 06:26:28 1999
From: Baker Stuart <stuart.baker@hyder.com>
Subject: RE: Jforum: Spaces in Windows NT Pathnames
Date: Mon, 7 Jun 1999 11:09:50 -0000
- sadly, it appears not, at least with scripts.ijs entries: I just tried it,
having happily used non-DOS style filenames explicitly within applications.
I think this is just because of the way that the buildpublic verb treats
these entries, nothing fundamental.
> ----------
> From: Kip Murray[SMTP:CMurray@UH.EDU]
> Reply To: forum@jsoftware.com
> Sent: 04 June 1999 00:37
> To: forum@jsoftware.com
> Subject: Jforum: Spaces in Windows NT Pathnames
>
> Can J's facilities for reading Windows NT pathnames deal with a
> pathname that has a space in it?
>
> I'm thinking, first of all, of the routine that picks up the user's
> profile from the startup icon; and secondly, of names added to
> scripts.ijs.
>
> Kip Murray <CMurray@UH.EDU>
> Math, University of Houston
>
> --------------------------------------------------------------------------
> ------
> 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 Jun 7 13:39:55 1999
From: "Oleg Kobchenko" <gccinc@usa.net>
Subject: Re: Jforum: J403a missing system\examples\winapi\winapi.ijs
Date: Mon, 7 Jun 1999 20:24:02 +0300
charset="iso-8859-1"
Actually it shouldn't matter where the new location is.
The 'scripts.ijs' should take care of that.
Then
require 'winapi'
would work in all old files without updating paths in each
of them.
-----Original Message-----
From: Murray Eisenberg <murray@math.umass.edu>
Date: Sunday, June 06, 1999 18:49
Subject: Re: Jforum: J403a missing system\examples\winapi\winapi.ijs
Actually, your proposed workaround
require 'winapi'
does NOT work. What does work is typing (or changing the .ijt script to
read):
require 'system\packages\winapi\winapi.ijs'
Chris Burke wrote:
>
> The winapi stuff was moved to system\packages in J403, but this
> reference was not updated. Thanks for pointing it out.
>
> You could copy it from J402, or just ignore the error message and
> enter:
>
> require 'winapi'
>
> The lab will run from then on.
>
> ----- Original Message -----
> From: arneson <arneson@pkb.mega.net.id>
> To: <forum@jsoftware.com>
> Sent: Sunday, June 06, 1999 1:15 AM
> Subject: Jforum: J403a missing system\examples\winapi\winapi.ijs
>
> > I tried to run Lab: DLL: Using System DLLs (file examples) from
> J403a.
> >
> > It failed on:
> >
> > require 'system\examples\winapi\winapi.ijs'
> >
> > because system\examples\winapi\ did not exist. Can I copy it from
> the J402 directory?
> >
>
> --------------------------------------------------------------------------
------
> J Forum: for information about this list, see
http://www.jsoftware.com/forum.htm
--
Murray Eisenberg murray@math.umass.edu
Mathematics & Statistics Dept. phone 413 549-1020 (H)
Univ. of Massachusetts 413 545-2859 (W)
Amherst, MA 01003-4515
----------------------------------------------------------------------------
----
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 Jun 7 17:13:28 1999
From: mellemf@nimo.com
X-Lotus-FromDomain: NMPC
Date: Mon, 7 Jun 1999 16:15:05 -0400
Subject: Jforum: Evaluating J4.03 ; with Bold ISIJ font the box chars are garbled
Content-type: text/plain; charset=us-ascii
Content-Disposition: inline
horizontal box characters become A below 2 dots. vertical become superscript
(little) threes.
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Mon Jun 7 22:52:23 1999
From: "Chris Burke" <cdburke@interlog.com>
References: <85256789.00731DA3.00@mail.nimo.com>
Subject: Re: Jforum: Evaluating J4.03 ; with Bold ISIJ font the box chars are garbled
Date: Mon, 7 Jun 1999 22:43:08 -0400
charset="iso-8859-1"
Select OEM, which should give you the proper linedraw characters.
----- Original Message -----
From: <mellemf@nimo.com>
Sent: Monday, June 07, 1999 4:15 PM
Subject: Jforum: Evaluating J4.03 ; with Bold ISIJ font the box chars
are garbled
> horizontal box characters become A below 2 dots. vertical become
superscript
> (little) threes.
>
>
>
>
> --------------------------------------------------------------------
------------
> 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 Jun 8 04:53:08 1999
From: Baker Stuart <stuart.baker@hyder.com>
Subject: Jforum: fit-specified fill
Date: Tue, 8 Jun 1999 09:24:15 -0000
I'd like to be able to do the following:
>!._ vnv NB. just open a [vnv] with fill specified to be infinity.
but this doesn't seem to work: the way I'm currently doing it is:
> ( >./;# &.> vnv) {.!._ &.> vnv
this does not look very nice for what should be quite simple - is it the
best that can be done?
Stuart Baker
HCS Testing Support Team
5th Floor Trafalgar House
x 43205
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Tue Jun 8 07:01:40 1999
From: "Seymour Glass" <glasss@mindspring.com>
Subject: RE: Jforum: fit-specified fill
Date: Tue, 8 Jun 1999 06:46:21 -0400
charset="iso-8859-1"
Importance: Normal
In-Reply-To: <0488C2686254D111A5C70000F847E8C5030DB33B@LUSIS_EXCH_1>
If a 'vnv' is nonnull, and the opened items are
of rank 1, you could use
(,!._&:>/) @: (,&(<i. 0 0)) vnv
but I don't know that this is any better than
your version (it looks slower than the original,
but I haven't checked).
Would be nice to have
;!._ @: (,&(<i. 0 0)) vnv
but ; does not support !. .
The original can be rewritten with a little less boxing as
fillvnv =. ({.!._ >)"0~ >./ @: (#S:0)
which can at least be buried in a library somewhere.
Henry Rich
> -----Original Message-----
> From: owner-jsoftware@lists.interlog.com
> [mailto:owner-jsoftware@lists.interlog.com]On Behalf Of Baker Stuart
> Sent: Tuesday, June 08, 1999 05:24
> To: 'forum@jsoftware.com'
> Subject: Jforum: fit-specified fill
>
>
> I'd like to be able to do the following:
>
> >!._ vnv NB. just open a [vnv] with fill specified to be infinity.
>
> but this doesn't seem to work: the way I'm currently doing it is:
>
> > ( >./;# &.> vnv) {.!._ &.> vnv
>
> this does not look very nice for what should be quite simple - is it the
> best that can be done?
>
> Stuart Baker
>
> HCS Testing Support Team
> 5th Floor Trafalgar House
> x 43205
>
> ------------------------------------------------------------------
> --------------
> 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 Jun 8 07:58:50 1999
From: M.Day@fscii.maff.gov.uk
Tue, 8 Jun 1999 12:23:18 +0100
Tue, 8 Jun 1999 12:20:01 +0100
Tue, 8 Jun 1999 12:18:02 +0100
Date: Tue, 8 Jun 1999 12:18:02 +0100
Content-Identifier: m1210608121744aa
Alternate-Recipient: Allowed
In-Reply-To: <0488C2686254D111A5C70000F847E8C5030DB33B@LUSIS_EXCH_1>
Subject: Re: Jforum: fit-specified fill
Stuart
If you always need to fill with infinity,
then
% > % each vnv NB. or
% > % &. > vnv
looks quite neat in explicit form.
However, its tacit form is pretty awful:
([: % [: > [: %&.> ]) vnv NB. !!!
I suppose these forms might save you
storage/time for large or frequent
operations, if that's important.
Mike Day
8 vi 99
Stuart Baker wrote:
>I'd like to be able to do the following:
>
> >!._ vnv NB. just open a [vnv] with fill specified to be infinity.
>
>but this doesn't seem to work: the way I'm currently doing it is:
>
> > ( >./;# &.> vnv) {.!._ &.> vnv
>
>this does not look very nice for what should be quite simple - is it the
>best that can be done?
>
> Stuart Baker
>
> HCS Testing Support Team
> 5th Floor Trafalgar House
> x 43205
>
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Tue Jun 8 09:33:37 1999
From: Baker Stuart <stuart.baker@hyder.com>
Subject: RE: Jforum: fit-specified fill
Date: Tue, 8 Jun 1999 14:14:51 -0000
Thank you both for your suggestions - I particularly like "%>% each" as I
can remember it for more than 2 minutes (so far). I needed to do various
table look-ups involving index-sets (ie non-negative variable length numeric
vectors) and wanted to try out working with filled level-zero arrays as well
as vnv's, to see which worked best for my particular data... I'll now go
away and experiment a bit.
> ----------
> From: M.Day@fscii.maff.gov.uk[SMTP:M.Day@fscii.maff.gov.uk]
> Reply To: forum@jsoftware.com
> Sent: 08 June 1999 11:18
> To: forum@jsoftware.com; stuart.baker@hyder.com
> Subject: Re: Jforum: fit-specified fill
>
> From: Seymour Glass[SMTP:glasss@mindspring.com]
> Reply To: forum@jsoftware.com
> Sent: 08 June 1999 10:46
> To: forum@jsoftware.com
> Subject: RE: Jforum: fit-specified fill
>
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Tue Jun 8 23:22:49 1999
Date: Tue, 08 Jun 1999 23:13:11 -0400
From: Kirk B Iverson <kbi@interlog.com>
Subject: Re: Jforum: HTML publish is not working in J403a
In-Reply-To: <4.1.19990602125137.00bcd100@pop.kolumbus.fi>
References: <c=FR%a=_%p=DynaSys_-_Schilt%l=DYNASYS01-990602065241Z-9@ma il.dynasys.tm.fr>
At 13:04 1999-06-02 +0300, Anssi Seppala wrote:
>I just loaded J403a and tried to make html publish but it does not work!
The HTML publisher uses an Internet Explorer OCX control in order to
render the generated HTML. In some environments the name of the control
I try to use is wrong. I'll prepare a permanent fix for this.
In the meantime, if you encounter this problem try making the following
change in system\extras\config\pubprefs.ijs:
OCX=: 'Shell.Explorer.1'
OLEMETHOD=: 'base navigate'
Anssi: Please let me know if this works for you.
/K
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Wed Jun 9 02:19:45 1999
Date: Wed, 09 Jun 1999 08:05:44 +0300
From: Anssi Seppala <anssi.seppala@enease.fi>
Subject: Re: Jforum: HTML publish is not working in J403a
In-Reply-To: <3.0.5.32.19990608231311.008fa7d0@mail.interlog.com>
References: <4.1.19990602125137.00bcd100@pop.kolumbus.fi>
<c=FR%a=_%p=DynaSys_-_Schilt%l=DYNASYS01-990602065241Z-9@ma il.dynasys.tm.fr>
Yes, HTML publish works now with these settings!
Thanks,
Anssi
At 23:13 8.6.1999 -0400, you wrote:
>At 13:04 1999-06-02 +0300, Anssi Seppala wrote:
>
>>I just loaded J403a and tried to make html publish but it does not work!
>
>
> The HTML publisher uses an Internet Explorer OCX control in order to
>render the generated HTML. In some environments the name of the control
>I try to use is wrong. I'll prepare a permanent fix for this.
>
> In the meantime, if you encounter this problem try making the following
>change in system\extras\config\pubprefs.ijs:
>
> OCX=: 'Shell.Explorer.1'
> OLEMETHOD=: 'base navigate'
>
>Anssi: Please let me know if this works for you.
>
>/K
>
>
>----------------------------------------------------------------------------
>----
>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 Jun 9 08:55:12 1999
Date: Wed, 09 Jun 1999 08:43:33 -0400
From: Cliff Reiter <reiterc@lafvax.lafayette.edu>
Subject: Re: Jforum: HTML publish is not working in J403a
Organization: Lafayette College
Content-type: text/plain; charset=us-ascii
References: <3.0.5.32.19990608231311.008fa7d0@mail.interlog.com>
> The HTML publisher uses an Internet Explorer OCX control in order to
> render the generated HTML. In some environments the name of the control
> I try to use is wrong. I'll prepare a permanent fix for this.
>
> In the meantime, if you encounter this problem try making the following
> change in system\extras\config\pubprefs.ijs:
>
> OCX=: 'Shell.Explorer.1'
> OLEMETHOD=: 'base navigate'
>
> Anssi: Please let me know if this works for you.
>
It works for me. Nice way to publish scripts.
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 Wed Jun 9 10:11:57 1999
From: "Chris Burke" <cdburke@interlog.com>
References: <375a62385840001@blik.skima.is>
Subject: Re: Jforum: J403
Date: Wed, 9 Jun 1999 09:59:29 -0400
charset="iso-8859-1"
You can use http://www.deja.com/home_ps.shtml to see newsgroup
history. A quick check of J403 on c.l.a. reveals two items, both
announcements of J403.
06/09/99 J Colour, Debug and FTP comp.lang.apl Bjorn Helgason
05/31/99 J403 comp.lang.apl Chris Burke
----- Original Message -----
From: <gosi@centrum.is>
Sent: Sunday, June 06, 1999 7:57 AM
Subject: Re: Jforum: J403
> I have not noticed the new release J403 announced
> in c.l.a.
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Wed Jun 9 16:06:22 1999
Date: Tue, 8 Jun 1999 21:58:55 +0200 (CEST)
From: Martin Neitzel <neitzel@is.gaertner.de>
Subject: Re: Jforum: fit-specified fill
Baker Stuart <stuart.baker@hyder.com>:
>
> I'd like to be able to do the following:
> >!._ vnv NB. just open a [vnv] with fill specified to be infinity.
> but this doesn't seem to work.
Yes, and it really cannot work. An explanation for it can be found
in my APL95 paper "Untying the Gordian Knot". I'm appending the
relevant part at the end.
As luck would have it, Roger's "Uniform Rank" paper is right next
to it in the proceedings. It gives all the J code necessary to
model "rank" on your own -- so if you want to solve "the fitted
fill" problem in a general way, you have a good head start to
actually implement some of the thoughts below.
Martin
Controlled Recombinations
The section on Dictionary APL contained the example
> (1 ; 2 3 ; 4 5 6)
1 0 0
2 3 0
4 5 6
In certain circumstances it might be useful to pad the vec-
tors with other values than 0. One might think that the
__fit__ conjunction __!.__ could be used to customize the
fill elements for __open__, just as it can be done for over-
takes:
>!.9 (1 ; 2 3 ; 4 5 6)
1 9 9
2 3 9
4 5 6
This might seem attractive at first but is not a viable way
to go. As already explained, the zeroes slip in during step
3, the recombination of the partial results. During step 2,
the __open__ of an individual cell has no way to ``know''
what the common result shape will be. So it is not possible
to get flexible padding by making __open__ smarter.
The general solution would be to attach the control over
fillings not to individual verbs, but to the shape assimila-
tion process in step 3.
What I would consider to be a step in the right direction is
having a conjunction at the level of the rank __"__. Just
like __"__ gives control over argument dissection and cell
pairings prior to verb evaluation, it could allow for con-
trol of the fill value for shape assimilation during the
reassembly of the intermediate results. Without compromis-
ing existing behavior, the rank conjunction could accept a
boxed pair containing rank and fill specifications. For
example:
> "(0; 9) 1 ; 2 3 ; 4 5 6
1 9 9
2 3 9
4 5 6
Details like the order of that boxed pair and the defaulting
of specifications are an issue left to the language lawyers
and designers. The important point is that it should really
be __rank__ itself that should be extended. A separate, new
__fill conjunction__ defined to control step 3 could not do
the job in concert with __rank__: applying both conjunctions
to a verb would result in nested verbs so that steps 1 and 3
would lie on different levels. No matter what the order of
the nesting were, there would never be anything to fill.
Why should one be content with mere fill __values__? A pro-
grammer could provide a fill __function__ that should be
triggered in the context of II.B's shape assimilation. The
function would be called with the short partial result and
the requested shape. A programmer could designed it to have
any desired effect, as long as results of a common shape are
generated. Common choices are:
padding with a constant
padding by repeating the last item
replicating the entire value
padding on the right instead of the left
Notice that, while these fill functions are easily pro-
grammed per se, the context-dependent target shape is not
known a priori. For this reason, it cannot be done within
the current language. (The overtake is an example of the
case where the programmer explicitly requests the target
shape.)
A fill function could be specified in two ways: as a gerund
in place of the fill value in the rank extension proposed
above, or as an intrinsic characteristic of a verb, associ-
ated with a verb just like __obverses__ and __identity func-
tions__.
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Wed Jun 9 16:38:31 1999
From: Andrew Nikitin <anikitin@fastenal.com>
Subject: Jforum: fit
Date: Wed, 9 Jun 1999 15:28:31 -0500
Is there a way to use fit conjunction with user defined verbs?
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Wed Jun 9 17:12:02 1999
From: "Roger Hui" <RHui@interlog.com>
References: <ED32805AF20ED21186AE00805FA615340131CC7E@NUT>
Subject: Re: Jforum: fit
Date: Wed, 9 Jun 1999 14:00:56 -0400
charset="iso-8859-1"
> Is there a way to use fit conjunction with user defined verbs?
No there is not. Another name for the conjunction is "ad hoc",
which describes it pretty well.
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Wed Jun 9 20:12:02 1999
From: "Seymour Glass" <glasss@mindspring.com>
Subject: RE: Jforum: Re: Boxed Arrays
Date: Wed, 9 Jun 1999 19:54:59 -0400
charset="iso-8859-1"
In-Reply-To: <008001bea6fe$a7d4d5c0$fbb1ba89@f3nbp>
Importance: Normal
The topic was why ; is not the same as ,&< , and Roger said:
> ...
> I should point out that in J ; is a verb that is syntactically like
> any other. In particular, in a phrase such as x;y;z;a;b;c;d
> the instances of ; can not look forward or look back;
> each instance of ; is given only its left and right arguments.
>
> This is not the case with the ; in APL.
I had always wondered why so many J users were put out by the
asymmetry of ; . This must be the reason.
They could be soothed a bit by an expansion of the definition
of =. and =: . J could define that, for boxed x, the statements
x =. y and x =: y 'match up' the boxing levels of x and y,
recursively if necessary.
I can give an example of the idea for =:, but not for =., as follows:
link =: ,&<
NB. 'globassign' replaces =:, with hierarchical x and y
globassign =: 4 : 0
if. 32 = 3!:0 x. do.
x. 0:@globassign&> y.
else.
(x.) =: y.
end.
y.
)
('a' link 'b' link 'c') NB. Note the boxing is not the same as with ;
+-+-----+
|a|+-+-+|
| ||b|c||
| |+-+-+|
+-+-----+
('a' link 'b' link 'c') globassign 1 link 2 link (<3)
a
1
b
2
c
+-+
|3|
+-+
The assignments work regardless of the boxed status of the
last linked item.
I personally would keep using ;, I think, but such a definition
of =. would help something I do a lot: Often my variables are
collections of stuff so the collection can travel as a unit, but
when I get to a verb I need to open the collection so I
can use the components. When several such collections are passed
in to a verb, the verb starts out like:
verb =: 3 : 0
'c1 c2 c3' =. y.
'v11 v12' =. c1
'v21 v22 v23' =. c2
'v31 v32' =. c3
etc.
and it can get worse if the collections have multiple boxing levels.
I am proposing that I be able to write
('v11 v12';'v21 v22 v23';'v31 v32') =. y.
and have the whole hierarchy flattened at one go.
Henry Rich
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Wed Jun 9 21:48:15 1999
Date: Wed, 09 Jun 1999 21:35:42 -0400
From: "Kenneth E. Iverson" <kei@interlog.com>
Organization: Iverson Software Inc.
Subject: Re: Jforum: Re: Boxed Arrays
References: <000f01beb2d3$7bb42640$d79bf7a5@seymourg>
Would c=:,&(<^:(L.=0:)) do what you want?
Seymour Glass wrote:
>
> The topic was why ; is not the same as ,&< , and Roger said:
> > ...
> > I should point out that in J ; is a verb that is syntactically like
> > any other. In particular, in a phrase such as x;y;z;a;b;c;d
> > the instances of ; can not look forward or look back;
> > each instance of ; is given only its left and right arguments.
> >
> > This is not the case with the ; in APL.
>
> I had always wondered why so many J users were put out by the
> asymmetry of ; . This must be the reason.
>
> They could be soothed a bit by an expansion of the definition
> of =. and =: . J could define that, for boxed x, the statements
> x =. y and x =: y 'match up' the boxing levels of x and y,
> recursively if necessary.
>
> I can give an example of the idea for =:, but not for =., as follows:
>
> link =: ,&<
>
> NB. 'globassign' replaces =:, with hierarchical x and y
> globassign =: 4 : 0
> if. 32 = 3!:0 x. do.
> x. 0:@globassign&> y.
> else.
> (x.) =: y.
> end.
> y.
> )
>
> ('a' link 'b' link 'c') NB. Note the boxing is not the same as with ;
> +-+-----+
> |a|+-+-+|
> | ||b|c||
> | |+-+-+|
> +-+-----+
> ('a' link 'b' link 'c') globassign 1 link 2 link (<3)
> a
> 1
> b
> 2
> c
> +-+
> |3|
> +-+
>
> The assignments work regardless of the boxed status of the
> last linked item.
>
> I personally would keep using ;, I think, but such a definition
> of =. would help something I do a lot: Often my variables are
> collections of stuff so the collection can travel as a unit, but
> when I get to a verb I need to open the collection so I
> can use the components. When several such collections are passed
> in to a verb, the verb starts out like:
>
> verb =: 3 : 0
> 'c1 c2 c3' =. y.
> 'v11 v12' =. c1
> 'v21 v22 v23' =. c2
> 'v31 v32' =. c3
> etc.
>
> and it can get worse if the collections have multiple boxing levels.
>
> I am proposing that I be able to write
>
> ('v11 v12';'v21 v22 v23';'v31 v32') =. y.
>
> and have the whole hierarchy flattened at one go.
>
> Henry Rich
>
> --------------------------------------------------------------------------------
> 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 Jun 9 23:49:37 1999
From: "Seymour Glass" <glasss@mindspring.com>
Subject: RE: Jforum: Re: Boxed Arrays
Date: Wed, 9 Jun 1999 23:37:47 -0400
charset="iso-8859-1"
In-Reply-To: <375F166E.5D75@interlog.com>
Importance: Normal
Let me say first that what I really want is the recursive
assignment, for the reasons given in the second half of my
post.
Here you are asking, is c=:,&(<^:(L.=0:)) a suitable
substitute for ; or ,&< ?
I say, no. We want to be able to join dissimilar items
and then to unbox the atoms and reconstitute the
original values. But 5 c 6 is the same as (<5) c (<6),
so the boxed/unboxed character of the operands
will be lost in the conversion. Since boxing is
important, c is not suitable for a general linking verb.
It is really the same problem that afflicts the right-hand
argument to ; , except that with c the problem applies
to both arguments.
Henry Rich
KEI wrote:
>
> Would c=:,&(<^:(L.=0:)) do what you want?
>
> Seymour Glass wrote:
> >
> > The topic was why ; is not the same as ,&< , and Roger said:
> > > ...
> > > I should point out that in J ; is a verb that is syntactically like
> > > any other. In particular, in a phrase such as x;y;z;a;b;c;d
> > > the instances of ; can not look forward or look back;
> > > each instance of ; is given only its left and right arguments.
> > >
> > > This is not the case with the ; in APL.
> >
> > I had always wondered why so many J users were put out by the
> > asymmetry of ; . This must be the reason.
> >
> > They could be soothed a bit by an expansion of the definition
> > of =. and =: . J could define that, for boxed x, the statements
> > x =. y and x =: y 'match up' the boxing levels of x and y,
> > recursively if necessary.
> >
> > I can give an example of the idea for =:, but not for =., as follows:
> >
> > link =: ,&<
> >
> > NB. 'globassign' replaces =:, with hierarchical x and y
> > globassign =: 4 : 0
> > if. 32 = 3!:0 x. do.
> > x. 0:@globassign&> y.
> > else.
> > (x.) =: y.
> > end.
> > y.
> > )
> >
> > ('a' link 'b' link 'c') NB. Note the boxing is not the same
> as with ;
> > +-+-----+
> > |a|+-+-+|
> > | ||b|c||
> > | |+-+-+|
> > +-+-----+
> > ('a' link 'b' link 'c') globassign 1 link 2 link (<3)
> > a
> > 1
> > b
> > 2
> > c
> > +-+
> > |3|
> > +-+
> >
> > The assignments work regardless of the boxed status of the
> > last linked item.
> >
> > I personally would keep using ;, I think, but such a definition
> > of =. would help something I do a lot: Often my variables are
> > collections of stuff so the collection can travel as a unit, but
> > when I get to a verb I need to open the collection so I
> > can use the components. When several such collections are passed
> > in to a verb, the verb starts out like:
> >
> > verb =: 3 : 0
> > 'c1 c2 c3' =. y.
> > 'v11 v12' =. c1
> > 'v21 v22 v23' =. c2
> > 'v31 v32' =. c3
> > etc.
> >
> > and it can get worse if the collections have multiple boxing levels.
> >
> > I am proposing that I be able to write
> >
> > ('v11 v12';'v21 v22 v23';'v31 v32') =. y.
> >
> > and have the whole hierarchy flattened at one go.
> >
> > Henry Rich
> >
> >
> ------------------------------------------------------------------
> --------------
> > 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 Fri Jun 11 07:38:14 1999
From: "Chris Burke" <cdburke@interlog.com>
Subject: Jforum: Math for the Layman
Date: Fri, 11 Jun 1999 07:24:51 -0400
charset="Windows-1252"
"Math for the Layman" by Ken Iverson is now available at
www.jsoftware.com/resource/papers.htm. From the Preface:
In 1936, Lancelot Hogben published his still-popular Mathematics for
the Million, stating his objective as follows:
"The view which we shall explore is that mathematics is the language
of size, shape and order and that it is an essential part of the
equipment of an intelligent citizen to understand this language. If
the rules of mathematics are the rules of grammar, there is no
stupidity involved when we fail to see that a mathematical truth is
obvious. The rules of ordinary grammar are not obvious. They have to
be learned. They are not eternal truths. They are conveniences
without whose aid truths about the sorts of things in the world
cannot be communicated from one person to another. "
Our objective is similar, but we now have new tools: the development
of computer programming has provided languages with grammars that
are simpler and more tractable than that of conventional
mathematical notation. Moreover, the general availability of the
computer makes possible convenient and accurate experimentation with
mathematical ideas.
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Fri Jun 11 12:47:44 1999
From: "Oleg Kobchenko" <gccinc@usa.net>
Subject: Jforum: Numerical solution of ODE systems
Date: Fri, 11 Jun 1999 19:29:40 +0300
charset="Windows-1252"
I want to try using J for numerical
solution of ODE systems. In particular,
the concern is about non-linear autonomous
systems with (constant, lagging) deviating
argument.
I wonder, if there is any existing (accessible)
code already in J or in a form that may be
easily ported.
My current best ball park is Runge-Kutta (etc.)
implementations from Matlab, which are open.
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Fri Jun 11 16:47:07 1999
Date: Fri, 11 Jun 1999 13:13:04 -0400 (EDT)
From: "J. Patrick Harrington" <jph@astro.umd.edu>
Subject: Jforum: Re: Numerical solution of ODE systems
I sometimes want to integrate an ODE, so I
transcribed some routines from "Numerical Recipes"
into J. They have the form of adverbs which work
on a verb which provides the derivatives. Here,
without comment:
---------------------------------------------------------
NB. Example of simple differential equation.
NB. f=. [: - */ NB. dy/dx = - x*y
NB. Calling params: (x0,y0;x_end;err,htry,hmin)
arg=. ((0 10);1;(1e_8,0.05,1e_8))
kmax=: 100 NB. max number of saved intermediate values
dxsav=: 0.05 NB. minimum save interval
NB. Usage: f odeint arg -> x_end, y(x_end)
odeint=. 1 : 0
x=. x1=. 0{ xy=. xy0=. >0{y.
x2=. >1{y.
param=. >2{y.
eps=. 0{ param
h1=. 1{ param
hmin=. 2{ param
tiny=. 1e_30
maxstp=. 10000
h=. h1 F_sign x2-x1
nok=: nbad=: kount=: 0
xyp=: xy0
if. kmax > 0 do. xsav=. x - 2*dxsav end.
nstp=. 0
while. (nstp=. >:nstp)<: maxstp do.
dydx=. u. xy
yscal=. (|}. xy)+(|h*dydx)+tiny
if. kmax > 0 do.
if. (|x-xsav)>|dxsav do.
if. kount<(kmax-1) do.
kount=. >: kount
xyp=: xyp,xy
xsav=x
end.
end.
end.
if. ((x+h-x2)*(x+h-x1))>0 do. h=. x2-x end.
out=. u. rkqs (xy;dydx;h;eps;yscal)
x=. 0{ xy=. >0{ out
hdid=. >1{ out
hnext=. >2{ out
if. hdid = h do. nok=: >: nok else. nbad=: >: nbad end.
if. ((x-x2)*(x2-x1))>: 0 do.
if. kmax ~: 0 do.
kount=: >: kount
xyp=: xyp,xy
end.
xy0=. xy
return.
end.
if. (|hnext)<hmin do.
z=. 'Stepsize smaller than minimum in odeint.'
return.
end.
h=. hnext
end.
z=. 'Too many steps in odeint.'
)
NB. Cash-Karp Runge-Kutta quality control step
NB. Numerical Recipes, 2nd ed., p 713.
rkqs=. 1 : 0
safety=. 0.9
pgrow=. _0.2
pshrink=. _0.25
errcon=. 1.89e_4
xy=. >0{y.
dydx=. >1{y.
h=. >2{y.
eps=. >3{y.
yscal=. >4{y.
whilst. 1 do.
out=. u. rkck (xy;dydx;h)
yerr=. >1{out
errmax=:(%eps)* >./ | yerr%yscal
if. errmax > 1 do.
htemp=. safety*h*(errmax^pshrink)
h=.((|htemp)>. (0.1*|h)) F_sign h
xnew=. (0{xy)+h
if. xnew = (0{xy) do.
r=.'Step too small!'
return.
end.
continue.
else.
if. errmax > errcon do.
hnext=. safety*h*(errmax^pgrow)
else.
hnext=. 5*h
end.
z=. ((>0{out);h;hnext)
end.
return.
end.
)
NB. Cash-Karp Runge-Kutta (5th order) adverb. Usage:
NB. f rkck (xy0;dydx0;h) ->(xy1[=x+h,y(x+h)];error)
rkck=. 1 : 0
xy0=. >0{y.
h=. >2{y.
k1=. h*>1{y.
k2=. h*u. xy0+(h,k1)%5
k3=. h*u. xy0+((0.3*h),((3*k1%40)+(9*k2%40)))
dy=. ((3*k1)+(_9*k2)+(12*k3))%10
k4=. h*u. xy0+((0.6*h),dy)
dy=. (_11*k1%54)+(5*k2%2)+(_70*k3%27)+(35*k4%27)
k5=. h*u. xy0+(h,dy)
dy=. (1631*k1%55296)+(175*k2%512)+(575*k3%13824)+(44275*k4%110592)+(253*k5%4096)
k6=. h*u. xy0+((0.875*h),dy)
dy=. (37*k1%378)+(250*k3%621)+(125*k4%594)+(512*k6%1771)
xy=. xy0+(h,dy)
dy4=. (2825*k1%27648)+(18575*k3%48384)+(13525*k4%55296)+(277*k5%14336)+(k6%4)
out=. xy;(dy-dy4)
)
NB. Model of FORTRAN "SIGN(A,B)" function:
F_sign=. ([: | [)* 1: -[: +: ]< 0:
--------------------------------------------------------------
NB. Example of simple differential equation.
NB. f=. [: - */ NB. dy/dx = - x*y
NB. Calling params: (x0,y0;x_end;err,htry,hmin)
arg=. ((0 10);1;(1e_8,0.05,1e_8))
kmax=: 100 NB. max number of saved intermediate values
dxsav=: 0.05 NB. minimum save interval
NB. Usage: f sdeint arg -> x_end, y(x_end)
sdeint=. 1 : 0
x=. x1=. 0{ xy=. xy0=. >0{y.
x2=. >1{y.
param=. >2{y.
eps=. 0{ param
h1=. 1{ param
hmin=. 2{ param
tiny=. 1e_30
maxstp=. 2000
h=. h1 F_sign x2-x1
nok=: nbad=: kount=: 0
xyp=: xy0
if. kmax > 0 do. xsav=. x - 2*dxsav end.
nstp=. 0
while. (nstp=. >:nstp)<: maxstp do.
dydx=. u. xy
yscal=. (|}. xy)+(|h*dydx)+tiny
if. kmax > 0 do.
if. (|x-xsav)>|dxsav do.
if. kount<(kmax-1) do.
kount=. >: kount
xyp=: xyp,xy
xsav=x
end.
end.
end.
if. ((x+h-x2)*(x+h-x1))>0 do. h=. x2-x end.
out=. u. stiff (xy;dydx;h;eps;yscal)
x=. 0{ xy=. >0{ out
hdid=. >1{ out
hnext=. >2{ out
if. hdid = h do. nok=: >: nok else. nbad=: >: nbad end.
if. ((x-x2)*(x2-x1))>: 0 do.
if. kmax ~: 0 do.
kount=: >: kount
xyp=: xyp,xy
end.
xy0=. xy
return.
end.
if. (|hnext)<hmin do.
z=. 'Stepsize smaller than minimum in odeint.'
return.
end.
h=. hnext
end.
z=. 'Too many steps in odeint.'
)
NB. Rosenbrock method for stiff equations.
NB. Numerical Recipes, 2nd ed., p 732.
stiff=. 1 : 0
safety=. 0.9
grow=. 1.5
pgrow=. %_4
shrnk=. 0.5
pshrnk=. %_3
errcon=. 0.1296
maxtry=. 40
x=. 0{ xysav=. xy=. >0{y.
n=. # dysav=. dydx=. >1{y.
h=. >2{y.
eps=. >3{y.
yscal=. >4{y.
jout=. jacobn xysav
dfdx=. >0{ jout
dfdy=. >1{ jout
jtry=. 0
while. (jtry=. >: jtry) <: maxtry do.
ai=. %. ((2%h)* = i. n) - dfdy NB. inverse of matrix
g1=. ai +/ .* dysav + h*dfdx%2
xy=. xysav + (h,(2*g1))
dydx=. u. xy
g2=. ai +/ .* dydx + (h*_1.5*dfdx) + _8*g1%h
xy=. xysav + ((0.6*h),((48*g1%25)+(6*g2%25)))
dydx=. u. xy
g3=. dydx + (h*121*dfdx%50) + ((372*g1%25)+(12*g2%5))%h
g3=. ai +/ .* g3
g4=. dydx+(h*29*dfdx%250)+((_112*g1%125)+(_54*g2%125)+(_2*g3%5))%h
g4=. ai +/ .* g4
xy=. xysav +(h,((19*g1%9)+(g2%2)+(25*g3%108)+(125*g4%108)))
err=. (17*g1%54)+(7*g2%36)+(125*g4%108)
if. (0{xy) = (0{xysav) do.
r=.'Step too small!'
return.
end.
errmax=.(%eps)* >./ | err%yscal
if. errmax <: 1 do.
if. errmax > errcon do.
hnext=. safety*h*(errmax^pgrow)
else.
hnext=. grow*h
end.
z=. (xy;h;hnext)
return.
else.
hnext=. safety*h*(errmax^pshrnk)
h=.((|hnext)>. (shrnk*|h)) F_sign h
end.
end.
z=. 'Exceeded MAXTRY in stiff.'
)
NB. Model of FORTRAN "SIGN(A,B)" function:
F_sign=. ([: | [)* 1: -[: +: ]< 0:
NB. A simple stiff system:
A=. 2 2 $ 998 1998 _999 _1999
L=. [ +/ .* }.@]
f =. A&L
jacobn=. 3 : 0
xy=. y. NB. not used - const. coeff.
dfdx=. 2 # 0
dfdy=. 2 2 $ 998 1998 _999 _1999
z=. (dfdx;dfdy)
)
-----------------------------------------------------
NB. G is the set of stiff ODE's
G=. 3 : 0
y1=. 1{y.
y2=. 2{y.
y3=. 3{y.
dy1=. (_0.013*y1)+(_1000*y1*y3)
dy2=. _2500*y2*y3
dy3=. (_0.013*y1)+(_1000*y1*y3)+(_2500*y2*y3)
z=. dy1,dy2,dy3
)
jacobn=. 3 : 0
dfdx=. 3 # 0
y1=. 1{y.
y2=. 2{y.
y3=. 3{y.
dfdy=. 3 3 $ 0
dfdy=. ((_0.013-1000*y3),0,(_1000*y1)) 0}dfdy
dfdy=. (0,(_2500*y3),(_2500*y2)) 1}dfdy
a1=. _0.013-1000*y3
a2=. _2500*y3
a3=. (_1000*y1)+(_2500*y2)
dfdy=. (a1,a2,a3) 2}dfdy
z=. (dfdx;dfdy)
)
G sdeint ((0 1 1 0);5;(1e_4 2.9e_4 1e_10))
nok
nbad
G odeint ((0 1 1 0);5;(1e_4 2.9e_4 1e_10))
nok
nbad
-----------------------------------------------------------
Cheers,
Patrick Harrington
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Fri Jun 11 16:56:43 1999
Date: Fri, 11 Jun 1999 16:54:16 -0400
From: David Ness <DNess@home.com>
Subject: Re: Jforum: Math for the Layman
References: <000901beb3fd$441e8060$3c0114d1@Pcdburke>
Sneaking in the http:// makes all the difference for easing our
access to http://www.jsoftware.com/resource/papers.htm
And, of course, many thanks for making it available.
Chris Burke wrote:
>
> "Math for the Layman" by Ken Iverson is now available at
> www.jsoftware.com/resource/papers.htm. From the Preface:
>
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Sat Jun 12 13:46:12 1999
From: "David Vincent-Jones" <geomap@galaxynet.com>
Subject: Jforum: Fonts Again
Date: Sat, 12 Jun 1999 10:28:17 -0700
When printing out Resource Papers the problem arises again of finding a font
that will show the line draw chars correctly.
Andale Mono (OEM) works well in J but is unhelpful for instance in Word.
How do we solve that problem?
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Mon Jun 14 06:45:18 1999
From: <paul_gauthier@dynasys.tm.fr>
Subject: RE: Jforum: Numerical solution of ODE systems
Date: Mon, 14 Jun 1999 12:41:42 +0200
Content-Transfer-Encoding: 8bit
X-MIME-Autoconverted: from quoted-printable to 8bit by plus.interlog.com id GAA05694
I don't know if your french is fair, but since most of the litterature
is math anyway, here is an interesting link with lot's of source code in
C.
http://www.univ-lille1.fr/~eudil/jbeuneu/equadif/integr.html
>-----Message d'origine-----
>De: Oleg Kobchenko [SMTP:gccinc@usa.net]
>Date: vendredi 11 juin 1999 18:30
>�: forum@jsoftware.com
>Objet: Jforum: Numerical solution of ODE systems
>
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Mon Jun 14 09:21:20 1999
From: mellemf@nimo.com
X-Lotus-FromDomain: NMPC
Date: Mon, 14 Jun 1999 09:14:21 -0400
Subject: Jforum: How many releases of J where made last year?
Content-type: text/plain; charset=us-ascii
Content-Disposition: inline
I believe ther were 2, not counting J403. I am pondering the advantages/cost
of the Professional edition vs the
Standard edition, as I pitch buying the J403 level. Any help in salesmanship
wordsmithing would be appreciated.
Thanks.
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Mon Jun 14 10:53:21 1999
From: Ajith Prasad <burap@nus.edu.sg>
Subject: Jforum: Memory Mapped Files i n J - Article by Chris Burke in Vector Vol.
15 No.487
Date: Mon, 14 Jun 1999 22:48:39 +0800
Will a copy of this article be available at the J website or is this
prevented by copyright considerations?
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Mon Jun 14 13:18:19 1999
Date: 13 Jun 99 09:00:53 -0800
Subject: Jforum: Passing a verb to a verb
From: "Piet de Jong" <piet@dejong.u-net.com>
What is a clean solution to the problem below ?
I suspect there must be an easy solution. Any help gratefully rec'd.
-------
I have a verb u say that repeatedly makes call to another verb, v say.
The verb v is not known at the time u is defined. Not even a list of
potential candidates.
I don't want to redefine u everytime I have a new v.
So I need to pass v at "runtime" that is pass v as an argument to u.
At the moment I pass the script m of v as a noun to u and then have
u "compile it" with the (3 : m) explicit definition conjunction.
But is there a cleaner way? In particular what do I do for tacit v?
Thanks for the help.
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Mon Jun 14 13:36:17 1999
From: "Seymour Glass" <glasss@mindspring.com>
Subject: RE: Jforum: Passing a verb to a verb
Date: Mon, 14 Jun 1999 13:33:11 -0400
charset="iso-8859-1"
Importance: Normal
In-Reply-To: <B38927BE-60BC4@204.244.233.45>
Some ideas:
1. Make u an adverb. Then, once the verb v is defined, you can
invoke them both using
v u
where, within u, you use the system symbol u. to invoke the verb
v .
2. Pass the name of v (string) as an argument to u, and then invoke it
using
((,<argument) `: 6)
which is syntactically a verb (what you are doing is converting the
string to a gerund by hand, and then evoking it.)
3. Pass the name of v as an argument to u, and then use ". to
execute the verb. The problem with this is that the arguments must
also be part of the execute string, so you will have to use 5!:5
to convert the arguments to ASCII. That's why I use `:6 for this sort
of thing.
4. What you are doing works too, and is reasonable.
You can invoke the verb using
args_to_v (13 : argument) args_to_v
(i. e. you don't have to assign the result of : to a name), and it
works for explicit or tacit v.
Henry Rich
> -----Original Message-----
> From: owner-jsoftware@lists.interlog.com
> [mailto:owner-jsoftware@lists.interlog.com]On Behalf Of Piet de Jong
> Sent: Sunday, June 13, 1999 13:01
> To: forum@jsoftware.com
> Subject: Jforum: Passing a verb to a verb
>
>
> What is a clean solution to the problem below ?
>
> I suspect there must be an easy solution. Any help gratefully rec'd.
>
> -------
>
> I have a verb u say that repeatedly makes call to another verb, v say.
>
> The verb v is not known at the time u is defined. Not even a list of
> potential candidates.
>
> I don't want to redefine u everytime I have a new v.
>
> So I need to pass v at "runtime" that is pass v as an argument to u.
>
> At the moment I pass the script m of v as a noun to u and then have
> u "compile it" with the (3 : m) explicit definition conjunction.
>
> But is there a cleaner way? In particular what do I do for tacit v?
>
> Thanks for the help.
>
>
>
>
> ------------------------------------------------------------------
> --------------
> 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 Jun 14 13:41:51 1999
From: Andrew Nikitin <anikitin@fastenal.com>
Subject: RE: Jforum: Passing a verb to a verb
Date: Mon, 14 Jun 1999 12:38:43 -0500
> -----Original Message-----
> From: Piet de Jong [SMTP:piet@dejong.u-net.com]
> Sent: Sunday, June 13, 1999 12:01
>
> At the moment I pass the script m of v as a noun to u and then have
> u "compile it" with the (3 : m) explicit definition conjunction.
>
> But is there a cleaner way? In particular what do I do for tacit v?
>
You can pass the name of the verb and apply it using ~ operator.
For example:
m=.+/ % #
m 1 10 5
5.33333
'm'~ 1 10 5
5.33333
In any case, I would say that your u should be adverb or conjunction
rather then verb.
nsg
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Mon Jun 14 14:40:16 1999
From: "Chris Burke" <cdburke@interlog.com>
References: <A12A4ACD69B4D011BB960020AFFBF108029E9128@exs02.ex.nus.edu.sg>
Subject: Re: Jforum: Memory Mapped Files i n J
Date: Mon, 14 Jun 1999 14:37:37 -0400
charset="iso-8859-1"
Done, see
http://www.jsoftware.com/pubs/mmf.htm
----- Original Message -----
From: Ajith Prasad <burap@nus.edu.sg>
Sent: Monday, June 14, 1999 10:48 AM
Subject: Jforum: Memory Mapped Files i n J - Article by Chris Burke in
Vector Vol.15 No.487
> Will a copy of this article be available at the J website or is this
> prevented by copyright considerations?
>
> --------------------------------------------------------------------
------------
> 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 Jun 14 18:27:16 1999
Date: Tue, 15 Jun 1999 00:13:36 +0200 (CEST)
From: Martin Neitzel <neitzel@is.gaertner.de>
In-Reply-To: <B38927BE-60BC4@204.244.233.45>
Organization: Gaertner Datensysteme, Braunschweig, Germany
Subject: Re: Jforum: Passing a verb to a verb
> So I need to pass v at "runtime" that is pass v as an argument to u.
Congrats, you hit the core of "functional programming": functions
passed as arguments. The "u" will then actually have to be not a
verb but a "meta-verb". (Greek "meta" means "with", BTW.)
As others have already pointed out, J's meta verbs are adverbs (taking
one function as argument) and conjunctions (two functions as arguments).
To get you started, here is an example using _explicit_ definition.
The placeholders u. v. m. n. in addition to x. y. are all it takes
to get rolling:
twice =: adverb define
u. u. y.
:
x. u. x. |. y.
)
>: twice 5
7
However, even as a beginner, one should be acutely aware of some
details which are easy to interpret incorrectly from this example:
(1) "twice" does NOT take two arguments. As an adverb, only the thing
to the left is its argument.
(2) The placeholders "u. u. y." in the definition are resolved at
two different times.
That is, the implicit binding is:
(>: twice) 5
7
And it is "twice"' job to derive a new verb from its >: argument
on-the-fly, as in:
(>: twice) 5 NB. you type this
tmp_verb =. verb define NB. twice's result is a verb.
>: >: y. NB. substitute u. only!
:
x. >: x. :> |. y.
)
tmp_verb 5 NB. J executes the verb and
7 NB. retracts it again.
NB. other applications of twice:
quadruple =: +: twice
quadruple 100
400
100 +twice 5 6
206 205
The Dictionary entry for Explicit Definition ":" gives the gory
details about other placeholders than "u.". It also tells the
entire truth about noun arguments which can make it a bit hard
for the beginner to spot adv/conjs quickly. I skipped that
here to keep things as simple as possible.
My point is: creating adverbs and conjunctions can be very easy
and straightforward, even if one avoids anything "tacit" like the
plague.
Martin Neitzel
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Mon Jun 14 21:53:15 1999
Date: Tue, 15 Jun 1999 03:48:43 +0200
From: "d.alis" <dalis@balcab.ch>
Subject: Re: Jforum: Passing a verb to a verb
References: <B38927BE-60BC4@204.244.233.45>
Contribution 3:
And dont forget the even simpler way of relying on side-effects, (i.e. globally
defined names)..
foo =. 3 : 0
v y.
)
v=.+/
foo i.10
45
v=.+/@: *:
foo i.10
285
In other words, fix the verb in the environment in a level above the one in which it
will be invoked. This can simplify things when you need to debug/trace during
initial developement. However, by introducing hidden dependencies it will make
debug/trace rather more difficult later on.
David
Piet de Jong wrote:
> What is a clean solution to the problem below ?
>
> I suspect there must be an easy solution. Any help gratefully rec'd.
>
> -------
>
> I have a verb u say that repeatedly makes call to another verb, v say.
>
> The verb v is not known at the time u is defined. Not even a list of
> potential candidates.
>
> I don't want to redefine u everytime I have a new v.
>
> So I need to pass v at "runtime" that is pass v as an argument to u.
>
> At the moment I pass the script m of v as a noun to u and then have
> u "compile it" with the (3 : m) explicit definition conjunction.
>
> But is there a cleaner way? In particular what do I do for tacit v?
>
> Thanks for the help.
>
> --------------------------------------------------------------------------------
> 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 Jun 15 03:25:48 1999
From: <paul_gauthier@dynasys.tm.fr>
Subject: RE: Jforum: Passing a verb to a verb
Date: Tue, 15 Jun 1999 09:17:38 +0200
By using the answers from Andrew Nikitin, Martin Neitzel and D. Alis, I
arrived to the following:
twice =: adverb define
u.~ u.~ y.
:
x. u. x. |. y.
)
v=: >:
'v'twice 5
7
Thus using all three concepts...Have a nice day/Paul
>
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Tue Jun 15 10:05:53 1999
From: Andrew Nikitin <anikitin@fastenal.com>
Subject: RE: Jforum: Passing a verb to a verb
Date: Tue, 15 Jun 1999 09:02:12 -0500
> From: paul_gauthier@dynasys.tm.fr [SMTP:paul_gauthier@dynasys.tm.fr]
> Sent: Tuesday, June 15, 1999 02:18
> twice =: adverb define
> u.~ u.~ y.
> :
> x. u. x. |. y.
> )
> v=: >:
> 'v'twice 5
> 7
>
> Thus using all three concepts...Have a nice day/Paul
>
I cannot see the third concept, i.e. passing parameter as a global
variable. You should use something like this to involve all three
concepts:
twice =: adverb define
FN~ FN~ y.
:
x. u. x. |. y.
)
v=: >:
FN=:'v'
? twice 5
or even
[: twice 5
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Tue Jun 15 11:12:18 1999
Date: Tue, 15 Jun 1999 11:07:31 -0400
From: Alain Miville de =?ISO-8859-1?Q?Ch=EAne?= <Infodev@compuserve.com>
Subject: Jforum: debug shouldn't stop
charset=ISO-8859-1
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
X-MIME-Autoconverted: from quoted-printable to 8bit by plus.interlog.com id LAA07566
While debugging a line with an adverb I got this:
|stop
| ((-.b)&#)OnNames RoutesNouns
|SplitUp[7]
dbrun''
|stop
| (y.)=:u. y.~
dbrun''
|stop
| ''
dbrun''
|stop
| (y.)=:u. y.~
dbrun''
|stop
| ''
dbrun''
|stop
| (y.)=:u. y.~
etc... where the debugger stops in the adverb OnNames. I didn't specify any
stops in the adverb. I didn't expect it to stop in there.
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Tue Jun 15 11:22:48 1999
Date: Tue, 15 Jun 1999 11:18:56 -0400
From: Alain Miville de =?ISO-8859-1?Q?Ch=EAne?= <Infodev@compuserve.com>
Subject: Jforum: Big Bug in debug
charset=ISO-8859-1
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
X-MIME-Autoconverted: from quoted-printable to 8bit by plus.interlog.com id LAA11786
The bug is bigger than I thought:
Here is the verb:
SplitUp=: 3 : 0
NB. --- split up routes which include alternates.
NB. --- adjusts tables directly.
yRoutesNouns=.'y',.RoutesNouns
zRoutesNouns=.'z',.RoutesNouns
if. -.+./1 2 3 4 5 6 7 8 9 e. Branches=.~.Branch=.RoutesBranch #~
b=.RoutesNo e. y. do. a: return. end.
(yRoutesNouns)=:(b&#) each ValuesOf RoutesNouns NB. --- get this
route's data. Global for ValuesOf!
((-.b)&#) OnNames RoutesNouns NB. --- clear it out of the original.
br=.Branch e."_ 1 ]0 10,"_ 0 Branches=. 100,Branches-. 0 10 NB. ---
one line per alternate, including the main one (100).
br=.(x=.-.({.{.br)*.{:{.br)}.br NB.--- remove 100 if starts or ends
with alternates.
Branches=.x}.Branches
for_i. i.#br do.
(zRoutesNouns)=:((i{br)&#) each (yRoutesNo+100|i{Branches);
ValuesOf }. yRoutesNouns
((>:zRoutesStop i. {:zRoutesStop)&{.) OnNames zRoutesNouns
(RoutesNouns)=:(ValuesOf RoutesNouns) (([:>[),]) each
ValuesOf zRoutesNouns
end.
a:
)
and here is the debug sequence...
dbrun''
|stop
| zRoutesNouns=.'z',.RoutesNouns
|SplitUp[3]
dbrun''
|stop
| -.+./1 2 3 4 5 6 7 8 9
e.Branches=.~.Branch=.RoutesBranch#~b=.RoutesNo e.y.
|SplitUp[4]
dbrun''
|stop
| (yRoutesNouns)=:(b&#)each ValuesOf RoutesNouns
|SplitUp[6]
dbrun''
|stop
| ((-.b)&#)OnNames RoutesNouns
|SplitUp[7]
dbrun''
|stop
| (y.)=:u. y.~
dbrun''
|stop
| ''
dbrun''
|stop
| (y.)=:u. y.~
dbrun''
|stop
| ''
dbrun''
|stop
| (y.)=:u. y.~
dbrun''
|stop
| ''
dbrun''
|stop
| (y.)=:u. y.~
dbrun''
|stop
| ''
dbrun''
|stop
| (y.)=:u. y.~
dbrun''
|stop
| ''
dbrun''
|stop
| yRoutesNouns=.'y',.RoutesNouns
|SplitUp[2]
Hey! I'm back to line 2 instead of line 8. What happened?
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Tue Jun 15 18:42:07 1999
From: "HARRIS,BILL (HP-LakeStevens,ex1)" <bill_harris@am.exch.hp.com>
Subject: Jforum: System implemented in J
Date: Tue, 15 Jun 1999 16:37:28 -0600
charset="iso-8859-1"
If you are interested, there is a description of a financial system
implemented in J in "Pipeline Inventory: The Missing Factor in
Organizational Expense Management," published in "National Productivity
Review," Summer, 1999, pp. 33-38. The focus of the article is on the
business side, but the underlying system was done in J 3.05C.
Bill
--
Bill Harris Year 2000 Program Office
mailto: bill_harris@am.exch.hp.com Hewlett-Packard Company
phone: (425) 335-2200 M/S 330C
fax: (425) 335-2483 8600 Soper Hill Road
web: http://hpweb.lsid.hp.com:8080/~billh/ Everett, WA 98205-1298
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Tue Jun 15 19:30:31 1999
From: Ajith Prasad <burap@nus.edu.sg>
Subject: RE: Jforum: System implemented in J
Date: Wed, 16 Jun 1999 07:29:19 +0800
charset="ISO-8859-1"
Does the article explain the reasons for the choice of J as the development
tool? What were the alternatives and why was J preferred? Thanks.
-----Original Message-----
From: HARRIS,BILL (HP-LakeStevens,ex1)
Sent: 6/16/99 6:37 AM
Subject: Jforum: System implemented in J
If you are interested, there is a description of a financial system
implemented in J in "Pipeline Inventory: The Missing Factor in
Organizational Expense Management," published in "National Productivity
Review," Summer, 1999, pp. 33-38. The focus of the article is on the
business side, but the underlying system was done in J 3.05C.
Bill
--
Bill Harris Year 2000 Program Office
mailto: bill_harris@am.exch.hp.com Hewlett-Packard Company
phone: (425) 335-2200 M/S 330C
fax: (425) 335-2483 8600 Soper Hill Road
web: http://hpweb.lsid.hp.com:8080/~billh/ Everett, WA 98205-1298
------------------------------------------------------------------------
--------
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 Jun 15 21:54:33 1999
Date: Tue, 15 Jun 1999 21:52:23 -0400
From: David Ness <DNess@home.com>
Subject: Re: Jforum: System implemented in J
References: <737091A8F919D31195E300A0C9968641059984@xboi06.boi.hp.com>
Is there a copy of the article on the net somewhere. If not, I doubt
if most of us will be able to see it...
HARRIS,BILL (HP-LakeStevens,ex1) wrote:
>
> If you are interested, there is a description of a financial system
> implemented in J in "Pipeline Inventory: The Missing Factor in
> Organizational Expense Management," published in "National Productivity
> Review," Summer, 1999, pp. 33-38. The focus of the article is on the
> business side, but the underlying system was done in J 3.05C.
>
> Bill Harris
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Wed Jun 16 02:32:44 1999
From: <paul_gauthier@dynasys.tm.fr>
Subject: RE: Jforum: System implemented in J
Date: Wed, 16 Jun 1999 08:28:52 +0200
Content-Transfer-Encoding: 8bit
X-MIME-Autoconverted: from quoted-printable to 8bit by plus.interlog.com id CAA19275
The link below, (15.1.220.16), can't be reached, any other link ?
thank's/Paul
>-----Message d'origine-----
>De: HARRIS,BILL (HP-LakeStevens,ex1) [SMTP:bill_harris@am.exch.hp.com]
>Date: mercredi 16 juin 1999 00:37
>�: 'forum@jsoftware.com'
>Objet: Jforum: System implemented in J
>
>If you are interested, there is a description of a financial system
>implemented in J in "Pipeline Inventory: The Missing Factor in
>Organizational Expense Management," published in "National Productivity
>Review," Summer, 1999, pp. 33-38. The focus of the article is on the
>business side, but the underlying system was done in J 3.05C.
>
>Bill
>--
>Bill Harris Year 2000 Program Office
>mailto: bill_harris@am.exch.hp.com Hewlett-Packard Company
>phone: (425) 335-2200 M/S 330C
>fax: (425) 335-2483 8600 Soper Hill Road
>web: http://hpweb.lsid.hp.com:8080/~billh/ Everett, WA 98205-1298
>
>-----------------------------------------------------------------------------
>---
>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 Jun 16 04:09:11 1999
From: <paul_gauthier@dynasys.tm.fr>
Subject: RE: Jforum: Passing a verb to a verb
Date: Wed, 16 Jun 1999 08:57:58 +0200
Content-Transfer-Encoding: 8bit
X-MIME-Autoconverted: from quoted-printable to 8bit by plus.interlog.com id CAA21386
Global or global variable ? that is the question...
The example I gave was meant to mean a global assignment of a verb, but
if one wants a global variable becoming a verb then one can use:
twice=: adverb define
(3 : u.) (3 : u.) y.
:
x. u. x. |. y.
)
v=:'>:y.'
v twice 5
7
But I though that the global assignment would be understood not in it's
restrictive "variable" flavor but in it's more general sense of globally
defining a label to be...
Thank's for the clarification and whichever flavor one has a use for, J
certainly can serve them.../Paul
>-----Message d'origine-----
>De: Andrew Nikitin [SMTP:anikitin@fastenal.com]
>Date: mardi 15 juin 1999 16:02
>�: 'forum@jsoftware.com'
>Objet: RE: Jforum: Passing a verb to a verb
>
>> From: paul_gauthier@dynasys.tm.fr [SMTP:paul_gauthier@dynasys.tm.fr]
>> Sent: Tuesday, June 15, 1999 02:18
>> twice =: adverb define
>> u.~ u.~ y.
>> :
>> x. u. x. |. y.
>> )
>> v=: >:
>> 'v'twice 5
>> 7
>>
>> Thus using all three concepts...Have a nice day/Paul
>>
>I cannot see the third concept, i.e. passing parameter as a global
>variable. You should use something like this to involve all three
>concepts:
>
>twice =: adverb define
> FN~ FN~ y.
>:
> x. u. x. |. y.
>)
>
>v=: >:
>FN=:'v'
>? twice 5
>
>or even
>
>[: twice 5
>
>
>
>-----------------------------------------------------------------------------
>---
>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 Jun 16 11:26:53 1999
From: "HARRIS,BILL (HP-LakeStevens,ex1)" <bill_harris@am.exch.hp.com>
Subject: RE: Jforum: System implemented in J
Date: Wed, 16 Jun 1999 09:23:44 -0600
charset="iso-8859-1"
David Ness asked if the article were available on the Web. Unfortunately,
it's not. I'll recheck the copyright transfer agreement I signed to see if
a copy could be posted, but I don't currently have an external Web site,
anyway. Paul Gauthier correctly noted that 15.x.x.x is inaccessible; the
Web site listed in my .sig is for internal HP use.
Ajith Prasad asked:
Does the article explain the reasons for the choice of J as the development
tool? What were the alternatives and why was J preferred? Thanks.
The article only gives a brief statement in that regard, mostly that an
earlier implementation with spreadsheets was error-prone. With a
spreadsheet being potentially an array of (hidden) formulas, it was too easy
for the person administering the system to modify the formula in one cell
for temporary purposes and forget to set it back.
I picked J for a number of reasons:
- pure perversity :-)
- the nature of the problem (arrays of numbers, with axes for departments,
categories of expenses, months, as well as the date the data was analyzed)
lends itself well to an array language
- J interfaced well (even better now, so it would seem -- I'm envious) with
Excel for comfortable data entry, with data from legacy systems (via csv
files), and with the Windows API
- the high level and conciseness of the language made it easier for one
person to do the work (so I presumed)
After getting going, I discovered other benefits. I stored weekly archives
of results. With an interpreted language, it was easy to run ad hoc
analyses of that data to meet various managers' needs. It was also quite
easy to scan along sequences of archives to see trends in projections.
Graphical analyses and presentations were quite easily done, although I
occasionally wished I had more of the facilities of XLISP-STAT readily
available in analyzing and presenting data. I could add a GUI to sysadmin
functions I did manually and pass the work off to others.
The worst thing was learning the language quickly with the material
available at the time. Chris' new "Math for Millions" (or whatever it is
called) is a great step forward. It would be nice to have a "Ranks for
Millions,.Millions" as well. :-)
Bill
--
Bill Harris Year 2000 Program Office
mailto: bill_harris@am.exch.hp.com Hewlett-Packard Company
phone: (425) 335-2200 M/S 330C
fax: (425) 335-2483 8600 Soper Hill Road
web: http://hpweb.lsid.hp.com:8080/~billh/ Everett, WA 98205-1298
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Wed Jun 16 12:59:03 1999
From: Ajith Prasad <burap@nus.edu.sg>
Subject: RE: Jforum: System implemented in J
Date: Thu, 17 Jun 1999 00:55:27 +0800
charset="ISO-8859-1"
Thanks. Fascinating.
One of the contributors to this forum mentioned to me in a private e-mail
that he thought that the "pedigree"/interests of the developers reflected
the niche for which their products were optimised. Thus, given Arthur
Whitney's backgound in the financial sector, K was optimised for business
problems, while J might be more suited for scientific/research work given
Ken Iverson's/Roger Hui's interests in such problems. Nice to see this
assumption put into question with this business application in J. Hope we
can see more such applications being described. The description of the
characteristics of the problem seems to indicate that it is the sort that
might be handled by multi-dimensional "OLAP packages", though not elegantly
given the cumbersome design of the rules languages of such packages.
Incidentally, "Math for the Layman" is by Ken Iverson and not Chris Burke.
Chris drew our attention to its availability. I agree that more such books
are needed to make J more accessible to a wider audience.
-----Original Message-----
From: HARRIS,BILL (HP-LakeStevens,ex1)
Sent: 6/16/99 11:23 PM
Subject: RE: Jforum: System implemented in J
Ajith Prasad asked:
Does the article explain the reasons for the choice of J as the
development
tool? What were the alternatives and why was J preferred? Thanks.
The article only gives a brief statement in that regard, mostly that an
earlier implementation with spreadsheets was error-prone. With a
spreadsheet being potentially an array of (hidden) formulas, it was too
easy
for the person administering the system to modify the formula in one
cell
for temporary purposes and forget to set it back.
I picked J for a number of reasons:
- pure perversity :-)
- the nature of the problem (arrays of numbers, with axes for
departments,
categories of expenses, months, as well as the date the data was
analyzed)
lends itself well to an array language
- J interfaced well (even better now, so it would seem -- I'm envious)
with
Excel for comfortable data entry, with data from legacy systems (via csv
files), and with the Windows API
- the high level and conciseness of the language made it easier for one
person to do the work (so I presumed)
After getting going, I discovered other benefits. I stored weekly
archives
of results. With an interpreted language, it was easy to run ad hoc
analyses of that data to meet various managers' needs. It was also
quite
easy to scan along sequences of archives to see trends in projections.
Graphical analyses and presentations were quite easily done, although I
occasionally wished I had more of the facilities of XLISP-STAT readily
available in analyzing and presenting data. I could add a GUI to
sysadmin
functions I did manually and pass the work off to others.
The worst thing was learning the language quickly with the material
available at the time. Chris' new "Math for Millions" (or whatever it
is
called) is a great step forward. It would be nice to have a "Ranks for
Millions,.Millions" as well. :-)
Bill
--
Bill Harris Year 2000 Program Office
mailto: bill_harris@am.exch.hp.com Hewlett-Packard Company
phone: (425) 335-2200 M/S 330C
fax: (425) 335-2483 8600 Soper Hill Road
web: http://hpweb.lsid.hp.com:8080/~billh/ Everett, WA 98205-1298
------------------------------------------------------------------------
--------
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 Jun 16 17:41:42 1999
Date: Wed, 16 Jun 1999 23:38:06 +0200
From: milan <milan.ice@bluewin.ch>
Subject: Jforum: attachments/reuters
boundary="------------0678FBA90437201D39622454"
--------------0678FBA90437201D39622454
I have 2 questions:
a) I have a J server in place to distribute data using FTP/ODBC etc.
that I use in a big company.
But there are situations where I should send attachments (which
might be Excel-Spreadsheets
or MsAccess databases) to specific E-mail addresses.
Does somebody know of a method or an already existing feature
accessible by J that would
allow me to send attachments to an E-mail address?
(Ps : the company uses outlook as its E-mail system and the MAPI
works fine with it ,but
I'm not sure if I can also use MAPI for sending attachments.)
b) are there any plans/ideas to access REUTERS market data by J ?
(I know such an access to REUTERS is not for free)
/Milan
--------------0678FBA90437201D39622454
<!doctype html public "-//w3c//dtd html 4.0 transitional//en">
<html>
<br> <font face="EuralDB"><font color="#000066">I have 2 questions:</font></font><font face="EuralDB"><font color="#000066"></font></font>
<p><font face="EuralDB"><font color="#000066"> a) I have a J server
in place to distribute data using FTP/ODBC etc. that I use in a big company.</font></font>
<br><font face="EuralDB"><font color="#000066">
But there are situations where I should send attachments (which might be
Excel-Spreadsheets</font></font>
<br><font face="EuralDB"><font color="#000066">
or MsAccess databases) to specific E-mail addresses.</font></font>
<br><font face="EuralDB"><font color="#000066"> </font></font>
<br><font face="EuralDB"><font color="#000066"> Does
somebody know of a method or an already existing feature accessible by
J that would</font></font>
<br><font face="EuralDB"><font color="#000066"> allow
me to send attachments to an E-mail address?</font></font>
<br><font face="EuralDB"><font color="#000066"> (Ps :
the company uses outlook as its E-mail system and the MAPI works fine with
it ,but</font></font>
<br><font face="EuralDB"><font color="#000066">
I'm not sure if I can also use MAPI for sending attachments.)</font></font><font face="EuralDB"><font color="#000066"></font></font>
<p><font face="EuralDB"><font color="#000066"> b) are there any plans/ideas
to access REUTERS market data by J ?</font></font>
<br><font face="EuralDB"><font color="#000066"> (I know
such an access to REUTERS is not for free)</font></font><font face="EuralDB"><font color="#000066"></font></font>
<p><font face="EuralDB"><font color="#000066"> /Milan</font></font>
<br><font face="EuralDB"><font color="#000066"> </font></font><font face="EuralDB"><font color="#000066"></font></font>
<p><font face="EuralDB"><font color="#000066"> </font></font>
<br><font face="EuralDB"><font color="#000066"> </font></font>
<p> </html>
--------------0678FBA90437201D39622454--
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Wed Jun 16 18:56:32 1999
From: "Alex Kornilovski" <kornal@idirect.com>
Subject: Re: Jforum: attachments/reuters
Date: Wed, 16 Jun 1999 19:00:47 -0400
Content-Type: multipart/mixed;
boundary="----=_NextPart_000_0011_01BEB82A.8B1B1E20"
This is a multi-part message in MIME format.
------=_NextPart_000_0011_01BEB82A.8B1B1E20
charset="iso-8859-1"
Milan wrote:
> Does somebody know of a method or an already existing feature
accessible by J that would
> allow me to send attachments to an E-mail address?
> (Ps : the company uses outlook as its E-mail system and the MAPI works
fine with it ,but
> I'm not sure if I can also use MAPI for sending attachments.)
Attached a bit modified version of Henry's MAPI interface that allows you to
send attachment (I was able to send attachment to myself and my
zdnetmail.com accounts). I am using Outlook express.
Alex
------=_NextPart_000_0011_01BEB82A.8B1B1E20
Content-Type: application/octet-stream;
name="Mapi.ijs"
Content-Transfer-Encoding: quoted-printable
Content-Disposition: attachment;
filename="Mapi.ijs"
NB. Simple MAPI Interface for Windows.
NB.
NB. Installs into _z_ locale.
NB.
NB. Supports one file as an attachment.=20
NB. Does not request dialog boxes.
NB. Does not support non-IPMs. Does not support credentialed messaging =
systems.
NB. Supports only 64-byte message IDs (because I couldn't find the =
definition of the
NB. flag to support 512-byte IDs).
NB.
NB. You log on to establish a session, do mail transactions, and log =
off.
NB. Sample sequence to process messages with qualifying subject lines =
might be:
NB.
NB. sess =3D. logonmail '' NB. Establish session.
NB. subjects =3D. sess collectmail 192 NB. Get subjects only, don't =
mark as read
NB. desiredmsgs =3D. 0 {"1 (#~ qualifying_verb) subjects NB. Get IDs of =
messages to read
NB. texts =3D. sess readmail desiredmsgs ,"0 _ <2048 NB. Get full text, =
no attachments
NB. sess deletemail desiredmsgs NB. Delete messages after processing
NB. logoffmail sess NB. end session
NB.
NB. sendmail is used to send outbound mail, with To, Cc, and Bcc =
recipients optional.
NB. An example, with only a To recipient:
NB. sess sendmail (1 ; < ,: 'Henry Rich';'SMTP:glasss@mindspring.com') ; =
'wow' ; 'Great Work!'
NB.
NB. Henry H. Rich (glasss@mindspring.com), February 1999
NB. Alex Kornilovski <kornal@idirect.com>, Marh 1999
require 'dll'
NB. Utility verbs
NB. Convert a set of bytes, MSB first, into a memory address
cpuorder_z_ =3D: |. NB. Intel CPUs are little-endian; use ] for =
big-endian
NB. Extract the return code from a DLL: if DLL error, use it, else the =
return is the first element
dllretcode_z_ =3D: (0&{::)`] @. (1: =3D #)
NB. y. is character string, result is memory address of a copy of the =
string, null-terminated.
NB. But if the string is null, use 0 instead of making an allocation
allostring_z_ =3D: 0: ` ( (] [ [ memw (, 0&,@>:@#)~) mema@>:@# ) @. =
(*@#)
NB. y. is result from allostring; it is freed (if nonzero)
deallostring_z_ =3D: (''"_)`((''"_)@memf) @. (0&~:) "0
NB. Log on a mail session. Use shared session if there is one. y. is =
name of profile to use
NB. (default '', which means use default profile). In Windows 95, the =
profile is set by
NB. clicking on Control Panel|Mail.
NB. Result is session handle, 0 if no session
logonmail_z_ =3D: 3 : 0
logonrc =3D. 'Mapi32 MAPILogon i i i i i i *i' cd 0;(s =3D. allostring =
y.);0;0;0;(,_1)
deallostring s
if. 0 =3D dllretcode logonrc do.
{. 6 {:: logonrc NB. Convert response to scalar for use as memory =
address
else.
0 NB. If error, return the error code
end.
)
NB. Log off a mail session. y. is the session id. Return the DLL =
retcode (0 if success).
logoffmail_z_ =3D: 3 : 0 "0
dllretcode 'Mapi32 MAPILogoff i i i i i' cd y.;0;0;0
)
NB. Send email.
NB. x. is session id returned from logonmail
NB. y. is recipient_list ; subject_text (string) ; message_text =
(string);attachment_file_name
NB. recipient_list is list of recipient_type , < list of name (string) =
; address (string)
NB. where recipient_type is 0 (originator), 1 (to), 2 (cc) , or 3 (bcc)
NB. You may use your address book to translate the name, and leave the =
address empty.
NB. You may NOT leave the name empty, even though the spec says you may =
- pick a harmless
NB. nonnull. If you use an Internet address, be sure to prefix it with =
'SMTP:'
NB. Example: sess sendmail (1 ; < ,: 'Alex';'SMTP:kornal@idirect.com') ; =
'attachment' ; 'text';'c:\bootlog.txt'
sendmail_z_ =3D: 4 : 0 "0 1
NB. y. is (possibly list of) dwords, result is byte string for them
longwords =3D. 13 : ', cpuorder"1 (256 256 256 256 #: y.) { a.'
att=3D.;{:y=3D.4{.y.
y=3D.}: y
NB. y. is (possibly list of) dwords, result is byte string for them
longwords =3D. 13 : ', cpuorder"1 (256 256 256 256 #: y.) { a.'
nrecip =3D: _2 { 1 , $ recipsallolist =3D. ; recipsallo =3D. =
allostring@> L:_1 (1) {"1 reciplist =3D. 0 {:: y
recipallo =3D. allostring longwords ; (0 {"1 reciplist) (0: , [ , ] , 0 =
0"_)"_ 1 L:_1 recipsallo
ct=3D.adr=3D.str=3D.0
if. 0~:$att do.
adr=3D.mema {.$att
att memw adr,0,{.$att
str=3D.allostring longwords 0,0,_1,adr,0,0
ct=3D.1
end.
msgallo =3D. allostring longwords 0,(textallo =3D. allostring@> 1 2 { =
y),0,0,0,0,0,nrecip,recipallo,ct,str
mailrc =3D. dllretcode 'Mapi32 MAPISendMail i i i i i i' cd =
x.;0;msgallo;1;0
deallostring msgallo,recipallo,textallo, ,recipsallolist ,adr,str NB. =
Release storage for nonnull pointers
mailrc NB. Return DLL errcode for error calling DLL, otherwise the DLL =
retcode
)
NB. Get next message id. y. is list of messages read so far (boxed), or =
null to start with first message.
NB. x. is session (the return from logonmail)
NB. result is list of boxed strings: new list of messages read
findnextmail_z_ =3D: 4 : 0 "0 1
retarea =3D. mema 512
if. #y. do.
mailrc =3D. 'Mapi32 MAPIFindNext i i i i *c i i i' cd =
x.;0;0;({:y.),0;0;retarea
else.
mailrc =3D. 'Mapi32 MAPIFindNext i i i i i i i i' cd =
x.;0;0;0;0;0;retarea
end.
if. 0 =3D dllretcode mailrc do.
newresult =3D. <memr retarea,0,_1
else.
newresult =3D. 0 $ <''
end.
memf retarea
y.,newresult
)
NB. Read the header for a message. y. is the message_id;flags. Flags =
defaults to 0.
NB. (2048=3Dno attachments, 512=3Dmake text an attachment, =
128=3Ddon't mark
NB. the messages as read, 64=3Dread headers only, no text)
NB. x. is the session id returned from logonmail
NB. result is originator_name;originator_addr;subject;text;date for the =
message
readmail_z_ =3D: 4 : 0 "0 1
NB. y. is addr,offset,len; result is pointer value
loadint =3D. 13 : '256 #. a. i. cpuorder^:_1 memr y.'
NB. y. is structure_address,offset
NB. Result is the string pointed to by the pointer at that offset, or '' =
if pointer is null
loadstring =3D. ( (''"_) ` (memr@(,&0 _1)) @. (0&~:) ) @ loadint @ (,&4) =
"1
'mid flgs' =3D. 2 {. y. , <0
if. mid -: '' do. '';'';'';'';'' return. end.
if. 0 =3D dllretcode mailrc =3D. 'Mapi32 MAPIReadMail i i i *c i i *i' =
cd x.;0;mid;flgs;0;(,_1) do.
pmsgstr =3D. {. 6 {:: mailrc
if. 0 ~: precipstr =3D. loadint pmsgstr,28,4 do.
onmaddr =3D. <@loadstring precipstr ,"_ 0 (8 12)
else.
onmaddr =3D. '';''
end.
osubtxdt =3D. <@loadstring pmsgstr ,"_ 0 (4 8 16)
'Mapi32 MAPIFreeBuffer i i' cd pmsgstr NB. Free the message and all =
its strings
onmaddr , osubtxdt
else.
'';'';'';'';''
end.
)
NB. y. is message id (boxed). x. is session returned from logonmail. =
The message is deleted.
NB. Result is return code from delete
deletemail_z_ =3D: 4 : 0 "0
dllretcode 'Mapi32 MAPIDeleteMail i i i *c i i ' cd x.;0;y.,0;0
)
NB. x. is session id, y. is flags (2048=3Dno attachments, 512=3Dmake =
text an attachment, 128=3Ddon't mark
NB. the messages as read, 64=3Dread headers only, no text)
NB. Result is message_id;origin_name;origin_addr;subject;text;rcvdate =
for all mail messages
collectmail_z_ =3D: 4 : '(,. x.&readmail@(,&(<y.))"0) x.&findnextmail^:_ =
(0$a:)' "0 0
------=_NextPart_000_0011_01BEB82A.8B1B1E20--
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Wed Jun 16 21:37:14 1999
From: k-list@iname.com
Date: Wed, 16 Jun 1999 21:35:00 -0400 (EDT)
Content-Type: Text/Plain
Subject: Jforum: double-byte & ocx
In j4.03, I can't enter Chinese (double-btye) text into ocx from keyboard, (more specifically, farpoint input pro 3.0.6 Edit control). However, the same can be done using VB.
By the way, double-byte text can be read or written to ocx using wd'oleset.
---------------------------------------------------
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 Jun 17 10:05:27 1999
From: "Seymour Glass" <glasss@mindspring.com>
Subject: RE: Jforum: attachments/reuters
Date: Thu, 17 Jun 1999 10:01:24 -0400
Content-Type: multipart/mixed;
boundary="----=_NextPart_000_0001_01BEB8A8.5BE7FFA0"
Importance: Normal
In-Reply-To: <001401beb84c$137ae900$38f3a1d1@alex>
This is a multi-part message in MIME format.
------=_NextPart_000_0001_01BEB8A8.5BE7FFA0
charset="iso-8859-1"
> -----Original Message-----
> From: owner-jsoftware@lists.interlog.com
> [mailto:owner-jsoftware@lists.interlog.com]On Behalf Of Alex Kornilovski
> Sent: Wednesday, June 16, 1999 19:01
> To: forum@jsoftware.com
> Subject: Re: Jforum: attachments/reuters
>
>
> Milan wrote:
>
> > Does somebody know of a method or an already existing feature
> accessible by J that would
> > allow me to send attachments to an E-mail address?
> > (Ps : the company uses outlook as its E-mail system and the
> MAPI works
> fine with it ,but
> > I'm not sure if I can also use MAPI for sending attachments.)
>
> Attached a bit modified version of Henry's MAPI interface that
> allows you to
> send attachment (I was able to send attachment to myself and my
> zdnetmail.com accounts). I am using Outlook express.
Thanks, Alex. I have taken your changes and inserted them into
my current version of mapi.ijs (which also includes support for
inbound attachments).
I modified the attachment specification from a single character
string to an array with shape (n,1). The n is the number of attachments -
you can send as many as you want. I made each attachment specification
a 1-element boxed list (rather than a single box) so that we can
add support for attachment position & other stuff later if we want
to.
This format is incompatible with your use of a single character
string, but I have put in code to support the single character
string as a special case so your current applications will still work.
I have attached the current version in case anyone wants it right away.
------=_NextPart_000_0001_01BEB8A8.5BE7FFA0
Content-Type: application/octet-stream;
name="mapi.ijs"
Content-Transfer-Encoding: quoted-printable
Content-Disposition: attachment;
filename="mapi.ijs"
NB. Simple MAPI Interface for Windows.
NB.
NB. Installs into _z_ locale.
NB.
NB. Does not request dialog boxes. Does not support non-IPMs.
NB. Does not support credentialed messaging systems.
NB.
NB. You log on to establish a session, do mail transactions, and log =
off.
NB. Sample sequence to process messages with qualifying subject lines =
might be:
NB.
NB. sess =3D. logonmail '' NB. Establish session. On some systems the =
argument (a profile name) is required
NB. subjects =3D. sess collectmail 192 NB. Get subjects only, don't =
mark as read
NB. desiredmsgs =3D. 0 {"1 (#~ qualifying_verb) subjects NB. Get IDs of =
messages to read
NB. texts =3D. sess readmail desiredmsgs ,"0 _ <2048 NB. Get full text, =
no attachments
NB. sess deletemail desiredmsgs NB. Delete messages after processing
NB. logoffmail sess NB. end session
NB.
NB. sendmail is used to send outbound mail, with To, Cc, and Bcc =
recipients optional.
NB. Attachments are optional.
NB. A session is required.
NB. An example, with only a To recipient:
NB. sess sendmail (1 ; < ,: 'Henry Rich';'SMTP:glasss@mindspring.com') ; =
'wow' ; 'Great Work!'
NB.
NB. Henry H. Rich (glasss@mindspring.com), March 1999
NB. Alex Kornilovski <kornal@idirect.com>, March 1999
require 'dll'
NB. Utility verbs
NB. Extract the return code from a DLL: if DLL error, use it, else the =
return is the first element
dllretcode_z_ =3D: (0&{::)`] @. (1: =3D #)
NB. y. is character string, result is memory address of a copy of the =
string, null-terminated.
NB. But if the string is null, use 0 instead of making an allocation
allostring_z_ =3D: 0: ` ( (] [ [ memw (, 0&,@>:@#)~) mema@>:@# ) @. =
(*@#)
NB. y. is result from allostring; it is freed (if nonzero)
deallostring_z_ =3D: (''"_)`((''"_)@memf) @. (0&~:) "0
NB.*logonmail v-- Establish mail session
NB. Log on a mail session. Use shared session if there is one. y. is =
name of profile to use
NB. (default '', which means use default profile). In Windows 95, the =
profile is set by
NB. clicking on Control Panel|Mail.
NB. Result is session handle, 0 if no session
logonmail_z_ =3D: 3 : 0
logonrc =3D. 'Mapi32 MAPILogon i i i i i i *i' cd 0;(s =3D. allostring =
y.);0;0;0;(,_1)
deallostring s
if. 0 =3D dllretcode logonrc do.
{. 6 {:: logonrc NB. Convert response to scalar for use as memory =
address
else.
0 NB. If error, return the error code
end.
)
NB.*logoffmail v-- Terminate mail session
NB. Log off a mail session. y. is the session id. Return the DLL =
retcode (0 if success).
logoffmail_z_ =3D: 3 : 0 "0
dllretcode 'Mapi32 MAPILogoff i i i i i' cd y.;0;0;0
)
NB.*sendmail v-- Send email
NB. x. is session id returned from logonmail
NB. y. is recipient_list ; subject_text (string) ; message_text (string) =
; (optional) attachments=20
NB. recipient_list is list of recipient_type , < list of name (string) =
; address (string)
NB. where recipient_type is 0 (originator), 1 (to), 2 (cc) , or 3 (bcc)
NB. attachments is (possibly a list of) boxed lists representing =
attachment files
NB. each list has just one element: the filename (other elements will =
be added in future)
NB. You may use your address book to translate the name, and leave the =
address empty.
NB. You may NOT leave the name empty, even though the spec says you may =
- pick a harmless
NB. nonnull. If you use an Internet address, be sure to prefix it with =
'SMTP:'
NB. Example: sess sendmail (1 ; < ,: 'name';'SMTP:addr') ; 'subject' ; =
'text' ;< (1 1 {body}lt;'c:att1.txt'),(1 1 {body}lt;'c:att2.txt')
sendmail_z_ =3D: 4 : 0 "0 1
NB. y. is (possibly list of) dwords, result is byte string for them
longwords =3D. 2&(3!:4)
NB. Allocate the message test
nrecip =3D. _2 { 1 , $ recipsallolist =3D. ; recipsallo =3D. =
allostring@> L:_1 (1) {"1 reciplist =3D. 0 {:: y.
recipallo =3D. allostring longwords , ; (0 {"1 reciplist) (0: , [ , ] , =
0 0"_)"_ 1 L:_1 recipsallo
NB. Allocate the attachments
if. natt =3D. _2 { 0 1 , $ att=3D. ,@<^:(2:=3D(3!:0)) 3 {:: y. , (<a:) =
do. NB. get attachment list (default null)
NB. the ,@<^:(2:=3D(3!:0)) is to support single character strings, =
now deprecated
attsallo =3D. allostring@> , 0 {"1 att NB. allocate attachment =
strings
attallo =3D. allostring longwords , (0 0 _1"_ , ] , 0 0"_)"0 attsallo =
NB. allocate file structures
else.
attsallo =3D. attallo =3D. 0
end.
msgallo =3D. allostring longwords 0,(textallo =3D. allostring@> 1 2 { =
y.),0,0,0,0,0,nrecip,recipallo,natt,attallo
mailrc =3D. dllretcode 'Mapi32 MAPISendMail i i i i i i' cd =
x.;0;msgallo;1;0
deallostring msgallo,recipallo,textallo,attallo,attsallo, =
,recipsallolist NB. Release storage for nonnull pointers
mailrc NB. Return DLL errcode for error calling DLL, otherwise the DLL =
retcode
)
NB. Get next message id. y. is list of messages read so far (boxed), or =
null to start with first message.
NB. x. is session (the return from logonmail)
NB. result is list of boxed strings: new list of messages read
findnextmail_z_ =3D: 4 : 0 "0 1
retarea =3D. mema 512
if. #y. do.
mailrc =3D. 'Mapi32 MAPIFindNext i i i i *c i i i' cd =
x.;0;0;({:y.),16384;0;retarea
else.
mailrc =3D. 'Mapi32 MAPIFindNext i i i i i i i i' cd =
x.;0;0;0;16384;0;retarea
end.
if. 0 =3D dllretcode mailrc do.
newresult =3D. <memr retarea,0,_1
else.
newresult =3D. 0 $ <''
end.
memf retarea
y.,newresult
)
NB.*readmail v-- Read an email message
NB. y. is the message_id;flags. Flags defaults to 0.
NB. (2048=3Dno attachments, 512=3Dmake text an attachment, =
128=3Ddon't mark
NB. the messages as read, 64=3Dread headers only, no text)
NB. x. is the session id returned from logonmail
NB. result is =
originator_name;originator_addr;subject;text;date;attchments for the =
message
NB. attachments are a list of =
character_position;filename;original_filename
NB. position=3D_1 means end-of-file
NB. NOTE: future releases may add items to the returned value
readmail_z_ =3D: 4 : 0 "0 1
NB. y. is addr,offset,len; result is pointer value
loadint =3D. 13 : '_2 (3!:4) memr y.'
NB. y. is structure_address,offset
NB. Result is the string pointed to by the pointer at that offset, or '' =
if pointer is null
loadstring =3D. ( (''"_) ` (memr@(,&0 _1)) @. ({.@(0&~:)) ) @ loadint @ =
(,&4) "1
'mid flgs' =3D. 2 {. y. , <0
if. mid -: '' do. '';'';'';'';'' return. end.
if. 0 =3D dllretcode mailrc =3D. 'Mapi32 MAPIReadMail i i i *c i i *i' =
cd x.;0;mid;flgs;0;(,_1) do.
pmsgstr =3D. {. 6 {:: mailrc
if. 0 ~: precipstr =3D. {. loadint pmsgstr,28,4 do.
onmaddr =3D. <@loadstring precipstr ,"_ 0 (8 12)
else.
onmaddr =3D. '';''
end.
if. 0 ~: *./ 'natt aatt' =3D. loadint pmsgstr,40,8 do.
cpos =3D. <@{.@loadint"1 aatt ,"_ 1 (8 + 24 * i. natt) ,"0 (4) NB. =
character position
cfn =3D. <@loadstring aatt ,"_ 0 (12 16 +/~ 24 * i. natt) NB. full =
filename, original filename
onatt =3D. cpos ,. cfn
else.
onatt =3D. 0 3 $ a:
end.
osubtxdt =3D. <@loadstring pmsgstr ,"_ 0 (4 8 16)
'Mapi32 MAPIFreeBuffer i i' cd pmsgstr NB. Free the message and all =
its strings
onmaddr , osubtxdt , <onatt
else.
'';'';'';'';''
end.
)
NB.*deletemail v-- Delete a mail message.
NB. y. is message id (boxed). x. is session returned from logonmail.
NB. Result is return code from delete
deletemail_z_ =3D: 4 : 0 "0
dllretcode 'Mapi32 MAPIDeleteMail i i i *c i i ' cd x.;0;y.,0;0
)
NB.*collectmail v-- Retrieve all mail, or just the headers
NB. x. is session id, y. is flags (2048=3Dno attachments, 512=3Dmake =
text an attachment, 128=3Ddon't mark
NB. the messages as read, 64=3Dread headers only, no text)
NB. Result is =
message_id;origin_name;origin_addr;subject;text;rcvdate;attachments for =
all mail messages
NB. attachments are a list of =
character_position;filename;original_filename
NB. position=3D_1 means end-of-file
NB. NOTE: future releases may add items to the returned value
collectmail_z_ =3D: 4 : '(,. x.&readmail@(,&(<y.))"0) x.&findnextmail^:_ =
(0$a:)' "0 0
------=_NextPart_000_0001_01BEB8A8.5BE7FFA0--
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Thu Jun 17 13:50:01 1999
Date: 16 Jun 99 09:47:20 -0800
Subject: RE: Jforum: Passing a verb to a verb
From: "Piet de Jong" <piet@dejong.u-net.com>
Thanks for all the suggestions regarding "passing a verb to a verb". They
were all very helpful. Still digesting some of the suggestions.
The topic has led me to a wider understanding of the power of J. Despite
the frustrations, I am once again reaffirming my vows to the language.
One source of frustration is a little portion of the Dictionary which I
often stare at, trying to understand :
"An adverb is produced according to the following definition (using nv to
denote noun or verb):
x (a1 a2) is x a1 a2
x (c nv) is x c nv
x (nv c) is nv c x
x (a c) is x a c x "
That's it. The nagging question is always "What is x?" A noun, a verb or
what? Probably a verb or an appropriate noun, that is provided it works
with say the conjunction.
Presumably a1, a2 and a are adverbs. Somehow I don't have an
"understanding" of these constructs, at least sufficient to use them
effectively.
Similar learning difficulties occur with the description of "conjunction
trains". Difficulties appear to vanish once explicit constructs are used.
But it would be nice to really "understand" (in an effective operational
sense) the tacit constructs. This is because part of J appears to be a
push to reformulate mathematical notation in ways that are more consistent
with computational constraints and demands. If Iverson thinks the tacit
constructs are useful, that is enough motivation for me to try and
understand and see why they are useful.
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Thu Jun 17 14:44:37 1999
From: "Seymour Glass" <glasss@mindspring.com>
Subject: RE: Jforum: Passing a verb to a verb
Date: Thu, 17 Jun 1999 14:41:47 -0400
charset="iso-8859-1"
Importance: Normal
In-Reply-To: <B38D236E-BB359@204.244.233.44>
> -----Original Message-----
> From: owner-jsoftware@lists.interlog.com
> [mailto:owner-jsoftware@lists.interlog.com]On Behalf Of Piet de Jong
> Sent: Wednesday, June 16, 1999 13:47
> To: forum@jsoftware.com
> Subject: RE: Jforum: Passing a verb to a verb
>
>
> ...One source of frustration is a little portion of the Dictionary which I
> often stare at, trying to understand :
>
> "An adverb is produced according to the following definition (using nv to
> denote noun or verb):
>
> x (a1 a2) is x a1 a2
> x (c nv) is x c nv
> x (nv c) is nv c x
> x (a c) is x a c x "
>
> That's it. The nagging question is always "What is x?" A noun, a verb or
> what? Probably a verb or an appropriate noun, that is provided it works
> with say the conjunction.
>
> Presumably a1, a2 and a are adverbs. Somehow I don't have an
> "understanding" of these constructs, at least sufficient to use them
> effectively.
I can't find the text you are referring to. The complete definition
is in Dictionary II.F, the parse table, which you should have taped
to your mirror for study while shaving each morning. It shows several
tridents that produce adverbs too.
You have it right, that x is a verb or a noun acceptable in the context.
You create an adverb when you want to modify the behavior of many
different verbs in the same way. Adverbs are less often used with
noun left arguments; } is the most important case.
If you have any specific questions, ask. Here are a few adverbs from
my utilities; you can use them as exercises. After you master adverbs,
you can move up to conjunctions!
NB. Adverb: like @], but performs u. before ], which is faster if the result
is smaller than y.
onyd_z_ =: ((]`]:) `:6)
NB. Adverb: like @], but avoids the ] if monadic
ony_z_ =: onyd (]. : [.)
NB. Adverb: like @[, but performs u. before [, which is faster if the result
is smaller than x.
onx_z_ =: ((]`]:) `:6) ~
NB. Adverb. Do u., but skip it if y. is null (result then is '')
onlyifnonnull_z_ =: ((''"_)`]:) @. (*@# onyd)
NB. Adverb. Monad, converts y. to rank m. by adding axes
enrank_z_ =: ,: ^: ((0&>.)@(]:&-)@#@$)
NB. adverb: purge memory, then do u.
afterpurging_z_ =: ]: [ (7!:4 @ (($0)"_))
>
> Similar learning difficulties occur with the description of "conjunction
> trains". Difficulties appear to vanish once explicit constructs
> are used.
> But it would be nice to really "understand" (in an effective operational
> sense) the tacit constructs. This is because part of J appears to be a
> push to reformulate mathematical notation in ways that are more consistent
> with computational constraints and demands. If Iverson thinks the tacit
> constructs are useful, that is enough motivation for me to try and
> understand and see why they are useful.
Enough for me, too. That reminds me of a question. The definition
of & says that 'Use of the bond conjunction is often called
Currying in honour of Haskell Curry'. That meant nothing to me, but
I recently finished reading an old book that mentions Curry and
his logic of 'combinators' which provided a notation for introducing
constants into functional programs, an advance (the book said) over Church's
lambda-calculus.
So, is KEI's homage to Curry a recognition of this important idea, or
for Curry's other contributions to functional programming, or what?
Henry Rich
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Thu Jun 17 16:02:33 1999
From: Andrew Nikitin <anikitin@fastenal.com>
Subject: Jforum: editor utilities
Date: Thu, 17 Jun 1999 14:58:48 -0500
The new version of editor utilities is available from
http://members.xoom.com:80/nsg21/jedit.zip
It uses new editor features like wd 'setreplace' and wd 'getscroll'. Old
code is retained for comptibility though.
This utilities are being bound to function keys and perform the
following functions:
F2 - show list of definitions
F4 - comment/decomment block
F5 - indent block
F6 - unindent block
F9 - select entire explicit definition for running/deleting/moving
BTW, I use this expression to choose actual advrb code depending on J
version:
Selection=: 1 : (>(4.03>:".>{.<;._2 wd'qabout'){(0 :0);(0 :0))
--- new explicit code (for versions 4.03 and above)
)
--- old explicit code(for versions less than 4.03)
)
and it seems pretty ugly to me. (In addition this will not work for beta
versions of J).
Could anybody suggest better solution?
nsg
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Thu Jun 17 16:53:40 1999
From: <paul_gauthier@dynasys.tm.fr>
Subject: RE: Jforum: editor utilities
Date: Thu, 17 Jun 1999 22:50:39 +0200
Content-Transfer-Encoding: 8bit
X-MIME-Autoconverted: from quoted-printable to 8bit by plus.interlog.com id QAA18902
Not much of an improvement but I am curious to know if it covers beta
versions, I can't test it.
Selection=: 1 :(>(4.03>:".>0{;:9!:14''){(0 :0);(0 :0))
--- new explicit code (for versions 4.03 and above)
)
--- old explicit code(for versions less than 4.03)
)
>-----Message d'origine-----
>De: Andrew Nikitin [SMTP:anikitin@fastenal.com]
>Date: jeudi 17 juin 1999 21:59
>�: 'forum@jsoftware.com'
>Objet: Jforum: editor utilities
>BTW, I use this expression to choose actual advrb code depending on J
>version:
>
>Selection=: 1 : (>(4.03>:".>{.<;._2 wd'qabout'){(0 :0);(0 :0))
>--- new explicit code (for versions 4.03 and above)
>)
>--- old explicit code(for versions less than 4.03)
>)
>
>and it seems pretty ugly to me. (In addition this will not work for beta
>versions of J).
>Could anybody suggest better solution?
>
>
>nsg
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Thu Jun 17 18:23:42 1999
Date: Thu, 17 Jun 1999 18:20:33 -0400
From: Fraser Jackson <Fraser_Jackson@compuserve.com>
Subject: Jforum: Displaying tables and constructing cross tabulations
charset=ISO-8859-1
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
X-MIME-Autoconverted: from quoted-printable to 8bit by plus.interlog.com id SAA07784
Displaying arrays with labels.
It is often valuable to have arrays printed with labels for the rows,
columns, planes, etc. The function Tables lists an array with
descriptive information for the locations along each dimension. For a
three dimensional array it has as right argument dim0;dim1;dim2;array
where dimi is a character matrix or numeric vector with one item for
each item along the i'th dimension of the array.
The expression
Tables dim0;dim1;dim2;array
displays the array with all rows and columns labelled, and separate
displays with an appropriate heading for all rank two objects.
Tabulation
Often for a database we want to count the items which have each
combination of values on several characteristics.
The function Tabulate constructs such a table, and lists of the values of
the characteristics. If the characteristics are in var0, var1, var2 then
the expression
Tabulate var0;var1;var2
will construct the required table of counts and display the result
dim0;dim1;dim2;counts
with dimi the sorted nub of the items in the corresponding variable, and
the array counts containing the number of cases for each combination of the
values of the variables.
Tables Tabulate var0;var1;var2
will list the result in a convenient and readable form.
If you are using mapped files see the note in the script. I have found it
better to construct an index file and use the function TabulateI instead.
Download the script tabulate.zip from //www. jsoftware.com. Any
comments are welcome to Fraser_Jackson@compuserve.com
Fraser
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Thu Jun 17 19:21:26 1999
From: "Seymour Glass" <glasss@mindspring.com>
Subject: RE: Jforum: editor utilities
Date: Thu, 17 Jun 1999 19:17:45 -0400
charset="iso-8859-1"
Content-Transfer-Encoding: 8bit
Importance: Normal
In-Reply-To: <c=FR%a=_%p=DynaSys_-_Schilt%l=DYNASYS01-990617205039Z-394@mail.dynasys.tm.fr>
That was my solution too (maybe with 4{. before the ". if that would
work on betas). But I'm curious: what changes in 4.03 to make a
different version required?
Henry Rich
> -----Original Message-----
> From: owner-jsoftware@lists.interlog.com
> [mailto:owner-jsoftware@lists.interlog.com]On Behalf Of
> paul_gauthier@dynasys.tm.fr
> Sent: Thursday, June 17, 1999 16:51
> To: forum@jsoftware.com
> Subject: RE: Jforum: editor utilities
>
>
> Not much of an improvement but I am curious to know if it covers beta
> versions, I can't test it.
> Selection=: 1 :(>(4.03>:".>0{;:9!:14''){(0 :0);(0 :0))
> --- new explicit code (for versions 4.03 and above)
> )
> --- old explicit code(for versions less than 4.03)
> )
>
> >-----Message d'origine-----
> >De: Andrew Nikitin [SMTP:anikitin@fastenal.com]
> >Date: jeudi 17 juin 1999 21:59
> >�: 'forum@jsoftware.com'
> >Objet: Jforum: editor utilities
> >BTW, I use this expression to choose actual advrb code depending on J
> >version:
> >
> >Selection=: 1 : (>(4.03>:".>{.<;._2 wd'qabout'){(0 :0);(0 :0))
> >--- new explicit code (for versions 4.03 and above)
> >)
> >--- old explicit code(for versions less than 4.03)
> >)
> >
> >and it seems pretty ugly to me. (In addition this will not work for beta
> >versions of J).
> >Could anybody suggest better solution?
> >
> >
> >nsg
>
> ------------------------------------------------------------------
> --------------
> 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 Jun 17 19:25:42 1999
From: Ajith Prasad <burap@nus.edu.sg>
Subject: RE: Jforum: editor utilities
Date: Fri, 18 Jun 1999 07:23:45 +0800
charset="windows-1252"
Whic directory should it be unzipped into? Thanks.
-----Original Message-----
From: Andrew Nikitin
Sent: 6/18/99 3:58 AM
Subject: Jforum: editor utilities
The new version of editor utilities is available from
http://members.xoom.com:80/nsg21/jedit.zip
It uses new editor features like wd 'setreplace' and wd 'getscroll'. Old
code is retained for comptibility though.
This utilities are being bound to function keys and perform the
following functions:
F2 - show list of definitions
F4 - comment/decomment block
F5 - indent block
F6 - unindent block
F9 - select entire explicit definition for running/deleting/moving
BTW, I use this expression to choose actual advrb code depending on J
version:
Selection=: 1 : (>(4.03>:".>{.<;._2 wd'qabout'){(0 :0);(0 :0))
--- new explicit code (for versions 4.03 and above)
)
--- old explicit code(for versions less than 4.03)
)
and it seems pretty ugly to me. (In addition this will not work for beta
versions of J).
Could anybody suggest better solution?
nsg
------------------------------------------------------------------------
--------
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 Jun 17 19:33:22 1999
From: Andrew Nikitin <anikitin@fastenal.com>
Subject: RE: Jforum: editor utilities
Date: Thu, 17 Jun 1999 18:29:11 -0500
> -----Original Message-----
> From: Seymour Glass [SMTP:glasss@mindspring.com]
> Sent: Thursday, June 17, 1999 18:18
>
> That was my solution too (maybe with 4{. before the ". if that would
> work on betas). But I'm curious: what changes in 4.03 to make a
> different version required?
>
New version takes benefit of two new wd commands: getscroll and
setreplace. The first allows to restore scroll position after performing
desired changes. (Old version of J cannot do this and that causes screen
flickering). Second smoothes thing a bit in large files (that is not
really nesessary, but why not to use useful feature?).
nsg
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Thu Jun 17 19:42:38 1999
From: Andrew Nikitin <anikitin@fastenal.com>
Subject: RE: Jforum: editor utilities
Date: Thu, 17 Jun 1999 18:40:28 -0500
> From: Ajith Prasad [SMTP:burap@nus.edu.sg]
> Sent: Thursday, June 17, 1999 18:24
>
> Whic directory should it be unzipped into? Thanks.
>
Wherever you wish. As for me, I unzipped it into 'user/' directory and
added line
load 'user/jedit'
to the end of system\extras\config\profile.ijs
nsg
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Fri Jun 18 13:32:25 1999
From: Andromeda <milan.ice@bluewin.ch>
Subject: Jforum: RE:
Date: Fri, 18 Jun 1999 19:32:06 +0200
Content-Type: multipart/mixed; boundary="---- =_NextPart_000_01BEB9C1.40B84F40"
------ =_NextPart_000_01BEB9C1.40B84F40
Martin,
the mapi.ijs from Alex is a good solution so I must not look for sth else
but thanks anyway for the proposed alternative.
/regards - Milan
----------
From: Martin Neitzel[SMTP:neitzel@gaertner.de]
Sent: Donnerstag, 17. Juni 1999 05:02
From: neitzel@gaertner.de (Martin Neitzel)
Subject: Re: Jforum: attachments/reuters
Organization: Gaertner Datensysteme, Braunschweig, Germany
References: <3768193D.F6F05740@bluewin.ch>
Date: Thu, 17 Jun 1999 03:02:39 GMT
milan <milan.ice@bluewin.ch> wrote:
>
>--------------0678FBA90437201D39622454
>Content-Type: text/plain; charset=us-ascii
>Content-Transfer-Encoding: 7bit
>
>
> I have 2 questions:
>
> a) I have a J server in place to distribute data using FTP/ODBC etc.
>that I use in a big company.
> But there are situations where I should send attachments (which
>might be Excel-Spreadsheets
> or MsAccess databases) to specific E-mail addresses.
>
> Does somebody know of a method or an already existing feature
>accessible by J that would
> allow me to send attachments to an E-mail address?
> (Ps : the company uses outlook as its E-mail system and the MAPI
>works fine with it ,but
> I'm not sure if I can also use MAPI for sending attachments.)
>
I'm not familiar with the MS MAPI but Henry's code looks like
attachments are possible.
Making attachments on your own is easy and just a matter two things:
(1) You have to create the very straightforward MIME headers and
sub part headers (for an example, I quoted your original question
as sent with your original boundary line and sub part header)
(2) Binary subparts such as an excel file must be ascii encoded
using "base64" encoding.
I have written simple base64 encoding and decoding utilities in J three
years ago. Should I make them available?
Martin
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
------ =_NextPart_000_01BEB9C1.40B84F40
Content-Type: application/ms-tnef
Content-Transfer-Encoding: base64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------ =_NextPart_000_01BEB9C1.40B84F40--
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Fri Jun 18 14:47:10 1999
From: Andromeda <milan.ice@bluewin.ch>
Subject: RE: Jforum: attachments/reuters
Date: Fri, 18 Jun 1999 19:27:07 +0200
Content-Type: multipart/mixed; boundary="---- =_NextPart_000_01BEB9C0.8F1DB220"
------ =_NextPart_000_01BEB9C0.8F1DB220
Hi Alex ,
thank you very much for your fast answer.The script does exactly what I wanted.
/cheers - Milan
----------
From: Alex Kornilovski[SMTP:kornal@idirect.com]
Sent: Donnerstag, 17. Juni 1999 01:00
Subject: Re: Jforum: attachments/reuters
<<File: Mapi.ijs>>
Milan wrote:
> Does somebody know of a method or an already existing feature
accessible by J that would
> allow me to send attachments to an E-mail address?
> (Ps : the company uses outlook as its E-mail system and the MAPI works
fine with it ,but
> I'm not sure if I can also use MAPI for sending attachments.)
Attached a bit modified version of Henry's MAPI interface that allows you to
send attachment (I was able to send attachment to myself and my
zdnetmail.com accounts). I am using Outlook express.
Alex
------ =_NextPart_000_01BEB9C0.8F1DB220
Content-Type: application/ms-tnef
Content-Transfer-Encoding: base64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------ =_NextPart_000_01BEB9C0.8F1DB220--
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Fri Jun 18 19:14:51 1999
From: Andrew Nikitin <anikitin@fastenal.com>
Subject: Jforum: scriptdoc fails on gl2
Date: Fri, 18 Jun 1999 18:11:23 -0500
Content-Type: multipart/mixed;
boundary="---- =_NextPart_000_01BEB9B5.F959BEA0"
This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.
------ =_NextPart_000_01BEB9B5.F959BEA0
I use the following replacement to the standard 'gl2' library (see
attached zip). When I run scriptdoc on this file it shows help, but if i
check on summary button, then scriptdoc fails with
error in: sdoc_sum_button
result error: wd
wd'set view *',dat
messagebox.
What is wrong with this file?
nsg
------ =_NextPart_000_01BEB9B5.F959BEA0
Content-Type: application/octet-stream;
name="gl2.zip"
Content-Transfer-Encoding: base64
Content-Disposition: attachment;
filename="gl2.zip"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=
------ =_NextPart_000_01BEB9B5.F959BEA0--
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Sat Jun 19 03:36:33 1999
Date: Sat, 19 Jun 1999 09:31:17 +0200
From: "d.alis" <dalis@balcab.ch>
Subject: Re: Jforum: J for Windows CE
References: <002201be252f$43f9fd60$2d0114d1@xps200>
Which WinCE machines have, or can be connected to, a hard disk? I have looked at a
couple of HP web pages and distributer sites, but I havent found a site that gives
me a clear idea one way or the other.
Thanks for any advice.
David
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Sat Jun 19 13:25:03 1999
From: "David Vincent-Jones" <geomap@galaxynet.com>
References: <002201be252f$43f9fd60$2d0114d1@xps200> <376B4744.BF7E8E48@balcab.ch>
Subject: Re: Jforum: J for Windows CE
Date: Sat, 19 Jun 1999 09:53:33 -0700
I know that there have been several messages on WinCE but it would really be
great if somebody could put together a comprehensive blurb on the whole
subject as it relates to J.
>From some prior messages it sounded like even installation really was a
hassle!
What are optimal hardware needs?
Does the chipset make a big difference?
How is it installed?
Are there operating limitations?
How will performance compare?
----- Original Message -----
From: d.alis <dalis@balcab.ch>
Sent: June 19, 1999 12:31 AM
Subject: Re: Jforum: J for Windows CE
> Which WinCE machines have, or can be connected to, a hard disk? I have
looked at a
> couple of HP web pages and distributer sites, but I havent found a site
that gives
> me a clear idea one way or the other.
> Thanks for any advice.
> David
>
>
>
> --------------------------------------------------------------------------
------
> J Forum: for information about this list, see
http://www.jsoftware.com/forum.htm
>
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Sat Jun 19 17:27:30 1999
Date: Sun, 20 Jun 1999 00:33:27 +0300
From: Anssi Seppala <anssi.seppala@enease.fi>
Subject: Re: Jforum: J for Windows CE
In-Reply-To: <001c01beba76$bb44ad20$490881ce@desktop>
References: <002201be252f$43f9fd60$2d0114d1@xps200>
<376B4744.BF7E8E48@balcab.ch>
Content-Transfer-Encoding: 8bit
X-MIME-Autoconverted: from quoted-printable to 8bit by plus.interlog.com id RAA28882
I have hp620lx. You do not need a hard disk attached. I have a PCMCIA flash
memory card 15MB and my PC can read/write such memory. With the memory card
I can move a lot of data.
Optimal WINCE is system you can afford memory, performance, display cost
$$ and you need all of them with J. With J there must be also a keyboard.
The J language is quite similar but GUI is not. The CE GUI is very limited
and PC GUI is only partly protable from PC to CE.
Anssi
At 09:53 19.6.1999 -0700, you wrote:
>I know that there have been several messages on WinCE but it would really be
>great if somebody could put together a comprehensive blurb on the whole
>subject as it relates to J.
>>From some prior messages it sounded like even installation really was a
>hassle!
>
>What are optimal hardware needs?
>
>Does the chipset make a big difference?
>
>How is it installed?
>
>Are there operating limitations?
>
>How will performance compare?
>
>
>----- Original Message -----
>From: d.alis <dalis@balcab.ch>
>To: <forum@JSoftware.Com>
>Sent: June 19, 1999 12:31 AM
>Subject: Re: Jforum: J for Windows CE
>
>
>> Which WinCE machines have, or can be connected to, a hard disk? I have
>looked at a
>> couple of HP web pages and distributer sites, but I havent found a site
>that gives
>> me a clear idea one way or the other.
>> Thanks for any advice.
>> David
>>
>>
>>
>> --------------------------------------------------------------------------
>------
>> 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
-------------------------------------------------------------------
Anssi Sepp�l�
Konsultti, s�hk�markkinat ja s�hk�njakelu: Sovellukset ja ohjelmistot
e-mail: anssi.seppala@enease.fi tel: +358 9 45400550
Enease Oy fax: +358 9 45400551
Mannerheimintie 66 A mob: +358 400 760077
FIN - 00260 HELSINKI, Finland http://www.kolumbus.fi/enease/
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Sun Jun 20 01:15:12 1999
Date: Sun, 20 Jun 1999 07:10:51 +0200
From: "d.alis" <dalis@balcab.ch>
Subject: Re: Jforum: J for Windows CE
References: <002201be252f$43f9fd60$2d0114d1@xps200>
<376B4744.BF7E8E48@balcab.ch> <4.1.19990620002605.00a62510@pop.kolumbus.fi>
Content-Transfer-Encoding: 8bit
I ask about hard disks because the applications I have in mind involve around
200-400Meg of data.
Anssi Seppala wrote:
> I have hp620lx. You do not need a hard disk attached. I have a PCMCIA flash
> memory card 15MB and my PC can read/write such memory. With the memory card
> I can move a lot of data.
>
> Optimal WINCE is system you can afford memory, performance, display cost
> $$ and you need all of them with J. With J there must be also a keyboard.
>
> The J language is quite similar but GUI is not. The CE GUI is very limited
> and PC GUI is only partly protable from PC to CE.
>
> Anssi
>
> At 09:53 19.6.1999 -0700, you wrote:
> >I know that there have been several messages on WinCE but it would really be
> >great if somebody could put together a comprehensive blurb on the whole
> >subject as it relates to J.
> >>From some prior messages it sounded like even installation really was a
> >hassle!
> >
> >What are optimal hardware needs?
> >
> >Does the chipset make a big difference?
> >
> >How is it installed?
> >
> >Are there operating limitations?
> >
> >How will performance compare?
> >
> >
> >----- Original Message -----
> >From: d.alis <dalis@balcab.ch>
> >To: <forum@JSoftware.Com>
> >Sent: June 19, 1999 12:31 AM
> >Subject: Re: Jforum: J for Windows CE
> >
> >
> >> Which WinCE machines have, or can be connected to, a hard disk? I have
> >looked at a
> >> couple of HP web pages and distributer sites, but I havent found a site
> >that gives
> >> me a clear idea one way or the other.
> >> Thanks for any advice.
> >> David
> >>
> >>
> >>
> >> --------------------------------------------------------------------------
> >------
> >> 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
>
> -------------------------------------------------------------------
> Anssi Sepp�l�
> Konsultti, s�hk�markkinat ja s�hk�njakelu: Sovellukset ja ohjelmistot
>
> e-mail: anssi.seppala@enease.fi tel: +358 9 45400550
> Enease Oy fax: +358 9 45400551
> Mannerheimintie 66 A mob: +358 400 760077
> FIN - 00260 HELSINKI, Finland http://www.kolumbus.fi/enease/
>
> --------------------------------------------------------------------------------
> 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 Jun 20 04:05:39 1999
Date: Sun, 20 Jun 1999 11:12:53 +0300
From: Anssi Seppala <anssi.seppala@enease.fi>
Subject: Re: Jforum: J for Windows CE
In-Reply-To: <376C77DB.98CD1E6D@balcab.ch>
References: <002201be252f$43f9fd60$2d0114d1@xps200>
<376B4744.BF7E8E48@balcab.ch>
<4.1.19990620002605.00a62510@pop.kolumbus.fi>
Why not ordinary PC or laptop? Hard disk requires power and CE computers
are designed for batteries.
Flash memories can be very big >80 mB and you may have 2 in some machines.
Maybe you can use data compression?
Anssi
Flash memory may be 80MB or even more.
At 07:10 20.6.1999 +0200, you wrote:
>I ask about hard disks because the applications I have in mind involve around
>200-400Meg of data.
>
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Sun Jun 20 12:18:31 1999
From: "Chris Burke" <cdburke@interlog.com>
Subject: Jforum: J403b
Date: Sun, 20 Jun 1999 12:18:34 -0400
charset="iso-8859-1"
J4.03b is now available at the web site. This release fixes some
problems in the 'a' release, including the Excel OLE server crash and
OCX cut/paste support. We recommend that serious J users download this
new version.
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Sun Jun 20 12:22:32 1999
From: "Chris Burke" <cdburke@interlog.com>
References: <ED32805AF20ED21186AE00805FA615340131CCAD@NUT>
Subject: Re: Jforum: scriptdoc fails on gl2
Date: Sun, 20 Jun 1999 12:22:38 -0400
charset="iso-8859-1"
The problem was that dat was too large (~80K) to be displayed in an
edit control. The reason why your file increased to this size, from
about 20K, was that in formatting the scriptdoc output, the data was
made into a matrix, and so each row was padded out with blanks.
This problem is fixed in J403b.
> I use the following replacement to the standard 'gl2' library (see
> attached zip). When I run scriptdoc on this file it shows help, but
if i
> check on summary button, then scriptdoc fails with
>
> error in: sdoc_sum_button
>
> result error: wd
> wd'set view *',dat
>
> messagebox.
>
> What is wrong with this file?
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Sun Jun 20 12:25:25 1999
From: "Chris Burke" <cdburke@interlog.com>
References: <A12A4ACD69B4D011BB960020AFFBF108029E90D8@exs02.ex.nus.edu.sg> <001f01beaa04$b2650ab0$9e15bac3@oti.com>
Subject: Re: Jforum: DICTIONARIES IN J
Date: Sun, 20 Jun 1999 12:25:55 -0400
charset="iso-8859-1"
Simon's dictionary class is in J403b, see
system\classes\dict\dictdemo.ijs
----- Original Message -----
From: Simon Garland <simon_garland@bluewin.ch>
Sent: Saturday, May 29, 1999 2:54 PM
Subject: Re: Jforum: DICTIONARIES IN J
> I've got a neat implementation of dictionaries (3 different classes,
for
> general text keys, for tokens, and for numbers) I'll sort out the
code and
> send it shortly.
>
> ----- Original Message -----
> From: Ajith Prasad <burap@nus.edu.sg>
> To: <forum@jsoftware.com>
> Sent: Saturday, May 29, 1999 6:53 PM
> Subject: Jforum: DICTIONARIES IN J
>
>
> > Python and K directly support "dictionaries". For example, in
Python, I
> can
> > write a salary lookup table for, say, 3 employees as a dictionary
of the
> > form:
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Sun Jun 20 17:07:32 1999
Date: Sun, 20 Jun 1999 23:02:23 +0200
From: "p.smith" <smith@ibm.net>
Subject: Re: Jforum: J for Windows CE
References: <002201be252f$43f9fd60$2d0114d1@xps200>
<376B4744.BF7E8E48@balcab.ch>
<4.1.19990620002605.00a62510@pop.kolumbus.fi> <4.1.19990620110534.009ec6c0@pop.kolumbus.fi>
Thanks for the suggestion.
Anssi Seppala wrote:
> Why not ordinary PC or laptop? Hard disk requires power and CE computers
> are designed for batteries.
>
> Flash memories can be very big >80 mB and you may have 2 in some machines.
> Maybe you can use data compression?
>
> Anssi
>
> Flash memory may be 80MB or even more.
> At 07:10 20.6.1999 +0200, you wrote:
> >I ask about hard disks because the applications I have in mind involve around
> >200-400Meg of data.
> >
>
> --------------------------------------------------------------------------------
> 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 Jun 21 01:59:07 1999
From: k-list@iname.com
Date: Mon, 21 Jun 1999 01:56:13 -0400 (EDT)
Content-Type: Text/Plain
Subject: Jforum: test results: double-byte & OCX
I agree that it is not cost effective to make J core to support double-byte. But it will be sufficient for most applications if double-byte i/o works.
I downloaded j403b last night, and did some testing on
OCX: input pro 3.0, VB textbox, Farpoint Spreadsheet
here what I got:-
environment 1:
English version Win98, ACP/OEMCP=1252
Njstar commuicator 2.0
(plug-in for viewing and input Chinese, Japanese and Korean)
oleset oleget keyboard-input: all work. \(^o^)/
environment 2:
Traditional Chinese version Win98 , ACP/OEMCP=950
(use native shell for viewing and input Chinese)
oleset oleget : both failed. (+.+)
keyboard-input: work. (^o^)
Will it be the choice of codepage in the MultiByte<->WideChar conversion API that causes error?
Moreover,
window's generic edit/text control works in all cases.
VB+OCX also work in all cases.
I do not prefer window's generic controls because there lacks "wd'get" command to retrieve text (although q can do the job).
---------------------------------------------------
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 Jun 21 11:35:59 1999
From: "Nichols, Peter" <pnichols@sprinc.com>
Subject: RE: Jforum: J403b
Date: Mon, 21 Jun 1999 10:21:56 -0500
Chris, the website only shows J4.03a and gets an error if you try to
download.
Peter Nichols
> -----Original Message-----
> From: Chris Burke [SMTP:cdburke@interlog.com]
> Sent: Sunday, June 20, 1999 11:19 AM
> To: JForum
> Subject: Jforum: J403b
>
> J4.03b is now available at the web site. This release fixes some
> problems in the 'a' release, including the Excel OLE server crash and
> OCX cut/paste support. We recommend that serious J users download this
> new version.
>
>
>
>
>
> --------------------------------------------------------------------------
> ------
> 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 Jun 21 11:49:23 1999
From: "Chris Burke" <cdburke@interlog.com>
References: <01074BD4EEC4D1118E3200805F6542A7177F15@SPRDALLAS2>
Subject: Re: Jforum: J403b
Date: Mon, 21 Jun 1999 11:46:26 -0400
charset="iso-8859-1"
We occasionally get complaints of this type.
J403b is definitely at the website (we have plenty of downloads so
far), and I suspect the problem is that your browser needs to be
refreshed. If this is the problem, does anyone know if there is any
way we can force the page to be refreshed?
----- Original Message -----
From: Nichols, Peter <pnichols@sprinc.com>
Sent: Monday, June 21, 1999 11:21 AM
Subject: RE: Jforum: J403b
> Chris, the website only shows J4.03a and gets an error if you try to
> download.
>
> Peter Nichols
>
> > -----Original Message-----
> > From: Chris Burke [SMTP:cdburke@interlog.com]
> > Sent: Sunday, June 20, 1999 11:19 AM
> > To: JForum
> > Subject: Jforum: J403b
> >
> > J4.03b is now available at the web site. This release fixes some
> > problems in the 'a' release, including the Excel OLE server crash
and
> > OCX cut/paste support. We recommend that serious J users download
this
> > new version.
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Mon Jun 21 12:45:15 1999
From: "Nichols, Peter" <pnichols@sprinc.com>
Subject: RE: Jforum: J403b
Date: Mon, 21 Jun 1999 11:38:06 -0500
Ok, Got it. Clicking Refresh in Internet Exploder gets it. Thanks.
> -----Original Message-----
> From: Chris Burke [SMTP:cdburke@interlog.com]
> Sent: Monday, June 21, 1999 10:46 AM
> To: forum@jsoftware.com
> Subject: Re: Jforum: J403b
>
> We occasionally get complaints of this type.
>
> J403b is definitely at the website (we have plenty of downloads so
> far), and I suspect the problem is that your browser needs to be
> refreshed. If this is the problem, does anyone know if there is any
> way we can force the page to be refreshed?
>
> ----- Original Message -----
> From: Nichols, Peter <pnichols@sprinc.com>
> To: <forum@jsoftware.com>
> Sent: Monday, June 21, 1999 11:21 AM
> Subject: RE: Jforum: J403b
>
>
> > Chris, the website only shows J4.03a and gets an error if you try to
> > download.
> >
> > Peter Nichols
> >
> > > -----Original Message-----
> > > From: Chris Burke [SMTP:cdburke@interlog.com]
> > > Sent: Sunday, June 20, 1999 11:19 AM
> > > To: JForum
> > > Subject: Jforum: J403b
> > >
> > > J4.03b is now available at the web site. This release fixes some
> > > problems in the 'a' release, including the Excel OLE server crash
> and
> > > OCX cut/paste support. We recommend that serious J users download
> this
> > > new version.
>
>
>
>
> --------------------------------------------------------------------------
> ------
> 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 Jun 21 13:41:10 1999
From: "Gilles Kirouac" <gkirouac@riq.qc.ca>
Date: Mon, 21 Jun 99 13:28:47 -0500
Subject: Jforum: Uninstalling J
Attempting to uninstall J403a with Add/Remove Prog
(in order to install elsewhere) leads to:
"An error occurred while trying to remove J folder...
Uninstallation has been canceled."
Using NT4 US English no SP.
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Mon Jun 21 21:35:21 1999
X-Sent: 22 Jun 1999 01:30:41 GMT
From: "Barry Savage" <bv@bsavage.net>
References: <01074BD4EEC4D1118E3200805F6542A7177F15@SPRDALLAS2>
Subject: Re: Jforum: J403b
Date: Mon, 21 Jun 1999 21:31:39 -0400
charset="iso-8859-1"
It was reported by at least one party
that the J website seems to contain
J403a rather than j403b (for NT/98/95).
I got the same error!
But in fact it is not an error and I was
able to download J403b.exe with no problem.
The solution lies in refreshing/reloading the page.
Unfortunately just pressing the appropriate button may
reload the page from the browser's cache rather than
update from the web site. To be sure to get the updated page ...
(1) press the Ctrl (control) key and Refesh button at the same
time in Internet Explorer
(2) press the Shift key and Reload button at the same time
in Netscape Navigator
Barry Savage
bv@bsavage.net
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Tue Jun 22 10:23:33 1999
Date: Tue, 22 Jun 1999 10:20:52 -0400
From: Brian Bambrough <b.bambrough@worldnet.att.net>
Subject: Jforum: Grids without OOP
When I leave a client with a system that I have written, one of my top
priorities is that the client's personnel can maintain it. In my case
this means actuarial students.
Actuarial students do not want to be programmers. They want to pass
the exams and become actuaries. On the way to this goal they do a lot
of projects, preferably with their beloved Excel spreadsheet program.
However, if their boss tells them to learn enough APL or J or whatever
to modify a system, they will attempt to do it.
So the system I leave has to have a very clear structure. I try to
use only a subset of the language. I write short lines of code with
almost all lines having a comment. I comment both the programming and
business aspects. I have a small subset of idioms and I use them
consistently. These the other practices minimize what they need to
know and makes the learning curve manageable.
To further minimize the learning curve it is my strong instinct to
stay away from OOP. OOP adds another layer of abstraction. It takes
a lot of effort to learn, especially for people who do not intend to
become programmers.
In the current implementation of J, the primary tool for data entry of
tables seems to be the grid. In all the documentation, labs and
scripts provided with J, the only way to access grids is with OOP. It
seems to me that it makes for a faster learning curve to dispense with
OOP and to program the grid directly. After all, there are only 22
glgrid commands plus a few other gl commands to learn.
Of course, I could be completely wrong in this. Maybe OOP is easier.
To find out, I have been experimenting with OOPless grids. I have
duplicated the "Report" example from the Lab "Grid Basic Examples" in
a script that is completely stand-alone except for "require 'gl2'".
I would be interested to know what others think of this attempt. If
anybody wants to check out my report script, I will be happy to e-mail
it.
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Tue Jun 22 13:58:54 1999
From: "Donald B. McIntyre" <donald.mcintyre@kinfauns.demon.co.uk>
References: <376F9BC4.E4979458@worldnet.att.net>
Subject: Re: Jforum: Grids without OOP
Date: Tue, 22 Jun 1999 17:43:46 +0100
charset="iso-8859-1"
Brian:
I am very interested in what you report and would be grateful if you would
e-mail a copy of your report script.
Donald B. McIntyre
Beaumont House, Apt.17
15 St John's Place
Perth PH1 5SZ
Scotland -- U.K.
Tel: (44) 1738-451-331
Fax: (44) 1738-451-332
donald.mcintyre@kinfauns.demon.co.uk
----- Original Message -----
From: Brian Bambrough <b.bambrough@worldnet.att.net>
Sent: 22 June 1999 15:20
Subject: Jforum: Grids without OOP
> When I leave a client with a system that I have written, one of my top
> priorities is that the client's personnel can maintain it. In my case
> this means actuarial students.
>
> Actuarial students do not want to be programmers. They want to pass
> the exams and become actuaries. On the way to this goal they do a lot
> of projects, preferably with their beloved Excel spreadsheet program.
> However, if their boss tells them to learn enough APL or J or whatever
> to modify a system, they will attempt to do it.
>
> So the system I leave has to have a very clear structure. I try to
> use only a subset of the language. I write short lines of code with
> almost all lines having a comment. I comment both the programming and
> business aspects. I have a small subset of idioms and I use them
> consistently. These the other practices minimize what they need to
> know and makes the learning curve manageable.
>
> To further minimize the learning curve it is my strong instinct to
> stay away from OOP. OOP adds another layer of abstraction. It takes
> a lot of effort to learn, especially for people who do not intend to
> become programmers.
>
> In the current implementation of J, the primary tool for data entry of
> tables seems to be the grid. In all the documentation, labs and
> scripts provided with J, the only way to access grids is with OOP. It
> seems to me that it makes for a faster learning curve to dispense with
> OOP and to program the grid directly. After all, there are only 22
> glgrid commands plus a few other gl commands to learn.
>
> Of course, I could be completely wrong in this. Maybe OOP is easier.
> To find out, I have been experimenting with OOPless grids. I have
> duplicated the "Report" example from the Lab "Grid Basic Examples" in
> a script that is completely stand-alone except for "require 'gl2'".
>
> I would be interested to know what others think of this attempt. If
> anybody wants to check out my report script, I will be happy to e-mail
> it.
>
>
>
> --------------------------------------------------------------------------
------
> 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 Jun 22 15:08:32 1999
Date: Tue, 22 Jun 1999 15:05:46 -0400
From: David Ness <DNess@home.com>
Subject: Jforum: I have a trivial problem that ...
... I think I remember from discussion here, but can't find my
notes:
v =: 0 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 0 1 0 1 0 1 1 1
representing, and in order, start (0) and end (1) of `strings', which
can clearly contain strings. The first begins at 0 and ends at 23.
The next begins at 1 and ends at 2, ...
I can get `string begins' easily with
beg =: (1-v)*+/\0=v
which produces
1 2 0 3 0 4 0 5 0 6 0 7 0 8 0 9 10 0 11 0 12 0 0 0
but I am having `mind storms' trying to produce
0 0 2 0 3 0 4 0 5 0 6 0 7 0 8 0 0 10 0 11 0 12 9 1
which indicates where the given strings end...
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Tue Jun 22 16:01:48 1999
From: "Steven J. Mandula" <smandula@idirect.com>
Subject: RE: Jforum: Grids without OOP
Date: Tue, 22 Jun 1999 13:55:42 -0400
charset="iso-8859-1"
Importance: Normal
In-Reply-To: <376F9BC4.E4979458@worldnet.att.net>
I would be interested receiving a copy of your report script
With thanks
Steve Mandula
smandula@idirect.com
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Tue Jun 22 16:01:50 1999
From: "Donald B. McIntyre" <donald.mcintyre@kinfauns.demon.co.uk>
References: <01074BD4EEC4D1118E3200805F6542A7177F15@SPRDALLAS2> <000201bebc4e$fa65d040$6c7bbfa8@palmtop>
Subject: Re: Jforum: J403b
Date: Tue, 22 Jun 1999 18:52:58 +0100
charset="iso-8859-1"
Many thanks for this clarification. I have a friend who had run into the
problem and he was grateful to learn your solution.
----- Original Message -----
From: Barry Savage <bv@bsavage.net>
Sent: 22 June 1999 02:31
Subject: Re: Jforum: J403b
> It was reported by at least one party
> that the J website seems to contain
> J403a rather than j403b (for NT/98/95).
>
> I got the same error!
> But in fact it is not an error and I was
> able to download J403b.exe with no problem.
>
> The solution lies in refreshing/reloading the page.
>
> Unfortunately just pressing the appropriate button may
> reload the page from the browser's cache rather than
> update from the web site. To be sure to get the updated page ...
>
> (1) press the Ctrl (control) key and Refesh button at the same
> time in Internet Explorer
>
> (2) press the Shift key and Reload button at the same time
> in Netscape Navigator
>
> Barry Savage
> bv@bsavage.net
>
>
>
>
> --------------------------------------------------------------------------
------
> 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 Jun 22 17:08:48 1999
From: arneson@pkb.mega.net.id (arneson)
Subject: Jforum: Read disk label?
Date: Tue, 22 Jun 1999 07:57:53 +0700
How can I read a disk label from J?
--------------------------------------------------------------------------------
J Forum: for information about this list, see http://www.jsoftware.com/forum.htm
From owner-jsoftware@lists.interlog.com Tue Jun 22 19:02:56 1999
From: "Alex Kornilovski" <kornal@idirect.com>
Subject: Re: Jforum: Read disk label?
Date: Tue, 22 Jun 1999 19:07:10 -0400
charset="iso-8859-1"
>How can I read a disk label from J?
load 'dll'
r=.'kernel32 GetVolumeInformationA i *c *c i i i i *c i' cd('a:\');(100