Whither Scientific & Engineering APL?

> Is there a simple way to convince Firefox to display this correctly?

Doug, two questions to this:

1. Which operating system do you use?

2. Is the "APL385 Unicode" font installed on your system?


Kai

On Jun 3, 11:13*pm, gwh...@alum.mit.edu (Doug White) wrote:
> Keywords:
> In article <26f9d9f6-8162-4492-b453-4570f2110...@a70g2000hsh.googlegroups.com>, "Stephen Taylor <edi...@vector.org.uk>" <StephenTaylorF...@googlemail.com> wrote:
>>
> >You might find helpful Beau Webber's review of APLX v4 at
>
> >http://www.vector.org.uk/archive/v233/webber.htm
>
> OK, that got my attention. *I have a couple of questiosn for APLX users:

Hi Doug,
Answers below :
>
> There is a LOT of web/HTML/.NET stuff that I will probably never have
> much use for. *Does it get in the way?

Not at all - don't use it, you do not see it.

>
> They now have a far more complex product than APL68000, which I had
> issues with in the past. *So:
>
> 1) How good is the documentation

Excellent off-line manuals - I used to print them out as ring-bound
books.
I still read the upgrade documentation very carefully, but :
Now I mainly use the on-line help - right click on a construct and
select help - excellent.

Also most features have simple example code that you can run, and then
hack - many of my functions and applications based on the newer
features started that way.
See : http://www.lab-tools.com/instrumenta...tereoPanorama/ which
was initially just based on example code to display an image (needs
updating - APLX now can use calls on ImageMagick to do really fancy
graphics operations).

>
> 2) How bug-free is the code?

APLX seems to be totaly robust - A few small bugs crept in with the
Beta version of V4, but were dealt with immediately - I run multi-
hundred MByte workspaces, sometimes simultaneously, some dating back
decades, just usually suspending at night, until XP keels over from
lack of resources (suspect IE7 ?). I have not had a problem with the
current version in the months I have been running it. With version 4 I
did do a )copy into empty clean workspaces - first setting )symbols,
with some of my decases old workspaces, to make sure I had clean
copies.

>
> 3) How good is the support?

Excellent - make a good suggestion (well I belive that including
things like Regular Expression parsing is a good idea) and with luck
you may get it at the next upgrade.
More simply - something not clear from the documentation, an answer
usually within hours (occasionally day or two).

>
> If I don't want to mess with Linux, the price is still pretty steep for
> an individual non-commercial user.

I agree but so are todays' prices for music and beer, and I can't do
without either (and I would have much less time for either of those
without APLX).

>
> Thanks!
>
> Doug White

cheers,
Beau Webber

On Jun 3, 11:25 pm, gwh...@alum.mit.edu (Doug White) wrote:
> Keywords:
> In article <cd950453-0cca-42ee-b56e-76239cc5b...@b1g2000hsg.googlegroups.com>, "Stephen Taylor <edi...@vector.org.uk>" <StephenTaylorF...@googlemail.com> wrote:
>
>
>
> >On Jun 3, 12:08 am, gwh...@alum.mit.edu (Doug White) wrote:
> >....
>
> >> I'd be interested in comparison's of how the various APL's handle things
> >> like executing Windows graphics operations. I know roughly how APL+WIN
> >> does it, and I _think_ I understand IBM's approach, but I have no clue
> >> how Dyalog deals with this.
>
> >You can see on the APL Wiki at
>
> >http://aplteam2.com/aplwiki/moin.cgi/SolitaireGame
>
> >the complete code for the game of Solitaire. It's written as a class,
> >which might be more of a jump into the future than you want; but one
> >doesn't have to use classes to write GUI in Dyalog.
>
> >(You can download and run the script in Dyalog 12, or even paste it
> >from the web page into the editor.)
>
> >You should be able to get a fair idea of how the GUI is built and
> >callbacks set and processed. I'd be glad to answer questions about it.
>
> Yikes! I'll start with a simplified version my original question
> "Whither APL?".
>
> There are (occasional) APL operators scattered about, but 90% of this is
> unrecognizable to someone who has been stuck working with the APL of old.
> Upon further examination, I think I must have a font issue, because there
> are lots of question marks where I would expect "lamp" comment symbols.
>
> Is there a simple way to convince Firefox to display this correctly?
>
> Thanks!
>
> Doug White

An article on the Vector site

http://www.vector.org.uk/archive/display.htm

offers advice on browser viewing problems.

I don't know if you find this encouraging, but I returned to APL in
2001 after a 15-year break from programming. One can catch up. And I
enjoy the writing better than I ever did before. Not only is the
development environment better, the language is even more expressive.
Enjoy!

Stephen

Keywords:
In article <677e2e87-1fc4-4a2c-a554-edf17daa8e74@d45g2000hsc.googlegroups.com>, kai <kaithomasmax@googlemail.com> wrote:
>> Is there a simple way to convince Firefox to display this correctly?
>
>Doug, two questions to this:
>
>1. Which operating system do you use?

Win XP, SP2

>2. Is the "APL385 Unicode" font installed on your system?

Not unless it got installed while I wasn't looking.

I found the one on the Vector site, and installed it. That works MUCH
better. I can't say I can read the code very well, but it is much more
recognizable as being APL than yesterday.

Thanks!

Doug White

Keywords:
In article <49d6b9bf-e6c6-4520-b522-594ff38c1881@l42g2000hsc.googlegroups.com>, "Stephen Taylor <editor@vector.org.uk>" <StephenTaylorFRSA@googlemail.com> wrote:
>On Jun 3, 11:25 pm, gwh...@alum.mit.edu (Doug White) wrote:
>> Keywords:
>> In article
> <cd950453-0cca-42ee-b56e-76239cc5b...@b1g2000hsg.googlegroups.com>, "Stephen
> Taylor <edi...@vector.org.uk>" <StephenTaylorF...@googlemail.com> wrote:
>>
>>
>>
>> >On Jun 3, 12:08 am, gwh...@alum.mit.edu (Doug White) wrote:
>> >....
>>
>> >> I'd be interested in comparison's of how the various APL's handle things
>> >> like executing Windows graphics operations. I know roughly how APL+WIN
>> >> does it, and I _think_ I understand IBM's approach, but I have no clue
>> >> how Dyalog deals with this.
>>
>> >You can see on the APL Wiki at
>>
>> >http://aplteam2.com/aplwiki/moin.cgi/SolitaireGame
>>
>> >the complete code for the game of Solitaire. It's written as a class,
>> >which might be more of a jump into the future than you want; but one
>> >doesn't have to use classes to write GUI in Dyalog.
>>
>> >(You can download and run the script in Dyalog 12, or even paste it
>> >from the web page into the editor.)
>>
>> >You should be able to get a fair idea of how the GUI is built and
>> >callbacks set and processed. I'd be glad to answer questions about it.
>>
>> Yikes! I'll start with a simplified version my original question
>> "Whither APL?".
>>
>> There are (occasional) APL operators scattered about, but 90% of this is
>> unrecognizable to someone who has been stuck working with the APL of old.
>> Upon further examination, I think I must have a font issue, because there
>> are lots of question marks where I would expect "lamp" comment symbols.
>>
>> Is there a simple way to convince Firefox to display this correctly?
>>
>> Thanks!
>>
>> Doug White
>
>An article on the Vector site
>
>http://www.vector.org.uk/archive/display.htm
>
>offers advice on browser viewing problems.
>
>I don't know if you find this encouraging, but I returned to APL in
>2001 after a 15-year break from programming. One can catch up. And I
>enjoy the writing better than I ever did before. Not only is the
>development environment better, the language is even more expressive.
>Enjoy!

Thanks! I got the Unicode font installed, and it looks a lot more
familiar now.

Doug White

Under Dyalog APL, line-based graphics is based on their Windows model
and works more or less the same way form composition does. It's quite
easy and yields good results. In general, you open a form, draw
lines, draw text, fill a pattern, and so on. You can set different
fonts, line thicknesses, and dashing patterns. Best of all, if you
want the graph to appear on say a Word document, you do the same thing
but write to a metafile object instead, then link the metafile to the
document when done.

Alternatively, a higher quality approach would be to write APL code to
generate Post Script. It's a little bit of a learning curve to be
able to produce Post Script but is no harder than driving a plotter at
the device level. (HPGL comes to mind) If you don't have Post Script
printers, there are a lot of software products which deal with Post
Script, notably recent versions of Word have a built-in EPS (extended
post script) converter, Ghost Script, and so on. By definition, using
Post Script from APL is system independent, as all you are doing is
writing a text file. Post Script itself can be fairly human readable.

Lastly, OpenGL is the end of the rainbow, or close to it, when it
comes to graphics. OpenGL is a forerunner of Microsoft DirectX.
(DirectX is a good example of Microsoft marketing practice, it
duplicates the functionality of OpenGL and bundles in features to play
sound, read the mouse and keyboard, etc). Evidently there are 2D
instructions for OpenGL suitable for doing more usual charting, but
the problem is what to do with the picture once you have it on the
screen - probably you can extract the bitmap and print and save it,
but for dealing with graphics as part of a document, usually you have
better results with vector graphics. I don't thing OpenGL will dump
out a metafile, but have never used this part.

For 3D graphics, OpenGL requires you to create a 3D mesh, wrap a skin
around it, set lighting sources, set a camera location, move the
camera position, move the object, and so on. All of this is handled
by the OpenGL API, from APL, you need to know how to call the
functions, usually with something like []NA. Except for the
particulars in calling external library functions from APL, OpenGL
would be APL system independent.

On Jun 7, 6:48*am, Ibeam2000 <Ibeam2...@gmail.com> wrote:
> Under Dyalog APL, line-based graphics is based on their Windows model
.....
>*Best of all, if you
> want the graph to appear on say a Word document, you do the same thing
> but write to a metafile object instead, then link the metafile to the
> document when done.
>
> Alternatively, a higher quality approach would be to write APL code to
> generate Post Script. .....

I have found the postscript approach most useful - I have a graph
drawer with a script driven digram drawer (2d/3d - just looks at the
data structure it is given) - that has a number of output formats
including metafile, but the only ones I have bothered to turn on in a
while are the screen and postscript ones.

If you want graphs in MS Word then I reccomend importps - there is a
free download to test it, but you need to pay a few pounds for the
full version - or get pstoedit for a whole set of tranformations :
www.geocities.com/wglunz/importps/
www.pstoedit.net/pstoedit

Best installed as part of ghostscript and ghostview/gsview (both
free), if you want to have postscript viewers etc. :
pages.cs.wisc.edu/~ghost/
www.ghostgum.com.au/

In general publishers like postscript graphs, and I find that I have
no grief from them, they appear in my papers (which I prepare in
Latex) just as I intended them. Also, pstoedit transforms them to pdf
so they are easy to share with people who do not have postscript
tools.

cheers,
Beau

Doug and . . .,

My background is also engineering (mechanical) where the ability to
conveniently use complex numbers can be prized. My core focus is
building a software package for gear design. To date, I gauge my package
to be somewhat involved with geometry and rating. As you describe, I too
want to make my efforts available for use as I think I have a new, useful,
and valuable approach to a classical and mature subject.

My original development was with classical APL (Harris). I used APL to
help develop and demonstrate some mathematical principles. I continued
development over the years and now have many APL functions (500). With
numerical results being my overarching objective, I never ventured outside
the workspace or interfaced with other packages. In order to make my
efforts available for use today, I need to extend my programming abilities
beyond classical APL. Over the past years, I have worked on a GUI to help
showcase my results (over 300 functions). I am using APL2 for
development. APL2 is available without cost for teaching and research
since I am with a university. I am very satisfied with the product and
the service has been terrific. To illustrate my satisfaction, I purchased
a personal license to continue GUI development! Like you, on my own nickel
(teacher's salary). I anticipate that you can generate your entire package
with APL2 and make available for distribution (without user costs) using
runtime APL2.

As for Matlab, I am amid Matlab fanatics (professors and students). C is
the default required programming language for engineering students in the
university I am with, yet many students default to Matlab for engineering
homework/research assignments whereas some students use Excel, Mathcad, or
Mathmatica (ok, shows the level of assignment). In fairness to Matlab, the
language is advertised in American Society of Engineering Education and
American Society of Mechanical Engineers; I anticipate likewise for other
engineering disciplines. Matlab seminars are given regularly on campus
each semester.

I feel APL has a lot to offer engineers today and in the future. I would
like to see the APL community grow and I am willing to orient students
towards APL. Being results driven, I do not mandate any language and
allow students to decide individually the language of choice to ascertain
results. Granted, I do make some unnecessary comments on language of
choice when students struggle. Wise or not, APL is my choice of languages
for coding. As you know, you can lead a horse to water but you cannot make
it drink.

David


--
Message posted using http://www.talkaboutprogramming.com/...comp.lang.apl/
More information at http://www.talkaboutprogramming.com/faq.html

Doug and . . .,

My background is also engineering (mechanical) where the ability to
conveniently use complex numbers can be prized. My core focus is
building a software package for gear design. To date, I gauge my package
to be somewhat involved with geometry and rating. As you describe, I too
want to make my efforts available for use as I think I have a new, useful,
and valuable approach to a classical and mature subject.

My original development was with classical APL (Harris). I used APL to
help develop and demonstrate some mathematical principles. I continued
development over the years and now have many APL functions (500). With
numerical results being my overarching objective, I never ventured outside
the workspace or interfaced with other packages. In order to make my
efforts available for use today, I need to extend my programming abilities
beyond classical APL. Over the past years, I have worked on a GUI to help
showcase my results (over 300 functions). I am using APL2 for
development. APL2 is available without cost for teaching and research
since I am with a university. I am very satisfied with the product and
the service has been terrific. To illustrate my satisfaction, I purchased
a personal license to continue GUI development! Like you, on my own nickel
(teacher's salary). I anticipate that you can generate your entire package
with APL2 and make available for distribution (without user costs) using
runtime APL2.

As for Matlab, I am amid Matlab fanatics (professors and students). C is
the default required programming language for engineering students in the
university I am with, yet many students default to Matlab for engineering
homework/research assignments whereas some students use Excel, Mathcad, or
Mathmatica (ok, shows the level of assignment). In fairness to Matlab, the
language is advertised in American Society of Engineering Education and
American Society of Mechanical Engineers; I anticipate likewise for other
engineering disciplines. Matlab seminars are given regularly on campus
each semester.

I feel APL has a lot to offer engineers today and in the future. I would
like to see the APL community grow and I am willing to orient students
towards APL. Being results driven, I do not mandate any language and
allow students to decide individually the language of choice to ascertain
results. Granted, I do make some unnecessary comments on language of
choice when students struggle. Wise or not, APL is my choice of languages
for coding. As you know, you can lead a horse to water but you cannot make
it drink.

David


--
Message posted using http://www.talkaboutprogramming.com/...comp.lang.apl/
More information at http://www.talkaboutprogramming.com/faq.html

Keywords:
In article <e381dd49-d6de-4abd-8818-3f265ae92d91@a1g2000hsb.googlegroups.com>, J.B.W.Webber@kent.ac.uk wrote:
>On Jun 7, 6:48=A0am, Ibeam2000 <Ibeam2...@gmail.com> wrote:
>> Under Dyalog APL, line-based graphics is based on their Windows model
>=2E....
>>=A0Best of all, if you
>> want the graph to appear on say a Word document, you do the same thing
>> but write to a metafile object instead, then link the metafile to the
>> document when done.
>>
>> Alternatively, a higher quality approach would be to write APL code to
>> generate Post Script. .....
>
>I have found the postscript approach most useful - I have a graph
>drawer with a script driven digram drawer (2d/3d - just looks at the
>data structure it is given) - that has a number of output formats
>including metafile, but the only ones I have bothered to turn on in a
>while are the screen and postscript ones.
>
>If you want graphs in MS Word then I reccomend importps - there is a
>free download to test it, but you need to pay a few pounds for the
>full version - or get pstoedit for a whole set of tranformations :
> www.geocities.com/wglunz/importps/
> www.pstoedit.net/pstoedit
>
>Best installed as part of ghostscript and ghostview/gsview (both
>free), if you want to have postscript viewers etc. :
> pages.cs.wisc.edu/~ghost/
> www.ghostgum.com.au/
>
>In general publishers like postscript graphs, and I find that I have
>no grief from them, they appear in my papers (which I prepare in
>Latex) just as I intended them. Also, pstoedit transforms them to pdf
>so they are easy to share with people who do not have postscript
>tools.

I haven't had to use it in a while, but I believe "pstoedit" isn't
compatible with Office 2007. The IT nazis at work decided that we should
all "upgrade" to Office 2007. That broke at least a half dozen things I
do regularly to get my work done. Productivity as a whole has dropped
considerably while folks hunt for commands in the new menu structure.

Doug White

Doug White