[Gcl-devel] Re: Axiom for Windows and TeXmacs

[Dan Martens]

Hi Bill,

I've been very busy here trying to keep up with my emails.  I will whip up
an axiom script tomorrow.  Where would you like me to post the build when
done?

Dan
----- Original Message ----- 
From: "Page, Bill" <address@hidden>
To: "'Mike Thomas'" <address@hidden>
Cc: <address@hidden>; <address@hidden>;
<address@hidden>; "'Dan Martens'" <address@hidden>;
<address@hidden>
Sent: Monday, December 06, 2004 3:49 PM
Subject: Axiom for Windows and TeXmacs


> Mike,
>
> On Monday, December 06, 2004 2:39 AM you wrote:
> > ...
> > I've already found one infelicity and need to find out your
> > standard process for generating those Makefile.pamphlet
> > patches to the GCL source tree - specifically, what "diff"
> > command line arguments do you use to make each patch?
>
> I believe it is diff -Naur
>
> >
> > Bill, I'm interested in seeing that extra information if
> > you can still be bothered to chase your own GCL HEAD build.
> > I think it would be good to try uncovering the problem so
> > that you can gain some faith in the process.
>
> I will get back to that later today. So far the only
> difference I can see is that last week I had not updated
> to the latest versions of MinGW runtime 3.35 and Win32 3.1.
> I have now done this and I will let you know the results
> later.
>
> >
> > I spent odd moments over the weekend (text mode) and today
> > (Windows TeXmacs - no cygwin involved at all) trying to
> > crash Axiom and or TeXmacs. I'm very pleased to report that
> > I failed completely.  My main attack has been to type in
> > code snippets which solve specific problems from the Axiom
> > book and various web sites.  Axiom is certainly a stunning
> > program.
>
> Wonderful. I was *very impressed* with the speed of Axiom
> under Windows and I also have not found any peculiarities
> in the output compared to Axiom on Linux. My first impression
> is that Axiom on Windows is faster than Axiom under Linux on
> the same hardware. But that seems improbable.  Perhaps it has
> something to do with the changes in GCL head? Certainly the
> combination of TeXmacs for Windows and Axiom is faster than
> TeXmacs for Linux and Axiom for Linux but I think that has
> to do with the threaded interface program.
>
> >
> > I'm also hoping that someone on the Maxima development
> > team will build a tm_maxima equivalent for Windows based
> > on Bill's superb weekend efforts.
>
> I would be very happy to work with someone who knows Maxima
> to produce a new native Windows tm_maxima. The version of
> TeXmacs for Windows is still very new and only a few plug-ins
> have been tested with it - most require a rewrite specifically
> for windows because Windows process handling is different from
> Linux. I think that example of threaded process handling that
> Mike Thomas sent me is superior to the older TeXmacs interface
> code (on Windows at least and possibly also under linux).
>
> If someone volunteers to help with testing, I could probably
> produce a first attempt at a tm_maxima by merging the new
> process code with the conversions embedded in the old tm_maxima
> for Linux. This might take a 1/2 day work.
>
> About Maxima and TeXmacs for Windows, see:
>
> http://www.math.utexas.edu/pipermail/maxima/2004/006925.html
> http://www.math.utexas.edu/pipermail/maxima/2003/005702.html
>
> This tutorial should be updated for the new native Windows
> version:
>
> http://www.fmi.uni-passau.de/~seidl/texmacs/tmtour.html
>
> I am also mildly motivated to add Maxima support to MathAction
> but so far only Bob McElrath has shown any direct interest
> and unfortunately he has had other priorities and not
> enough time.
>
> I am very impressed by Dan Martin's work with porting TeXmacs
> to windows. From the description at
>
> http://www.texmacs.org/tmweb/download/windows.en.html
>
> I think it is currently rather undervalued by the original
> TeXmacs developers perhaps because Windows is viewed as a
> less desirable target platform by most open source
> developers? But the reality is that the number of potential
> windows users still far exceeds the number linux desktop
> users. I think that it is interesting to see from the
> MathAction web stats that in just 5 days, the new Windows
> version of Axiom has already by downloaded more that 125
> times!
>
> > I just wish we had saveable worksheets for both CAS systems -
> > TeXmacs seems not to be able to do that yet.
>
> I am not sure I understand your comment. It is already
> possible now to save a TeXmacs document that includes any
> kind of session output (including Axiom) as if it was a
> "worksheet". My only quibble with TeXmacs is that it currently
> has no easy way to re-execute all of the session commands
> embedded in a document with one click. One must visit each
> command in the document and press Enter to cause it to be
> re-executed.
>
> Or do you mean simply that you cannot access Maxima from
> TeXmacs for Windows? I think both the Cygwin and Linux
> versions of TeXmacs can do this.
>
> >
> > Attached is a picture of Windows Axiom in TeXmacs.
> >
> > Bill, I can send you an up to date PDF copy of the Axiom
> > book built from CVS if you would like to include it in
> > your web zip file.
> >
>
> That's a cool idea. I think we need to give some thought to
> how to package Axiom for Windows. As C Y (subject: crazy
> idea) has pointed out, Windows users tend to have greater
> expectations when in comes to the completeness. C Y's
> "crazy idea" was that maybe we should include TeXmacs for
> Windows together with the Axiom for Windows - assuming
> that TeXmacs would be the preferred interface. Even if
> we don't do that, then at least I think the Windows package
> should include the tm_axiom interface and as you suggest,
> the Axiom book.
>
> Dan Martin has offered to provide a auto-install script
> for Axiom for Windows using the Nullsoft installer system.
>
> Is your PDF version of the book built with working internal
> links? The method that I used most recently (dvipdfm under
> Debian) did not produce working links.
>
> Regards,
> Bill Page.
>