Hi Jaime!
> OK, I will change the name of that file. By the way, the init
> file was not really working because it was being saved as
> netmath.ini and read as xmaxima.ini at start up.
> I also changed the handling of that file to something
> that I find more convenient: when xmaxima is closed, the
> init file is saved silently, with the current configuration.
> That way, when xmaxima is started it will keep the same
> configuration that it had in the last session.
>
Good!
> P.S. I'll be out of town (Caceres, Spain) this week, but when
> I return I will have a big bunch of changes to xmaxima to
> commit.
>
I'm really glad you taking care of xmaxima!
But, please, be careful. IMHO crucial patches
should not be committed just a few days before
release. More testing is required. The point is that
some xmaxima parts, xmaxima<->maxima interaction in particular,
are extremely fragile. For example right now cvs xmaxima is
broken due to last patch to RunMaxima.tcl. Now xmaxima produces
output with lots of empty lines in various places, even
inside expressions. On clisp it is broken entirely -
it prints EOL after each single character!
Behavior is somewhat erratic since place where spurious
EOLs are produced depends on platform (Windows XP is
different from Windows Me) and even vary from run to run.
It seems like it depends on timing.
I know that xmaxima is fragile by experience.
Have you noticed that on Windows xmaxima prints Maxima banner
and doesn't print it on Unix/Linux? Historically xmaxima
used to cut Maxima banner (first line of the session is
(%i1) prompt). But two years ago AFAIR banner was turned on.
Unfortunatly it caused xmaxima to hang at the very start
of session or after first command on Linux (but not on Windows!).
Details can be found in this list archive. Nobody understood
what is going on and banner was rolled back.
Best wishes,
Vadim
--
Vadim V. Zhytnikov
<vvzhy at mail.ru>
<vvzhy at netorn.ru>