On Sun, Apr 02, 2006 at 07:14:05PM +0200, steven mestdagh wrote:
> Tobias Ulmer [2006-04-02, 17:25:36]:
> > > here is a port of scilab, "a scientific software package for numerical
> > > computations providing a powerful open computing environment for
> > > engineering and scientific applications." (see http://www.scilab.org)
> > > 
> > > The port was tested under i386 and sparc64 and works well so far.
> > > 
> > > * no PVM (paralell scilab) support, because it's a lot of work and
> > >   I have no environment to test it. If anybody needs this, feel free to
> > >   send patches (i386 should work already).
> > > 
> > > * someone please countercheck the license, i skimmed over it and maybe
> > >   i'm missing something!
> > > 
> > > * i'm not happy with the package location (/usr/local/lib/scilab...) but
> > >   it's the default location and changing this looks like a lot of
> > >   trouble.
> > > 
> > > * so, here we go. please test test test and send diffs ;)
> > > 
> > > Tobias
> > 
> > Oops, first update, there was still some test code in the Makefile,
> > gtk2 is not usable. And the subject should be math/scilab of course.
> 
> i'll take care of this.  a first, very quick glance:
> why do you have patch-configure_in and patch-configure? can you just
> make the port run autoconf?

using autoconf to generate the configure script ist not a problem.

However there are custom macros, #defines and #includes that don't
belong there in the *.h.in files sprinkled all over the place and
autoheader removes them all. Fixing this seems to be a too large task
to me.

I can think of three solutions:

* Use CONFIGURE_STYLE=autoconf and prevent autoheader from running,
  somehow... (How?)

* Use CONFIGURE_STYLE=gnu and run autoconf after 'patch'

* Just provide the (large) patch-configure diff

> is it possible to skip the version number in scilab-4.0 in PLIST? if
> not, can you replace it by a variable (through SUBST_VARS)?
>

I think i have patches ready to move the whole thing to /usr/local/scilab.
Imho the best location.
 
> steven
> 
> Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm
> 
> 

Tobias

Reply via email to