[Bug 326421] Review Request: xmds - the eXtensible Multi-Dimensional Simulator

[bugzilla at redhat.com]

Please do not reply directly to this email. All additional
comments should be made in the comments box of this bug report.

Summary: Review Request: xmds - the eXtensible Multi-Dimensional Simulator


https://bugzilla.redhat.com/show_bug.cgi?id=326421





------- Additional Comments From paultcochrane at gmail.com  2007-10-16 11:55 EST -------
Hi Patrice,

Thanks for reviewing this submission!  :-)

(In reply to comment #1)
> The source should be an url to the upstream source.
> See:
> http://fedoraproject.org/wiki/Packaging/SourceURL

I've updated this in the .spec and uploaded the files again:

Spec URL: http://downloads.sourceforge.net/xmds/xmds.spec
SRPM URL: http://downloads.sourceforge.net/xmds/xmds-1.6-3.src.rpm

> A buildrequires for fftw is missing. It seems to be fftw2-devel,
> more precisely.

Added.

> loadxsil.m shouldn't be in %_bindir, it may better be in 
> %doc since there is no specific directory for matlab scripts
> currently in fedora.

I've added a patch which stops loadxsil.m from being installed in %_bindir and
added code to put it into %_docdir

> Docs are missing. Without doc, the package is not very useful. 
> There is an example directory, it should be shipped in %doc.
> Also the manual would be a must, if it is covered by a free
> documentation license.

They're actually in a separate part of the repository.  Should they be a
separate package?  Say xmds-doc?

I've added the example xmds files, but I'm not 100% sure if I've done things
correctly.

> scilab and matlab are not part of fedora. But unless I am 
> wrong, the .dat created by xsil2graphics -s are simple ascii
> output file that may be used with any plotting program? Maybe
> this could be said somewhere?

Actually the output should work fine with Octave which is properly free.  In the
 current svn version of xmds we've got R and Gnuplot output working as well,
however, Octave should work "out of the box" with this version of xmds.

> In summary it is unneeded to repeat the package name.

Oops, I just realised I missed this.  The package still needs more review, so
I'll fix this with the next set of problems which I'm sure about to turn up...

> Unless I am wrong, xsil2graphics and loadxsil are useful by 
> themselves, maybe they could be in a subpackage, since
> as far as I can tell from a quick browsing, there aren't many
> other xsil tools.

Atm they only make sense with xmds.  AFAIK xsil is a format which isn't likely
to stick around much longer, so having separately packaged tools isn't going to
be worth the effort, I think.

> I suggest using %dist since this is a binary package. Also 
> why use a release of 3 in the submission, why not begin with 
> 1?

That's because this is xmds version 1.6 revision 3.  Am I supposed to do this
another way?  Is the revision number the number of the xmds.spec file or something?

> It seems to me that a requires on gcc-c++ would be in 
> order, since it is called during model compilation. Same
> for fftw2-devel.

Added these dependencies to the new .spec file.

> Is libxmds.a meant to be used separately from xmds?

Um, no.  Should we be putting this somewhere else?

Thanks again for your help!

Paul

-- 
Configure bugmail: https://bugzilla.redhat.com/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.