[Debichem-devel] [Outreachy] pymol

Andreas Tille andreas at fam-tille.de
Wed Jul 27 20:42:05 UTC 2016


Hi again,

I commited two changes to SVN (fix changelog, remove strange file
monkeypatch_distutils.pyc no idea how to do this better).  When trying
to build in pbuilder this (expectedly) failed due to the pinning to a
certain libpng12-0 version which does not exist in debian unstable.  I
stumbled upon this after sending my previous mail and now I'm wondering
what might have been your motivation for this.

Kind regards

      Andreas.

On Wed, Jul 27, 2016 at 09:57:29PM +0200, Andreas Tille wrote:
> Hi Tanya,
> 
> I noticed you have created changelog entries for 1.8.2.1 so you seem to
> have managed to update the patches and build the new version, right?
> 
> Could you please merge both entries for 1.8.2.1-1 (UNRELEASED) and
> 1.8.2.1-2?
> 
> On Wed, Jul 27, 2016 at 07:44:51PM +0300, merlettaia wrote:
> > Hi Andreas,
> > 
> > For now I wrote simple test which runs over all .py and .pml scripts inside
> > specific directory and tries to run them with pymol with -c flag (for now
> > this script doesn't care if python errors appear during run - it only
> > checks that pymol closes as expected). This is made because some of these
> > scripts (located in `data` and `examples` directories in upstream source
> > tree) use data from other directories in relative paths and write to the
> > folder where they are located in. I symlink them to temporary directory. I
> > haven't solved usage of files from relative paths yet, that's why this
> > leads to python exceptions, but pymol still closes as expected.
> > Incorrect paths are present in installed example scripts, for example, in
> > /usr/share/doc/pymol/examples/generate_amber.py there is a line:
> > 
> >     model= io.pdb.fromFile("../../test/dat/pept.pdb")
> > 
> > Since in installation this path is incorrect, this script also raises
> > python exception (but pymol command line closes as expected).
> 
> I admit I'm not fully sure if I understand all this but for me this
> sounds like another quilt patch would be needed to fix the pathes.
> 
> > I don't know
> > how to deal with these relative files. I can fix as many examples as I
> > could to make them find input parameters correctly when they are installed,
> > or write several examples by myself. What is a better solution?
> 
> Hard to say.  May be users might know some of these examples from
> anywhere else than Debian and want to run these.  Moreover if you manage
> to fix these examples this could be forwarded upstream which is in the
> interest of all users.
>  
> > I've updated patches 22 and 31 and fixed several lintian warnings.
> 
> Very nice!
>  
> > Is this bug in the same state as it was?
> > https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=506202 - I can fix it
> > too, if no one fixed it yet.
> 
> Cool, I filed a bug 8 years ago.  I do not even remember. :-)  If there
> is nothing in the bug log and neither in the packaging it is probably
> not fixed.  I would not really call it a show stoper for an upload but
> if you would fix this it would be nice.
>  
> > For now I'll add a testsuite for apbs (it will be very simple, but probably
> > it would help make it work with pdb2pqr).
> 
> Very good job so far.
> 
> Thanks a lot for your work on this
> 
>       Andreas.
> 
> -- 
> http://fam-tille.de
> 
> 

-- 
http://fam-tille.de



More information about the Debichem-devel mailing list