I looked at the eclass/python.eclass and found the PYTHON_OVERRIDE_PYVER variable which could be used to compile VMD. VMD-1.8.6 cannot compile with python-2.5.x, so it would be useful to use something like: PYTHON_OVERRIDE_PYVER="2.4" emerge vmd This currently doesn't work! However changing PYVER in the ebuild itself works OK!
Not intended to be used like that, get vmd fixed. *** This bug has been marked as a duplicate of bug 192902 ***
Now that I look at this bug report I have another idea. How about having something like /etc/portage/package.env where one could specify environment variables that go with the specific package. Specifically I have a problem with emerge --empty world which doesn't compile netcdf because it uses fortran and fortran is sensitive to the order how things get compiled. It is not always a problem (sometime I am lucky) but I usually have to do MAKEOPTS="-j1" emerge netcdf. Wouldn't it be neat to specify this in /etc/portage/package.env ?
(In reply to comment #1) > Not intended to be used like that, get vmd fixed. > > > *** This bug has been marked as a duplicate of bug 192902 *** > No, this is the bug for eclass/python.eclass and has nothing to do with the sci-chemistry/vmd-1.8.6.ebuild I used vmd only as an example!
(In reply to comment #3) > (In reply to comment #1) > > Not intended to be used like that, get vmd fixed. > > > > > > *** This bug has been marked as a duplicate of bug 192902 *** > > > > No, this is the bug for eclass/python.eclass and has nothing to do with the > sci-chemistry/vmd-1.8.6.ebuild I used vmd only as an example! > OK, this is maybe not a bug, but more like a request for extending the functionality. I know it is not a fix for vmd but brings to the flexibility of portage itself!
Please, read what the ebuild does before filing bugs about an eclass.
