here's an ebuild and patch for pymol-0.99_rc6 that includes the very useful sidechain_check wizard written by Michael Lernel
Created attachment 81704 [details] ebuild to add sidechain_check
Created attachment 81705 [details, diff] patch to add "Sidechain Check" to the Wizard menu.
Created attachment 81706 [details] sidechain_check script
How are we supposed to know when he updates his scripts? They're completely unversioned, and I can't even access a directory listing to find last modified date.
hmm....I'm not in contact with him, but I doubt it will need updating unless pymol changes drastically. If that happens I can look for an updated script, or disable it. What do you suggest?
Created attachment 81735 [details] new pymol ebuild to depend on >=apbs-0.4.0 I found out >=apbs-0.4.0 is required...modified ebuild.
I suggest contacting him and asking if he can add version numbers to his script names so distributions are able to package them properly.
(In reply to comment #7) > I suggest contacting him and asking if he can add version numbers to his script > names so distributions are able to package them properly. > I added my CVS log to the top of the script. Is that good enough? I also added some contact information. http://www.umich.edu/~mlerner/PyMOL/sidechain_check.py
(In reply to comment #8) > I added my CVS log to the top of the script. Is that good enough? I also > added some contact information. > http://www.umich.edu/~mlerner/PyMOL/sidechain_check.py Hi there, thanks for taking the time to reply. What we really need is a URL for a given file that will never change contents (e.g., the same URL is always the same md5sum). So something like http://www.umich.edu/~mlerner/PyMOL/sidechain_check-1.4.py (note the embedded version number). (Also noticed you play ultimate, I did the same in my undergrad days.)
(In reply to comment #9) > (In reply to comment #8) > > I added my CVS log to the top of the script. Is that good enough? I also > > added some contact information. > > http://www.umich.edu/~mlerner/PyMOL/sidechain_check.py > > Hi there, thanks for taking the time to reply. What we really need is a URL for > a given file that will never change contents (e.g., the same URL is always the > same md5sum). So something like > http://www.umich.edu/~mlerner/PyMOL/sidechain_check-1.4.py (note the embedded > version number). > > (Also noticed you play ultimate, I did the same in my undergrad days.) > Hmmn .. I'm a grad student, which means that I'll *eventually* leave www.umich.edu, but this should work as long as I'm here: http://www.umich.edu/~mlerner/PyMOL/sidechain_check-1.4.py (I'd *love* to have some eligibility left so I could keep playing college ultimate. These days, I'm playing on a club team and in a league. Ann Arbor has a ton of frisbee .. not as much as the northwest, but way better than the northeast where I went to college.)
Created attachment 82824 [details] ebuild to include sidechain_check plugin this works as expected on ~amd64 thanks!
Hey there, this would actually work better as its own package. That way, if Lerner puts out a new plugin, people don't need to recompile all of pymol. Could you attach an ebuild to do that? Should be pretty straightforward.
Aboslutely, give me a few days...got something important coming up. :D
(In reply to comment #13) > Aboslutely, give me a few days...got something important coming up. :D ping
mmkay...working on it :D
Created attachment 89522 [details] ebuild to add sidechain_check to pymol Since sidechain_check modifies /usr/lib/python${PYVER}/site-packages/pmg_tk/skins/normal/__init__.py, I save a copy to restore in case the package is ever unmerged. If you can think of a better way to handle it (yeah, it's ugly) please let me know. Thanks, je_fro
I'd actually suggest telling users to make the changes manually, or give a command to apply the patch in FILESDIR instead of doing that hacky stuff. Shouldn't need your own pymol tarball if you do this. I didn't realize that sidechain_check required apbs, that seems weird. Also you're missing multilib intelligence, replace all mentions of lib with $(get_libdir) and inherit multilib.eclass. Looks like you could add python_mod_optimize() in postinst and python_mod_cleanup in postrm().
Thanks for the input...I'll implement those tonight. As for apbs, I suspect that most people who want this are going to use the apbs plugin next....
(In reply to comment #18) > As for apbs, I suspect that most people who want this are going to use the apbs > plugin next.... I personally see no need to do so. Just because I'm fixing flips doesn't mean I want to look at an electrostatic surface.
How would you feel about a pymol "plugins" flag that will handle apbs _and_ sidechain_check?
apbs plugin is already distributed with pymol, by upstream. That makes it totally different from this one.
Yes, but you still need to install apbs for the plugin to be of any value.