Mon 28 May 2007 04:19:00 PM UTC, comment #12:
closed after 0.8.0 release
|
Wed 09 May 2007 06:37:56 PM UTC, comment #11:
Yes, I saw, and it seems that #19840 is related to that. We need a way to change the position of the hydrogen atoms, may be left, right or auto instead of just auto. Feel free to open a new bug report, since the auto position is not always what is wanted. Seems 0.8.0 will be a bit late ;-)
|
Wed 09 May 2007 06:30:00 PM UTC, comment #10:
There is one problem with the templates: I had to input some of the hydroxyl groups as groups of atoms instead of single atoms, as gchempaint was putting OH/HO in opposite and thus masking the bonds of the ring. As a result, to attach something to the hydroxyl group you have to delete the atom group and insert the atom. Hope this is clear enough :)
I think an option in which order to put atoms and hydrogen, something like right click -> flip atoms order would be a much nicer solution. I'll be happy to fix the templates then.
|
Wed 09 May 2007 05:46:16 PM UTC, comment #9:
I was frightened myself :) I got one reproducible case, report will follow shortly. I think I might be hitting the same problem every time, as I am drawing quite big molecules.
|
Wed 09 May 2007 05:39:20 PM UTC, comment #8:
Fixed the typo.
I don't get so many crashes. You are frightening me.
|
Wed 09 May 2007 04:48:17 PM UTC, comment #7:
There is a wee typo in α-D-glucopyranose name - I translated it from Polish, but left k instead of changing it to c.
As for the crashes, I am getting some random ones quite frequently, so I'll just shake the program in gdb and try to obtain some (or hopefully one reproducible) backtrace.
|
Wed 09 May 2007 04:41:14 PM UTC, comment #6:
I intend to add the templates.xml to the tarball as is (just change its name to haworth.xml.in to avoid collisions and build issues). The names will not be added to the po files at the moment, may be they should, but I don't want to add new strings to translate at the moment.
I fixed the bond length issue. We might have similar issues when importing data from other programs, but this is not really easy to fix (needs hacking in openbabel).
The bad news is that gchempaint crashed but I'm unable to reproduce, so I fear we'll have to live with it until somebody finds how to reproduce this crash (or another one).
If you want to add other templates, just send me the file, no need to open an new bug or reopen this one.
|
Wed 09 May 2007 02:50:53 PM UTC, comment #5:
FYI: furanoses are 90, 140 and 185 in length.
|
Wed 09 May 2007 02:49:43 PM UTC, comment #4:
I feared this might happen. I should find something better than the median bond length for scaling. Needs some thinking.
|
Wed 09 May 2007 02:45:08 PM UTC, comment #3:
After adding to the templates, pyranoses keep their dimensions (140 for ring, 90 for hydroxyl groups). Furanoses, on the other hand, become far too big.
|
Wed 09 May 2007 02:29:06 PM UTC, comment #2:
Great thanks.
Anyway, you should add them as templates. Next we'll ship those templates with gchempaint (or gchemutils in future releases).
|
Wed 09 May 2007 02:22:18 PM UTC, comment #1:
I think this will require a feature for the templates to contain bonds of different width to work well.
|
Wed 09 May 2007 02:12:27 PM UTC, original submission:
I was kinda dissatisfied with the way carbohydrates are presented, so I started wrapping up some haworth projections manually. For starters, here is glucose (pyranose) and ribose (furanose). Given enough time and manpower, we should be able to wrap up all the common ones. Keep in mind that some bond length fiddling was necessary.
|