Sat 08 Nov 2008 10:35:49 AM UTC, comment #16:
closed after 0.10.0 release.
|
Tue 29 Jul 2008 02:35:19 PM UTC, comment #15:
For the essential, this bug seems fixed in svn trunk version. Things are not absolutely perfect, but good enough for 0.10 and the samples attached to this bug. I consider it fixed and I'll close it after 0.10.0 release.
If other files fail to open correctly with the development branch, please file as many bugs reports as necessary.
(file #16209)
|
Mon 28 Apr 2008 07:58:15 AM UTC, comment #14:
The current screenshot shows that things now mostly work. I'll not close this bug because it is not perfect yet and some chemistry is still lost.
I think that cdx export will not be possible for 0.10 because of time constraints unless the release waits until 2009. cdxml export will still use the openbabel version with much data loss.
(file #15536)
|
Sat 23 Feb 2008 09:46:31 PM UTC, comment #13:
I am so sorry for not getting back to you sooner. Thank you, Jean, for your patience and your work.
Here are the samples:
Hope they help!
(file #15100)
|
Mon 18 Feb 2008 08:43:52 PM UTC, comment #12:
As can be seen in the screenshot, the new cdx plugin code is already somewhat better than the previous openbabel based one. At the moment, all fragments are replaced by a carbon atom as can be seen, for example, in the second molecule where OBn is missing. This is my priority at the moment, but it is all but simple.
Next, I'll have to address text decorations and alignment.
The two reaction schemes are not still perceived as such, because GChemPaint support is not yet sufficient, and because there is no pagination either, so that, if I tried to make them reactions in GChempaint, they would end on one very large line, clearly inappropriate.
So even if things are in progress, we are still far from perfection.
I'm still interested by the samples I asked in comment #11.
(file #15072)
|
Sun 30 Dec 2007 10:14:01 AM UTC, comment #11:
I'd like to have the same samples but in cdxml format if possible to be certain that it will work too, also a sample with a chemical nick (as Ph-Cl) or so would be useful (in oth format).
Current state is a correct loading of simple structures with no stereochemical bonds and basic text (no rich text) import. No export yet.
|
Wed 28 Nov 2007 05:04:12 PM UTC, comment #10:
Thanks. We evacuated to be cautious, but our neighborhood was completely intact.
Let me know if you need any more ChemDraw files.
|
Wed 07 Nov 2007 08:48:54 AM UTC, comment #9:
Thanks for the samples. I hope fires had only minor consequences for you.
|
Tue 06 Nov 2007 05:27:20 PM UTC, comment #8:
Sorry it's been taking so long.
I've been busy trying to catch up after the San Diego fires (in California, USA, which is where I live).
I made ChemDraw files and PNG files of what the CD files should look like. Hope this helps. If you want more complicated examples, let me know. Thanks.
Tak
(file #14330)
|
Sun 28 Oct 2007 02:07:45 PM UTC, comment #7:
Yes, I can help with that for sure. I'll be on it later this afternoon or tomorrow. Thank you for your work!
|
Sun 28 Oct 2007 08:41:55 AM UTC, comment #6:
In the end the issue is more critical than I initially thought. May be Openbabel is just inappropriate to import cdx or cdxml files in GChemPaint. I proposed a first patch to the openbabel-devel list adne I'm waiting for comments, but, most probably, I'll have to write a direct import/export procedure.
You might help by providing little sample files illustrating only a few features at once, hopefully in both formats and with an associated screenshot.
|
Fri 21 Sep 2007 10:14:01 AM UTC, comment #5:
The triple -> quadruple bond issue is an OpenBabel issue too (fixed in svn, both trunk ans 2.1 branch).
|
Thu 20 Sep 2007 12:45:03 PM UTC, comment #4:
Partly understand what happens: only molecules are imported, so reaction arrows and text are discarded; abbreviations are replaced at some point (not in gchempaint) by a semi-developed formula (the case for Bn, Tf, TMSO,...), but the new atoms are put sometimes at silly places. If this is fixable, it is at the OpenBabel level (I partly maintain the cdk format plugin there).
I do not understand at the moment how we end with quadruple bonds instead of triple.
Seems this bug will remain opened for a long time. The attached sample is a very interesting use case. I diminish the severity to wish, as it concerns essentially missing features.
|
Tue 18 Sep 2007 05:45:01 PM UTC, comment #3:
Thanks. I'll first fix the molecules issue, then see how it is possible to load the reactions too, but this will need enhancements in gchempaint (no support for multi steps reactions, planned for 0.10) and a large change in OpenBabel.
|
Tue 18 Sep 2007 05:19:14 PM UTC, comment #2:
Here is the screenshot in WinXP
(file #13983)
|
Tue 18 Sep 2007 05:04:18 PM UTC, comment #1:
Thanks for reporting. Do you have access to chemdraw ? If so, please attach a screenshot of that file. I'm not sure it is a gchempaint or an openbabel bug (or both). Anyway, I'll manage to fix it.
|
Tue 18 Sep 2007 04:33:17 PM UTC, original submission:
Hi Jean,
Thank you for 0.8.3! It fixed all the problems mentioned earlier. It opens .gchempaint files perfectly.
It does not open .cdx (chemdraw) files correctly though. The structures look like they've been shaken in a bag and the texts are gone. I vaguely remember that you mentioned that installing the latest openbabel may solve the problem?
I have uninstalled everything and recompiled openbabel 2.1.1, goffice 0.4.3, etc. I still have this problem.
sample .cdx file attached
Thank you!
|