John Mayfield
John Mayfield
What are they...? We should be API stable, in the past though there was issues with classes duplicated across JChemPaint and CDK.
See Gotha's https://github.com/cdk/cdk/wiki/AtomContainer2. Quick fix should be ```-DCdkUseLegacyAtomContainer=false```
Sorry other way... ``-DCdkUseLegacyAtomContainer=true``
Yep, ``java -DCdkUseLegacyAtomContainer=true -jar app-jar/target/JChemPaint.jar`` stops the exception as expected but there still a lot of other things that are broken in general to the point it's unusable.
Are you getting a NPE? If so the bug is still present in the current dev: https://github.com/JChemPaint/jchempaint/blob/master/core/src/main/java/org/openscience/jchempaint/application/JChemPaint.java#L195 One (or both) of those nulls should not be null, that method expects...
Sounds good, maybe a week or so for comment on the API - until release though it’s still changeable. I will fix the code smells and try and get some...
@uli-f incase you didn’t get pinged on it
> > maybe a week or so for comment on the API > > Shall I send out a tweet? No I will add some more doc first then do...
I'll look into it, but doing RDKit UGM talk this week and next (presenting). You can't really have X and v on the right hand side... SMIRKS doesn't work like...
Fixed... the bond ordering swapping was not handled correctly. Can you add this as a test case. Something simpler like: `` [C:1]-[O:2]>>[O:2]=[C:1] `` would also have shown the issue.