Greg Landrum
Greg Landrum
I reverted this because @rvianello was not, in fact, done reviewing it. That was my mistake. Sorry for the confusion.
> Still relevant. If there is a consensus that Brenk is left and Dundee can be removed, I can try making a PR for this. How about we deprecate both...
Confirmed. That documentation is in dire need of updating.
This happens because we removed the allowed oxidation state of 6 from the Al atom. So the hypervalent AlCl3 dimer as is present in that mol file is not supported...
I think so. That change was really just intended to make it easier to figure out what the valence should be. If you have other ideas, we can definitely discuss
I agree (and @ptosco definitely agrees) that we need to have a discussion about how dative bonds are handled, but that's not directly connected to the removal of the valence...
@ricrogz I understand the point you are making, but you want a change in dative bond behavior, like what @ptosco is working on. I think the current behavior: to reject...
It definitely changes things, and is documented as such, but this is not related to dative bonds. The reason your Al example is behaving differently is because we removed the...
This one looks good to me. @ricrogz: you've done some work on the 2D vs 3D in mol blocks too, could you please also take a look?
[This comment was updated and corrected on 21.08.2025] Here's a summary of the current status (I tested using v2025.03.5) of what the RDKit does: | flag | coords | wedging...