Sigh. Oh InChI. At a recent meeting, 'we' tried to encourage InChI to move
to Github and an open development model, but the bogeyman of people
modifying the code and still calling it InChI was brought up. As Roger
pointed out, you don't need the InChI code to do this, you can just make up
an InChI out of your head and put it into Wikipedia, e.g. InChI=1S/CCC/Noel.
Anyway, I have no particular objection to this. Although I would still like
to maintain the ability to link against an existing inchi library if
present - mainly just to speed up the build process as a developer, but
maybe also useful for Linux distros if they already bundle it.
- Noel
On 7 June 2018 at 21:48, Matthew Swain <m.sw...@me.com> wrote:
> There are actually a fair few small modifications to the InChI 1.04 code
> bundled in Open Babel. Most look superficial, but one that actually
> slightly changes the behaviour is the increase of MAX_ATOMS to 65534
> in ichisize.h. The more recent InChI 1.05 implements a similar change, but
> uses a value of 32766 and outputs InChIs starting with “1B” (for beta?)
> instead of “1S” above 1024 atoms.
>
> I think in future it would be better to download the unmodified InChI
> release on-demand when building, instead of bundling it. Like what is done
> for RapidJSON:
> https://github.com/openbabel/openbabel/blob/master/
> CMakeLists.txt#L463-L490
>
> Matt
>
> On 6 June 2018 at 15:13:26, Geoffrey Hutchison (geoff.hutchi...@gmail.com)
> wrote:
>
> > Not sure how relevant this is to OpenBabel, but I assume we modify InChi
> > in order to integrate it?
>
> No, we do not modify the InChI code in any way in the integration.
>
> > Did somebody bring this up with the Inchi Trust?
>
> I would guess 'no.' I don't think IUPAC or the InChI Trust really
> understand software licensing issues. IANAL, but I know that similar issues
> have been brought up over the lifetime of InChI.
>
> I think their basic concern is someone modifying the InChI code (such as
> it is) so that it no longer produces correct InChI - which then breaks the
> entire point of an InChI. So I understand the intent of the clause, but
> that shouldn't be in the license - just a trademark on InChI, such that
> claiming it implies correctness.
>
> I'd suggest posting on the InChI mailing list.
>
> -Geoff
> ------------------------------------------------------------------------------
>
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> _______________________________________________
> OpenBabel-Devel mailing list
> OpenBabel-Devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/openbabel-devel
>
>
> ------------------------------------------------------------
> ------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> _______________________________________________
> OpenBabel-Devel mailing list
> OpenBabel-Devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/openbabel-devel
>
>
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
OpenBabel-Devel mailing list
OpenBabel-Devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/openbabel-devel
