2025-02-19T17:14:13+0000 Stuart Henderson <s...@spacehopper.org>:
> On 2025/02/17 17:43, Johannes Thyssen Tishman wrote:
> > While I usually prefer to enable most options in a port, I decided to
> > start with a smaller vtk by not including all it's modules. In the
> > future, if there is an interest for other modules, we can enable them.
>
> good call.
>
> On 2025/02/19 15:31, Johannes Thyssen Tishman wrote:
> > 2025-02-18T20:00:16-0400 j...@bitminer.ca:
> > > A couple of suggestions:
> > >
> > > 1. change all those 0.0 in the SHARED_LIBS lines to a symbol like
> > > ${LIBVIR}
> > >
> > > 2. consider turning on wrap-python to allow python to have access to the
> > > libraries. Perhaps this could come later.
> > >
> > > I successfully ran an 8.2 version for a while, but that goes back a long
> > > time ago.
> > >
> > > (sorry I don't have the time to test this.)
> >
> > Thank you for the suggestions. Please find an updated tarball attached
> > with python support enabled.
>
> I approve of the ".for LIB in" in this situation, much more compact than
> a huge number of SHARED_LIBS lines. (With this many libraries I don't
> think we can meaningfully check for ABI issues, and as it's C++ there's
> an extremely high chance that there will be symbols removed so a major
> bump each time anyway).
>
> I've not looked further yet, but from reading the port, this seems good
> to me so far.
>
> How are we doing on the path to freecad, are there many ports left?
For FreeCAD-1.0.0, vtk should be the last one (considering I didn't miss
anything). Thank you Stuart.
