*ping*

Sep 17, 2023 22:08:53 Johannes Thyssen Tishman <li...@thyssentishman.com>:

> Apologies for the poor formatting. I had 'textwidth' set to 80
> characters. Fixed below.
>
> Johannes Thyssen Tishman <li...@thyssentishman.com> wrote:
>> Hello everyone,
>>
>> attached is both an update for CGAL from version 4.14.2 to the latest
>> 5.6 and a revision to OpenSCAD to work with the CGAL update. Note that
>> these are to be imported together to keep OpenSCAD from breaking.
>>
>> Despite being actively developed, OpenSCAD hasn't seen a release since
>> 2021. Until now, OpenSCAD is CGAL's only consumer, which has allowed
>> CGAL to become outdated without causing any issues. However the latest
>> version of the recently imported PrusaSlicer also requires CGAL, but
>> will not work with the old version we have in ports.
>>
>> Regarding the OpenSCAD revision, the master branch already depends on
>> newer versions of CGAL which means that we can simply backport some
>> commits to update CGAL. That is basically all what the attached diff
>> does. Most if not all new patches should be fine to drop once OpenSCAD
>> makes a release.
>>
>> Now for the CGAL update, since version 5.0 CGAL is a header-only
>> library which can be installed without compiling. I've included only
>> the essential dependencies[1] and left all other optional dependencies
>> up to the software using CGAL that depends on them. Besides building a
>> couple of the examples provided in CGAL's source, I'm not sure how to
>> judge the quality of the port. I also blindly assumed that the
>> 'BROKEN-mips64' line could be removed as there is no compilation
>> needed anymore. I'd be interested to know if this update unbreaks
>> OpenSCAD in sparc64.
>>
>> Any feedback or OK's would be very appreciated. Both 'portcheck' and
>> 'make port-lib-depends-check' are happy with the new changes in both
>> ports (the latter has some complains about OpenSCAD's port, but I
>> assumed it could be ignored since it's from a previous update). I
>> tested some examples from OpenSCAD and they seem to work just fine,
>> but I'd appreciate some testing from someone who actually uses it.
>>
>> Thanks!
>>
>> Kind regards,
>> Johannes
>>
>> [1] https://doc.cgal.org/latest/Manual/thirdparty.html

Reply via email to