Re: scratch/igc renamed ad feature/igc [was: scratch/igc as a feature branch]

[Michael Albinus]

Gerd Möllmann <gerd.moellmann@gmail.com> writes:

Hi Gerd,

> An open question is on which systems we don't want to test igc for some 
> reason.
> I've left a placeholder in test/Makefile.in for that (TEST_IGC should be
> set to "no" for these).

For the time being, we test on emba. This uses Debian bookworm as basis.

In Dockerfile.emba, I've added a now target emacs-igc. It installs
libmps-dev and libmps3, and it configures emacs --with-mps.

In gitlab-ci.yml, I've replaced the target emacs-inotify by
emacs-igc. This should be sufficient as starting point.

We can adapt as we want if there are further needs. And of course, once
merged with master, we need a more sophisticated integration. But this
can wait.

>> One question: will be '--with-mps' still a confuration option after
>> merge to master? I.e., do we need to distinguish different testc
>> scenarios?
>
> My understanding is that some systems don't support MPS, but I don't
> know which ones that are. ISTR that Po Lu once said that Android is one,
> but I'm not sure. So, I think --without-mps will still be needed.

This I don't know. Must be discussed on emacs-devel@.

Best regards, Michael.