Am Sa., 15. Juni 2024 um 19:57 Uhr schrieb David Bremner <david@tethera.net>:
>
> Marc Fargas <telenieko@telenieko.com> writes:
>
> > Hi again,
> >
> > El dom. 24 de mar. 2024, Marc escribió:
> >> El mié. 13 de mar. 2024, Carl escribió:
> >>> (...)
> >>> Could you put your fix together in the form of a git-appliable patch?
> >>> Such as by applying it to the notmuch source, running `git commit` and
> >>> then `git format-patch HEAD~` or similar.
> >>
> >> Please disregard the previous patch, consider the one attached
> >> here. Hope I did the "format-patch" thing properly.
> >>
> >> I had to move the `without-restriction` call to the top of the
> >> `with-temporary-notmuch-message-buffer` defmacro. It does work properly
> >> now.
>
> Applied to master. Sorry for the long-ish silence.
>

Does this depend on emacs version by any chance, i.e. is
`without-restriction` defined on all emacsen? In Fedora's copr
infrastructure, all builds succeed but some tests fail with
`Symbol'€™s function definition is void: without-restriction`. The
picture is the following:

PASS:
centos-stream+epel-next-8 with emacs 26.1-11.el8
rhel+epel-8 with emacs 26.1-11.el8
fedora-39 with emacs 29.3-1.fc39
(plus all newer fedoras)

FAIL:
centos-stream+epel-next-9 with emacs 27.2-9.el9
rhel+epel-9 with emacs 27.2-9.el9

Yes, RHEL 8 is EOled.

I find it strange that tests pass on both emacs 26 and 29 but not
27.2. But said code path is not executed on emacs 26 at all?

The failing 42 tests are mostly around json and crypto, but not all.

Cheers
Michael
_______________________________________________
notmuch mailing list -- notmuch@notmuchmail.org
To unsubscribe send an email to notmuch-leave@notmuchmail.org