Mercurial 6.1rc0 tagged

[Gregory Szorc]

On Sat, Feb 19, 2022 at 8:54 PM Manuel Jacob <me at manueljacob.de> wrote:

> On 20/02/2022 02.03, Gregory Szorc wrote:
> > I found a crashing regression when using 6.1rc0 locally. Nearly all of
> > my commands threw errors when I had obsolescence enabled.
>
> What do you mean by “obsolescence enabled”? I do not get the error if
> the core obsolescence features (config experimental.evolution=all) are
> enabled but the evolve extension is disabled. However, I can confirm the
> error if the evolve extension is enabled (when the latest released
> version 10.4.1 is installed). It works for me with evolve
> e314f6207e5f04da921cc890cf67bb7b270c076d or later.
>
> > The fix /
> > revert is https://phab.mercurial-scm.org/D12203
> > <https://phab.mercurial-scm.org/D12203>.
> >
> > Should we ignore 6.1rc0 and tag a 6.1rc1 with the fix?
> >
> > Also, I'm somewhat surprised this didn't trip any tests. Perhaps our
> > test coverage with obsolescence enabled isn't too great?
>
> The evolve CI was failing:
> https://foss.heptapod.net/mercurial/evolve/-/pipelines/34785
>

Following up on this, I'm convinced it was an evolve extension issue. Sorry
for the noise. I'll try to get Windows 6.1rc0 artifacts published.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mercurial-scm.org/pipermail/mercurial-packaging/attachments/20220227/f3a180b1/attachment-0002.html>