Hi!

Normally, I would try to fix problems rather than backing out of a commit.
However, today it was not possible for me to give the problem any
attention. Given the severity (it broke the build process), I decided to
revert it so that I could look at in piece and quiet.

When the fix to ucs-normalize has been pushed, I will verify that Emacs
builds from scratch with my patch is installed (on multiple systems),
before publishing it again.

Anyway, I will post further messages on this to the "22169" issue, to avoid
filling emacs-devel and personal mail boxes with (possibly) irrelevant
information.

Finally, thanks to everybody reporting this.

    -- Anders


On Thu, Dec 24, 2015 at 6:01 PM, Eli Zaretskii <eliz@gnu.org> wrote:

> > Date: Thu, 24 Dec 2015 10:06:00 +0100
> > From: Anders Lindgren <andlind@gmail.com>
> > Cc: Kaushal Modi <kaushal.modi@gmail.com>,
> >       Emacs developers <emacs-devel@gnu.org>,
> >       Angelo Graziosi <angelo.graziosi@alice.it>
> >
> > The commit has been reverted. Sorry about the inconvenience.
>
> Thanks, Anders.
>
> For the future, there's no need to revert a commit, unless it's clear
> that there's a bug in the commit itself.  In your case, the commit is
> correct and bug-free, and it fixes a real problem.  In such cases, the
> way to fix any fallout is to find the new problem exposed by the
> commit and fix that new problem.  If people are annoyed by the
> temporary breakage of the branch, they can always reset their HEAD to
> a prior commit, until the problem is fixed.  They can even try
> debugging the new problem and fixing it ;-)
>
> Anyway, I found the reason for these weird messages.  It was a subtle
> issue with the code in ucs-normalize.el that was always there, and was
> simply exposed by promoting it to be compiled by bootstrap-emacs
> earlier in the bootstrap process.
>
> So please revert the revert, and I will follow up with the changes
> that fix the broken bootstrap.
>
> Thanks.
>