[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Bisecting display bugs

From: Óscar Fuentes
Subject: Re: Bisecting display bugs
Date: Mon, 11 Jul 2016 19:46:42 +0200
User-agent: Gnus/5.13 (Gnus v5.13) Emacs/25.1.50 (gnu/linux)

Eli Zaretskii <address@hidden> writes:


>> +Some caveats:
>> +
>> +- This script cleans Emacs' source directory with ‘git clean -xfd’, so
>> +  make sure your uncommitted changes are saved somewhere else.
> While bootstrapping after each step of bisect is safe, it makes the
> run much longer, so I'd try bisecting without a bootstrap first,
> especially if the range of commits to bisect is relatively small.

Although that sounds good in principle, my experience tells me that
taking the short path does waste more time than it saves on this case.
YMMV, but too many times the bisection stopped with a build failure that
wouldn't happen with bootstrap, or simply gave wrong results because the
build didn't work correctly.

Speaking of this, the sample script could offer bug failure detection
and `git bisect skip'.


>>                                              but some are only
>> +apparent through visual inspection.  Since building Emacs takes a long
>> +time, it can be a pain to debug these manually.
> I don't follow this logic: building is an automated process, while
> visual inspection is usually very fast; the automated comparison
> you are about to suggest doesn't decrease the build time.  So how does
> the conclusion follow?

Because I have to be there in front of the terminal. And I can make
mistakes. Anything that makes the bisection automatic is a big


reply via email to

[Prev in Thread] Current Thread [Next in Thread]