Re: xassert in dispextern.h

[Kim F. Storm]

Jason Rumney <address@hidden> writes:

> David Kastrup <address@hidden> writes:
>
>> That's completely backwards.
>
> What's completely backwards is to turn off a feature that helps us
> find bugs so that people can treat the CVS HEAD as a stable release of
> Emacs.

At least half of the xasserts that have been reported to fail so far
have been false alarms -- or at least too harmless to cause an abort.

Before Miles enabled xasserts unconditionally, emacs had IMO been
running rock solid for a long, long time -- except for a few
unexplainable crashes. 

Now we have a more unstable and slower emacs -- but none of the aborts
so far have given a clue to any of the unexplainable crashes.

I agree with David that the benefits from the xasserts do not justify
the costs at this stage of development.  IIRC, even 21.1 was released
without having xasserts defined by default during pretest.

-- 
Kim F. Storm <address@hidden> http://www.cua.dk