httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From (David Robinson)
Subject Re: NEXTSTEP bug fixes (fwd)
Date Tue, 29 Aug 1995 18:48:00 GMT
>> > 
>> > Unfortunately, this *is* past deadline, 
>> for the record, what counts as a show stopper bug?
>> this bug can cause big problems. If it had been on a more popular
>> flavor of Unix I'm sure it would count as a show stopper. So where
>> does the line get drawn ?
>> if it means an extra day before 0.8.12, that's fine by me.
>At the risk of pissing someone off..
>I would place my vote with Rob H on this one. 
>I understand the need to have a tight voting/release procedure,
>but I still think that democracy should be in effect for changes
>in that schedule. There would be no harm in delaying the release
>long enough to include this fix. (IMHO)

I'll take the time to reply to this one, despite being very busy.

There are good arguments for not delaying 0.8.12. It is _much_ easier
for us to develop, test and distribute patches if we all have a consistent
code. It is difficult to develop a new patch when others are testing (voting);
either you waste your time in having to re-patch the next release, or the
others waste their time in having to re-evaluate the current distribution.

However, this only applies if a new version is NOT automatically released
to the public, i.e. if 0.8.12 is an internal only version. If we intend
to release 0.8.12, then it is senseless to not including patches which fix
known bugs.

Too frequent releases is a sign of laziness, pure and simple. We show that
we do not bother to think about the significance of the known bugs in the
code, and we do not mind wasting peoples time when then re-compile
the latest version only to discover that their favourite bug has yet to
be fixed.

+1 on building 0.8.12.
-1 on releasing 0.8.12 to the public.
-1 on releasing a version with these bugs:

 NeXT showstopper
 Unable to compile under SCO
 trailing slash alias bug


View raw message