bison-patches
[Top][All Lists]
Advanced

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

Re: testsuite: work around autoconf 2.63b bug


From: Joel E. Denny
Subject: Re: testsuite: work around autoconf 2.63b bug
Date: Tue, 7 Apr 2009 15:38:17 -0400 (EDT)

Hi Eric,

On Tue, 7 Apr 2009, Eric Blake wrote:

> I noticed this started failing when I upgraded autoconf to the recently-
> released beta 2.63b:
> 
> $ dbgmake check TESTSUITEFLAGS='-k \#'
> 
> The bug (actually, two related bugs) has since been fixed in autoconf, per 
> this 
> thread: 
> http://lists.gnu.org/archive/html/autoconf-patches/2009-04/msg00002.html
> 
> But since others may be bit by this, would you like me to apply this patch to 
> all the active branches to avoid tickling the bug in the first place?

Yes.  I don't usually install betas of autoconf, but it might help someone 
else.  Thanks for continuing to look out for us.

> Is branch-2.4.2 still active?

Yes.  The other active branches are master and branch-2.5.

> Do you prefer me applying topic branches and then merging up from oldest to 
> newest (how automake.git and git.git behave), or just linear history with 
> cherry-picking (how findutils.git and m4.git have been behaving)?  gnulib.git 
> also uses linear history, but unless you have multiple active branches, the 
> question is moot.  I kind of like the idea of merging topic branches.

I've been cherry picking because I can easily understand that approach.  
I haven't studied merging topic branches enough to decide whether it's 
worthwhile.  Will that approach quickly accumulate dead branches in 
bison.git?  Will commits with multiple parents make future use of 
bison.git more confusing?  Anyway, if you feel it's advantageous, go for 
it, and maybe I'll learn something new.

Thanks.




reply via email to

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