deleted DELETE

Bron Gondwana brong at
Thu Aug 30 22:42:22 EDT 2007

On Thu, Aug 30, 2007 at 10:19:33AM +0100, David Carter wrote:
> On Wed, 29 Aug 2007, Ken Murchison wrote:
>>> Small bug fix (the p++ line) attached.
>> Good catch.  I didn't look at the cyr_expire patch real close, since I 
>> assumed Fastmail was using it in production.  I know, I know, I shouldn't 
>> assume anything and test it myself.  Mea culpa.
> Well, if Bron can spot bugs and blatant omissions in my patches and I can 
> spot bugs in his code, then peer review seems to be working. It would 
> probably work faster if we were all running Cyrus CVS rather than our own 
> private source trees. I hope to get there about this time next year.

It's amazing how much trying to understand how someone else's code works
forces you to build a brand new mental model and the bugs stand out.
Reading your own code of course, the model is already there and you
make stupid assuptions.  *sigh*

> There shouldn't be any embarrassment about broken code in CVS, so long as 
> we are able to generate a separate branch for emergency bug fix releases. 
> At the end of the day CMU's CVS repository is the canonical reference. 
> Trying to track third party patches (and patches to patches) is hard work.

It certainly is!  Quilt is helping a bit here, but not as much as I'd
like - there are a wide range of mistakes that can make your working
area bogus, so I spend too much time reading 'svn diff' output on the
patches directory.  Diffs of diffs are really annoying to read!

I don't suppose CMU is tempted to move to something a bit more
featureful than CVS any time soon?  Even SVN is an upgrade, but
there are a plethora of distributed systems out there too.


More information about the Cyrus-devel mailing list