On Nov 2, 5:12 pm, Dan McGee <[email protected]> wrote:
> Hey everyone,
>
> I have no idea what's going on with memcached development, and digging
> through the list doesn't make it apparent why there have only been 6
> commits since last February to the master branch.

  We've been doing all of our fixes and new features in the engine
branch.  There's been a couple of changes that have come in, but they
went a bit beyond maintenance, so they haven't gone in yet (the larger
the change, the more difficult it is to forward-port it where
development is happening).

> Since this is what
> gets shipped by distros, and GCC 4.5.X has been out for a while, it
> stinks that things aren't compiling cleanly anymore. The errors below
> are due to pointer magic that isn't allowed if following C
> strict-aliasing rules:

  Issue 60 was opened a while back for this error and we got it
building under 4.4.  Apparently 4.4.1 got a bit more strict.

> I've fixed this issue and two other small things noted along the way
> in my fork (http://github.com/toofishes/memcached), so wanted to give

  Looks pretty small and doesn't seem to upset any of the builders.
Thanks.

> an FYI for that, but was also trying to get a bit of a grasp on where
> development sits- http://contributing.appspot.com/memcached and 
> http://code.google.com/p/memcached/wiki/DevelopmentRepos really aren't
> all that helpful if you don't want to troll through mailing list

  Thanks for the feedback.  I'll update that a bit.

> archives to figure out what is actually relevant. I did pull down the
> other repos on the latter page and was quite surprised to see 300+
> commits there, so let me know if this work was wasted on a wrong
> branch.

  This is right for minor fixes to get the code compiling on other
platforms (BTW, if you could donate a build slave, that'd be awesome
-- right now, I can only verify it doesn't break other systems):
http://code.google.com/p/memcached/wiki/BuildFarm

Reply via email to