On 11 February 2014 12:17, Dirkjan Ochtman <dirk...@ochtman.nl> wrote: > On Tue, Feb 11, 2014 at 11:56 AM, Benoit Chesneau <bchesn...@gmail.com> wrote: >> Imo we can't know if COUCHDB-1986 is really fixed for now due to the >> comment, the fact that all tests pass using the combination of the 2 >> commits (bit not with one or the other alone) is mysterious. However >> they are fixing real issues so i would release now with them instead >> of waiting more. >> >> Thoughts? > > SGTM. > > Cheers, > > Dirkjan
Hi Dirkjan, all, Go for it. I don't think we should be delaying a release unless: - we have a clear regression that is not addressed - we can't build on a supported platform - we've got an accepted blocking ticket open i.e. shipping improved software regularly is to me more important than having all open JIRA tickets addressed[1]. However I don't think we have a successful `make distcheck` on FreeBSD 10 yet, & unfortunately I've not got time before mid March (!) to do further work on FreeBSD 10 issues - you'll notice I'm even hiding from IRC! There are a number of other recent & potential fixes in ibrowse, which is not yet tagged v4.0.3 they may also be of use. BTW Alex has access to the box and I'm happy to provide that to any community member who wishes to help out. [1]: http://www.hyrumwright.org/papers/floss2009.pdf BTW this paper from the svn team in 2009 is worth a read A+ Dave
