2011/2/27 Beman Dawes <bdawes@acm.org>
From my viewpoint, the process of putting together a release is now smooth enough that I'm willing to do a point release, but...
(1) Do we have enough fixes that will be ready by, say, a week from today to make a point release worthwhile?
For my library Boost.Icl, which is the single new library addition of 1.46 it would be nice to have a point release. John Reid, who is writing a wrapper application around ICL, detected a bunch of compilation problems shortly before the release of 1.46.0 so I wasn't able to merge those fixes into the release branch. Moreover there are a couple of documentation flaws that were detected within the last few days. The fixes of the bugs detected by John are already in the trunk. Should more problems arise from the release of 1.46.0 I'd be eager to fix them during the one week period. So yes, from my perspective, it would be nice to have a point release and I have improvements on the ICL that are already done or can be done within one week.
(2) We would need to be sure that we have a procedure that ensures that fixes for the point release don't get dropped from the next full release.
For ICL I see no problem here. Best regards, Joachim -- Interval Container Library [Boost.Icl] http://www.joachim-faulhaber.de