From: Robert Ramey (ramey_at_[hidden])
Date: 2004-01-09 12:31:25
I have to say I was surprised at the amount of time that
has elapsed between the anouncement of boost 1.31 (1 November ?)
and the actual release date (TBD at 8 Jan).
I had assumed that this entailed freezing or branching the current
development version and getting out any pending bugs. This
seemed to me a short job - maybe 2 weeks.
The first warning is when I hear people wanting to get
in "one more feature"
In my view, I see the CVS development tree as a
"release candidate" and would contain code that that
as far as the author knows is ready for release. Of
course, this is never quite true due to the fact that
we test on more platforms than any one person has.
I know this might sound ridiculous but
how about a "surpise release" strategy?
Maintaines only check in to CVS code that they believe
is ready for release. Some designated individual would
"stop the music" and only bug fixes would be permited.
Maintainers could improve their as they see fit. Occasionally
or in small increments - but only checking in code they
feel comfortable releasing.
the "music would be stopped" (suspended) when there are enough
changes to make it worth the trouble.
developers using the cvs tree would be confident that
that thier efforts would be compatible with the next release.
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk