On Wed, 17 Oct 2012, Rainer Gerhards wrote:
Please see http://blog.gerhards.net/2012/10/new-stable-policy-for-rsyslog.html
As I posted, it makes sense. A lot of projects are finding that the develop-test-stable cycle doesn't really work. People avoid the .0 releases no matter how much testing they have had, and they are perfectly willing to jump 30 numbers in the second column rather than a single number in the first column.
I would say don't make changes to the first digit unless you are making backwards-incompatible changes (which should be very rare)
if you are doing a large or invasive new feature, create some -rc releases and post them here for people to try (the fact that you do a -rc will act as a heads up for people that this needs more testing than a normal release), and then just release 7.x.0 when it passes your tests, and fix bugs in the 7.x.y releases.
just don't get yourself trapped into supporting a lot of 7.x series, stop supporting 7.x.y around the time of 7.x+1.<small number> unless it looks like 7.x+1 changes are a lemon
Yes, this will sound similar to what the kernel developers are doing, where do you think I stole the ideas :)
David Lang _______________________________________________ rsyslog mailing list http://lists.adiscon.net/mailman/listinfo/rsyslog http://www.rsyslog.com/professional-services/ What's up with rsyslog? Follow https://twitter.com/rgerhards NOTE WELL: This is a PUBLIC mailing list, posts are ARCHIVED by a myriad of sites beyond our control. PLEASE UNSUBSCRIBE and DO NOT POST if you DON'T LIKE THAT.
