On Mon, Jan 27, 2020 at 02:17:13PM -0800, Quanah Gibson-Mount wrote:
No, not at all. I would say OpenLDAP has too few releases in a year (only 1-2 currently for most years, unfortunately), so having more frequent releases for it is probably a good thing. But a piece of software in general that is releasing constantly? Not a fan of it at all, and haven't seen it as a good thing as far as softare quality is concerned. There's plenty of software that releases much less frequently than OpenLDAP as well, because there isn't a driving need for it to have a new release.
There's always as much to do as many people you have on hand. With a large team, release schedules that seem to work best nowadays look like Linux/Python/PostgreSQL where you have a time based feature release schedule so whatever isn't ready yet just waits for the next merge window and bugfix releases for the currently supported version come as and when needed.
Let's see how quickly we can get from a first 2.5 release to more general adoption and see if we can maybe ride the wave a bit, renaming 2.5 to a 2.6 when people are more happy with it and starting 2.7 at the same time, which I think is what you meant by this proposal?
And as Howard noted, there's a balance to strike between stability and feature development. If we release every 2 weeks, but slapd core dumps 90% of the time, is that really better? Sure, the project looks more "active", but I wouldn't see that as a benefit/gain.