Ask HN: Why does FreeBSD do Releases the way the do?

26 points

4 years ago

Can someone help me understand how/why FreeBSD does releases. It almost seem user hostile (unless you're just a FreeBSD developer wanting to hack on FreeBSD).

What I find odd is:

(a) there are 3 simultaneous production releases (13.0, 12.3, 12.2) - which seems confusing and lack of focus

(b) the development process is CURRENT > STABLE > BETA* > RC* > RELEASE. But the problem is, all developers just work on CURRENT (not RELEASE). So if there is ever a problem on RELEASE, you basically won't get the fix until the next MAJOR release, not the next point release. This also create a weird side effect where major releases only have 2-3 point releases and never matures.

(c) it seems appropriate to only run RELEASE in production, but the problem is - RELEASE is only supported for 3 months after the next RELEASE. 3 months! [1]

(d) this creates a situation where companies basically have to fork FreeBSD just to use it in their production environment (and merge from CURRENT frequently, which by it's very nature is unstable). Netflix has to do this, but not all companies have the huge resources Netflix does [2]

(e) just the simple fact that freebsd.org list 4 releases on their homepage is alone confusing.

These problems has been long documented for over a decade (documented well by @rsync [3][4]), I just feel like I must be missing something obvious as to why FreeBSD does there releases so strangely.