Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 30 Nov 2005 13:15:00 -0600
From:      Greg Barniskis <nalists@scls.lib.wi.us>
To:        Sean Murphy <smurphy@calarts.edu>
Cc:        freebsd-questions@freebsd.org
Subject:   Re: Am I Right about Stable VS Point
Message-ID:  <438DFA34.2070308@scls.lib.wi.us>
In-Reply-To: <438DE8E8.1050906@calarts.edu>
References:  <438DE8E8.1050906@calarts.edu>

next in thread | previous in thread | raw e-mail | index | archive | help
Sean Murphy wrote:
> I have read the FreeBSD Handbook Chapter 20
> 
> http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/cutting-edge.html
> 
>     I understand what FreeBSD-Current is.
> 
>     FreeBSD-Stable is a little fogy for me.  Here is what I found and I 
> think.
> 
> quote "FreeBSD-STABLE is our development branch from which major 
> releases are made.  Changes go into this branch at a different pace, and 
> with the general assumption that they have first gone into 
> FreeBSD-CURRENT for testing.  This is still a development branch, 
> however, and this means that at any given time, the sources for 
> FreeBSD-STABLE may or may not be suitable for any particular purpose. It 
> is simply another engineering development track, not a resource for 
> end-users."
> 
>     So "Stable" is not really "Stable" it is still a branch for 
> development and security fixes that go into "Point Releases".  Which 
> means "Point Releases" are the real true "Stable" area. Right?

More or less, yes. -STABLE really is pretty stable in my experience, 
since the development code that gets checked in is ostensibly good, 
running code that has already been tested. But... sometimes it isn't 
good code, or sometimes one change conflicts with other recent 
changes that got checked in.

Also, features in -CURRENT may diverge from the last release point 
by a rather wide margin, so it isn't really the best testing 
environment for evaluating how a change will affect users when it is 
grafted onto the last release point -- that is more the job for 
-STABLE. Finally, some bugs just don't manifest until a wider range 
of users have tried out the new code.

Release points represent a junction where -STABLE really does prove 
to be very stable for a wide range of uses and platforms. A release 
might still have some bugs that didn't manifest yet, but that is 
much less likely than if you run -STABLE.


> So when I need a security update I should CVSup the tag line should be
> 
> RELENG_6_0 for the real stable version, also includes bug and security 
> fixes.

This will include the release point code plus /critical/ bug fixes. 
As I understand it, this is not all available bug fixes, just the 
fixes for clear operational threats such as security-related bugs or 
things that might lead to data loss. Minor bugs might not get fixed 
until the next major release point.

Yes, "RELENG_X_Y" is the recommended CVS setting for production 
servers and any non-expert use, and RELENG_6_0 is ostensibly the 
most stable and secure branch to be following today.


-- 
Greg Barniskis, Computer Systems Integrator
South Central Library System (SCLS)
Library Interchange Network (LINK)
<gregb at scls.lib.wi.us>, (608) 266-6348



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?438DFA34.2070308>