Subject: Re: Current versus 1.4.2?
To: Andrew Brown <firstname.lastname@example.org>
From: Greywolf <email@example.com>
Date: 05/23/2000 23:06:23
On Tue, 23 May 2000, Andrew Brown wrote:
# Date: Tue, 23 May 2000 16:03:50 -0400
# From: Andrew Brown <firstname.lastname@example.org>
# To: Jon Lindgren <email@example.com>
# Cc: firstname.lastname@example.org
# Subject: Re: Current versus 1.4.2?
# >> >http://www.netbsd.org/Releases/release-map.html
# >> that still doesn't answer the question of which is a subset of which.
# >> i always assumed that the release branch was a proper subset of the
# >> current branch (assuming of course that you don't want to try to mail
# >> jelly to a tree and call it a trunk).
# >Take a look at the diagram. 1.4 is a separate branch from
# >-current. Essentially, when 1.4 is released, two separate branches are
# >created. Bug fixes, perhaps some drivers, etc... are pulled from -current
# >into the 1.4 branch from time to time, but they are separate
# >branches. When sufficient bug fixes or little features are added to 1.4,
# >it becomes 1.4.1 (then 1.4.2). At that point, -current may be completely
# >different due to structure or archtecture changes. So it would be proper
# >to say that the 1.4 branch contains all of -current _at the time the
# >source was branched_. Past that, -current and 1.4 diverge; -current
# >remains the ever growing trunk of the tree, while 1.4 grows for a while,
# >then ends.
# certainly, in terms of the engineering of the tree itself (names
# change, things move, things get better, certain vestigial features are
# trimmed), but i'm speaking in slightly more gross terms.
# i always believed that, as far as support for devices goes, the
# current "trunk" (from which all releases are grown) contains support
# for all the stuff that any release ever contained, whereas current has
# support in it for things that no release (yet) has supported. that,
# to me at least, makes current a proper superset of the release, or
# conversely, the release a proper subset of the "trunk". no?
I think it would be a shame for development effort to have been wasted
to provide something in -current which will never see the light of
day in a release (well, something that works/is good/is worthwhile).
I realise that -current is in a state of constant flux, but how are
-current and the release path ever reconciled? Isn't there usually
a snap of -current just before version M.m is released which is then
the frozen public-current tree while some development happens in the
private-current tree, and about a week after M.m is released,
public-current gets sync'd with private-current, and what was public-
current remains as the M.m branch, and the only stuff to further come
from (release on?) an M.m branch are bugfixes in the form of M.m.r
# >Ne pas?
# je ne sais pas. :(
Ni moi aussi.
A _Real_ Operating System for _Real_ Hackers.
______ _ __ __ ____ _____ ____
______ / |_/ /__ / /_/ __ ) ___// __ \
______ / |/ / _ \/ __/ __ \__ \/ / / /
______ / /| / __/ /_/ /_/ /__/ / /_/ /
______ /_/ |_/\___/\__/_____/____/_____/
With many thanks to the core team and UCB CSRG.