Re: General Resolution: Force AMD64 into Sarge
On 2004-07-22 Noah Meyerhans <firstname.lastname@example.org> wrote:
> I've seen at least one message in this thread argue that a compromise
> between the two positions might be to get amd64 in to sid ASAP with the
> objective of waiting until some point release of sarge before including
> it there.
Iirc Goswin has already asked the stable release manager (Martin
Schulze) about this and has received a negative response.
> Obviously it's not our normal practice to add architectures with
> point releases, but I think it's pretty clear that real support for
> this architecture is too important to wait until sarge+1 is
> released, potentially two years from now.
This is less about if we consider it "normal practice" and are willing
to make exception but about what our infrastructure can handle. - Our
dot-releases usually feature about 50  changed source packages each
containing only minimal differences to the one they replace.
It is very important that there are really only minimal changes, first
because it is "stable", i.e. the behavior does not change, and second
because packages targeted for a dot-release are literally untested.
(No sane person simply plugs proposed-updates in sources.list as it
beats the whole point of running stable).
This basically means that due to amount of changes that are possible
for dot releases amd64 would need to have been ready for sarge_r0 to
be a valid candidate sarge_r1. If that was the case, why wouldn't we
we have released with sarge_r0?
 I did not do any research on this I just tried to come up with a
number that sounded right. ;-)
"See, I told you they'd listen to Reason," [SPOILER] Svfurlr fnlf,
fuhggvat qbja gur juveyvat tha.
Neal Stephenson in "Snow Crash"