Paul Dwerryhouse:
> On Thu, Feb 07, 2008 at 01:48:14PM +0800, jidanni@jidanni.org wrote:
>> Gentlemen, why not let the others get cut on the cutting edge of
>> Debian sid whilst we relax and wait oh, say 72 hours for the packages
>> we want to stabilize? Slightly stable unstable, but not too stable as
>> to be stale as stable.
>
> Isn't that what testing is for?
Yes. :) It's just that the rules for a package to move from unstable to
testing are quite tough. Which they should be, because every transition
to testing might be the last one for a package before the release.
The proposal is missing the most important part: what should prevent a
package from moving into the "slightly stable unstable"? This is the
really hard part and the current answer is the definition of testing.
> [Not that I often find unstable to actually be unstable].
It is, you are just not using the Debian definition of stability. :)
"Stable" in the Debian sense means that the archive doesn't change in
terms of version numbers, features and even (uncritical) bugs.
Therefore, sid is unstable. That just doesn't mean it crashes every day.
J.
--
I wish I could achieve a 'just stepped out of the salon' look more
often. Or at least once.
[Agree] [Disagree]
<http://www.slowlydownward.com/NODATA/data_enter2.html>
Attachment:
signature.asc
Description: Digital signature