We're going OT, however my experience based on last two Debian releases:
testing becomes quite "stable in means of usability" somewhere half year
before it's released as "stable". The sooner before the stable, the
rapidly increasing is the chance that the snapshot that You have will
not be installable at all, will have dependencies severely broken, etc.
That's why I suggest: focus on base platform, stabilise it, polish the
dependencies. Then compile software against it and release it, compile
newer version and release it, etc.. Desktop software itself shouldn't
break dependencies.
Peter
Frans Pop wrote / napísal(a):
On Tuesday 15 May 2007 14:44, Mgr. Peter Tuharsky wrote:
Yes, bugs are unavoidable. However, testing is often in situation
"whole system broken" or "nearly useless". I see difference here;
occassional bug in desktop app is acceptable. Whole system unreliable
is not acceptable.
Can you substantiate that? In my experience it is not true.
And even unstable is almost always usable if you know how to avoid
temporary uninstability of packages and how to downgrade a package
occasionally.
Though I'd not advice running unstable to end users, I would happily
suggest testing.
Cheers,
FJP
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
