[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Questions about testing

On Tue, 02 Jan 2001 05:38:36 Branden Robinson wrote:
> Consider that when I manage to hork up X, I know about it within
> hours of
> dinstall.  Likewise, a few days ago when Wichert busted the vim
> postinst,
> he was told about it quite fast indeed.
> I don't have any concrete recommendations for how to take this into
> account, but I certainly think that a 14-day waiting period for
> packages
> like these is excessive.

I agree with you. Lots of popular packages shouldn't need 2 weeks for
obvious bugs to be noticed, but I think that most packages should wait
2 weeks to get reasonable quality in testing. That means that not all
packages should have the same waiting time. Is there some way that
maintainers could take responsibility for deciding how long the
waiting period should be? I don't know very much about the details of
how the archive is managed. I'd suggest a default of 14 days and
probably some minimum time period as well.

On a slightly related topic, packages that are updated everyday are a
big headache for those of us that are living at the end of a modem,
because we have to update many 10's of packages a day == lots of
downloading. Its not a problem these days, but it was awkward just
after kde and xfree4 were added to woody. (I'm not trying to dictate
how people do their job, just pointing out my experiences.) I'd like
to see some kind of policy that says that in general packages
shouldn't be updated more than twice a week. I know I'm probably in a
minority here though and I'll be told to apt-get update twice a week
;) but I'd like to hear if there are good reasons why this would be
bad policy.


Reply to: