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

Re: DEP1: Clarifying policies and workflows for Non Maintainer Uploads (NMUs)

On Fri May 30 08:48, Sune Vuorela wrote:
> > This means that people can opt out using DELAYED, but must post something
> > in the BTS. I think that the problem is not whether the communication is
> > public in the BTS or private, it is that "something the BTS" does not
> > imply communication. One can send a patch to the BTS and upload a NMU
> > without ever asking if it is welcome. Therefore I would much prefer that
> > the DEP clarifies this by writing that the use of DELAYED is mandatory
> > if the NMUer does not ask if the upload is welcome.
> Yeah. let us delay bugfixes from reaching the users as long as
> possible.

Debian has a system where packages have a primarily responsible
maintainer, not one which is a free-for-all. You may disagree whether
this is the best solution, but that is a separate discussion.

Given that we have a primary maintainer there must be a balance between
getting fixes/new versions out as soon as possible and respecting the
autonomy of a maintainer. 

Requiring either authorization or notification and a delay is, IMO, the
least that we should do to keep this balance. Authorization may be in
the form of an explicit mail or presence on the LowThresholdNMU list and
notification/delay may be a post to the BTS and upload to DELAYED, which
makes it very simple for an NMUer to do (they should be posting to the
BTS _anyway_ and DELAYED doesn't require a separate upload action) and
in many cases means that they can just upload directly. 

If you think that significant fixes are being delayed by the maintainer
then by all means complain about specific cases, but this does not mean
we should be making NMUs without maintainers having a chance to respond.

In the vast majority of cases these uploads are being made to unstable
and will only be affecting users who have accepted some amount of
breakage and disruption by using pre-release versions. Another couple of
days is not going to cause any harm.


Matthew Johnson

Attachment: signature.asc
Description: Digital signature

Reply to: