Loïc Minier <lool+debian@via.ecp.fr> writes:
> On Mon, Apr 09, 2007, Marc 'HE' Brockschmidt wrote:
>> Loïc Minier <lool+debian@via.ecp.fr> writes:
>> > I don't want changelog entries to be lost as they represent actual
>> > uploads; hence, in the past, I've merged changelog entries between
>> > experimental and unstable so that all experimental and unstable uploads
>> > appear in the changelog of the latest version; this is sometimes weird
>> > as it can shows two versions doing the same set of changes when we did
>> > them separately.
>> Which I think is a problem. I believe that changelog entries should only
>> be merged when the related chages were merged. There is seldom more
>> development on a stable (ie, uploaded to unstable) branch of a package
>> than on the experimental version, so I think a useful way to handle
>> this problem is to merge all changes made to the unstable version into
>> the experimental version (and include the fitting changelog entries). If
>> a package is moved from experimental to unstable, it should be done with
>> svn mv [1].
>
> So you want to keep experimental uploads in the unstable changelog?
Yes.
>> I firmly believe that the changelog should only reflect development done
>> on the branch that is actually uploaded, and not contain what was done
>> to other branches. Remember, we do non-linear development, but
>> changelogs only provide a linear way to represent changes, so we should
>> document the actual line of development in the changelog of the package
>> that we are uploading.
> Sure, how does the proposal I made contradict the above?
Merging unstable changelog entries that were relevant to a version that
is not the base of the current version contradicts that.
Marc
--
Fachbegriffe der Informatik - Einfach erklärt
152: PORN
Poster Ohne Richtigen Namen
Attachment:
pgpZXqGkSE90D.pgp
Description: PGP signature