Terminating the append_revisions_only experiment, for now

[Martin Pool]

On 23 August 2011 17:41, vila <v.ladeuil+lp at free.fr> wrote:
>>>>>> Max Bowsher <_ at maxb.eu> writes:
>
>    > We currently have 97 packages failing with AppendRevisionsOnlyViolation
>    > - that's around a sixth of all failures.
>
>    > I think we should roll back the change to the importer to set
>    > append_revisions_only on UDD branches, and make the importer clear the
>    > flag when doing its own pushes to Launchpad.
>
> Very close to my own thoughts, thanks for bringing this up.

+1 from me too.

>
>    > It feels clear to me that what seemed like an obvious win at the
>    > time of introduction has revealed lots of interesting and
>    > unexpected edge cases.
>
> This is also very close to my feelings and I like to add that we should
> *really* write tests to capture them and make sure we don't regress
> again in the future.
>
> Overall, I feel we have far too much failures when landing *good* stuff
> *because* we lack a proper way to test. It's a long known issue that we
> miss a launchpad test server but *not* having tests to guard against
> regressions is clearly a path we don't want to pursue as demonstrated
> here (IMNSHO).
>
> Whether we address that by using staging or any sort of local launchpad
> test server is open to discussion but the sooner we start this
> discussion the better.

I think the easiest way to write and run the tests is probably to make
them run against the staging Launchpad, and then perhaps to have them
skip if Launchpad is unreachable.  At any rate I was going to try
doing that for the file_mp related bugs.  We'll see how it works out.

Martin