Web lists-archives.com

Re: [PATCH v5 00/16] Many promisor remotes




On Mon, Apr 15, 2019 at 11:27 AM Junio C Hamano <gitster@xxxxxxxxx> wrote:
>
> Christian Couder <christian.couder@xxxxxxxxx> writes:
>
> > This patch series is based on:
> >
> > 763fb763b8 (Merge branch 'jt/batch-fetch-blobs-in-diff' into jch, 2019-04-08)
> >
> > to avoid issues with jt/batch-fetch-blobs-in-diff.
>
> Yuck.  As an experienced contributor, you should know better than
> that by now to do that.  A merge into jch/pu are rebuilt at least
> once and often three times a day, and in no way a good solid base
> to build on top.

Sorry if it creates problems.

> If you really need to depend on another topic or two, please base
> your work on a merge between 'master' (or some well known ancestor
> of it) and the tips of the topics instead.

Ok I will do that then.

> Having said that, I thought that the semantic conflict has been
> corrected and the machinery to rebuild 'pu' has been replaying the
> correct resolution ever since, so there was no need for such a
> rebase?  Isn't it the case and do we still have the breakage due to
> semantic conflict with JTan's topic in 'pu'?

There is one patch in the series, Patch 8/16 (diff: use
promisor-remote.h instead of fetch-object.h), that fix the breakage,
so if the series is applied on top of jt/batch-fetch-blobs-in-diff, it
will apply correctly and if it is not applied on top of
jt/batch-fetch-blobs-in-diff then the patch can just be dropped and
everything else will apply correctly. I thought that it might be
better to make the fix explicit than to rely on the rebuild machinery.