Re: [PATCH 00/11] More preparatory work for multiparent tree-walker

On Tue, Feb 11, 2014 at 11:59:02AM -0800, Junio C Hamano wrote:

Kirill Smelkov [off-list ref] writes:

quoted

Sorry for the confusion. Could you please do the following:

Patches should be applied over to ks/tree-diff-walk
(74aa4a18). Before applying the patches, please cherry-pick

    c90483d9    (tree-diff: no need to manually verify that there is no
                 mode change for a path)

    ef4f0928    (tree-diff: no need to pass match to
                 skip_uninteresting())

into that branch, and drop them from ks/combine-diff - we'll have one
branch reworking the diff tree-walker, and the other taking advantage of
it for combine-diff.

As long as that does not lose changes to tests and clean-ups, I'm
fine with that direction.  For example, I do not know if you want to
lose e3f62d12 (diffcore-order: export generic ordering interface,
2014-01-20), which is part of the combine-diff topic.

Sorry for the confusion again, and please don't worry: we are not going
to lose anything - my only plea here was to transfer two of the patches
to more appropriate topic.

That couple touches tree-diff.c - they were some initial cleanups I've
noticed while working on separate combine-diff tree-walker, which we
decided to drop instead of generalizing diff tree-walker to handle all
cases. Only the cleanups are still relevant and needed as a base for
what I've sent you here.

And as to e3f62d12 (diffcore-order: export generic ordering interface,
2014-01-20) and other patches on ks/diff-c-with-diff-order topic - they
stay as they are - we do not need to rework them as ks/combine-diff
builds on top of the topic and generalizing diff tree-walker is
orthogonal work.


So in short: could you please transform the two "tree-diff" patches from
ks/combine-diff to ks/tree-diff-walk, then apply sent-here patches to
ks/tree-diff-walk; thats all.


Thanks,
Kirill