Jeff King [off-list ref] writes:
That seems worse to me, because "it is impossible" implies that this
can never be changed. But I don't think that's true. We might one day
output something useful for merges.
Fair enough. You are more optimistic than I am ;-)
I think one could argue that any merge information (including conflict
resolution) works against the root notion of format-patch, which is a
set of changes that can be applied on a range of basesa.
That's true and it was the primary motive for omiting merges.
But even that I
would be hesitant to commit to (since --base exists now).
I am not quite sure what --base has to throw into the equation. The
information --base gives is often useful when I want to learn where
the patches were taken from, but that does not restrict where the
patches are actually applied to in any meaningful way (iow, "on a
range of bases" part is not affected).
On Mon, Feb 22, 2021 at 05:24:16PM -0800, Junio C Hamano wrote:
quoted
I think one could argue that any merge information (including conflict
resolution) works against the root notion of format-patch, which is a
set of changes that can be applied on a range of basesa.
That's true and it was the primary motive for omiting merges.
quoted
But even that I
would be hesitant to commit to (since --base exists now).
I am not quite sure what --base has to throw into the equation. The
information --base gives is often useful when I want to learn where
the patches were taken from, but that does not restrict where the
patches are actually applied to in any meaningful way (iow, "on a
range of bases" part is not affected).
What I meant is that without "--base", telling somebody "here is the
merge you should replay on top of these other patches" is virtually
meaningless. You cannot know what the merge base would be! So you might
be merging in other random crap, and you might or might not see the same
conflicts.
But in a world with --base, I can imagine some people recreating whole
sequences of the history graph by using "--base" along with some (to be
invented) format for representing a merge via email. That mode would
certainly not be the default, but at least at that point it is
conceivably useful. Sort of like a bundle, but more human-readable (it
would also need committer info to recreate the commit ids perfectly, of
course).
All of which meant only to argue that "it is not possible or not useful
to represent a merge in an email" is something that could change in the
future. :)
-Peff