Merging with descendants

Stephen Lee sphen.lee at gmail.com
Mon Jul 22 00:55:48 UTC 2013


On Tue, Jul 16, 2013 at 4:50 PM, Matt Mackall <mpm at selenic.com> wrote:
> On Fri, 2013-07-12 at 18:37 -0700, eode wrote:
>> > hg merge <rev>
>>
>> Are there plans to implement Oedipus merges, whether via the 'hg merge
>> <rev>' method or otherwise?  Would it be useful to open up a feature request
>> on bugzilla, or would that just be more noise?
>
> Nope.

I allowed Oedipus merges in the custom Mercurial we installed at my
work place as described way back in this thread - I'm now regretting
this.

A problem that keeps arising is a developer cloning and creating a
bookmark to start working on a bug. Before committing anything they
get distracted by another bug, and after some time return to the
original bug. Since the clone is now out-of-date they run pull and
merge - adding a totally useless and confusing Oedipus merge to the
history.  They should have run update instead.

We have a requirement that every integration involves a merge so we
can audit when branches were merged.
Looking back, this requirement would have been better met by named
branches: have two branches 'default' and 'stable'.
Developers use bookmarks on the default branch, integration merges are
on stable.
Do "hg up stable ; hg merge bugXYZ" to integrate a branch - the
Oedipus merge is allowed in this case since the changesets are on
different named branches.

So agreeing with Matt here - Oedipus merges should not be in core.



More information about the Mercurial mailing list