[PATCH 2 of 3] issue1578: reproduce crash where synthetic revs break merge detection

[Greg Ward]

On Fri, Apr 3, 2009 at 1:53 PM, Matt Mackall <mpm at selenic.com> wrote:
> On Sat, 2009-03-28 at 12:28 -0400, Greg Ward wrote:
>> # HG changeset patch
>> # User Greg Ward <greg-hg at gerg.ca>
>> # Date 1238257630 14400
>> # Node ID 6fb6f7f83d0e2c24fb8dea9fec04fe801f1dcf51
>> # Parent  34d59e749d891be3a7133bec7a098584c013ef14
>> issue1578: reproduce crash where synthetic revs break merge detection
>> (N.B. this breaks the tests: fix coming next).
>
> Looks like this test no longer applies cleanly. I applied the fix itself
> though.

Oops.  That's probably  because the test changes for 1578 depend on
the test changes in my patch for 1577, which is why I sent the two
fixes as a 3-part patch series.  I should have been explicit about the
dependency: sorry.

Anyways, I'll repost the other two patches (fix for 1577 and tests for
1578) soon.  IMHO 1577 should be fixed on hg-stable, since the tests
are broken now.

Greg