Re: Bug in merge (repro included)

"Max Bowsher" <maxb@ukf.net> writes:

> Sander Striker wrote:
>>
>> GNU diff3 is not the same as Subversion. The results may differ.
>> I don't feel it is wrong per se.
>
> I agree we don't have to be exactly the same.
>
> But I think failing to apply several lines of changes, and not raising
> any kind of conflict is a definite bug?

If you choose to describe it in those terms then it sounds like a
bug. If I describe it as "determining that the change is already
present in the target ands so can be skipped" is it still a bug?

$ cat foo-B1
A
B
$ cat foo-B2
A
X
$ cat foo-T
A
X

Using GNU diff3:

$ diff3 -m foo-T foo-B1 foo-B2
A
<<<<<<< foo-B1
B
=======
X
>>>>>>> foo-B2

Using Subversion:

$ subversion/tests/libsvn_diff/diff3-test foo-T foo-B1 foo-B2
A
X

Subversion will do this for each hunk that makes up the diff, see one
of my earlier mails for an five-line example that is more like the
original report.

-- 
Philip Martin
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org