[MERGE] convert locking problems in selftest at win32 into known errors

Vincent Ladeuil v.ladeuil+lp at free.fr
Fri Dec 21 08:26:13 GMT 2007


>>>>> "bialix" == Alexander Belchenko <bialix at ukr.net> writes:

    bialix> John Arbash Meinel пишет:
    >> John Arbash Meinel has voted comment.
    >> Status is now: Semi-approved
    >> Comment:
    >> Alexander, did you have a response to Robert's comments?

    bialix> I sent response in November 12, 2007. Do you want
    bialix> another response?  Here is: I don't understand enough
    bialix> some tests and some code involved in tests and don't
    bialix> understand how and when lock acquired. If bzrlib
    bialix> could provide some logging for each lock (from which
    bialix> function it acquired) may be I can track down these
    bialix> problems. But for now I don't care much why it
    bialix> failed, if it not failed on linux, so I prefer mark
    bialix> it as XFAIL. Because having about 50 permanently
    bialix> failed tests reduce value of selftest on win32.

I think this is the right use for KnownFailure, we know it fails
and we want to be reminded that we ought to fix it, but "the show
must go on" !

It's a step in the right direction and it's not worse that what
we have today.

And if it makes the test suite more usable it's a net win.

    Vincent



More information about the bazaar mailing list