[Mailman-Developers] race condition in locking ?

[Barry A. Warsaw]

>>>>> "TW" == Thomas Wouters <thomas@xs4all.net> writes:

    TW> I'm busy rewriting LockFile.py to use the new locking
    TW> (actually, I've rewritten it, I've just not tested it yet ;)
    TW> and I should be able to post it tomorrow... Unless anyone has
    TW> objects to this locking method. Did I miss anything ? Is there
    TW> anyone I should talk to before changing things, like the
    TW> original authors ?

Thomas,

I've only skimmed your emails but I think your analysis is basically
correct, and the solution you propose here seems reasonable.  This
mailing list is the right place to discuss it, and although I've been
too busy to respond lately, I /will/ see your message, and your
changes.

My suggestions would be these: be sure to look at the latest version
of LockFile.py from the CVS repository.  It doesn't fix your problems
but it embodies the current semantics I want from the API.

Second: if you submit substantial changes (as I suspect you will),
we'll need to get FSF assignments from you.  I'll send you the
appropriate form when I see how big your changes are, but I wanted to
give you a heads up.

Thanks very much for looking into this.

-Barry