file:// + exclusive locking/external locking - what's right?

Is the issue the external lock file?

Or the exclusive lock itself?

Personally, when I first saw this, I did not understand the external lock file.

But given that it records the user and the time of the lock that was taken
out, it makes some sense to me now.

Maybe if we separated it out in terms of those two concerns and then decided
what would be useful to keep and not keep when faced with concurrency issues
across a networked filesystem w/o a server?

What's the right answer here?

Ignoring networked filesystems because they have to use a server implies that
we shouldn't even have the file:// protocol, since it should all be
client-server.

Shouldn't it?

-- 
Shamim Islam
BA BS
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org