Greg Hudson wrote:
>On Fri, 2004-11-05 at 10:54, kfogel@collab.net wrote:
>
>
>>I suspect Branko is probably right, that we will one day need an
>>opaque type instead of a 'char *'. That doesn't seem like a very
>>risky projection.
>>
>>
>
>Actually, I'm coming to the conclusion that it makes no sense. Are we
>really going to have a bunch of ACLs in the FS referring to a group
>database outside the FS? If two different callers use different group
>databases, they'll get different permission behavior?
>
>
We have exactly the same problem with user authentication today. The
three RA methods each have a different way for authenticating the user.
It's always been the repository admin's responsibility to keep them in sync.
>Maybe there's something I don't understand about the DAV ACL spec, but
>that seems bizarre.
>
>But this is precisely the kind of question we *shouldn't* be deciding
>now, because we're solving locks, not ACLs, and we shouldn't make a
>simple problem hard by tying it into a hard problem.
>
>
We're not deciding this question. We're simply putting in a mechanism
that can present the information, regardless of how we got it or where
it came from.
-- Brane
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Fri Nov 5 22:28:37 2004