Ben Collins-Sussman wrote:
>On Wed, 2004-05-19 at 16:04, Branko Čibej wrote:
>
>>ust to clarify things here, the ACL design I'm working on doesn't
>>seriously affect the locking design. I'm only working on the premise
>>that the changes needed for adding ACLs are very similar to the changes
>>needed for adding locks, so I'm looking at implementing locks in such a
>>way that ACLs can use the same infrastructure, that's all. There's no
>>radical departure from locking-plan.txt.
>>
>>
>
>Really? I had assumed that you were planning to implement locks *using*
>ACLs.
>
They're not quite the same. They're both almost, but not quite, exactly
unlike tea. :-p
>My larger concerns are:
>
> 1. if your ACL design is something that can be added "on top" of the
>locking changes, then should it be discussed in a separate thread, and
>treated as a follow-up project to locking?
>
>
Yes. And it doesn't have to be done now, because at this point I believe
that ACLs are too big for 1.1. I'm hoping for 1.1, of course, but I
won't push it because locking is more important and we do have a few
quite satisfactory workarounds for authz.
> 2. is it possible to add your ACLs in 1.1 and not 2.0?
>
>
1.x rather than 1.1, but yes, I believe that if we can add locking
before 2.0, the same should be true for ACLs. At least I've not found a
reason to the contrary yet.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Thu May 20 00:05:36 2004