Re: [PATCH] Proposed ra_svn protocol change

Greg Hudson <ghudson@MIT.EDU> writes:
> With this patch, the server always makes a trivial authentication
> request except at the beginning of the connection. When I add the
> CRAM-MD5 code, I will try to throw in some kind of simple authorization
> policy options (such as "anonymous reads and authorized writes"), and
> then commit and change-rev-prop will do more interesting authentication
> challenges.
>
> I'm just about ready to commit this patch (it passes make svncheck, but
> I need to do cross-version tests in both directions), but I thought I
> would send it here first in case anyone has opinions.

Just positive ones, which I realize are useless to you :-).

The Version 2 plan sounds great, and the Version 1 workaround sounds
as good as can be expected given the constraints.

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Wed Oct 22 17:08:17 2003