List Info

Thread: Re: svn commit: r27647 - in trunk/subversion: libsvn_ra_neon libsvn_ra_serf




Re: svn commit: r27647 - in trunk/subversion: libsvn_ra_neon libsvn_ra_serf
user name
2007-11-07 00:25:13
On 11/6/07, kfogeltigris.org <kfogeltigris.org> wrote:
> Author: kfogel
> Date: Tue Nov  6 21:52:49 2007
> New Revision: 27647
>
> Log:
> Do one capabilities exchange at the start of the DAV
session, instead
> of the client chattily repeating its capabilities with
every request.
> This follows up to r27613.
>
> Suggested by: sussman
>
> * subversion/libsvn_ra_neon/session.c
>   (exchange_capabilities): Renamed from
discover_capabilities.
>     Add capabilities announcement headers to request.
>   (svn_ra_neon__has_capability): Adjust for above, and
check the call
>     for error while we're at it.
>   (svn_ra_neon__open): Exchange capabilities.

Maybe I'm missing something (I only looked at the diff, not
the
context), but if we're exchanging capabilities during the
open call,
why do we need to even consider calling
exchange_capabilities again
during svn_ra_neon__has_capability?

(Ditto for serf.)

(Also, are there possible negative consequences to this
extra OPTIONS
request at the beginning... how does it interact with
authentication,
for example?

--dave


-- 
David Glasser | glasserdavidglasser.net | http://www.davidglasser.
net/

------------------------------------------------------------
---------
To unsubscribe, e-mail: dev-unsubscribesubversion.tigris.org
For additional commands, e-mail: dev-helpsubversion.tigris.org


[1]

about | contact  Other archives ( Real Estate discussion Medical topics )