On Sat, Dec 03, 2005 at 11:00:25AM -0800, Wilfredo S?nchez Vega wrote:
>
> Well, I'm a bit wary of trying to overload the simple provider too
> much, I used a first-class provider to avoid polluting the simple
> provider. I have no objection to leveraging the simple provider's
> helper, though.
>
Well, I've already done the polluting bits... :-)
> I do like the single entry in the KeyChain better than two
> entries. I picked that up from Ken's patch and didn't have a better
> idea. Using the config-file storage for the username is better; I
> think the username some come from the subversion client and not the
> KeyChain.
Okay, let's give me approach a try. Expect to see it in the
next few hours. I'm also only a few (non-concurrent) hours away
from GNOME keyring integration. :-)
>
> >2.) Configury.
> >
> >
I think I've got this nailed now.
>
> >3.) Support for --non-interactive.
> >
> > Is surrounding the SecKeychain calls with calls to
> > SecKeychainSetUserInteractionAllowed(FALSE) and then
> > SecKeychainSetUserInteractionAllowed(TRUE) the right way to
> >implement
> > --non-interactive? Could it be that easy?
>
> That sounds right, but again I'm also new to the KeyChain API.
>
Excellent. I'll go with that strategy.
--ben
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Sat Dec 3 21:57:26 2005