On Wed, Mar 28, 2001 at 03:06:22PM -0600, Ben Collins-Sussman wrote:
> Greg Stein <firstname.lastname@example.org> writes:
> > On Wed, Mar 28, 2001 at 12:38:22PM -0600, Ben Collins-Sussman wrote:
> > >...
> > > Irony #1: after working hard to make the commit-driver send things in
> > > postfix text-delta style, we're currently *ignoring* this
> > > optimization.
> > ra_dav will be doing checks before the text delta is ever sent, so this
> > "irony" doesn't really apply.
> > The most important check is that when a modification is marked for a
> > particular resource, ra_dav hits the server and says "I want to modify
> > <this>" and the server will respond at that point to say "that is out of
> > date. bugger off."
> Oh, so are you not building an fs transaction that looks like the
> working copy? Or is this extra checking just happening as you build
Sure, I create a transaction and make all the changes in there. But before
each change is transmitted, a CHECKOUT occurs, where I do the extra check.
The pattern is something like:
(where the PUT comes at the end because of the postfix stuff)
Each of those CHECKOUT methods will perform a check-for-latest.
Greg Stein, http://www.lyra.org/
Received on Sat Oct 21 14:36:26 2006