Re: CVS update: subversion/subversion/libsvn_fs tree.c

On Tue, Mar 27, 2001 at 11:40:52AM -0600, Karl Fogel wrote:
> Greg Stein <gstein@lyra.org> writes:
> > > + /* ### kff todo: is there any compelling reason to get the nodes in
> > > + one db transaction? Right now we don't; txn_body_get_root() gets
> > > + one node at a time. */
> >
> > I don't think so. They're all readonly, and the algorithm allows for them to
> > move/shift as they will. There isn't an invariant that they are all
> > synchronized in some fashion, so there isn't a need to ensure they are read
> > in one, synchronous shot.
>
> There's no guarantee that they're all readonly. They could all be txn
> roots, for all we know.

Sorry... I meant "you only read the things". If you read them at time T or
time T+1, it isn't really going to matter to the algorithm.

Descending... that's a different story, as Jim pointed out.

Hrm. Oh... actually, you could end up reading at time T, and then everything
could change in T+1 and your data is obsolete.

Never mind me... :-)

Cheers,
-g

-- 
Greg Stein, http://www.lyra.org/