[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: Delta directions (was Re: delta combiner sloooooooowwwwww)

From: Daniel Berlin <dberlin_at_dberlin.org>
Date: 2005-02-11 02:30:53 CET

On Thu, 2005-02-10 at 13:22 -0500, Greg Hudson wrote:
> I hadn't read part of the "make blame even faster" thread when I sent
> this. Having read it, I have one addendum:
> On Thu, 2005-02-10 at 13:08, Greg Hudson wrote:
> > (Obviously not before 2.0, but I'm not sure if we can
> > realize any of these benefits before 2.0.)
> As people pointed out in that other thread, we can realize the "faster
> delta combiner" benefit before 2.0, and it may be worth changing to
> xdelta in 1.2 in order to do so. My biggest concern is that FSFS
> repositories with very shallow file histories may become significantly
> larger, since we will lose all self-compression. Perhaps we can
> continue using vdelta for self-compression in FSFS?

You'd probably also want to do it for sending over the wire for
checkout, so it doesn't end up sending fulltexts over the wire.

> I'm not sure how much of the speed win (if any) we'd be giving up by doing that.

> At any rate, the more desirable end state is to ditch svndiff and switch
> to a format which doesn't even allow random access, and to have a
> compression layer on the output of the delta routine. That would let us
> recapture most or all of the space efficiency and would let us use
> deltas in other ways. But getting to that end state is much more
> difficult than simply changing our delta algorithm from vdelta to
> xdelta.

I was, of course, looking for some immediate gratification so i don't
have to wait a year or more to possibly move gcc over to subversion
(assuming people will move :P)
IE make it fast enough for now, make it even better later.

To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Fri Feb 11 02:32:08 2005

This is an archived mail posted to the Subversion Dev mailing list.

This site is subject to the Apache Privacy Policy and the Apache Public Forum Archive Policy.