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

Re: optimizing fsfs: reverse diffs?

From: Daniel Berlin <dberlin_at_dberlin.org>
Date: 2005-04-08 20:21:21 CEST

On Fri, 2005-04-08 at 13:13 -0400, Alan Grow wrote:
> On Fri, Apr 08, 2005 at 12:56:56PM -0400, Daniel Berlin wrote:
> > We don't want previous revisions to be modified, we want them completely
> > read-only and immutable.
>
> The intent of this makes a lot of sense. While I don't see that
> previous revisions would need to be "redeltified" under a
> reverse-diff-back-from-HEAD scheme, it's true that their delta chain
> changes as HEAD changes. Apparently, this must really matter under the
> svn client-server mo

If you don't reverse diff the previous against the new-head, you'll end
up with all full-texts, because by definition, before you do this, the
old-head was a full text:

IE in a reverse diff scheme, things look like this:

diff (1.9 to 1.8) 1.8
diff (1.10 to 1.9) 1.9
diff (1.11 to 1.10) 1.10
ft 1.11

If you add 1.12
you need to make

diff (1.9 to 1.8) 1.8
diff (1.10 to 1.9) 1.9
diff (1.11 to 1.10) 1.10
diff (1.12 to 1.11) 1.11 <----- this is the redeltification
ft 1.12

Otherwise, you'll just have

diff 1.8
diff 1.9
diff 1.10
ft 1.11
ft 1.12
ad infinitum

> > When somebody actually comes up with a case where the majority of the
> > checkout time is being spend trying to run down skip-deltas for the
> > HEAD, i'd be concerned.
> > This never really is the case, though, so it's all just talk. The
> > slowdown is not here, move along :)
> >
>
> You called me on it--I've done no profiling myself & have no numbers,
> only vague accusations. :) Just wanted to see if something like this
> has been considered. I can tell it has--thanks to you, Daniel, & Mark
> both.
>
:)
If you do, however, discover that it is a real percentage of the time
(IE more than 10% or 15%), *please*, let us know.

---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@subversion.tigris.org
For additional commands, e-mail: users-help@subversion.tigris.org
Received on Fri Apr 8 20:23:28 2005

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

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