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

Re: Subversion FSFS corruption

From: Daniel Berlin <dberlin_at_dberlin.org>
Date: 2005-08-24 02:24:55 CEST

On Tue, 2005-08-23 at 20:12 -0400, Greg Hudson wrote:
> On Tue, 2005-08-23 at 14:24 -0700, Ted Gould wrote:
> > This is odd to me, as I believe that this is a direct import of the
> > picture. I don't know anything about the Delta format, but it seems
> > that there should be very few instructions at head of each chunk. More
> > of a "just use this data" type command.
>
> One would imagine that would be the case for a jpeg file, but it looks
> to me like there are 310 bytes of instructions there. The other chunks
> (which verify just fine) also contain substantial numbers of
> instructions.
> That suggests an oddity in our delta algorithm and a
> possible space performance issue on compressed files.

What do you mean, "oddity"?
We hold inserts for as long as possible before issuing them, in hopes of
extending a later match back into them. We can't do much better than we
do.
If the matches are ending earlier than you think they should, then
either the rolling checksum isn't working quite right, or i screwed up
one of the loop conditions.

--Dan

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Wed Aug 24 02:25:41 2005

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