question about 2.6.3pre2's --link-by-hash behaviour

Wayne Davison wayned at
Mon Sep 27 18:57:11 GMT 2004

On Fri, Sep 24, 2004 at 03:47:03PM +0200, Paul Slootman wrote:
> If I then run it again, I get the following [a different hashed file]

I didn't see that in my just-run test.  I did notice a problem with the
code not removing an existing destination file prior to trying to hard
link a hashed file into place (when the hashed file already exists).  I
checked in a fix for this, along with the missing close() you found.
> It forces the checksum-seed to 12345, although I can't see why.

Good point.  I guess the thought at the time was that it was using the
actual rsync checksum, but since it is not, it doesn't need to force the
checksum_seed to a static value.  I've removed that code.

I haven't ever tried to make use of this patch, so feel free to let me
know if there are other problems with the code.


More information about the rsync mailing list