2.6.4 Potential protocol-corrupting bug

[Wayne Davison]

On Thu, Mar 31, 2005 at 01:08:26PM -0800, Steve Harris wrote:
> I was just wondering if the following bugfix in 2.6.4 was a serious
> problem that could cause a problem in a destination file and if it did
> would that get picked up by the checksum ?

The problem would have most likely caused the protocol to abort.  If
you're using 2.6.3, it should not have caused a corrupted destination
file unless you had --partial enabled (which tells rsync to keep a
corrupted file).  However, older rsync versions used to save the
failed-to-verify file in place of the destination file, so I suppose
there's a chance that rsync could have lived long enough to tell the
receiver to create a file with some bogus data and then died before
the resend fixed it.

..wayne..