History hash collisions
rra at stanford.edu
Fri Nov 12 16:29:54 UTC 1999
Forrest J Cavalier <mibsoft at mibsoftware.com> writes:
> Apparently, the non-tagged DBZ code assumes that the first 6 bytes of
> the md5 hash are unique for all articles present in the history file.
> If the 6 bytes are not unique, you will not be able to retrieve an
> article, and I think you will get a duplicate hash message during
Yeah, I wasn looking at this the other day too, and I think this explains
some of the duplicate article messages that people get. We should
consider increasing it as part of the history API rework.
Russ Allbery (rra at stanford.edu) <URL:http://www.eyrie.org/~eagle/>
More information about the inn-workers