Re: Vdelta answers

[Karl Fogel <kf...@galois.collab.net>]

Greg Hudson <gh...@mit.edu> writes:
> Third, he clued me in on an interesting hashing trick used in their
> vdelta implementation: vdelta allocates an array of hash table entries
> equal to the length of the string being compressed.  The entry used
> for a given index into the string is just &array[index]; thus, the
> index can be implicit, saving four bytes in the hash entries.  Thus,
> the hash table looks a bit like:

Wow, that's quite clever!  I like it. :-)

> Fourth, he said that he was finishing up an open-source library for
> data transformations called Vcode, which would contain a Vdelta
> generator with vcdiff output.  This is good news; even if the license
> is bogus somehow, it gives us something to refer to (e.g. for settling
> the second point).

You might want to ask him which licence and when he's planning on
releasing it (am saying "you might want to" only because you've
already had contact with him).