-=> Quoting David Roper to Frank Cox <=-
DR> Frank, I know you didn't write it but do you have any idea why
  > numbers don't compress so good? The example you gave in the
  > example compresses real well, and so does a random sentence of my
  > choice, but when I try to use numbers like
  > 243569120023450015000113000000000000000000000 It doesn't give
  > much compression. Maybe loses 2 or 3 bytes or so.
The 'dictionary' as it is misnamed is the key, here. It plays on the
rule that certain sequences of characters exhibit a high frequeny of
occurence. It is simple and elegant, but with numbers all bets are
off. 
I'm glad you tested it out, I have use for such a routine.
Now, if I can just find what I ded with it. 
--- Maximus/2 3.01
---------------