RE: OFFTOPIC: Re: bzip2 for kernel dists? (fwd)

James Mastros (root@jennifer-unix.dyn.ml.org)
Fri, 12 Dec 1997 16:21:14 -0500 (EST)


Sorry to send two nearly identical mails to linux-kernel, but I noticed that
I mispelled the e-mail of the author of bzip2, and relised that I should
have mentioned the following URLs:
http://www.murona.demon.co.uk/ -- bzip2's offical web page, to which I am
having troube connecting to (don't worry, admin, I'm having trouble
connecting to many sites...)
OK, on second thought, worry. I'm getting squid errors (from the remote
end). Looks like demon's problem, but I let jstewart handle it <G>.

On Fri, 12 Dec 1997, Adam McKee wrote:
[snip all the part vaugely related to the linux kernel]
> The succession of LZ77-based compression by BWT-based
> compression is inevitable as machines get faster. But it will take some
> time to replace a standard as strongly entrenched as gzip. gzip is an
> absolutely superb LZ77-based compressor - maybe even the best of its kind.
>
> How about integrating the BWT-based compression in bzip2 into gzip in
> order to speed the transition and make it easier?
>
> gzip -bwt? Just an idea.
Sounds good to me, but I would implement this in the reverse way -- have
bzip2 understand .bzip, .bzip2, .gz, and .Z archives (possibly just by
forking gzip), and have a -z and -Z option. Also, default to -z when
executed as g{un+""}zip, and -Z when executed as {de|""}compress. (bzip
only on decompress, due to copyright/tm problems).

(Note: I'm adding a CC to the author of bzip2)

-=- James Mastros

-- 
When the annals of distributed computing are written, and the name 'Bovine'
appears in there, I can say "Hey, I was a part of that, I checked .0012% of
the keyspace".
	-=- Brian Wilson <wilsonb@mindspring.net>
Go to www.distributed.net before I come make you!