home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!cs.utexas.edu!zaphod.mps.ohio-state.edu!darwin.sura.net!sgiblab!adagio.panasonic.com!nntp-server.caltech.edu!madler
- From: madler@cco.caltech.edu (Mark Adler)
- Newsgroups: comp.compression
- Subject: Re: pkzip 2.04c / unzip v5.0 compatibility
- Date: 8 Jan 1993 18:03:05 GMT
- Organization: California Institute of Technology, Pasadena
- Lines: 26
- Message-ID: <1ikfkpINN4me@gap.caltech.edu>
- References: <1ied29INN37p@butler.cc.tut.fi> <1if4d6INNpgs@gap.caltech.edu> <1993Jan8.142251.17607@murdoch.acc.Virginia.EDU>
- NNTP-Posting-Host: sandman.caltech.edu
-
-
- >> I thought PKZip's and Info-Zip's compression engines were identical,
- >> but maybe PKWare sacrificed some compression efficiency for speed, as
-
- No, they are certainly not identical. At least on the assumption that
- the Katz patent has something to do with his compression algorithm.
- However both algorithms can trade speed for compression by changing
- some simple parameters.
-
- What is identical is the *format* of the compressed data, which
- essentially defines what a zip decompressor has to do. It does not
- put any restrictions on how that compressed data stream is generated.
- In particular, it does not say how matching strings are found.
-
- In this way you can have compressors with very different speed and
- compression performance, yet a single decompressor can work with all
- of them.
-
- This basic principle is what allows competing companies to agree on
- standard compressed data streams, like JPEG, JBIG, MPEG, etc. The
- standard allows common decoders for the format, but the companies
- compete with more efficient and more effective *encoders*, as wellas
- cheaper or better implementations of the decoders.
-
- Mark Adler
- madler@cco.caltech.edu
-