🚀 go-pugleaf

In article <DJOHNSON.95Jan9142045@seuss.ucsd.edu>,
Darin Johnson <djohnson@seuss.ucsd.edu> wrote:
>> The title of the patent is:
>>
>> 	High speed data compression and decompression apparatus and method
>>
>> A layman's reading of this is that it covers decompression as well as
>> compression.
>
>Er, you can't base patents upon a layman's reading.  That would
>be disaster.  Second, a layman's reading of the *title* doesn't
>say anything about the contents.
>
>As a non-layman would understand it, the decompression is fast
>because of the method of compression.  The real guts of the
>patent is in the compression.  Just explaining the file format
>gives a programmer enough information to decompress, but little
>information on how to compress.

Not true-- the implementation of an encoder
and decoder for LZW look virtually identical.

LZ and LZW are adaptive codes; that is, as encoding progresses,
the encoder modifies the code table by adding code words
corresponding to input sequences it deems likely to
occur (because they have occurred before) in the input stream.
The code table itself is never explicitly transmitted;
the decoder deduces it on the fly by simulating the encoder.
--
Don Hatch  hatch@sgi.com  (415) 390-5150  Silicon Graphics, Inc.