[vorbis] New decode chip

[Alejandro G. Belluscio]

Hello fungus,

Tuesday, August 6, 2002, 12:19:22 PM, you wrote:

f> Shawn wrote:
>> A functionality must have a certain level of specificity in order to
>> evolve a solution.
>> 

f> This thread is total rubbish, you can't "evolve" things
f> like ogg decoders any more than you can "evolve" an
f> algorithm to add two numbers. Adding numbers is a
f> precisely defined algorithm with no fuzziness, you
f> can't vary the algorithm a bit to see if the numbers
f> add together "better".

Please remember that you can evolve the memory distribution, the timing
of instruction and the inlining of code. And that's just without
changing a single instruction (and you can multiply by 6 or sum a shift
and a double shift, and believe me, it's usually faster).
And the fact that you are forgeting is that this i a FPGA. So it´s not a
program that we are trying to improve but a transistor implementation.
Which could be evolved from a reference implementation.

f> Genetic algorithms are used for tasks where there's a lot
f> of fuzziness and the "correct" answer would otherwise be
f> found by trial and error, eg. creating the shape of a
f> racing boat hull.
For shape designing you use simulated annealing. Don't confuse the
algorithms.

<p><p>--- >8 ----
List archives:  http://www.xiph.org/archives/
Ogg project homepage: http://www.xiph.org/ogg/
To unsubscribe from this list, send a message to 'vorbis-request at xiph.org'
containing only the word 'unsubscribe' in the body.  No subject is needed.
Unsubscribe messages sent to the list will be ignored/filtered.