Developer Forum

Hallo,

I think that everybody will agree that InfoRelation is a very useful
proposal. Thanks to this service I was able to finish my academic
project in field of biology. Also I can recommend InfoRelation for those
who would like to earn some extra money (I have checked it and it works
;-) )!

Sally Finch

On 20 Oct 2004 01:54:53 -0700, XXXX@XXXXX.COM



Hardly. Looks mostly like a bunch of students and journalist wanabes
who want questions asked for free. Most of which could be anwered by
simply learning how to use a search engine or library. (Well, you
might make $5 if you know a lot about OPEC, or $10 for information on
the cost of the EURO. )

Surprisingly there are no questions on compressing random data. UFO's,
alien food, and manned missions to mars, yes. But nothing on random
data :(

-Tim

how about "least probability" data, eg:
you allways pick the values least likely with the current model, and then
update the model with the picked value.
I would guess this would exhibit patterns similar to random data (fairly
even statistical distribution, fairly chaotic values, ...), however, it
would differ in that it could be compressed by making a predictor that takes
the model into account, but would compress poorly with a normal compressor.
likely random variation would be needed though, as the basic algo I was
imagining on its own would result in repeating patterns (which would
compress to easily, eg, with lz77).
other algos could be possible to generate data that does not compress well
with lz77 (eg: based on unused or least-used hash entries).

that or I am just being stupid...

Now that's an original type of spam...
http://www.yqcomputer.com/ :inforelation55%40yahoo.com+

But anyway, what I wanted to say :-) Anyone here who intents to make
money on these wonderful loseless compression algorithms of random
data (Jules, Erik, etc) here is your way to make money - just put your
idea on inforelation.com and watch the money coming!


Thanks, Paul