tag:blogger.com,1999:blog-146982802011-04-21T20:17:14.829-07:00jaxonixhttp://www.blogger.com/profile/01603459945877792770noreply@blogger.comBlogger11100tag:blogger.com,1999:blog-14698280.post-1121969806215829882005-07-21T11:14:00.000-07:002005-07-21T11:16:46.220-07:00i was trying to post this to janonix knox blog and suddenly the blog which is about infinite style compression and perpetual motion went down. i wonder why?<br /><br />i think the differnce between the data compression camps are: <br /><br />PIGEON CAMP <br />you have a couple of examples of 32 bit numbers, which are the result <br />of hip top compress algo no1 <br /><br />#FFFF and #AC64 which come at 1/2^32 chance each, both with self <br />information of 32 bits, and as all the set of symbols of 32 bits <br />equally spaced in probability then these numbers will never decompress <br />to somthing with more than a self information of 32 bits. and so <br />on...... <br /><br />PERIGRIN CAMP <br />you have a couple of 32 bit numbers, which are the result of being <br />sampled from the running seed of RADAR gimper special edition <br />compression vapour ware number 1, as each number does not have the same <br />probability of occurance with #AC64 being 1/2^30 chance and #FFFF <br />having 1/2^33 chance the the self information of one is 30 bits and the <br />other is 33 bits, then knowing that a modulation between the two states <br />would supply 3 bits to the seed, with such running of seed giving 3m <br />bits of information to the running stream above its uniterfered <br />preferance for self information. so 3m bits of extra information can be <br />modulated. This results in a storage capacity of 3m supported on a <br />stream of ratio n:1 giving n-2:1 storage bits due to skip coding and <br />with 3/4 of a bit per bit information rate difference due to loss of <br />coding symbol giving 3/4 * (n-2)/(n+1) bits per bit to be modulated. <br />So the information efficiency rate would be 3/4 * (n-2)/(n+1) * i/(3m) <br />, with i being the bits absorbed for your data to be compressed. <br /><br />that would be self information mesured bits of 50:50 random nature, if <br />the bits in were not random then the 3/4 would have to be <br />1*p(0)+p(1)/2. with the 11 symbol used in 110 uniquely decodeable <br />symbol as seperate from 0 and 10 symbols used to represent 0 and 1. <br /><br />and explination for any "wierd" language could be that verbelizing patterns often result in <br /><br />onomatapeic <br />words, but sometimes a better more engineering vocabulary can only be <br />expressed when a mistake in short expression happens. <br /><br />and so follows a tautology<br />restricting language makes the population more ameanable to ones will. <br /><br /><br />with both meanings demonstrating a statement meaning height of at least 2. i.e. there can be <br /><br />two interpretations of the facts.<div class="blogger-post-footer"><img width='1' height='1' src='https://blogger.googleusercontent.com/tracker/14698280-112196980621582988?l=sorryish.blogspot.com' alt='' /></div>jaxonixhttp://www.blogger.com/profile/01603459945877792770noreply@blogger.com3