disclaimer: i am not professional programmer.
As you may know mandela effect in full swing, to point that even bible no more exemption(but of course if you read it, you know it was main goal).
https://www.youtube.com/watch?v=6I-GMVpinpQ
word "matrix" now in bible, marvelous isnt it? i made little search and find that text of new testament in original greek at least for now is intact. peoples can say text is changed only because their remember it, but because peoples can memorize only small portions of it, and not without mistakes, so, my question is: how to save in mind full text? i check it and turn out size of file is approximately 2mb/unicode. obviously, you cant put in brain 2mb file, so it needs to be archived, in winrar it 340kb, better, but still not enough. any ideas?
that what i thinking: lets take two examples string from Kolmogorov complexity article abababababababababababababababab and 4c1j5b2p0cv4w1x8rx2y39umgw5q85s7, while first is compressed quite easy, write ab 16 times, it is not obvious how to compress second, as i get it, nobody telling its not possible, only not obvious. my idea: let there be some method A, and -A, that map set of all strings with length n into another set of all strings with length n, ie bijection, so that if we put second string in it we get first, so it would look like this "write ab 16 times, then apply -A". question, how we get A? as it seems it as not obvious as second string itself. but who cares, take any method to generate A, like reversible pseudorandom number generator, apply to string, if new string is still uncompressible, apply again and add to counter of how many times we apply X one, repeat until size of new string, ie counter plus string, become bigger than original, or time of findings become to astronomical. if we dint find compression with A, dont upset, simply use method B, then C, more times we try, bigger probability of success become. but still, we want our method A be fast and efficient, so all strings that we perceive as uncompressible become compressible, like it has many repetition or it became sort in order. what properties A must have to became good? i dont know. only i can add to all of this, is this algorithm: take some file, now look at it only as binary sequence, with 0 and 1, now look at it as ball and boxes, like you have n - size of file in bits, boxes, and m - amount of 1, balls. in that case you can rewrite your binary sequence as triplet of numbers, number of bits, number of 1, and indexing number of it decomposition, like 00111 become 5,3,1; or 0101 become 4,2,2(note: i actually get formulas for that, but i forget them at my relatives house and i dont want to redo them again, i think peoples here smart enough to derive them faster then i go to relatives and back; also while formula binary->triplet is parallel, formula for triplet->binary is not, maybe you can get better formula?). problem: when number of 1 is equal roughly from 1/4 to 1/2 to number of bits algorithm dont compress. my solution: map all uncompressible sequences n-size into n+c-size(i dont know what c equals; also maybe someone find method to map n into n-c, that would be nice) but with less 1 in them, like 100111001 became 00000000011(also this method is good for energy conservation, in above example if to transmit 1 we need say n and for 0 n/2, then for first sequence we get 7n energy and for second only 6,5n) or something like that. also problem - you need somehow distinguish numbers, ie symbol or method for comma, like putting in middle of file bitness of n in reverse, forward, reverse, forward order. and because it work on binary you can run it how long as you like. overall it work like this: get binary, write at start 0, it would mean it original sequence, turn it into triplet, if number of 1-s is (1/4;1/2) of n map n, write 1 at start of file and get new triplet, if not write 0, now put 1 at start, it would mean it non-original sequence, now you can run it again or reverse it to original. your thoughts?
also while writing notice funny mapping rule, 0->0, 1->1, but 10->00, 11->01, 100->10 etc. how much you can squeeze out of it?
As you may know mandela effect in full swing, to point that even bible no more exemption(but of course if you read it, you know it was main goal).
https://www.youtube.com/watch?v=6I-GMVpinpQ
word "matrix" now in bible, marvelous isnt it? i made little search and find that text of new testament in original greek at least for now is intact. peoples can say text is changed only because their remember it, but because peoples can memorize only small portions of it, and not without mistakes, so, my question is: how to save in mind full text? i check it and turn out size of file is approximately 2mb/unicode. obviously, you cant put in brain 2mb file, so it needs to be archived, in winrar it 340kb, better, but still not enough. any ideas?
that what i thinking: lets take two examples string from Kolmogorov complexity article abababababababababababababababab and 4c1j5b2p0cv4w1x8rx2y39umgw5q85s7, while first is compressed quite easy, write ab 16 times, it is not obvious how to compress second, as i get it, nobody telling its not possible, only not obvious. my idea: let there be some method A, and -A, that map set of all strings with length n into another set of all strings with length n, ie bijection, so that if we put second string in it we get first, so it would look like this "write ab 16 times, then apply -A". question, how we get A? as it seems it as not obvious as second string itself. but who cares, take any method to generate A, like reversible pseudorandom number generator, apply to string, if new string is still uncompressible, apply again and add to counter of how many times we apply X one, repeat until size of new string, ie counter plus string, become bigger than original, or time of findings become to astronomical. if we dint find compression with A, dont upset, simply use method B, then C, more times we try, bigger probability of success become. but still, we want our method A be fast and efficient, so all strings that we perceive as uncompressible become compressible, like it has many repetition or it became sort in order. what properties A must have to became good? i dont know. only i can add to all of this, is this algorithm: take some file, now look at it only as binary sequence, with 0 and 1, now look at it as ball and boxes, like you have n - size of file in bits, boxes, and m - amount of 1, balls. in that case you can rewrite your binary sequence as triplet of numbers, number of bits, number of 1, and indexing number of it decomposition, like 00111 become 5,3,1; or 0101 become 4,2,2(note: i actually get formulas for that, but i forget them at my relatives house and i dont want to redo them again, i think peoples here smart enough to derive them faster then i go to relatives and back; also while formula binary->triplet is parallel, formula for triplet->binary is not, maybe you can get better formula?). problem: when number of 1 is equal roughly from 1/4 to 1/2 to number of bits algorithm dont compress. my solution: map all uncompressible sequences n-size into n+c-size(i dont know what c equals; also maybe someone find method to map n into n-c, that would be nice) but with less 1 in them, like 100111001 became 00000000011(also this method is good for energy conservation, in above example if to transmit 1 we need say n and for 0 n/2, then for first sequence we get 7n energy and for second only 6,5n) or something like that. also problem - you need somehow distinguish numbers, ie symbol or method for comma, like putting in middle of file bitness of n in reverse, forward, reverse, forward order. and because it work on binary you can run it how long as you like. overall it work like this: get binary, write at start 0, it would mean it original sequence, turn it into triplet, if number of 1-s is (1/4;1/2) of n map n, write 1 at start of file and get new triplet, if not write 0, now put 1 at start, it would mean it non-original sequence, now you can run it again or reverse it to original. your thoughts?
also while writing notice funny mapping rule, 0->0, 1->1, but 10->00, 11->01, 100->10 etc. how much you can squeeze out of it?