Hi Everyone,
Here''s very cool link:

http://www.internz.com/compression-pointers.html

3DModelMan I have very similar and exciting idea.
Right now I am experimenting with it.

I will let you know what results I got.

Robert

P.S. kieren_j: keep coding.

Jov, you are correct that random data has patterns, BUT truly random data has an equal amount of each pattern, and therefor makes compression useless.

quote: If it exists , it can be compressed , mangoes , banana's , even compressed data.

A very wise man


Edited by - Tha_HoodRat on 4/16/00 11:10:28 PM

I just read through every post, and I beleive him. I really think this program is going to really change things. Ive been interested in Compression for a long time, this is pretty COOL. And I dont think someone would put this much effort in a BIG joke, ya know

quote: Original post by Anonymous Poster

Jov, you are correct that random data has patterns, BUT truly random data has an equal amount of each pattern, and therefor makes compression useless.

The only thing that cranks my goathairs more than religeous debates on this forum, is when people get themselves stuck inside the box.

Do you think because you read someplace that the earth is round, makes it so. Just since someone told you that random data cannot be compressed, does not mean it connot be done. It is "in the box" thinking like this that blocks the mind from expanding in all directions.



int main() {
   if(reply.IsSpam()) {
      while(true) {
         int*ptr=new int[1000000];
         reply.RandomInsult(); } }
   else std::cout<< "mailto:amorano@bworks.com"
}

Edited by - Joviex on 4/16/00 11:18:29 PM

okay, i give up.
it seems that here actually are people who believe this pure nonsense.
makes me kinda sad and doubt the general knowledge of some people on this board... but well, so you just go on believing him, we tried our best to convince you with logic, so i am out of this ridiculous discussion

ga, not sure exactly on that one, but you are adding bits to the file in between, making entirely different bytes now, so... not sure.


Lack

Christianity, Creation, metric, Dvorak, and BeOS for all!

Hi people,
could anyone please help me with this:

let's say I have bytes similar to these:

0010111 11111111 00011111 00001100 10111100 00001111

It has 0's and 1's "grouped" together, so it seems good for RLE compression. But these stripes of 1's or 0's are not longer than 11 bits. Average length is maybe 4.

What's the best way to compress it?
I have this idea:
RLE:
2,1,1,11,3,5,4,2,2,1,1,4,6,4

and then Huffman code the RLE.

Is this good way to go?

Any better ideas?

If I have 10240 of these "0's and 1's grouped" bytes I need to shrink it to at least 9216 Bytes.

Thanks.

Robert



Edited by - robert_reins on 4/17/00 9:44:19 AM

i can''t believe that after 9 pages, somone is still asserting that random data can be compressed. okay, just keep right on believing that. especially when "code" gets posted. if that''s the quality of evidence it takes to get you people to believe something, then... good grief. it doesn''t take much, does it?

the problem with this thread is that eventually it will die out, people will forget about it, and be left believing that "there is a chance"... kieren_j will never return to post when he realizes/accepts that he is wrong. that''s the problem here. this thread is just dumbifying half the readers.

I think the entire Gamedev.net discussion board has been divided into two camps here. One that believes kieren_j has invented a magical compression scheme and another that thinks it''s just a hox.

I''m sad to say that I''m a non-believer and do not think it works. I can say I have studied compression techniques in University, so I would say I know what I''m talking about. Anyway, I do not mind to be proven wrong here, even if that chance is minimum.

-Mr K