Return Styles: Pseud0ch, Terminal, Valhalla, NES, Geocities, Blue Moon.

Pages: 1-

Synth Sort -> metaCompress v0.1a

Name: Anonymous 2014-01-19 20:59

8-bit test
11010010 :: +4
1101 :: +3 // 0010 :: (+1)
[11] :: +2 // 01 :: (+1) \\ [00] :: +0 // 10 :: (+1)
0 :: +0 // 1 :: (+1) // 1 :: +1 // 0 :: (+0)

output +4+3+2+0+0+1

binary out 0100 011 10 00 0 1

Name: Anonymous 2014-01-19 21:18


>██▓▓██▒▒▒▒▒▓▓▓▓▓▓▒▓████▓██▓▓██▓
>▓██▓▓▒░░░░▒▒▒▒▓█████████▓██▓▓██
>▓▓██░ ░░░░░░░░▒▓█████████▓██▓▓█
>█▓▓▒ ░░░░░░░░░░░▓████████▓▓██▓▓
>██▓ ░▒▒░░░░░░░░░░▒████████▓▓██▓
>▓█▒░▒▒▒▒▒▒░░░░░░░░▒████████▓▓██
>▓▓▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒░░▓█████▒██▓▓█
>█▓▒▓▓▓▓▒▒▒▒▒▒▒▒▒▒░░▒▒▓██▒▒▒██▓▓
>██▒▒▓▒▒▒▒▒▒▒░░▒▒▒▒▒▒░▒█▓▒▓▒▓██▓
>▓█▒▓▓▓▓▓▓▓▓▓▓▓▓▒▒▒▒▒░▒█▓░▒▒▓▓█
>▓▓▓▒██▓▓▓▓▒▓██▓▓▓▒▒▒░▒▒▒▒▒░▓▓█
>█▓▓▒▓▓▓▓▓▒▒▒▒▒▒▒▒▒░░▒▒░▒▒▒░ ░▒▓
>██▓▓▒▒▓▓▓▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒░
>▓██▓▓▓▓▓▓▒░▒▓▒▒▓▓▒▒▒▒▒▒▒▒▒░ ░░
>▓▓██▓▓▓▓▓▓▓▓▒▒▒▓█▓▒▒▒▒▒▓▒░ ░▒░
>█▓▓███▓▓▓█▓▓▓▓▒▒▓▓▓▒▒▒▓▓▒░░▒▒
>██▓▓▓█▓▓▓▓▓▓▓▓▒▒▒▒▒▒▓▓▓▒░░▓▒░
>▓██▓▓░░░▓▓▓▓▓▒▒▒▒▒▓▓▓▒▒░░▒▓░
>▓▓█░ ░░░░▓▓▓▓▒▒▓▓██▓▒▒░▒▒▓▒

Name: Anonymous 2014-01-19 21:34

>>1
What language is this?

Name: Anonymous 2014-01-19 21:48

mC-404 Notation

Name: Anonymous 2014-01-20 15:58

install BSD/Gentoo

Name: Anonymous 2014-01-20 20:35

hmm...
Or You can compress three bits to two 25% of the time, but it hasn't got much room for control structures...

Three bit sliding window gives you an up-down encoding though =)

eg. 1010
101 :: +2 (10) // 010 :: +1 (01)

but it doesn't know if the middle bits are 01 or 10

Name: Anonymous 2014-01-20 21:01

What is ``metaCompress v0.1a''? The only Google result is this thread. Is this some FrozenVoid infinite compression thing?

Name: Anonymous 2014-01-20 21:25

Kind of, though i'd be happy enough if it didn't expand the data, tricksy hobbitses =D

Name: Anonymous 2014-01-20 22:15

[101]101 / [011]011 / [110]110
1[011]01 / 0[110]11 / 1[101]10
10[110]1 / 01[101]1 / 11[011]0
101[101] / 011[011] / 110[110]

+2 / +2 / +2
+2 / +2 / +2
+2 / +2 / +2
+2 / +2 / +2

6->8 +temporary ambiguous sequences ^^

Name: Anonymous 2014-01-20 22:34

If it hits enough three-bit runs in the data it might be okay though =)
three bit runs have maybe 1 in four odds... so 6->8 is probably pretty typical..? long runs drop n/3 bits rounded down =)

Name: Anonymous 2014-01-20 22:44

It can compress 3 bytes to two if its 1x 24bit run or 8x 3bit runs... which is, uh, 4 / 2^24

Name: Anonymous 2014-01-20 22:49

oh, and combinations of 3-bit multiples ^^ so a few more

Name: Anonymous 2014-01-20 23:15

Where can we mere mortals obtain this software?

Name: Anonymous 2014-01-20 23:20

And you can just squeeze in a single secondary control sequence in 2-bits, but it can't follow a run =P

=D this might actually work!

Name: Anonymous 2014-01-20 23:31

>>13
i haven't coded it yet... still tinkering =)

Name: Anonymous 2014-01-20 23:38

What's the best use for control structures?
Markers and repeats?

Name: Anonymous 2014-01-21 1:59

And how to find the longest repeating sequence of bitcode?
or most repeated sequence?

Name: Anonymous 2014-01-21 2:49

>>1
Whatever the scheme short sequences of bits cannot be compressed in any way. Common sense.
There is not way to compress 2 bits to 1, 3 to 2, 4 to 3, but as bit count increases there is a larger chance to compress a single bit/byte. Taking a 4GB file there is very high chance some form of compression will compress it to 4GB-1byte.

Name: Anonymous 2014-01-21 7:08

>>18
Taking a 4GB file there is very high chance some form of compression will compress it to 4GB-1byte.
There are 2^2^32 different 4GB files. There are 2^(2^32 - 1) different 4GB-1byte files. Therefore no more than half of 4GB files can be compressed by 1 byte (and even in that case you have to do something with all 4GB-1byte files). Note that the file length is actually irrelevant.

Name: Anonymous 2014-01-21 8:04

>>19
4GB-1bit isn't it...

>>18
There is no way to compress/fit all eight 3-bit combinations in the four 2-bit spaces..

But actually, in a four gig file there might be fairly long repeated sequences, random or not... then it's just how efficiently you can find, mark and repeat them =)

Name: Anonymous 2014-01-21 8:16

I think this is going to be able to repeat ambiguous sequences too ^^ so that'll be interesting

Name: Anonymous 2014-01-21 9:43

>>19
>le pedophile sage

Name: Anonymous 2014-01-21 12:12

Name: Anonymous 2014-01-21 19:49

He's back on youtube?

Name: Anonymous 2014-01-21 20:04

Another source described as a current NSA analyst said: ‘In a world where I would not be restricted from killing an American, I personally would go and kill him myself.’

What about the other 96% ?

Name: Anonymous 2014-01-21 20:26

Such a typical american attitude...
The only reason that some american can't kill snowden is that He is American too!

Name: Anonymous 2014-01-21 23:46

Burn ^^

Name: Anonymous 2014-01-21 23:48

>>25
Another source?
WHAT WAS THE FIRST SOURCE?
And what was this source?
Holy shit, man.

Name: Anonymous 2014-01-22 0:00

le pentagon sage Everywhere!

Name: Anonymous 2014-01-22 0:38

Hmm...
there seems to be a little issue with ambiguous control structures..
And probably dorks trying to patent other people's things before they even work xD

Name: Anonymous 2014-01-22 0:45

Silly dregs...
This ain't the jewish olympics ^^

Name: Anonymous 2014-01-22 3:23

So, anyway, spec 0.2
First two bits of a chain (bitcode) determine number of true/on bits in the first three spaces (data), and if they are all zero (+0) or all one (+3), the chain terminates with instant 3->2 bit compression..
If not [ie, one or two true/on bits], then the sliding window kicks in, encoding -1/+0/+1 changes in the window sum.. Only the +0 code is properly Ambiguous, -1/+1 codes always disambiguate (up/down encoding), and can also terminate a chain by reaching either +0/+3 window sum..
Since chained bitcodes only use 3/4 of the two-bit space, there is room for a single control sequence, but it needs to follow either a +1 or +2, so we get 2x 4-bit control sequences. That's probably still not going to be enough to be useful, so i think i'm going with another extensible 4-bits for mark & repeat address codes, bringing the control structures up to 1-byte per piece, or 3 bytes minimum for a mark/mark/repeat set.
To be effective, data being repeated only once needs to be longer than 3 bytes, but if repeated multiple times it could be as short as 10 bits.

Name: Anonymous 2014-01-22 3:36

pretty snazzy so far eh? =)

Name: Anonymous 2014-01-22 7:45

Luke what about the nsa

Name: Anonymous 2014-01-23 6:23

110321112211011101100201221110111221110121120311021010111002001211112121010110020122111122110111011003110200121110211101120311021010111002001212222100121211

Name: Anonymous 2014-01-23 9:00

>>35
>le pedophile base 4

Name: Anonymous 2014-01-24 0:45

doesn't that make base 256 >le pentagon base^4

However, three of the five panel members concluded that the NSA spying programme "lacks a viable legal foundation" under the Patriot Act.

It "represents an unsustainable attempt to shoehorn a pre-existing surveillance programme into the text of a statute with which it is not compatible", they said.

The programme also raised constitutional concerns, including "serious threats to privacy and civil liberties as a policy matter, and has shown only limited value".

"As a result, the board recommends that the government end the programme," said the report.

However two panel members - both lawyers from the Bush administration - strongly dissented, saying the issue of legality should be left to the courts to decide.

One, Rachel Brand, also argued that declaring the process illegal could affect the morale of intelligence agencies and make them overly cautious.

saying the issue of legality should be left to the courts to decide.

But declaring the process legal could affect the morality of intelligence agencies and make them even more overly ambitious.

Name: Anonymous 2014-01-24 0:52

How can the courts decide if something is legal without knowing what it is, in it's entirety?
And if the public disagrees, does that make the public illegitimate, or the other?

Don't change these.
Name: Email:
Entire Thread Thread List