Best results in compression of Random Million digit bin file
Paul
bin file? 1 byte? 2 bytes? more?
Paul
Tom St Denis
With or without including the size of the decompressor, filenames, and
any related metadata?
Tom
Paul
"Tom St Denis" <t...@iahu.ca> wrote in message
news:6bb1f580-8f91-4145...@e20g2000vbc.googlegroups.com...
Yes Tom, I just want to know the highest level of compression accomplished
in bytes for just the data itself, excluding all else.
Paul
crisgoogle
> "Tom St Denis" <t...@iahu.ca> wrote in messagenews:6bb1f580-8f91-4145...@e20g2000vbc.googlegroups.com...
"Yes" what? You failed to answer the question.
This is important because I can easily reduce the million random
digits file
to exactly 0 (yes, zero) "data" bytes if you don't count all that
other stuff.
Paul
"Tom St Denis" <t...@iahu.ca> wrote in message
news:6bb1f580-8f91-4145...@e20g2000vbc.googlegroups.com...
Without including the size of the decompressor, filenames and any related
metadata.
Paul
Alan Morgan
I can compress it down to one byte (all other files get expanded by one byte,
which isn't bad). The decompressor is pretty big, however.
Alan
--
Defendit numerus
Thomas Richter
>>> What is the best results achieved in compressing the Random Million
>>> digit
>>> bin file? 1 byte? 2 bytes? more?
>>>
>>> Paul
>>
>> With or without including the size of the decompressor, filenames, and
>> any related metadata?
>>
>> Tom
>
> Without including the size of the decompressor, filenames and any
> related metadata.
0 bytes. Literally: ZERO. I can write a decompressor for that easily.
Thus, I afraid, your question is rather meaningless.
Greetings,
Thomas
Paul
"Thomas Richter" <th...@math.tu-berlin.de> wrote in message
news:h1tt61$g9h$1...@infosun2.rus.uni-stuttgart.de...
yes, I guess I don't know how to explain what I'm looking for. Oh well. I
won't reveal anything then until I get someone who can understand what I'm
looking for.
Paul
Paul
"Paul" <pa...@tretbase.com> wrote in message
news:x7r0m.40738$QS.1...@newsfe18.iad...
> What is the best results achieved in compressing the Random Million digit
> bin file? 1 byte? 2 bytes? more?
>
> Paul
I'll start another thread as people couldn't understand what I was asking.
Paul
Thomas Richter
>>> Without including the size of the decompressor, filenames and any
>>> related metadata.
>>
>> 0 bytes. Literally: ZERO. I can write a decompressor for that easily.
>>
>> Thus, I afraid, your question is rather meaningless.
>>
>> Greetings,
>> Thomas
>
> yes, I guess I don't know how to explain what I'm looking for. Oh
> well. I won't reveal anything then until I get someone who can
> understand what I'm looking for.
Well, I afraid you first need to understand the question. The problem is
that you cannot measure the efficiency of a compression scheme on *one
specific* source. You can only measure it on a *class of sources* of
which the one you picked is a typical one.
For example, a more meaningful question would be:
a) what is the minimal size of the compressed file and the decompressor
(two files only, file names do not matter): This would be interpreting
the compressor in the sense of the Kolmogorov complexity (one method).
b) What is the minimal size a gaussian i.i.d. source with uniform
probability density can be compressed to? (This would be a model for the
given file). The answer here is known: The size of the file itself (on
average).
Thanks,
Thomas
Earl_Colby_Pottinger
> bin file? 1 byte? 2 bytes? more?
Basicly, you can prove your compressor/decompressor works if the all
of the resulting compressed_files, metadata and decompressor can be
tarballed into a single file that is 50 bytes or smaller than the
binary version of the 'Million-Digits-File' that any of us can
download and test on our own trusted systems.
Your claims are meaningless, they are just words. Your computer code
running on your system is untrusted because of possible cheats and/or
possible mistakes on your part when testing. Your witnesses can not
be trusted until it is clear they have not been brided or in-fact do
not know how to properly test such a compression program.
If you want to prove it, you have to show it.
And the easy way is simple. Send us your decompressor, send us your
compressed version of the 'Million-Digits-File' and tarballed together
would be best.
A Tarball that is 50 bytes or more smaller than the binary version of
the 'Million-Digits-File' is what will really make us sit up and
notice.
And ***WE*** will run it on hardware that we can trust. Hardware with
no internet connections, hardware with no hidden version of the
'Million-Digits-File', hardware under our own personal control. If
you can then pass all of the above, I will personally kiss your shoe.
Hech, I will post a video of it on YouTube if you want.
But the Tarball of *ALL* the needed files must be 50 bytes or more
smaller than the binary version of the 'Million-Digits-File' if you
want my beautiful lips on your smelly shoes.
PS. It's been 15 years of baiting people who make claims like you, and
I have not lost a single bet yet.
Fibonacci Code
Like that,
I determined to get your kiss too.
Regards,
Fibonacci.
-p.s. joke