Progress specifics in regards to Compression of the Random Million digit bin file.
Paul
he has on his website. Within his contest specifications, what is the best
achieved compression on that file to date?
Paul
Willem
) Mark Nelson has a contest for compression of the Million digit bin file that
) he has on his website. Within his contest specifications, what is the best
) achieved compression on that file to date?
His contest *includes* all metadata in the size. Including decompressor size.
As far as I know, within those specifications, there has been no
compression. IIRC, there were a couple dozen bits of redundancy, but
no decompressor has been found that is short enough to take advantage of
this.
SaSW, Willem
--
Disclaimer: I am in no way responsible for any of the statements
made in the above text. For all I know I might be
drugged or something..
No I'm not paranoid. You all think I'm paranoid, don't you !
#EOT
Paul
"Willem" <wil...@snail.stack.nl> wrote in message
news:slrnh44vuq....@snail.stack.nl...
Thanks Willem.
Paul
Mark Nelson
Because there is no prize of any sort for a failure, regardless of how
close, there have been no submissions.
Clearly, within the bounds of the contest, a submission of 415,241 +
delta is possible with a very small delta. A C source file could be
cooked up that is probably only a few dozen bytes. I'm sure there is a
language somewhere that can do it one byte ;-)
Any entrant along these lines will be more or less a bogus submission,
because it might be the smallest entrant in the contest, but there is
no interesting work going on.
Anyone doing interesting work is going to have a hard time competing
with the bogus entries. A program that really, truly manages to get
something close to 1.0 compression on the file while actually putting
it through a compression algorithm will be rather large.
Of course Jules has claimed 90% compression with full reversibility on
that file, but nobody will ever take that seriously. He is the only
one, and it appears to me that there are no readers of
comp.compression who do not view him as A) a crank/troll, B) a
lunatic, C) a fraud, or D) some combination of these.
- Mark
Paul
"Mark Nelson" <snork...@gmail.com> wrote in message
news:92a7cb36-ae10-4a76...@o18g2000pra.googlegroups.com...
I appreciate the response Mark. I don't know Jules enough to know if he is
being a troll or really thinks it can be done. I can only speak for myself.
I believe it can be done. That is why I'm trying to gather this
information. In my case, I'm using a language that is higher level than C
so that means less code needed in the source file. However, since it is
interpreted that means the decompressor would be large when joined with the
interpreter. But as I understand it your challenge only looks at the size
of the source file and doesn't concern with the interpreter or other
compiler fluff.
Paul
Tom St Denis
> "Mark Nelson" <snorkel...@gmail.com> wrote in message
I think it just matters as to what you're getting out of the
interpreter. If all you're doing is making use of functionality like
trees, lists, etc.. it's probably fair to exclude that because it's
really not "part of" the message. But if you are extracting data from
the interpreter [like days of the week strings or whatever] that get
stored directly into the decompressed output, it should be included.
The point here, in case you haven't figured it out, is to push any
message specific entropy out of the decompressor. i.e. your
decompressor should work on ANY compressed strings and not just the
one specific to the challenge. And in fairness of the contest, really
your compressor should be able to compress an entire class of strings
[depending on how you model and extract redundancy]. And not just
this one.
I think I can speak for most readers here though, we'd be entirely
thrilled if you'd just go away and shut the hell up about your ideas
and claims until you can prove a single one of them. It's getting
really boring and old to read about "yet another random data
compressor." We all know you can't do what you're alluding to, and
simple math/logic proves us right. Yet you insist on laying claim to
idiotic impossible inventions. Stop it. Get a life, honestly.
Tom
Pete Fraser
news:92a7cb36-ae10-4a76...@o18g2000pra.googlegroups.com...
> Of course Jules has claimed 90% compression
> with full reversibility on that file, but nobody will
> ever take that seriously.
He seriously weaseled on the reversibility.
About three weeks ago he posted a sequence of files
showing recursive compression to about 5K.
I commented:
The one thing you have failed to mention is that
you can't use the final 5k byte file to re-construct
the original.
Jules replied:
Not true, Pete. I have decompressed files using this method. It is
true that I don't bother to decompress every single file I compress,
but the decompressor works -- but I'll give you this much, the
decompressor isn't -- at this moment in time -- as robust as the
compressor. I'd have to disclose info I don't want to discuss were I
to say more.
I replied:
So you didn't test the decompression on the example you just
gave us. You've repeatedly complained that we're a rude
and skeptical audience. It surprises me that you wouldn't
have verified the example that you chose to give us. Why don't
you write a simple shell script to decompress
the resultant file at each compression stage, and perform a
comparison / verification?
That was three weeks ago, and I haven't heard from Jules yet.
I assume he can't do it, is running out of excuses, and doesn't
want to respond with a bald-faced lie.
Paul
"Tom St Denis" <t...@iahu.ca> wrote in message
news:3316dbac-b72e-4ec7...@l32g2000vba.googlegroups.com...
So Tom, if I accomplish (oh let's say) Mark's challenge where does that
leave your comments?
Paul
Paul
"Tom St Denis" <t...@iahu.ca> wrote in message
news:3316dbac-b72e-4ec7...@l32g2000vba.googlegroups.com...
Tom, on another note, I can't continue to reply to you because of the amount
of rudeness that comes from your posts. I won't reciprocate it either.
Paul
Pete Fraser
news:0%N0m.14753$AL2....@newsfe10.iad...
> So Tom, if I accomplish (oh let's say) Mark's challenge where does that
> leave your comments?
Too many hypotheticals here.
Just accomplish it, and leave all of us (including Tom)
looking very foolish.
Jim Leonard
> Tom, on another note, I can't continue to reply to you because of the amount
> of rudeness that comes from your posts. I won't reciprocate it either.
Even though you just did?
Don't come back to this group, for any reason, until you've written
and tested your decompressor. That act alone will answer all of your
questions.
Tom St Denis
Not only that but it's a serious lack of respect to ignore the plain
obvious truth. The counting arguments aren't really up for debate,
it's fairly straight forward and undeniable logic that proves these
ideas true.
Yet he, and people like him, keep propagating their for lack of a
better term "noise" because they want attention. They're no better
than the 4 yr old kid grabbing at your jacket constantly to get your
attention when the adults are talking. At some point the adult lashes
back and tells the kid to go shut up and sit in the corner.
And it's not even a question of "oh gosh I hope I'm right so Paul
doesn't make a fool out of me..." it's about as certain a fact in my
mind as say I'm living and typing this message.
Paul
"Jim Leonard" <Moby...@gmail.com> wrote in message
news:d9d60355-2481-4a12...@g20g2000vba.googlegroups.com...
I'm going to take your advice and stop brining communications here. But you
may want to watch the www.tretbase.com/forum over the next few days and
maybe even just a bit later today. ;-)
Paul
Pete Fraser
> I'm going to take your advice and stop brining communications here. But
> you may want to watch the www.tretbase.com/forum over the next few days
> and maybe even just a bit later today. ;-)
You're starting to sound a lot like jg.
Your hints of revelations perhaps later today ring
a bit thin when, in your forum, you suggested on 5/21
that you should know much more by 5/25. Then silence.
You also suggested that you were struggling with Excel.
Harold showed you how to do what it seemed you
wanted with 6 lines of C code, but no reply from you.
Pete
Brine turkeys, not communications.
MichaelHH
to small? However it might be close enough to foil efforts.
As for the guy asking us to check his website...
"Must be logged in to view"
Thats cheap. I am not making a new account, accepting spam galore, or
having to memorize a new password, for a few posts.