Best way to bring compression software to market

[Blindtech]

Good day all,

Myself and another developer have created a new compression model that we have not seen anywhere and would like advice on the best way to bring it to market.

Summary of the compression:

Without divulging to much here are the high points:

We can compress almost any size data to under 1mb completely lossless. File type does not matter

Example we can compress 1TB to under 50kb in just under 10 seconds and decompress it in about the same amount of time.

This is done using advanced calculus. And math that has not been discovered until we did about 6 months ago.

We were thinking about licening the software to data centers and charging for a percent of savings so they pay only for implementation upfront

Would be happy to here everyone's thoughts.

Also, yes we have a working prototype.

[Shelwien]

Start by winning http://prize.hutter1.net/

[comp1]

We can compress almost any size data to under 1mb completely lossless. File type does not matter

[Blindtech]

Start by winning http://prize.hutter1.net/

Nice that's easy cash thank you!

[Bulat Ziganshin]

It's impossible. large corporations already patented limitless compression, prohibiting to use it in any new products. instead, on order to maximize their incomes, they improve compression ratios by only 5-10% each year, so users need to renew their licenses periodically. i suggest you to go into free country such as Venezuela or North Korea, which abandoned patents allowing developers to work as they wish. good luck!

[Blindtech]

It's impossible. large corporations already patented limitless compression, prohibiting to use it in any new products. instead, on order to maximize their incomes, they improve compression ratios by only 5-10% each year, so users need to renew their licenses periodically. i suggest you to go into free country such as Venezuela or North Korea, which abandoned patents allowing developers to work as they wish. good luck!

While I completely understand. I also have to say that our method uses new math that hasn't been discovered or published. It is a completely different way of compressing data that doesn't have a hint of the current methods so I do doubt that this is patented already.

[introspec]

While I completely understand. I also have to say that our method uses new math that hasn't been discovered or published. It is a completely different way of compressing data that doesn't have a hint of the current methods so I do doubt that this is patented already.

That is not how these people work. They found clever ways to patent addition and multiplication, which means their patent base covers pretty much all mathematics to be done during the next century or so.

[Sebastian]

Also, yes we have a working prototype.

No you don't

[thorfdbg]

While I completely understand. I also have to say that our method uses new math that hasn't been discovered or published.

Could we please stop this nonsense? Lossless Compression is the injective mapping of one set of items to another. It does not get more general than that. And even in this very elementary setup, one cannot compress every input sequence to a limited size. This is called the "counting theorem", and is "discovered and published math", which follows from elementary logic.

If you want to sell snake oil, please go elsewhere. This is not the right place for "new math". Until then, I still stick to 2+2=4.

[xinix]

thorfdbg Dont touch him. Let him win the contest and everyone will understand who is worth it

[comp1]

Once a year, when we get a thread like this, two options are always presented:

1) Assume the thread creator has a degree of intellect far beyond any of our capabilities of understanding.

2) Laugh uncontrollably.

Option 2 always wins...

Hey everyone!

I just finished developing a new compression algorithm that allows users to compress Windows 10 64-bit to fit on a 3.5" floppy disk!

But gosh, I just don't know how to proceed with my inconceivable level of brilliance and I also don't plan on proving my assertions to you all. What shall I do?

@xinix - Sorry but I couldn't resist. If he truly has developed what he says, he will be compelled to prove us wrong.

[Sportman]

Example we can compress 1TB to under 50kb in just under 10 seconds and decompress it in about the same amount of time.

I make it easy for you, post a program that can read a random 100GB input file (10 times less then 1TB) and write from every 2,097,152 bytes read the last byte of it to an output file (what become 50KB in size), then I check if it can run under 10 seconds on my fastest computer.

[Blindtech]

All of the disbelief makes me smile. We are also entering Techstars. I get it. It was hard for me to believe when we first discovered it as well. I'll update as things progress.

[hexagone]

Good day all,

Myself and another developer have created a new compression model that we have not seen anywhere and would like advice on the best way to bring it to market.

Summary of the compression:

Without divulging to much here are the high points:

We can compress almost any size data to under 1mb completely lossless. File type does not matter

Example we can compress 1TB to under 50kb in just under 10 seconds and decompress it in about the same amount of time.

This is done using advanced calculus. And math that has not been discovered until we did about 6 months ago.

We were thinking about licening the software to data centers and charging for a percent of savings so they pay only for implementation upfront

Would be happy to here everyone's thoughts.

Also, yes we have a working prototype.

This is wrong on so many levels ...
The weird thing is that all these breakthrough compression posts follow a similar template: impossible compression ratios ? check. genius idea/new math ? check. wants money ? check.
Unless you want to hang around with the breakthrough/recursive/infinite compression folks who pop up periodically (and disappear quickly), I am afraid you landed in the wrong place.
<s>Your alias name is fitting BTW.</s>

[Sportman]

All of the disbelief makes me smile.

I wrote my request in .NET and it took 45.975 seconds (100GB input file to a 50KB output file both at RAMDISK).
How do you want to compress a 10 times bigger file (1TB) under 10 seconds with advanced calculus added?

So that part I do not believe.

[Sportman]

Would be happy to here everyone's thoughts.

You can contact Dr. Piotr Blass for advice, he was in 2001 and 2002 Chief Architect at Zeosync Inc., the company invented a similar math based random data compressor with "Multimillion dollar revenue for the company resulted" and "St George and others at the three-year-old company have put $10m into the work so far and the company has embarked on a funding round for about $20m. ZeoSync has about 30 staff now and has built up a team of mathematicians from around the world":

Dr. Piotr Blass July 25, 2014 Curriculum Vitae:
https://www.facebook.com/notes/piotr...2646646031747/

Zeosync Inc. January 7, 2002 press release:
https://web.archive.org/web/20020111...essrelease.htm

Peter St George (CEO of ZeoSync inc.) January 7, 2002 'No Shannon limit' for novel compression scheme:
https://web.archive.org/web/20020424...G20020107S0037

ZeoSync inc. July 19, 2001 website:
https://web.archive.org/web/20010719111638/zeosync.com

Zeosync Inc. US patent April 18, 2002 US20020044074A1:
https://patentimages.storage.googlea...20044074A1.pdf

Zeosync Inc. World patent March 7, 2002 WO2002019536:
https://patentscope.wipo.int/search/...d=WO2002019536

[hexagone]

St George and others at the three-year-old company have put $10m into the work so far and the company has embarked on a funding round for about $20m. ZeoSync has about 30 staff now and has built up a team of mathematicians from around the world

For context, this quote is from 2002, before they raised money and then shut down.

[moisesmcardona]

Why not provide a sample software to us to test?

[Gonzalo]

This is hilarious

[Sportman]

this quote is from 2002, before they raised money and then shut down

Indeed SEC filings are after their press release (January 7, 2002):
https://www.sec.gov/Archives/edgar/v...1/02011672.pdf (filing date: January 24, 2002, sign date: October 31, 2001)
https://www.sec.gov/Archives/edgar/v...1/02016795.pdf (filing date: February 19, 2002, sign date: February 1, 2002)
https://www.sec.gov/Archives/edgar/v...2/02024619.pdf (filing date: February 22, 2002, sign date: February 12, 2002)
https://www.sec.gov/Archives/edgar/v...3/02039903.pdf (filing date: June 6, 2002, sign date: April 29, 2002)

Peter St George (CEO) and Pope Richard (Director) are listed as beneficial owner (sign date: October 31, 2001).

I do not know anything about shutdown in 2002, maybe they sold their inventing or when looking to all the filings they had difficulty to get investors and moved to sleep mode, looks like they stopped their company in 2005 (after 6 years):

ZEOSYNC CORPORATION
Date Filed 08/27/1999
State FL
Status INACTIVE
Last Event ADMIN DISSOLUTION FOR ANNUAL REPORT
Event Date Filed 09/16/2005

Annual Reports

Report Year Filed Date
2002 05/06/2002
2003 04/30/2003
2004 04/30/2004

At May 6, 2002 Joseph Sarubbi is added as Officers and Directors (sign date: April 29, 2002) and at April 30, 2003 deleted (sign date: April 28, 2003).

[Blindtech]

Good day everyone. We are still chugging away at a few automation pieces. We estimate about 4-6 months out from the fully functional prototype. To be clear when I said we had the working prototype what I meant was the math checked out. No, we are not targeting the consumer market but for those of you that are interested once the actual software prototype is complete, we will be doing a live demo for everyone to see.


The format of the demo will be a live stream and will work as follows:


1. A viewer can submit any data to either be downloaded or via USB if they attend the demo
2. We will then compress the data one machine #1
3. When compression is complete it will generate a 6-20 digit alphanumeric code
4. We will enter this code into machine #2 which will reconstruct the contents of the original file


The machines will not have any internet connection or connection to each other to prove we are not just copying files.


We are based out of Denver Colorado our website is Artelec.io


For those of you that answered our original questions and helped answer thank you. For those of you that mocked and made jokes thank you for providing more motivation.


Once we have things lined up I will report the demo details here and on our website.


Cheers

[compgt]

Good day everyone. We are still chugging away at a few automation pieces. We estimate about 4-6 months out from the fully functional prototype. To be clear when I said we had the working prototype what I meant was the math checked out. No, we are not targeting the consumer market but for those of you that are interested once the actual software prototype is complete, we will be doing a live demo for everyone to see.

...

Cheers

4-6 months? That's too long i think. If things are clearly thought out, you can do it in less. The maths will come later. Practical compression ideas first that are easy to code.

When i was active coding compression programs up to 2009, i could finish a decent, perfectly working compressor in a day/week or two.

And if i may ask, how many million$ do you think is this new compression algorithm worth? I am thinking in terms of $120 million damages awarded from Microsoft-Stac debacle for DoubleSpace' infringement.

[compgt]

Well, Blindtech's compressor is clearly about compression of random data too. Since it promises to compress "any size data" with astounding compression ratios.

I've heard of JG's compression of "random appearing" data (RAD), and Zeosync's announcements.

But also of Datafiles/16, which somehow surprised me since i have thought of an algorithm before that is not a string-replacement algorithm. Datafiles/16 works on a matrix (at least that was the clue given in comp.compression FAQ); my algorithm (which was fascinating at first) uses a "frequency table" and just adds and subtracts. It's delta encoding of the symbol frequencies. Output is the frequency table (Q: highly compressible or not)?

[CompressMaster]

@Blindtech

Example we can compress 1TB to under 50kb in just under 10 seconds and decompress it in about the same amount of time.

Me too. I am able to compress 1TB to 0KB losslessly! Of course it´s possible ONLY if you have very simple and a lot of repeating patterns encoded inside files. Otherwise, you´re up against Information Theory and Shannon Information Entropy Limit.

But as you said, it´ll be good for us if your prototype will be available for testing (preferably EXE offline implementation - CMD or GUI) and therefore we could verify your claims. I´d be very glad if you could publish your software/algorithm.

[JamesB]

Good day everyone. We are still chugging away at a few automation pieces. We estimate about 4-6 months out from the fully functional prototype. To be clear when I said we had the working prototype what I meant was the math checked out. No, we are not targeting the consumer market but for those of you that are interested once the actual software prototype is complete, we will be doing a live demo for everyone to see.


The format of the demo will be a live stream and will work as follows:


1. A viewer can submit any data to either be downloaded or via USB if they attend the demo
2. We will then compress the data one machine #1
3. When compression is complete it will generate a 6-20 digit alphanumeric code
4. We will enter this code into machine #2 which will reconstruct the contents of the original file


The machines will not have any internet connection or connection to each other to prove we are not just copying files.


We are based out of Denver Colorado our website is Artelec.io


For those of you that answered our original questions and helped answer thank you. For those of you that mocked and made jokes thank you for providing more motivation.


Once we have things lined up I will report the demo details here and on our website.


Cheers

It's impossible. End of story. If you have a system that works then we can prove it's via smoke and mirrors, because, as I said before, *it's provably impossible*.

Claiming you can do it leaves yourself open to fraud lawsuits should anyone pay you money based on your faked experiment.

How many files do you permit people to submit? Can we submit 3 billion files to you and ask for a different 6-digit case-insensitive alphanumeric code for each?

[Conor]

Good day everyone. We are still chugging away at a few automation pieces. We estimate about 4-6 months out from the fully functional prototype. To be clear when I said we had the working prototype what I meant was the math checked out. No, we are not targeting the consumer market but for those of you that are interested once the actual software prototype is complete, we will be doing a live demo for everyone to see.

There are two kinds of impossible:
1. Thought to be impossible but too complex to prove for certain. E.g. the 19th century view on heavier-than-air flight.
2. Impossible without hope of workaround due to a logical contradiction.

Everyone wants to be the discoverer of a category 1 breakthrough, but algorithms which compress all possible inputs have always been, and shall always remain, in category 2, because of a simple logical contradiction. Each of your 20-digit alphanumeric codes will need to represent more than one possible input file, and the only way past this is to output more data. When you start adding extra bits in an attempt to make it work, the amount needed will turn out to be the same as a typical compression codec.

Many people have been in your position, rapt by the image of a future in which you're the wealthy genius designer of a data processing revolution. They come and go. Soon you, too, will go.

If you spend a few days honestly trying to prove why it will not work, you will save any further embarrassment and 4-6 months of effort, plus whatever money you and others may spend on it. Your feet probably float an inch off the ground when you rush to your computer each morning, and the last thing you want to do is burst your own bubble. But you make your choices and bear the cost of them. We've done our best to enlighten you.

You have no decoder, and until you have one you have no algorithm.

[Sportman]

1. A viewer can submit any data to either be downloaded or via USB if they attend the demo

You can create a version only working at 1MB files, with only 3 bytes or more compression gain and submit it in this thread to claim my reward:
https://encode.su/threads/1176-losel...ll=1#post59199

[AiZ]

Any news?

The "company" website's domain name has expired last month, I have a bad feeling...

[bwt]

i agree with you comp1

[anasalzuvix]

Hi Any Update? "3