f



Venture capitalist Tom Perkins about Jan Sloot

Part of article USA TODAY 11/18/2008:

There's the bizarre story of Jan Sloot, a Dutch engineer "who had no
academic credentials and who had been for most of his life a
television repair technician."

Sloot was working on fitting hours of video onto a smartcard and
claimed to have a prototype. A smartcard, at 75 kilobytes, holds about
1/50,000th the data of a DVD. Sloot's invention seemed to violate the
laws of physics.

Perkins is a graduate of MIT in physics, and a skeptic, but soon was
convinced that Sloot's breakthrough worked. Perkins flew to Amsterdam,
and Sloot recorded and played back on the spot a TV show Perkins
chose.

Perkins agreed to fund Sloot's venture and soon he, Sloot and another
investor were drinking champagne and eating chocolate cake. The next
morning, Perkins got a call that Sloot had died.

That left a problem: Sloot had disclosed his technology except for the
"compiler" program that translated the images into data. Fearing theft
of his invention, he kept this key part.

The compiler was never found, and Perkins' team of scientists could
not reverse-engineer it.

No one has come close to packing so much data into so little memory.
We'll likely never know if Sloot's invention was genius or a hoax.

http://www.usatoday.com/money/books/2007-11-18-valley-boy-Perkins_N.htm


I will add here that the software was located in chips, Dutch High
Tech Analysis a business magazine about IT published 4/17/2008 an
article where technology analist Henny van der Pluijm say that Roel
Pieper probably had the source code because it could be read from the
embedded chips because he had a working prototype.

Part of article High Tech Analysis:
http://www.admanager.nl/online/nieuws/7950/Roel_Pieper_kon_technologie_van_Jan_Sloot_w%E9l_kraken/

With the new Google translate you can translate this Dutch article in
23 other languages http://www.google.com/translate_t
0
Sportman
5/19/2008 9:58:30 AM
comp.compression 4696 articles. 0 followers. Post Follow

10 Replies
1054 Views

Similar Articles

[PageSpeed] 55

On May 19, 4:58=A0am, Sportman <sport...@gmail.com> wrote:
> Part of article USA TODAY 11/18/2008:
=2E..
>We'll likely never know if Sloot's invention was genius or a hoax.

I'm pretty sure we know already!

If Tom Perkins is really as smart as he ought to be, I have a feeling
this story is embellished a bit for the sake of selling the book.

I do enjoy this journalistic device: "Perkins is a graduate of MIT in
physics"

This undocumented claim is supposed to assure us of his genius, as if
MIT has never graduated an idiot.

|
| Mark Nelson
| http://marknelson.us
|
0
Mark
5/20/2008 12:05:00 PM
> This undocumented claim is supposed to assure us of his genius, as if
> MIT has never graduated an idiot.

 Wou, wou, wou. Getting hot in here. ROTFL

 Ciao
   Niels


0
spamtrap
5/21/2008 3:28:29 PM
Does Sloot's codec only work for video or any data? I can't see how 8
KB can fit an hour of video content. The number of seconds in two
hours just barely fit into 8192 bytes, so I'd assume it was a RAD
codec primarily designed for large files such as video.

Too bad these faggots just keep dying before the scheduled release of
their revolutionary shit...
0
Industrial
5/21/2008 5:47:30 PM
On May 21, 5:47 pm, Industrial One <industrial_...@hotmail.com> wrote:
> Does Sloot's codec only work for video or any data? I can't see how 8
> KB can fit an hour of video content. The number of seconds in two
> hours just barely fit into 8192 bytes, so I'd assume it was a RAD
> codec primarily designed for large files such as video.
>
> Too bad these faggots just keep dying before the scheduled release of
> their revolutionary shit...

Sometimes they did not die, they just made a run with all the money
leaving investors wondering why they can't make any sense with what
was left behind.
0
earlcolby
5/22/2008 7:13:26 AM
On May 21, 7:47=A0pm, Industrial One <industrial_...@hotmail.com> wrote:
> Does Sloot's codec only work for video or any data? I can't see how 8
> KB can fit an hour of video content. The number of seconds in two
> hours just barely fit into 8192 bytes, so I'd assume it was a RAD
> codec primarily designed for large files such as video.

Yes it worked for any data.

Jan Sloot even invented his algorithm for text and pictures. He wanted
to create a database (repa-base) with all info (description problems,
description fixes, part lists and scematics) to repair all brand TV's
but because that time the storage space was to little he tried
compression but that was not enough (saved round 50%). Later he found
a solution by creating a little database with a fixed amount unique
data and use that parts to build any text or picture with a unique
algorithm. Jan Sloot always said it's not compression but a different
way of storage/coding, he also said the principle is very simple and
easy to copy.

This repa-base got attention of a client (successful business man) of
his repair shop who became his first investor to start a company to
sell this repa-base to all repair shops so they could work faster. Jan
Sloot was already started with trying to use his repa-base algorithm
to store raw video signals the same way. Because the affords and
resources he put in his video project the repa-base project got to
less attention and his investor lost patient and stapped out.

His first working video storage device was very simple and little in
size and total solid state. One 2KB memory (video in) and two 4MB
flash memories (one of it was tempory storage) and a 128KB chipcard
(external storage) and two CPU's. At the chipcard he could generate
and store a final key what in combination with this same unit can
generate and playback a full length DVD lossless.

At 04/29/1997 he applied this device for a Dutch patent and 01/04/1999
it was approved. This patent was financed by that first investor.

Later more and more bigger investors started to take care about him
and his invention and he builded more complex devices. The latest
device had 5 CPU's and five 73MB memory units and could store many
hours of video as 1KB key including audio so one 128KB card could
contain 128 full DVD movies. It could playback 16 movies at the same
time and you could jump to any scene instantly without any sync
problem. This is the device Tom Perkins saw. A simpler version of this
device was applied for a Dutch patent 08/20/1998 and granted
07/03/2000 it could also store any book in 1KB. All his devices were
little in size (5 cigarette boxes big) and fully solid state.

> Too bad these faggots just keep dying before the scheduled release of
> their revolutionary shit...

He died in his garden when he was gardening because heart problems
(official release). Nobody saw it happening he was found dead by his
wife, he was hanging in his gate because his fall. His office was
total empty the day he died, before it was total full nobody know
where his stuff is.
0
Sportman
5/22/2008 4:18:17 PM
Sportman:

snip...
> His first working video storage device was very simple and little in
> size and total solid state. One 2KB memory (video in) and two 4MB
> flash memories (one of it was tempory storage) and a 128KB chipcard
> (external storage) and two CPU's. At the chipcard he could generate
> and store a final key what in combination with this same unit can
> generate and playback a full length DVD lossless.
>
> At 04/29/1997 he applied this device for a Dutch patent and 01/04/1999
> it was approved. This patent was financed by that first investor.
>
> Later more and more bigger investors started to take care about him
> and his invention and he builded more complex devices. The latest
> device had 5 CPU's and five 73MB memory units and could store many
> hours of video as 1KB key including audio so one 128KB card could
> contain 128 full DVD movies. It could playback 16 movies at the same
> time and you could jump to any scene instantly without any sync
> problem. This is the device Tom Perkins saw. A simpler version of this
> device was applied for a Dutch patent 08/20/1998 and granted
> 07/03/2000 it could also store any book in 1KB. All his devices were
> little in size (5 cigarette boxes big) and fully solid state.
>
>   

Uhmm... this doesn't cope with any consumer-grade hardware bus at that 
times.
Let alone the algorithm, how can you playback 16 DVD-resolution movies 
*at the same time*, presumably at 24 fps with that technology ? I don't 
think there exists a bus that can take that much data per second from/to 
memory and to a display device. Even if you got multiple buses towards 
the display devices, you would need a single one that serves the many to 
transmit the movies at the same time. (wouldn't you ?)
16 DVD frames at 24fps would take around 477MB/sec transfer rate... PCI 
buses were introduced in 1993 - according to wikipedia - and transfer 
133MB/sec. AGP - very specialized bus, and certainly not easy to 
interface for mere mortals - was introduced in 1997.
It's hard to believe he could use an AGP bus...or that he created a bus 
by himself that was almost five times faster than PCI.
There's some "history" issue here (?!).

Best,
E.

0
erpy
5/22/2008 10:49:58 PM
On May 23, 12:49=A0am, erpy <i...@forwardgames.com> wrote:
> Uhmm... this doesn't cope with any consumer-grade hardware bus at that
> times.
> Let alone the algorithm, how can you playback 16 DVD-resolution movies
> *at the same time*, presumably at 24 fps with that technology ?

This is a good question. Jan Sloot did most of his demonstrations at
an 24inc monitor, let say a PAL DVD has 720=D7576 pixel resolution then
his monitor needed at least 2880x2304 pixel to display 4x4=3D16 full
screen DVD movies. I have no idea of this resultion was possible at an
24inc monitor in 1998/1999.

What can explain the bandwidth problem is that Jan Sloot's system only
updated the pixels who changed, so it didn't need to transmit frames,
this is described at the Davoc website (http://www.davoc.nl/
technics.html) created by investor Leon Sterk before Roel Pieper
stepped in:

Part of the description:
That way a perfect picture can be quaranteed, in which every detail is
controlled by a reference code. No stripes, fading or other
disturbance will be shown in the picture. Of every pixel only one
pixel, if recognized as DDS, will be changed. And only if the pixel is
different then the one before. Therefore the picture will stand like a
photo on the screen. No flicker or any other disturbance will be
shown. Because of the fact that the screen will be no longer refreshed
50 times a second it is not necessary have to line, grid or screen
synchronization.
0
Sportman
5/23/2008 1:51:42 PM
On May 23, 8:51 am, Sportman <sport...@gmail.com> wrote:
> Part of the description:
> That way a perfect picture can be quaranteed, in which every detail is
> controlled by a reference code. No stripes, fading or other
> disturbance will be shown in the picture. Of every pixel only one
> pixel, if recognized as DDS, will be changed. And only if the pixel is
> different then the one before. Therefore the picture will stand like a
> photo on the screen. No flicker or any other disturbance will be
> shown. Because of the fact that the screen will be no longer refreshed
> 50 times a second it is not necessary have to line, grid or screen
> synchronization.

This makes no sense whatsoever, especially the last sentence.  It
describes a video compression as understood by a child ("only update
changed pixels") which is the tip of the proverbial iceberg in any
video compression system.

I think it's clear that Sloot was just another huckster.  The
different between him and Adam was that he was a tiny bit more clever
to build a "prototype" piece of hardware to distract people from the
fact that he was simply playing movies from another location.

A big indicator of snake oil is whether or not there are video records
of these demonstrations.  In all cases, there are none.
0
Jim
5/23/2008 2:36:07 PM
On May 23, 4:36=A0pm, Jim Leonard <MobyGa...@gmail.com> wrote:
> This makes no sense whatsoever, especially the last sentence. =A0It
> describes a video compression as understood by a child ("only update
> changed pixels") which is the tip of the proverbial iceberg in any
> video compression system.

If you take a 90 minutes raw DVD video 720=D7576 pixel at 25 fps and
convert every pixel to a value corresponding with the pixel
information then you can create for every of the 414,720 pixel
positions a streams with each 135,000 values. If a pixel value can be
0-1023 then there are 10 bits needed for a value so 168,750 bytes for
every stream so total round 70GB. If you can lossless compress every
stream 20 times then there is 3,5GB left. So it looks like an
impossible task to bring this down to 1KB.

But Jan Sloot did it, in worste case he used 2KB + 4MB + 4MB + 128KB =3D
8,13MB if he used all internal and external memory of his first
patented device to store and playback one DVD.

The 2KB temp memory he used to store a line of pixel values from a RGB
video signal and the 4MB temp memory he used to store 625 lines what's
one frame. This frame is stored in the other 4MB memory. Then he fill
again the 2KB temp memory with a line and again the 4MB temp memory
with 625 lines. Every time a line is equal as an early done line he
store not the line but a reference code of the line done before and he
do the same with a frame; every time a frame is equal as an early done
frame he do not store the frame but a reference code of the frame done
before. This comparing and assigning is done by the two CPU's one for
the line and one for the frame. So far I can follow it, by this way
all line and frame values are 100% unique.
Then he store this seconde frame also in the 4MB memory where he
already stored the first frame and he do that also for all the other
frames and when all frames are done the information in this 4MB memory
is stored in the external 128KB chipcard as last action. The playback
is the same but then in reverse.

If a line fit in 2KB then 625 lines=3D1 frame fit in 1250KB. Because
more then 3 frames do not fit in 4MB how he stored hundred thousand
frames in one 4MB memory? He leave us with a mystery.

He describe one other thing what he explain with a little example:

If we minimize a line to 16 pixels =3D 16 values between 0-1000 then we
can get for example:

200 500 100 400
090 750 640 360
190 360 870 100
850 030 390 920

If the next 16 pixels are:

220 520 120 420
110 770 660 380
210 380 890 120
870 050 410 940

Then he do not store this sequence but store the reference number of
the first one +20 this can also minus - or qual =3D
Probably he do this for full lines and frames (625 lines) to store
less because so he get more matches with early done lines and frames.

> I think it's clear that Sloot was just another huckster. =A0The
> different between him and Adam was that he was a tiny bit more clever
> to build a "prototype" piece of hardware to distract people from the
> fact that he was simply playing movies from another location.

It's impossible he played movies from other locations. First a metal
box act as a cage of faraday so you need an antenna what nobody saw
and some demonstations where done in gates villas with big gardens
with some distance between the road and house. Second his device was
demonstrated (at locations he didn't know) and opened (without his
permission) in the US without him. Third you need 16 TV tuners to
playback 16 movies at the same time, that's a lot of space and
bandwidth to send that from remote. Fourth he could jump to any point
in the movie instucted by the skeptic watchers. Fifth he demonstraded
live recording from a video camera of DVD/video or TV source and
playback for the chipcard. Six he took out the chipcard the movie
stopped and putting in an other showed an other movie. Seven also
after his dead the playback device was still playing videos from
chipcards. They say that the problem was that you could not create
chipcards, he hided that source code on a floppy. Jan Sloot said the
source code and description was laying in a safe, after his dead they
found a safe key but they never found the safe matching this key.

> A big indicator of snake oil is whether or not there are video records
> of these demonstrations. =A0In all cases, there are none.

There is at least one video made by a company where Adam Clark
demonstrated his technology, I have verified this fact. If there was a
camera for recording by some demonstration of Jan Sloot I assume that
that camera is also used for recording a demonstation. None of this
tapes is public available.
0
Sportman
5/24/2008 12:54:29 AM
On Mon, 19 May 2008 02:58:30 -0700 (PDT), Sportman
<sportman@gmail.com> wrote:

>Part of article USA TODAY 11/18/2008:
>
>There's the bizarre story of Jan Sloot, a Dutch engineer "who had no
>academic credentials and who had been for most of his life a
>television repair technician."

Not to mention a con man.
0
Chris
5/25/2008 12:02:09 AM
Reply: