Home > Archive > Compression > February 2008 > The Tolly Group independent tested MatrixView SWISH
You are viewing an archived Text-only version of the thread.
To view this thread in it's original format and/or if you want to reply to
this thread please [click here]
| Author |
The Tolly Group independent tested MatrixView SWISH
|
|
| Sportman 2008-02-01, 6:58 pm |
| MatrixView SWISH compared independent by The Tolly Group with GZIP and
no compression in web pages:
http://www.matrixview.com/TollySWISHReport_Jan2008.pdf
It's a nice idea to replace GZIP HTML compression but how they want to
arrange that?
They can give a decompressor free away so it can be added to Internet
Explorer (81,14% market share) and Mozilla/Firefox (13,81% market
share)
But then the problem start; maybe Microsoft is willing to pay for a
license but only 35,81% of the web servers use Internet Information
Services (IIS). I don't see how apache Software Foundation go to pay a
license bill to add it to their open source apache web server with
50,61% market.
Then there is the problem recent discuss here that hosting providers
don't like to switch on compression because it give higher CPU load so
they need more hardware (is more sy min's) and can only bill lower
bandwidth $ to their customers. Normal ISP's try to use as much
bandwidth as possible because how more they need how cheaper it become
and that make their price position better to ISP's who use less
bandwidth; price structure like this 1Mbit $50, 10Mbit $25 a Mbit,
100Mbit $10 a Mbit and 1000Mbit $5 a Mbit, etc.
Only if Microsoft buy an exclusive SWISH license to add it to Internet
Explorer (IE) and Internet Information Services (IIS) to use better
compression as selling argument to their not ISP customers (mainly
companies and websites with own servers) I see a little chance for
MatrixView to earn money with this technology.
Matrixview can better try to make a video compressor with better
compression and consume less CPU then the one used in Flash. I wonder
if MatrixView know that The Tolly Group also independent tested in
October 2003 Adam Clark; Adam's Platform 1.0 (AP) video compressor.
The Tolly Group tested between two Apple Power Mac G4 (400MHz
PowerPC) a raw video file 2,656,030,123 bytes streamed over 5Mbps,
1Mbps and 56Kbps with AP and MPEG-4. In all tests (5Mb/1Mb/56Kb) AP
used 731,172 bytes total network traffic and MPEG-4 60,075,469 bytes
for 5Mbps, 12,425721 bytes for 1Mbps and 7,251,087 bytes for 56Kbps.
The streaming video Quality for AP was in all three bandwidth
Excellent (Broadcast or DVD-like viewing experience) while MPEG-4
(using much more traffic) scored Excellent by 5 Mbps, Marginal (Severe
artifacts and/or jagged motion) by 1 Mbps and Poor (Unwatchable.
Severely degraded image with jagged or frozen motion) by 56 Kbps. This
report is later September 2004 removed from the The Tolly Group
website....
| |
| Jim Leonard 2008-02-01, 6:58 pm |
| Disclaimer: I run an internet-based business.
On Feb 1, 1:53 pm, Sportman <sport...@gmail.com> wrote:
> MatrixView SWISH compared independent by The Tolly Group with GZIP and
> no compression in web pages:http://www.matrixview.com/TollySWISHReport_Jan2008.pdf
>
> It's a nice idea to replace GZIP HTML compression but how they want to
> arrange that?
You missed the bigger picture: Most times in the report for gzip were
under 1.5 seconds, even on a 64kbps LAN. Most users don't notice a
1.0 second vs. 0.5 second difference. Unless they plan to make
whatever technology they stole from GPL'd code available for free,
there is no reason to pay for something nobody needs.
I'm amazed Tolly is still in business, let alone verifying
compression. The Adam's Platform debacle is the kind of thing that
crushes your reputation as a third-party verification service.
> Then there is the problem recent discuss here that hosting providers
> don't like to switch on compression because it give higher CPU load so
Most WAN accelerators are third-party black boxes that you integrate
into your datacenter. CPU time is a non-issue, as they don't run on a
general-purpose server.
And you're wrong, most hosting providers DO like to turn on
compression because they never have a problem selling bandwidth, they
have a problem running out of bandwidth due to overselling what they
have to their customers.
> they need more hardware (is more symin's) and can only bill lower
> bandwidth $ to their customers. Normal ISP's try to use as much
> bandwidth as possible because how more they need how cheaper it become
> and that make their price position better to ISP's who use less
> bandwidth; price structure like this 1Mbit $50, 10Mbit $25 a Mbit,
> 100Mbit $10 a Mbit and 1000Mbit $5 a Mbit, etc.
That is not how service tiers are priced. Please research with actual
providers before guessing.
> I see a little chance for
> MatrixView to earn money with this technology.
On that we can agree on.
> Matrixview can better try to make a video compressor with better
> compression and consume less CPU then the one used in Flash. I wonder
Flash 9 (might still be in beta) now supports H.264/AVC which is a
realistic upgrade in today's world. Users want decent quality at a
small size so that they can start watching immediately (no buffering)
and providers want small sizes to keep their bandwidth usage down.
CPU decoding complexity is not an issue considering that even the
cheapest computer you can buy for $300 at Walmart has nearly double
the performance necessary to decode for the typical 320x240 H.264 web
video.
> if MatrixView know that The Tolly Group also independent tested in
> October 2003 Adam Clark; Adam's Platform 1.0 (AP) video compressor.
I suppose it will take Adam himself coming to your door and telling
you to your face that it was a hoax. And even then, something tells
me you won't believe it.
> This
> report is later September 2004 removed from the The Tolly Group
> website....
Gee, I wonder why. It's a guv-urn-mental conspiracy!
| |
|
| On Fri, 1 Feb 2008 11:53:19 -0800 (PST), in comp.compression, Sportman
<sportman@gmail.com> wrote:
>MatrixView SWISH compared independent by The Tolly Group with GZIP and
>no compression in web pages:
>http://www.matrixview.com/TollySWISHReport_Jan2008.pdf
>
>It's a nice idea to replace GZIP HTML compression
Why? Seems to me like a solution in search of a problem.
Looking at the report, I see that their testing shows an average compression
ratio of 73% for GZip, and 85% for SWISH. That seems like an awfully
marginal improvement, once you figure that it's not standard and costs a lot
more money than GZip.
-Scott
| |
| earlcolby.pottinger@sympatico.ca 2008-02-03, 6:58 pm |
| On Feb 1, 9:45 pm, Jim Leonard <MobyGa...@gmail.com> wrote:
> Disclaimer: I run an internet-based business.
>
> On Feb 1, 1:53 pm, Sportman <sport...@gmail.com> wrote:
[color=darkred]
> I suppose it will take Adam himself coming to your door and telling
> you to your face that it was a hoax. And even then, something tells
> me you won't believe it.
[color=darkred]
> Gee, I wonder why. It's a guv-urn-mental conspiracy!
In Sportman's mind/ego it is easier to believe that thousands of
people around the entire world living/working with governments that
are actively hostile to each other are all working as one to hide this
secret than the to believe that one man was a con-artist and liar.
| |
| Sportman 2008-02-04, 6:58 pm |
| On Feb 3, 11:52=A0pm, earlcolby.pottin...@sympatico.ca wrote:
> In Sportman's mind/ego it is easier to believe that thousands of
> people around the entire world living/working with governments that
> are actively hostile to each other are all working as one to hide this
> secret than the to believe that one man was a con-artist and liar.
Maybe he know the answer:
Benjamin Schumacher is a U.S. theoretical physicist, working mostly in
the field of quantum information theory.
He discovered a way of interpreting quantum states as information. He
came up with a way of compressing the information in a state, and
storing the information in a smaller number of states. This is now
known as Schumacher compression. This was the quantum analog of
Shannon's noiseless coding theorem, and it helped to start the field
known as quantum information theory.
Schumacher is also credited with inventing the term qubit, which is to
quantum computation as a bit is to traditional computation.
He is the author of Physics in Spacetime, a textbook on Special
Relativity, and is also a professor at Kenyon College, a small liberal
arts college in rural Ohio.
http://physics.kenyon.edu/people/sc.../schumacher.htm
| |
| earlcolby.pottinger@sympatico.ca 2008-02-04, 10:03 pm |
| But standard computers/electronics do not store data in unique quantum
states.
Any time you find someone using the words 'data compression' and
'quantum' together *WITHOUT* a working quantum computer in the mix you
have another con-artist on the loose.
Why do you think he had to invent the word? It is meaningless on any
standard computer. And by the way have you priced the storage costs
of qubits? Does not match the original compression claims either.
Just because you wish so hard for it to be true, that will not make it
happen.
|
|
|
|
|