Many years ago, Adobe "encrypted" their Type 1 fonts. Adobe under considerable pressure finally released their format. In 1991, I wrote a decryption program, untype1, to extract outlines and posted it to comp.sources.misc. You can Google for it.
A couple of days after the post, I got a threatening cease and desist email from their legal department. I was amused and responded that, um, they had published the algorithm but that I would be willing to add any notice that they would write up.
Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof; or abridging the freedom of speech, or of the press; or the right of the people peaceably to assemble, and to petition the government for a redress of grievances.
Now exactly how is it that privately written blocking-software violates the first amendment? I use blocking software: my ISP blocks a large portion of spam, and more importantly, I use Slashdot Preferences. For example, I block Jon Katz.
znu is right. QuickTime is a full-blown media architecture (API, file format, codecs, plug-in interfaces for effects and hardware, and the list goes on and on) and porting it to Linux would require dealing with a lot of MacOS legacy issues. Blech.
A much better idea is to ask yourself, in the year 2000, what is it that you really want? Remember, QuickTime has evolved over the last 10 years and many of its original assumptions are not as important today. So, do you want an MPEG stack? Do you want streaming? Do you want to develop video post-production apps? Do you want video conferencing?
QuickTime is a huge system that Apple has to work hard at maintaining on their own OS. You're much better off looking at it as a menu and picking the top 3 or 4 features and working on (or asking for) those.
First, if Transmeta deserves a patent for a JIT, where there is considerable prior art, then this guy who has *improved* the state of the art, deserves one.
First, lets clear up some misinformation: JPEG and MP3 both use the Discrete Cosine Transform (DCT). They don't use the Discrete Fourier Transform (DFT). The elegant DCT is an inexpensive approximation to the holy grail, the Karhunen Loeve Transform. And it is a very good approximation at that.
MPEG-2 goes even further by having quantizer scale factors on a macroblock basis. This allows a coder to use more bits on the subjectively more important part of a picture. Read this as locality.
Wavelets are just a transform. That's all. The zerotree, which its inventor Shapiro says is very similar to the humble EOB symbol in JPEG, has been applied to DCTs just as easily, with trivial results. The same can be said for SPIHT.
Wavelets have some advantage in reducing blocking artifacts at high compression ratios, perhaps for video conferencing. But even at that they are on the wrong side of history. Bandwidth is going up not down.
OK, by now you've gotten the idea that I've got a bad attitude about wavelets. This is true. I was at a seminar at UCSC a few years back when a wet behind the ears grad student said that in five years time, wavelets would dominate the market. Oh, I'm sorry, that was five years ago. Time flys.
PNG is necessary. JPEG2000 is completely unnecessary.
Provided you reverse engineered it lawfully, it is no longer a trade secret. You can't burgle a factory, and there are issues with hiring trusted employees. But other than that they have to protect their trade secret.
However, they can have trade secrets, patents, copyrights and trademarks all at the same time:
Copyrights on the media Patents on the DVD CSS technology Trademark on DVD Trade Secret on stuff I don't know about (yet)
When Linus says, "remember: a compiler is a TOOL" people nod in agreement. When he says that if a job's criteria can't be met by sticking within a standard, the compliance to the standard should suffer, not the criteria. And people murmur their approval.
And Linux is completely dependent on gcc and very popular and useful and portable.
Now along comes Sun shadowed by the evil Microsoft. Frankly, Sun (and Microsoft) are in it to make a buck.
Java is very public and portable and open. But Sun was in a dogfight with Microsoft and saw the mire of the Standards Process as a way of shaking them. And it sort of worked.
Now when Sun's job criteria can't be met by sticking within a Standards Process, the compliance to the Standards Process should suffer, not the criteria.
Pragmatic sauce for the goose is pragmatic sauce for the gander.
XFree86 4.0 is being delayed two whole months. I'll bet it is also losing market share in the corporate segment as well. That's it. This must be the end of X. Game over.
Actually, I can't even remember what got released two months ago. XFree86 4.0 will come out. It will be good. Mozilla will come out. It will be good.
Slashdot Mirror
I always thought that the name shell was derived from command language:
command language -> cl -> shell
Many years ago, Adobe "encrypted" their Type 1 fonts. Adobe under considerable pressure finally released their format. In 1991, I wrote a decryption program, untype1, to extract outlines and posted it to comp.sources.misc. You can Google for it.
A couple of days after the post, I got a threatening cease and desist email from their legal department. I was amused and responded that, um, they had published the algorithm but that I would be willing to add any notice that they would write up.
I never heard from them again.
neither gzip nor bzip2 use arithmetic coding. Why?
Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof; or abridging the freedom of speech, or of the press; or the right of the people peaceably to assemble, and to petition the government for a redress of grievances.
Now exactly how is it that privately written blocking-software violates the first amendment? I use blocking software: my ISP blocks a large portion of spam, and more importantly, I use Slashdot Preferences. For example, I block Jon Katz.
I always read people calling BSD a cathedral model and Linux the bazaar model. I really do think you have it backwards.
BSD has forked. There is no single, one true BSD. I don't think you get more bazaar than that.
Linux on the other hand, has Linus on high, with his archangel Alan. It is true that you can send a patch, but I think a prayer has more of a chance.
I think Linux started out as a bazaar and has changed into a cathedral. And think that BSD started out as a cathedral and has changed into a bazaar.
znu is right. QuickTime is a full-blown media architecture (API, file format, codecs, plug-in interfaces for effects and hardware, and the list goes on and on) and porting it to Linux would require dealing with a lot of MacOS legacy issues. Blech.
A much better idea is to ask yourself, in the year 2000, what is it that you really want? Remember, QuickTime has evolved over the last 10 years and many of its original assumptions are not as important today. So, do you want an MPEG stack? Do you want streaming? Do you want to develop video post-production apps? Do you want video conferencing?
QuickTime is a huge system that Apple has to work hard at maintaining on their own OS. You're much better off looking at it as a menu and picking the top 3 or 4 features and working on (or asking for) those.
First, if Transmeta deserves a patent for a JIT, where there is considerable prior art, then this guy who has *improved* the state of the art, deserves one.
Hats off to an original idea.
First, lets clear up some misinformation: JPEG and MP3 both use the Discrete Cosine Transform (DCT). They don't use the Discrete Fourier Transform (DFT). The elegant DCT is an inexpensive approximation to the holy grail, the Karhunen Loeve Transform. And it is a very good approximation at that.
MPEG-2 goes even further by having quantizer scale factors on a macroblock basis. This allows a coder to use more bits on the subjectively more important part of a picture. Read this as locality.
Wavelets are just a transform. That's all. The zerotree, which its inventor Shapiro says is very similar to the humble EOB symbol in JPEG, has been applied to DCTs just as easily, with trivial results. The same can be said for SPIHT.
Wavelets have some advantage in reducing blocking artifacts at high compression ratios, perhaps for video conferencing. But even at that they are on the wrong side of history. Bandwidth is going up not down.
OK, by now you've gotten the idea that I've got a bad attitude about wavelets. This is true. I was at a seminar at UCSC a few years back when a wet behind the ears grad student said that in five years time, wavelets would dominate the market. Oh, I'm sorry, that was five years ago. Time flys.
PNG is necessary. JPEG2000 is completely unnecessary.
Provided you reverse engineered it lawfully,
it is no longer a trade secret. You can't
burgle a factory, and there are issues with
hiring trusted employees. But other than
that they have to protect their trade secret.
However, they can have trade secrets, patents,
copyrights and trademarks all at the same time:
Copyrights on the media
Patents on the DVD CSS technology
Trademark on DVD
Trade Secret on stuff I don't know about (yet)
Standards. Don't ya just love 'em.
When Linus says, "remember: a compiler is a TOOL"
people nod in agreement. When he says that if a
job's criteria can't be met by sticking within a
standard, the compliance to the standard should
suffer, not the criteria. And people murmur
their approval.
And Linux is completely dependent on gcc and
very popular and useful and portable.
Now along comes Sun shadowed by the evil
Microsoft. Frankly, Sun (and Microsoft) are
in it to make a buck.
Java is very public and portable and open.
But Sun was in a dogfight with Microsoft
and saw the mire of the Standards Process
as a way of shaking them. And it sort of worked.
Now when Sun's job criteria can't be met by
sticking within a Standards Process, the
compliance to the Standards Process should
suffer, not the criteria.
Pragmatic sauce for the goose
is pragmatic sauce for the gander.
XFree86 4.0 is being delayed two whole months.
I'll bet it is also losing market share in the
corporate segment as well. That's it. This
must be the end of X. Game over.
Actually, I can't even remember what got released
two months ago. XFree86 4.0 will come out.
It will be good. Mozilla will come out.
It will be good.