Slashdot Log In
Suddenly a JPEG Patent and Licensing Fee
Posted by
CmdrTaco
on Thu Jul 18, 2002 12:06 PM
from the where-have-I-heard-this-before. dept.
from the where-have-I-heard-this-before. dept.
Michael Long writes "Forgent Networks (www.forgentnetworks.com) has announced that it owns the software patent on JPEG compression technology, and has stated that it is "in contact" with computer, software, camera, and other digital imaging product manufacturers regarding licensing terms. This ambush of the digitial imaging industry will probably stand as the worst public relations nightmare a company can inflict upon itself."
This discussion has been archived.
No new comments can be posted.
Suddenly a JPEG Patent and Licensing Fee
|
Log In/Create an Account
| Top
| 1223 comments
(Spill at 50!) | Index Only
| Search Discussion
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
Sorry, But... (Score:4, Funny)
In the same vein, I was going to patent making claims about patents on
Or, hey, how about a patent on claiming prior use exemptions on a patent? Wouldn't this allow a corporation to patent anything and make money on either side of the patent fight? Oh, shit, now I've done it...
Virg
Read and discover they have no claim (Score:4, Informative)
First, the patent talks about 2 encoding schemes and applying them to various scenarios.
A) Run-length encoding the amplitude of digitally sampled signal. An idea older than time, but not used in JPEG, so who cares.
B) Huffman encoding the amplitude of a digitally sampled signal. David Huffman (at latest) came up with the encoding scheme in 1953 (basing off him being in grad school when making it and age at death [ucsc.edu]), so I think we can establish prior art.
But the real issue is JPEG, which is the lossy end of the coding scheme. This involves (excuse my math) a Discrete Cosine Transformation to translate the amplitues into the coefficients of the frequencies being encoded.
Huffman encoding doesn't come in until the lossless compression stage, which is technically not JPEG, but JFIF, the file system wrapped around the JPEG encoding scheme that makes JPEG encodning into a JPEG file we all know and love... a minor distinction, but again, any monkey can show prior art.
It's a good thing they can't do that.. (Score:4, Funny)
Let's Return It! (Score:5, Funny)
Send it all back to Forgent. Email a few to each employee.
(begin letter)
Dear Forgent Employee,
Attached is some of my porn collection. I am returning it to your company because I inadvertantly used your patented JPEG format.
The rest is still to follow. My printer is slow. Playmate Debby especially requires a lot of ink because of her unorthodox techniques and tools, as you can clearly see in image #4057.
Thank You for your patience and understanding,
[Slashdot User]"
(end letter)
They should do well with this... (Score:5, Funny)
Re:They should do well with this... (Score:4, Informative)
Re:They should do well with this... (Score:5, Interesting)
Re:They should do well with this... (Score:5, Funny)
Re:They should do well with this... (Score:4, Funny)
or
Re:They should do well with this... (Score:5, Informative)
The important part:
--xPhase
Re:They should do well with this... (Score:5, Interesting)
PNG is a royalty free community owned format, similar to Ogg Vorbis...
There is a lot of misunderstanding in the open source community as to how patents work. The claims made by Ogg Vorbis (i.e. it is patent free) are extremely unlikely to be true. Similarly, it is unlikely that the PNG format is not patented by someone.
The problem is that people tend to think of patents in much the same way that they think of copyrights. With copyrights, if a developer creates something without reference to the work of others, that developer is free and clear of other's copyrights, and can make it freely available. Not so with patents. A developer may create a new technology (PNG, Ogg Vorbis, etc), and that developer may choose to not patent it, but that technology is not free and clear of patents unless nobody has patented anything that is used in any part of the technology. If any part of your "new" idea has been thought of before, you're not clear of patent issues. Given the sheer number of software patents being filed and issued, given the incredibly broad claims that are being allowed, and given the fact that you don't have access to what patents are pending in the patent office (generally for a few years) just waiting to pop up, nobody can back up a statement such as, "I developed this, and it is patent-free."
I truly wish it were otherwise. As a former patent attorney, I have been watching the coming train wreck for a while now. It is only a matter of time before major chunks of what the open source community relies on turns out to be patented and owned by non-too-friendly people.
-Steve
Re:They should do well with this... (Score:5, Informative)
I'm not a lawyer, but I was one of the people in the working group that developed PNG. (I'm credited in the spec under the name Robert Poole, although I don't think they updated my contact info recently.) PNG uses the same compression scheme used in GNU gzip, and that scheme was chosen specifically because it had been well researched and found to not conflict with any current patents. It also gives fairly decent performance and compression ratios for highly entropic data.
That's not to say that some other aspect of the PNG spec won't come under fire -- the file format is similar enough to TIFF and the Amiga's IFF/ILBM that if there are some core patents on tagged file formats, we could be in trouble. But that's doubtful, since prior art would probably play a role in any defense against such a patent assault. Bottom line -- if PNG comes under fire, the FSF lawyers would be all over the situation.
Re:They should do well with this... (Score:5, Insightful)
Well, they have done what they can to make it more likely. Specifically, they have a staff of lawyers scrutinizing everything they do, specifically to make sure they don't run afoul of any patents. They would have been done by now if it weren't for the care they are taking about patents.
It's ironic: patents are supposed to spur innovation onward, but at the moment patents are a huge drag on the development of new software. If you want to make sure you don't get bitten by any patents, you need to go to a great deal of effort.
steveha
they've done more research than that (Score:5, Informative)
Re:They should do well with this... (Score:5, Informative)
Why?
JPG is a lossy encoding mechanism. It disacrds a significant amount of information in any given image to create smaller file size.
PNG is a lossless encoding mechanism. It uses several very intelligently designed formulas and structures to very efficiently encode an image to reduce its file-size without losing any image data.
Because of this difference, PNG files of all but the simplest images will *always* be larger than corresponding JPG files.
For simple graphics like logos, stylized text, and flat-shaded cartoons, PNG can be made to produce better looking images at lower filesize than JPG or even GIF. For this reason, PNG is idea for making simple graphics for websites such as blocks of color, logos, etc. For photographic or shaded images of any kind, JPG is simply better at producing better image quality at smaller filesizes.
Now, if you're on any kind of broadband connection, that point becomes pretty moot since the difference between downloading a 10k jpeg and a 100k PNG is less than a second. On modem connections, moving to all PNG would make the internet completely void of all but the simplest graphics.
IMHO, it's time to build a lossy format for storing graphics similar to Ogg Vorbis. Perhaps the video codec Ogg just released can be used to make reasonable single-framed movies? Anyone familiar with the format care to comment?
Re:They should do well with this... (Score:5, Insightful)
Gee, it shows you're not running a server.
Saving 90% bandwidth is a god-given when you do.
Re:Didn't apple try this? (Score:5, Informative)
http://www.gnu.org/philosophy/gif.html [gnu.org]
Apple is strangling Firewire adoption (IEEE 1394) with patent royalty fees.
Re:Didn't apple try this? (Score:5, Informative)
This is the oldest myth in the book, and one of the most oft repeated I see. .25 per unit is 'strangling' the adoption rate.
While Apple helped develop the FireWire spec, it doesn't collect all the licensing fees. The licensing fees go to the IEEE1394 consortium (of which Apple's a member) and it's the consortium that decides the division of the licensing monies on a patent-by-patent basis.
Also, I wouldn't say that
Re:Pantent? (Score:5, Insightful)
Misleading examples, anyone?
The judgement in the "coffee case", Liebeck v. McDonald's [atlanet.org], followed after over 700 other cases between 1982 and 1992 in which a McDonald's customer was burned by overheated coffee. Coffee is usually served around 140 degrees Fahrenheit; McDonald's was serving it at over 180. A liquid at 180 degrees F. will cause third-degree burns to human skin in between two and seven seconds. (A "third-degree burn" does not refer to the skin being burned away, but to the full thickness of the skin being burnt.) Coffee at 180 degrees is not fit for consumption, as it will severely burn the mouth and throat.
Stella Liebeck did not set out to mooch millions of dollars from McDonald's. She initially wanted a settlement of $20,000 to cover her medical costs -- which included eight days in the hospital and skin-grafting operations. A jury awarded her the $2.7 million dollars in punitive damages -- to punish McDonald's for knowingly continuing to put its customers in harm's way. The judge reduced punitive damages to $480,000 despite calling the company "reckless, callous, and willful" in its deliberate risking of customers' well-being in order to save costs.
See the link above for details. If you want to say that our society is too litigious, go ahead -- it is -- but please do not Ms. Liebeck for that. She was the victim of another of our society's problems -- corporations who believe it will be cheaper to pay off (or toss aside) victims of their recklessness rather than do the right thing in the first place.
Re:Coffee (Score:5, Funny)
That's quite true; one should make tea with boiling water, unless it is Chinese tea in which case one makes it with water around 180 degrees F. However, one does not serve it to one's guests at that temperature, since it loses some heat while steeping or brewing. One never leaves tea or coffee on a heater for hours, maintaining its temperature at 180 F until the moment of service; the subtle aromatics of either beverage will quickly evaporate, leaving a soulless and bitter brew.
Moreover, in proper society one does not serve tea or coffee in heat-insulating styrofoam cups. One serves both in china, which does retain heat but not quite as well as styrofoam. (It is because china takes on and dissipates some of the heat that teacups have handles whereas foam cups do not.)
One also serves coffee at table in an open cup, so one's guest can add milk or other adulterants. One does not expect one's guest to remove a tightly fitting lid first, nor to perform said operation without the stability and protection of a table. Presenting such a puzzle to one's guest -- especially a puzzle loaded with the gory surprise of a near-boiling liquid within, ready to scald the loser in this hideous parlor-game -- is beyond the pale of hospitality.
Thus, the standards of proper society for the preparation and serving of tea and coffee do not form a defense for McDonald's in this case.
Re:Pantent? (Score:5, Interesting)
Maeryk
Re:What have they actually patented? (Score:5, Informative)
Re:Pantent? (Score:5, Insightful)
If I don't want you walking in my back yard, I have to stop you within a reasonable amount of time and cannot wait until the path becomes a well worn and commonly accepted walkway.
At least that is what my business law instructor pounded into my head.
It seems like the same principle will kill this patent. They did not act within a reasonable time to mitigate the damages done to other companies who used the patent.
Re:Pantent? (Score:4, Funny)