Slashdot Mirror
U.S. Bans Some Cellphones For Patent Reasons
runner_one writes "According to the New York Times, A federal agency has banned imports of new cellphones made with Qualcomm semiconductors because the chips violate a patent held by Broadcom. The International Trade Commission said today that the import ban would not apply to mobile phone models that were imported on or before June 7." Update: 06/08 13:05 GMT by KD : Glenn Fleishman notes that Apple's iPhone will be allowed into the country, since it doesn't use any 3G chips. He adds that Apple "might have the most advanced smartphone on the market unless President Bush or his trade representative overturn the ruling (which they have the power to do)."
Play with patent fire and you're going to get burned. Remember Qualcomm suing Nokia?, Qualcomm suing GTE Wireless, Qualcomm suing Maxim, Qualcomm suing Motorolla, Qualcomm suing Ericsson, Qualcomm suing Broadcom?
Everytime a large corporation loses a big case like this, I feel we're a step closer to sane patent reform. Hopefully someone will win a patent against Broadcomm next.
There are shills on slashdot. Apparently, I'm one of them.
Every time they issue a patent that's later invalidated, they should pay compensation for issuing the patent.
The problem here, has and always will be the over willingness of the patent office to issue patents when the invention preexists but is not documented publicly, or where it's a minor increment of an existing variation. It's in the law that they have to test for obviousness and prior art, but they so narrowly define those terms as to remove the tests.
The free market will fix it, make them pay for their mistakes just like every other professional body.
There's also the new (pending?) revisions from SCOTUS on the "obviousness" clause that changes the meaning (I hope) to "if the new 'invention' is simply a combination of existing components that does what you would expect by combining those components, it doesn't count as a patentable invention."
One-click shouldn't be a patent, because it simply chains together existing components in a way that results in the simple sums of functionality. There is nothing "new" there.
A patent on a chip that simply sums together signal detection with wake-up logic to save batteries should also not be patentable, because both of those technologies have been around for a long time.
It's only when the sum of components does not simply result in an obvious function that a patent should be granted. My test would be something like: strip the patent application of the description of what it does and the submitter and give it to a group of engineers in that field. If that small group can't determine what the invention is supposed to do within, say, 8 hours (or some other 'reasonable' time frame), then it can get a patent. Otherwise, no patent.
The problem is that people are now using combinations of preexisting ideas as patents when they shouldn't be allowed. I wouldn't even grant a patent on the optical mouse, because it's an "obvious" combination of an optical motion sensor and a computer pointing device. Now, I might patent something related to how the optical motion sensor works, but only so long as it's not simply a combination of existing components.
"There are a dozen opinions on a matter until you know the truth. Then there is only one." - CS Lewis (paraprhase)
Qualcomm has been aggressive in promoting their patent rights. But to say
that they are unique in this field is completely ignoring one side of the story.
Every try to make something in the GSM/UMTS space? You will have about a dozen companies
approach you with their hands out. Nokia, Qualcomm, Ericcson, Motorola, Lucent, Samsung
and several others. CDMA royalties are about 5%, almost all paid to Qualcomm. GSM/UMTS
royalties sum up to about 18%. The only difference is that if you are one of the big guys,
you "cross-license" your patents so that you don't end up actually paying anything. If you
are a new entrant... well, you are out of luck.
What is pissing off the Nokias, Ericcsons and Broadcomms of the world is that in the CDMA space,
they have no patents at all. None. That is because they fought CDMA every step of the way until Qualcomm
demonstrated conclusively that it is a commercially practical technology. Then they
turned around and tried to claim it as their own, and tried to co-opt it by applying it with minor
modifications to the UMTS space.
Actually, I am amazed that Broadcomm is getting away with this. Their sum total of contributions to
the wireless space is close to zero. They have done some work in the wireline world in the early
years, but they have contributed zip, zilch, nada in the development of wireless IP. They are not
even a name in the industry. HOwever, it is possible that they have purchased some IPR of late.
I am quite happy to see cracks in the patent edifice as a whole, but making Qualcomm the villain in
this is not correct. Qualcomm laid many of the foundations for modern wireless communications technology;
Qualcomm corporate R&D is about as close as you can get to how Bell Labs used to be.
Lots of Qualcomm's IPR consists of non-trivial, non-obvious, fundamental contributions to communications
theory. Most other wireless companies, in particular Nokia, Motorola, Ericcson etc have done nothing
fundamental in the past 15 years. They are product companies whose forte consists of taking old technologies
and packaging them in crowd-pleasing form factors, or (in the case of Ericcson), maintaining relationships
with behemoth carriers.
Magnus
Why did this get rated troll? In all seriousness, Broadcom refuses to write drivers or help or anything to get their wireless cards working natively in Linux. They do not care. Currently the best way to get the BCM43xx chipsets working is to use bcm43xx-fwcutter, but that of course violates patents because it uses the actual firmware of the wireless card to work.