Patently-O Bits and Bytes No. 7

• Patent Reform: The Senate has now placed the Patent Reform Act (S.1145) on the senate Calendar. [Link].  “Minority views” have been filed but are not yet available.
• BPAI Composition: The CAFC has denied an en banc request in the Translogic case. That case challenged the legality of past BPAI appointment practices. [See Duffy]
• Harmonization & Work Sharing: The PTO & JPO have announced a new bit of work-sharing cooperation. Under the agreement, a filing in one patent office would be deemed a filing in the partner office as well.  If successful, this proposal could eventually replace the PCT system. [LINK]
• UK Patenting Computer Programs: A UK Court has now ordered the UK PTO to stop rejecting patents simply because they claim computer programs. [IPKAT]
• Valuing Patents: Accountants don’t know how to place a value on patents. The solution appears to be an arbitrary standard. [LINK]

 

Comments

[MaxDrei]

So, Europe allows patents on programs on a carrier, and on signals, but ONLY when the innovation falls in a TRIPS "field of technology". Seems coherent to me. How about it, as the clear frontier of patentability? Dispute it, anybody?

[Michael R. Thomas]

The concept of passage without major work is tragic for americas future. Without ajustments to inventor fairness other countries will jump ahead in attracting inventor,s to their country for manufacturing and central office employments.

[curious]

http://www.csmonitor.com/2008/0128/p09s02-coop.html

Proposed patent reforms mean less protection for the underdog.

[ironicslip]

driedmax :: care to play the feud?

200 people surveyed top 3 answers on the board:
whispers and squeals ... "good answer good answer"

1) in re Nuijten was about digital watermarks .. i think the court missed the boat and the ocean on which they sailed X X X ... whatever your interest in that technology watermarks act as audit trails or traceability elements in a signal ... in fact they are integral with the signal ... so is it fair to call it protection? well, not necessarily ... if you can make a copy so can the pirate ... so the digital watermark is what exactly? noise? or does it have some tangible, novel and nonobvious benefit? Is that why TRIPS sails that way?

2) we have DMCA -- 17 USC 1201 a2 relates to circumvention of copyright protection "technologies" -- namely anyone who is skilled to get around the copyright protections is in violatiuon and considered an "anti-circumvention device" (forget about research and academia -- remember SDMI -- Felten -- Blue Spike and Adobe and door clickers and and and ... all DMCA assertions) -- on top of all the other statutory protections for copyright -- this was the real dinger just in case all of that stuff failed (ironic that Rep Boucher wanted fair use incorporated in the Act but did not get it and here he is again introducing limits on patents for things like tax schemes) ...

ironically, watermarks were not thought to relate to anti-copying (even in light of the whole SDMI skididdle)-- only DRM (file wrappers or extensions or access controls) were heartily considered ... 10 years on ... how 'bout them DRM systems ... what is left is anti-circumvention to appease the copyright cabal ... but no wrappers for the content and no business model to be patented -- sincethe DRM patents even in all of those purchases clearly do not stop the copying

what to do what to do ... Richard Dawson turns with that glean in his eye and that big pinky ring and says ...

3) WALLA ... TRIPS ... directed more to copyright protection than patent protection but therein lies the irony :: patented technologies are essentially given a lesser value than the copyrighted material they protect ... without the watermarks the copyright owner has a very difficult time trying to measure things like piracy (in lieu of privacy -- which is more balanced in the US and is obviously going to meet a certain stand-off with wiretaps and packet tracing -- another form of traceability) ... with watermarks it is possible to show the receipt of the copy of the content and trace unauthorized sources ... and, perhaps, just maybe move the world of copyright into a cleaner set of rules such as -- anti-circumvention ... vis-a-vis DMCA ... copyright laws ... TRIPS ...

simple question ... if digital watermarks enable a new business model or simply save the existing players in the space (including Rep. Berman and the rest of the folks in favor of Patent Reform ... CFP) ... who should determine the value of the "intellectual property"? -0- is the watermark the lien or the receipt? what is the property being protected? is the copyright the point at which the watermark cannot be retreived? or what exactly? is that why TRIPS says yes and the US Court says no to signals being protected?

does anyone know how to measure intellectual property fairly? (besides simply saying do not circumvent)

DING DING DING

thanks for playing ...

[MaxDrei]

Oh dear, Ironic Slip, Michael Thomas, (Not the same person, are you?). I don't understand a word of what you write. Sorry. My point was that the proper protection of computer-implemented inventions should permit claims to software on a carrier, and signal claims, but not claims to business methods or mathematical procedures, as such. I was hoping that somebody (like Aharonian perhaps) would challenge that proposition.

[ironicslip]

"My point was that the proper protection of computer-implemented inventions should permit claims to software on a carrier, and signal claims, but not claims to business methods or mathematical procedures, as such. I was hoping that somebody (like Aharonian perhaps) would challenge that proposition"

i cannot speak for greg aharonian ...

let me offer the following example of how i am not clear on your question:

1) do you believe something like RSA should have been patented -- likely its "first major commercial use" was the White House to protect internal transmissions -- beepers, cellphones, etc.

2) RSA is a mathematical procedure based on a complex but "unprovable problem" (in a mathematical procedure sense) with certain properties that solve specific "business", security, privacy issues ...

examples: it made it possible to protect credit card transactions instead of those weekly books the card companies would send to merchants to check valid credit cards & eliminated the need for human couriers of money between banks ... would you consider RSA to be "signal related" but not necessarily fixed to a carrier?

if so,

3) how are you separating the invention of RSA from the business[es] that the invention enables-[ed]? or, at the very least, the cost of doing business that were reduced as a result of the invention?

where is the value to be protected? by what standard?

[MaxDrei]

Business uses technology. Innovation in technology ought to be patentable, whereas innovation in, say, accountancy, ought not to be. No idea what RSA is, but from your description it strikes me it brings a contribution to the technical field of telecommunication. In that case, there's room for valid issued claims that protect it.

[big hairy rat]

ironicslip:

RSA is an interesting example for the patentability of software. RSA is one of the few software inventions that I would concede might be so important and novel that they should be patentable. (Conjugate gradient, the simplex method, Karmakar's method and spreadsheets are other examples that I consider to have been sufficiently important and novel at the time they were introduced to warrant patent protection.)

On the other hand the RSA algorithm was invented (discovered?, developed?) independently both before and after the RSA team did its work. The high rate of independent discovery in the software field is a major reason why software patents are so contentious.

Other reasons why software patents are contentious include the broad reach and poor quality of many (most?) software patents.

[ironicslip]

Dried Max: If you dont know RSA ... then maybe look at it a little closer ... the TRIPs issue should be interpreted in the context of WTO / copyright protection -- my rant above was pointing out how quickly you can do institutional change if the real bullies of intangibles (namely, copyright) push for something ...

if you could patent software that did prevent copying (a la DRM) would that be reasonable? it is a business de facto, right? you stop copies before you are paid ... alas DRM has not worked, and in theory is impossible ("whatever i can see cant be taken from me" -- a phish song lyric) but many patents were issued with languiage such as "secure" "copy prevention" -- is this language appropriate if NOT possible -- even in theory? perpetual motion, anyone?

Big Hairy Rat : what was done in the UK is not completely relevant to the argument i present and perhaps you are misinterpreting that RSA is based on prime factorization though other public key algorithms are based on other hard problems (ie if one way functions do exist than public key exists ... even though there is no settled proof that this can be proven in a "hard" sense)

... but we can look at other problems in crypto that did come to blows in very significant ways (Micali's work and Schnorr's signature scheme versus Kravitz's DSA) ... so, change RSA to ECC (elliptic curve cryptography) ... for the example as the NSA appears to concede it was independently invented and i cant find a better example at moment ;-)

my issue with trying to carve out a "software patent" or a "business method patent" is that money being a tangible concrete result ... those methods that make it possible to get paid ... must be considered ... at the very least ... there are several patents that are as significant as RSA and that type of technology that are increasingly viewed this way ... ECC being the replacement potentially for RSA in many applications ... digital watermarking for auditability and traceability ... etc.

a foundational aspect of cryptography is akin to hypothesis testing in evaluating whether something is truly novel and nonobvious -- Kerckhoff's "Law"/"Principle"

assume the enemy knows the cipher and put all of the security in the key ... so, assume computers can do anything that can be defined in a function and except all variants that are obvious or anticipated ... from there, as with open source software ... it is very possible to create a system to evaluate the patentabilioty of code within some reasonable analysis akin to public comment and analysis of ciphers being tested by unknown persons ...

how to do this and how to ensure proper attribution is the issue that is more troubling than simply asserting that all software cannot be patented or all business methoids cannot be patented ...

quality as with harley davidsons is a different aspect of analysis, imho ... we have plenty of crappy trade secrets for software that do very real damage to the real world ... not to name names start with diebold's security or even breathalyzer source code and think of the benefits if this type of software had to be patented or subjected to reverse engineering in the interests of public good

(most of it has been subject to oracle attacks -- asking with an input for an output to see what the function looks like -- but the pesky dmca clouds any ability to post results or work acadmeically through sensitive issues)