It's unfortunate that Lemley doesn't offer any examples of IT inventions meriting patent protection.

My dear, this morning you critiqued a patent claim for its wordiness, and analogized the construction language rather than patent searches and licensing negotations to transactional costs in real property!!! Critiquing the novelty or non-obviousness of a patent is one thing, but you're stretching it by calling the time to read a patent claim transaction costs hampering the software industry...

Ok. Read this paper Lemley wrote with Kimberly Moore. They looked at a couple patents- Valuable Patents (July 2003).:

***We examined data on every patent that issued between 1963 and 1999 – 2,925,537 patents in all – and every patent lawsuit that terminated during 1999-2000 – 4,247 different cases covering 6,861 patents.***

That paper isn't specific to the IT industry, or to software patents. And figuring out which patents are valuable isn't necessarily related to figuring out which ones are worthwhile in a public policy sense. The NTP wireless email patent was worth more than half a billion dollars, but hardly anyone would claim that patent ought to have been approved.

Some hold the view that at least several more NTP patents would have been rescinded by the USPTO had RIM not been made an example of by a judge intent on inducing future patent litigants to settle.

This doesn't change my take on your approach which criticizes claim construction more than the merits of a patent. Are there patents sitting in a vacuum somewhere? You take as a per se indication of a patent's merit its claim construction without regarding its implementation, USPTO re-examination and court interpretation. You don't seem to consider that these factors affect, perhaps equally or more to the USPTO, what is addressed by public policy.

Based on this reasoning, you probably confuse "technological superiority" and market viability.

IBM has the most patents and generates the most in patent royalties in the industry. Is IBM using funny money from bad patents to float their open source business?

I see open source extending its motto from "we have close to marginal transaction costs and free source code" to "everything else must be free too since we'd rather claim ourselves as merely hobbyists and eliminate patents for R&D intensive companies than hire patent lawyers, and if these patent holders say anything we'll just reply that we make a superior product for next to free."

One important case that it seems to me these ten points do not address is the assertion of a patent after a standard has been agreed and implemented, by a company which purposely did /not/ take part in the standards setting process.

There have been a number of notorious recent examples of such post-facto "standards ambush"; and several software standards, that went through extensive work and negotiation in the standards-setting process with the intention of creating a freely-reimplentable final standard, that remain crippled to this day by subsequent patent assertions.

It seems to me that there is a strong case for legislating, at least with respect to the most prominent national and international SSOs, that if any entity does not declare relevant IP when a standard is being adopted, the entity should subsequently be estopped from asserting that IP to the extent that the IP is used and required in order to achieve a conforming implementation of the standard.

"Having previously described patents as the system that has “fostered American growth and entrepreneurship,” Lemley is careful to point out that abolishing patents for the technology industries as neither a viable nor desired step."

We are not advocating abolishing patents for the technology industries. We are advocating abolishing software patents only. Abolishing software patents is certainly viable. The software industry flourished for years with only copyright. We can flourish quit nicely without software patents.

"if what you’re trying to say as a standard setting organization is, “We don’t want to pay inventors anything for their technology,” I think you’re going too far."

There is no conflict between copyright and standards.
-------------------
Steve Stites

Professor Lemley is apparently unaware of the well-known fact that the first US patent was granted for a process of making potash - not a machine.

Too bad for him :)

I am wondering where he gets money for his "research" on patents - I wouldn't be too suprised if some of that funding comes from the biggest high-tech infringers, like MS, HP, Intel, Cisco etc. etc. - the biggest abusers of the US patent system and worst enemies of true invention.

Hypocrisy is the right word for such people

Professor Lemley sits on the advisory committee for EFF, and previously consulted for the DOJ in the US Microsoft antitrust case.

JH, check out this article: Overlapping Patent Exposure Can Deter Holdouts, by Doug Lichtman (May 2006). www.pff.org/issues-pubs/ip/bulletins/bulletin1.3patent.pdf

But how does it play in Paraguay?

"1. SSOs: impose RAND licensing terms."

OK this is Microsoft's wet dream.

Reality check: FOSS is here and RAND is not compatible with FOSS.

So explain why should standards be implemented which disallow FOSS (hint to the answer: in every country except the USA, Microsoft is a foreign corporation)

On page 5, Lemley states that although he's not a fan of royalty free licensing where SSO members forgo patent rights, "there are limited circumstances in which royalty free is appropriate," such as when a consortium works on open source.

I note a distinct lack of examples of good software patents from software patent supporters. I wonder if this is a function of the dirth of programmers among those who support them and/or because there really aren't any. I'm willing to go more with the latter, though the former plays into it.

I've seen some of the wizbang new languages that come out of Microsoft Research. They might seem innovative to someone at the PFF, but they aren't anything more than just interesting to a programmer. A judge, lacking programming experience, might not understand that C-Omega, a variant of C#, is actually just a minor evolution of C# rather than anything worthy of patent protection. C-Omega and F# are cool languages, for example, but nothing sufficiently novel that Microsoft deserves anything more than a pat on the head for making them. I don't know if Microsoft has actually applied for patents on them, I'm just using them as examples of what a non-programmer might find innovative, but a programmer would find to be just a natural step forward that is obvious.

Another example for you, Noel, is SMB. SMB is just a way of communicating file and print commands over a network. It is cool, but not something that would be extremely hard to duplicate in some form for a senior programmer. It represents no fundamental jump over other storage designs from the past. Yet the SAMBA team has been threatened by Microsoft on patent grounds before.

This is why most programmers hate software patents. Unlike patents in other industries, they almost invariably become government-granted monopolies on public knowledge of obvious things.

You raise the interesting issue that judges generally don't have technical backgrounds, and that as very influential players in the software patent policy space, this may negatively affect the landscape of innovation. Hopefully, you've followed this development: www.whitehouse.gov/infocus/judicialnominees/moore.html.

Your view on non-obvious patents, which in several discourses with you, I take as comprising most "bad" patents, seems consistent with PFF's previous KSR brief: www.pff.org/issues-pubs/filings/050511ksr-teleflex-amicus.pdf.

2MikeT:

Good software patents ? Plenty...

Diffie-Hellman , RSA, Schnorr (from encryption hall of fame)
Karmarkar, LZW, JPEG...

From more specialized field of speech signal processing for telecom and other apps:
cepstrum, all kinds of vocoders (CELP, MELP...), speech recognition algorithms,
From general audio - MP3

From communications: Townshend's patent on 56K modem technology
(YES, this is SOFTWARE patent, as well as HARDWARE patent, depending on how you implement your modem...)

How about Qualcom's CDMA patent's, GSM patents, MPEG patents, DVD patents, RFID patents etc. etc. etc.

Simple and obvious for your mythical "senior programmer" ?

I suggest you educate yourself better before writing stupid comments...

Anmd there is NO difference whatsoever between software and hardware, it's only a matter of implementation.
Write it on your forehead :)

And everyone one of those algorithm patents fall into cross-discipline categories not related purely to software, which has been my long-standing contention with Noel.

There is no such thing as "pure software"

Every (good) software piece has its purpose in life, when executed on a computer it does something useful and practical...

Computer can be a tiny DSP chip in your cell phone or a general purpose Intel Pentium desktop - as far as algorithms and software patents are concerned there is NO difference.
The FOSS crowd denying this simple fact of reality is stupid and would cause innumerable problems, both technical and legal.

ANY computer running some executable program embodying some algorithm for achieving some practical result is a specialized machine, much like any other mechanical or electrical machine subject to the patent law. Period.

Ok, so what would you call a patent on SMB, then?

RAND is an undefined term, so before using it, you should define it.