[The entry below is an issue of Greg Aharonian's PATNEWS email newsletter deconstructing IBM's latest patent publicity moves. Reproduced by permission of the author, as it is not posted on the Web. Email patnews at ns1 dot patenting-art dot com for your own subscription.]
Excuse me folks, but I would have appreciated it if one of you had let me known when I accidently tripped through the looking glass and joined Alice in Wonderland. IBM is now backing software patent reform, and expects everyone to forget their past? After all, this is the IBM that
- in the 1990s, made billions "enforcing" their software patent portfolio, often with old patents
- in the 1990s, helped undermine some software patent reforms (e.g., backing the impotent SPI)
- as of 2004, was still filing lots of (software) patents with paltry amounts of prior art submitted, to bulk up its portfolio
But now that IBM has made tons of money playing this game, it doesn't want other companies playing this game, other companies that learned from IBM how to play this game and now are being told not to play IBM's game, but rather to make the mythical big bucks from open source (which seem to flow much to IBM). So while I appreciate the raw business tactics of IBM (i.e., yet another reason to buy their stock), their history cannot be ignored - atonement is necessary.
An article in the Monday New York Times has some inconsistencies. A quote from John Kelly, a senior vice president at IBM (who heads an IP strategy group):
Another development in recent years that pushed IBM to reconsider its patent approach has been the surge in patent filings and lawsuits, including the rise of firms whose only business is to file patent infringement suits, known as "patent trolls".
"It seemed to us the pendulum has swung way too far in the direction of companies blindly chasing patents, and blindly chasing the enforcement of patents", Mr. Kelly said.
Excuse me, but where did these companies learn these tactics except at the feet of IBM patent lawyers? And what company has blindly filed the most software patents? IBM. And what company is still blindly filing the most software patents? IBM. After all, the IBM software patents that issued in 2004 (admittedly filed a few years earlier, maybe before their new strategies were being shaped) where no different than peer averages - about half cited no non-patent prior art, the average cited two or three non-patent art items.
[IBM] supports proposals in the United States to make software patents more difficult to obtain, hoping to help curb the patent and litigate frenzy.
Such support is toooooooooooooooo damn late. In the 1990s, IBM and others frustrated one of the best ways to make software patents more difficult to obtain - the building of a large software prior art archive - when it backed the utterly bogusly fraudulent Software Patent Institute (anyone really remember them anymore?). The SPI wasted a ton on money at the one point in time when a lot of money well spent by experts (not the chuckleheads in Michigan) would have produced a very useful resource not only for the Patent Office but the software industry as well. Ten years has been wasted due to the money flushed down the toilet at the SPI. Does IBM plan to apologize now for their involvement in this waste? And yes, I did resent not having a chance to build such a database in the 1990s with the support of others. I can build a killer database, which is probably why I wasn't given a chance to do so.
Next, from an article in ZDNET Australia:
"There has been a dramatic increase in the number of filings of patents recently, around the world, but particularly in the United States.", Jim Stallings (IBM IP VP) said.
Actually software patent apps steady-stated/linearized a few years ago (as did issuances leveling out at the 20,000 or so a year level) as people did learn their lessons from IBM's filing strategies. Besides, the time to act is before any dramatic increases, like in the 1990s.
"What's happened is it's challenged examiners' ability to inspect history, so the bias has been towards granting the patent.", [Stallings] said.
Funny, IBM wasn't complaining when the bias was in its favor in the 1990s when it was acquiring, and ENFORCING, its patents. (see a critical article in the March 17, 1997 Business Week about IBM's enforcements in Silicon Valley). And it would be easier for examiners to inspect history if they had access to something less pathetic than the Software Patent Institute. Or if IBM lead a public effort to get Congress to stop stealing PTO fees.
"There is a process for after [the patent] is granted to challenge it. It's a very weak process." ...
Actually the PTO has a good record for invalidating bad software patents under reexam. It may be a lengthy process in time, but not a weak process. If I had the money and time, I'd be glad to file lots of ex-parte reexam requests for IBM patents to start the ball rolling.
"We are saying that process needs to be enhanced. It should not not only be the job of the examiner, which is an individual, to grant that patent and to inspect prior art."
Oh slay me through my heart. IBM - please lead the world in reform - do more thorough non-patent prior art searches for all of your patent applications. Don't do what you do, but what you say. What you do is sampled next.
For example, consider U.S. patent 6,877,155 issued to IBM last week (filed in 1998 - a continuation from 1994) titled "System and method for generating target language code utilizing an object oriented code generator". It cites only 4 patents and 2 IBM TDBs, while ignoring all of the object oriented and code generator non-patent prior art from the 1980s and early 1990s. Or consider U.S. patent 6,877,081 issued to IBM, originally filed in February 2001, titled "System and method for managing memory compression transparent to an operating system" - and it cites NO non-patent prior art. Or my favorite, really-needs-to-be searched, U.S. patent 6,874,084 titled "M&A for establishing a secure communication connection between a java application and secure server", filed May 2000, citing, incredibly, only three U.S. patents and NO
non-patent prior art. Look at claim one:
1. A method for creating a secure connection to a server, comprising: within a platform independent applet, initializing a secure socketconnection request via HyperText Transport Protocol (HTTP); using a certificate database to authenticate the server; and creating the secure connection between the platform independent applet and the server using HyperText Transport Protocol Secure (HTTPS) if the server is authenticated, wherein the secure connection between the platform independent applet and the server is created without intervention by a Web browser.
They couldn't find any non-patent prior art prior to 2000 of relevance? IBM - pleaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaase hire me as your searcher.
Stallings called on industry to stop what he calls "bad behavior" by companies who either seek patents for unoriginal work or collect and hoard patents.
Yes, let's abolish all companies who get software patents, oh my my my, on reserving the use of a toilet in an airplane. And much like children, companies exhibiting "bad behavior" learned it from their elders. I wonder from whom for software patents? Is IBM chastizing Microsoft's current practices, the Microsoft that recently hired some sharp patent people from..... IBM? IBM's prior tactics helped create a climate where patent hoarding today is a needed strategy, especially when people start counting IBM's 40,000 or so patents.
"If you are a company and invent patents you should state your intent to use them and there should be a period of time in which you have to use them", he said. "There are companies that are in the business [of] simply collecting and want to sleep on it."
To a few of my clients, this is utterly hilarious. Let's just say I spend a fair amount of time dealing with priority dates in the mid-late 1980s. One solution would be to have an exponentially increasingmaintenance fee, which if nothing else, gives Congress more money to steal (IBM - a public call for Congress to stop stealing PTO fees?).
"We think software patents are important, but they should be granted for things that are new", he added. "We're open to sharing information about the patent itself to prove that it's new. And we think everyone should be held to that standard."
Yes, but are you open to the idea of spending some money to do some decent searching to see if your own patent applictions are really new? And let's be honest, when you are as big and rich as IBM, you can afford to be generous with your disclosures.
IBM's antidote to the problem is to increase the scope of the investigation into 'prior art' associated with software patents.
Why just software patents? How about hardware patents, and the whole patent system. Patent quality is problematic across all fields, and software is creeping into all fields of technology.
Stallings believes that sort of undertaking is something the academic community, volunteers and others are willing to help in.
Up until this suggestion, I would have a) chastized IBM, and b) bought more of their stock. This last suggestion will adversely affect the patents IBM cares more about, than its patents it doesn't care about or the patents of the bad-behaviors. IBM is calculating something I am missing (maybe a tactic aimed at companies like Sun). There is the scenario that if all software IP is eliminated, IBM should still be a dominating and large force in the marketplace. So maybe IBM is suggesting whatever, since whatever the whatever outcome is, IBM still benefits by being number one. What the heck - buy more of their stock.
The suggestion is also a logistics distraction. IBM knows that this collective effort is unnecessary. IBM, on its own (especially with my help :-) can easily help the Patent Offices greatly improve their capabilities to deal with prior art. IBM could have done so anytime in the last 15 years, but didn't. Groups efforts in this case are less efficient (e.g., public troubles dealing with claim interpretation and 102 dates). The odd thing is that this type of effort could easily be extended to all sorts of patents, especially the areas where IBM really cares about. Possibly not in their stockholders interest to back such an approach. As I said, kind of odd thing for them to suggest.
Not surprisingly, IBM is making a sharp distinction with its generosity, which is only being extended to interoperability-type patents of the type that are relevant to communication standards such as those for the Internet. But for the bulk of its patenting/business activities (hardware, application software, ecommerce) it is mostly sticking to its old ways. Again, a good reason to invest in IBM, as long as you don't mind their checkered history with regards to software IP (and in this day and age of corporate crimes, if they worse you can say about a company is that they have a checkered history, I suppose the distaste is washed out by the stock appreciation :-). Besides, interoperability is just as well handled by getting standards committees to pay more attention to their committee-members being honest with their patent intentions.
IBM's history in this area requires some atonements to be made.
Internet Patent News Service