Thursday, April 8, 2010

IBM confused and confusing - but definitely still hostile

Big Blue Confusion.

After Tuesday's revelation, the media wanted to know from IBM where it stood concerning the patent pledge.

IBM scrambled to come up with an answer. Late on Tuesday IBM issued a statement that was contradictory in itself. On the one hand, they said they stood by the patent pledge they made five years ago. On the other hand, they came up with some theories that weren't in the original pledge. In other words, if they were going to stand by anything, it would be something fundamentally different from the original pledge: something that would give them the unilateral right to decide which individuals and companies are deserving of the benefits of the pledge, and which ones are not.

eWeek was first to publish IBM's reply late on Tuesday. At the end of this article, you can find a whole paragraph that talks about how IBM might consider TurboHercules not to qualify for the benefits of the pledge.

Then both the Financial Times and the Wall Street Journal pointed out that the concept of "qualified" simply wasn't in the original pledge. The Wall Street Journal asked: "If TurboHercules doesn’t qualify, who does?" The Financial Times found that "[e]ven under a very generous reading of the case, IBM is stretching the definition considerably to defend its turf. There’s a clear message there for any other open source company rash enough to try to take on Big Blue with its own weapons."

In other words, the world's two leading financial papers, independently of each other, didn't put any stock into the most essential part of IBM's statement.

Eric Raymond, an open source luminary often referred to by his initials ESR, wrote in his blog that IBM was "digging itself in deeper" with its "retroactive attempt to deprive the pledge of actual effect" since the only criterion that the original pledge established was that a project's license had to be an open source license. ESR wonders "when the adult supervision at IBM is going to step in."

When ZDNet asked me what I thought of IBM's reply, I also pointed out that the word "qualified" was not in the language of the pledge. "The problem is that 'qualified' becomes an arbitrary and discriminatory decision on their end."

Actually, it's not only the word "qualified" that IBM used and that wasn't in the actual pledge. They also talked about TurboHercules' "motivations". That's a little bit of an Orwellian notion... IBM checking on what we think and feel and then exercising its discretion as to whether it stands by its promises or not.

IBM may solve 1% of the problem with a 180-degree turn

Ultimately, IBM may have realized that it couldn't get away with totally baseless redefinitions of things it published five years ago. So the next episode of the story: Jim Zemlin, the president of the IBM-sponsored Linux Foundation, published a statement yesterday that is signed by Daniel Frye, VP Open Systems Development, IBM Linux Technology Center.

In that statement, the part of the original pledge that refers to who can benefit from the pledge is quoted again.

What is not mentioned is that there is one and only one exception in the original pledge: if someone asserts patents or other intellectual property rights (copyright, for instance) against open source, IBM reserves the right to hold its patents against him. Again, that is the only exception provided for by the pledge. And TurboHercules has not asserted any patents, copyrights, trademarks or other intellectual property rights against anyone.

Some are getting confused because TurboHercules lodged a complaint with the European Commission over the anticompetitive impact of IBM's behavior. Would that mean TurboHercules sort of attacked and could no longer benefit from the pledge? Absolutely not. The pledge only talked about an intellectual property rights attack against open source. It didn't talk about reporting to the police an IBM car that is illegally parked somewhere. Reporting illegal parking simply isn't an assertion of intellectual property rights. Nor is the lodging of an antitrust complaint.

Getting back to the IBM letter, it ends with the promise that "IBM will not sue for the infringement of any of those 500 patents by any Open Source Software."

Actually, that concluding statement is again a redefinition of what IBM originally promised. The original promise was not just that IBM would not sue for patent infringement. It was that IBM would not assert those patents. That is an important difference because the original promise "not to assert" is a broader one than now saying "will not sue for the infringement". The letter IBM sent to TurboHercules was certainly an assertion, even though they haven't sued so far.

Jim Zemlin (Linux Foundation) acting as His Master's Voice

More importantly, Jim Zemlin, who published and commented on that IBM statement, is totally wrong to say that ”[f]ortunately all of us can breathe easy" because this is anything but meaningful progress on the actual issue.

All that has happened is that IBM has put out inconsistent statements and still has not said clearly and affirmatively that TurboHercules meets the one and only criterion defined in the pledge (the software is under a license that was formally recognized as an open source license at the relevant point in time, and still is, of course). Nor has IBM said clearly and affirmatively that TurboHercules has not done anything that would be related to the one and only exception the original pledge defined (asserting IPRs against open source).

More than anything else, IBM appears confused and with its own confusion it's confusing, maybe deliberately trying to confuse, everyone else.

This is now another part of IBM's business - now open source, previously mainframe - making a statement. Do we - the non-IBM part of the world - now have to figure out which division of IBM calls the shots? So far there are actually very strong indications that those open source people at IBM don't have much to say compared to their mainframe counterparts.

But even if this confusion were to be resolved with an unequivocal and definitive statement by IBM's CEO himself, it wouldn't mean, contrary to what Jim Zemlin says to please his biggest sponsor, that ”[f]ortunately all of us can breathe easy".

Those are 500 patents. What about the other 50,000 or so that IBM owns? Open source is safe from 1% of the threat, not from 99%?

The patent pledge was a drop in the ocean from the beginning.

Out of the 173 patents (67 of them in the application stage) IBM asserted against Hercules, 2 would be removed from the list. That would also just amount to 1% of the issue. 99% of the problem would still be there.

If Hercules infringed only a single one of those patents related to the mainframe CPU instruction set, it could no longer be used. It simply wouldn't be able to execute any significant mainframe software because every significant piece of software will typically make use of each machine language instruction somewhere.

They just need one patent out of those 173 (or 171, if the two pledged ones are subtracted) to shut down Hercules, if that one patent is indeed infringed.

IBM still hostile, dangerous and utterly hypocritical

The bottom line is that IBM continues with the strategy that was apparent five years ago with the patent pledge. Simply put, IBM wants to fool the FOSS community. It's hypocritical. It portrays as a generous gesture something that doesn't make FOSS developers and users safer in any noteworthy way.

IBM still has hostile intentions because it wants to protect its mainframe turf no matter what. IBM is still dangerous for all of FOSS because no one would be safe from IBM in areas where IBM has a business interest in keeping competitors out. And worse than that, if IBM gets away with this, think of what other big patent holders might do.

If you'd like to be updated on patent issues affecting free software and open source, please subscribe to my RSS feed (in the right-hand column) and/or follow me on Twitter @FOSSpatents.