Saturday, April 10, 2010

TurboHercules' entire correspondence with IBM available online

On Tuesday I published IBM's most recent letter to TurboHercules, dated March 11, 2010. By now, the three letters previously exchanged between the two companies have also been made available (on the TurboHercules website). This is useful material to see what has happened and in which context.

In its first official reaction to the publication of the latest letter, IBM stated the following:
We were merely responding to TurboHercules' surprise that IBM had intellectual property rights on a platform we've been developing for more than 40 years.

Now that all of the letters are available, it's easy to demonstrate that the sentence quoted above is wrong. A few commentators, who were unfortunately misled by IBM's statement, suggested that IBM didn't threaten TurboHercules with patents but basically just answered a request for information. This fallacy can be dealt with very quickly.

In TurboHercules' first letter to IBM, TurboHercules inquired about a way to let customers run IBM's operating systems, especially z/OS, on Hercules. Neither the word "patent" nor the broad and general term "intellectual property" comes up. It's particularly easy to check because that document can be searched automatically with the PDF Reader.

Four months later, but still four months before the letter I published on Tuesday, IBM then replied with this letter. That document is not searchable because it's a digital image, but it's just one page.

The second paragraph of that letter acknowledges what TurboHercules really wanted to know:
In your letter, you ask whether IBM would consider licensing its operating systems for use on the TurboHercules platform in order to help TurboHercules SAS (Turbo) establish a commercial business.

The third paragraph then begins with the following infringement assertion:
First, you state that Turbo "implements the instruction set of IBM mainframes on Intel-based servers". We think that mimicking IBM's proprietary, 64-bit System z architecture requires IBM intellectual property, and you will understand that IBM could not reasonably be asked to consider licensing its operating systems for use on infringing platforms.

It's not too hard to interpret, but let me put it in colloquial terms to make it even clearer what IBM said: "You (TurboHercules) say you emulate our mainframe CPU on Intel-based computers. If what you (TurboHercules) say is true, then you infringe IBM intellectual property."

IBM asserted an infringement, not in a very specific form initially, but without a doubt, IBM brought up an "intellectual property" infringement assertion against Hercules, literally and proverbially, out of the Blue. IBM started to bully.

IBM used a very broad and general term. "Intellectual property" in connection with software can mean patents, copyright, trade marks, trade secrets, design patterns. The term "intellectual property" is so broad and unspecific that Richard Stallman, the founder of the software freedom movement, even rejects it entirely as a "seductive mirage".

Therefore, TurboHercules had no idea what exactly IBM meant. Considering that Hercules had been around for a long time without any complaints over any infringement, Roger Bowler was surprised in the sense that he couldn't see why his project would all of a sudden infringe any rights if no such claim had been made during all of that time (during which there can be no doubt that IBM was well aware of it; IBM had temporarily even recommended the Hercules emulator in one of its so-called Redbooks).

That's why Roger wanted to know more about this. So he wrote another letter to IBM, and the key passage for the patent issue is this one:
We also were surprised at the suggestion that our TurboHercules product - which merely relies on Hercules open source emulation software to run z/OS on Intel-based servers - might infringe certain IBM intellectual property. Hercules has been widely used in the development community, as well as within IBM itself, over the past ten years. Prior to receiving your letter, we were not aware of any claim that Hercules might infringe IBM's intellectual property. If you believe that the Hercules open source project infringes any IBM intellectual property, please identify it so we can investigate that claim.

Let me repeat a previously quoted one of IBM's gross distortions of the facts and compare it to what actually happened:
"We were merely responding to TurboHercules' surprise that IBM had intellectual property rights on a platform we've been developing for more than 40 years."

But there's nothing in TurboHercules's second letter that would enable IBM to reasonably say that TurboHercules was surprised that IBM had "intellectual property rights on [the mainframe] platform".

TurboHercules of course assumed that the world's largest patent holder would own mainframe-related patents. TurboHercules of course knew that a company the size of IBM would protect software copyright and other rights.

TurboHercules was, however, quite understandably surprised that a FOSS project that had been around since 1999 would now suddenly infringe rights (and not before), especially since this had not been under cover but IBM had been well aware of it all the time, about which there can be no doubt. And TurboHercules couldn't figure out what IBM meant by "intellectual property" given the vague nature of the term.

Everyone following this issue should look at how IBM tries to fool commentators and the FOSS community. IBM should admit what it has done.

To sum it up, the facts are:
  • TurboHercules never said "intellectual property" or "patent" or anything like it before IBM said "intellectual property" and "infringing".

  • TurboHercules, contrary to IBM's statement, was never surprised that IBM would own mainframe-related rights. They knew IBM would always do its homework.

  • TurboHercules was surprised that an open source project started in 1999 and even mentioned in an IBM Redbook a long time ago would suddenly, in 2009, be the object of an infringement assertion.

  • TurboHercules didn't, contrary to what a small minority of commentators believes, beg IBM to come forward with a patent list. IBM made an unspecified infringement assertion. In my opinion there was nothing unclear about the fact that an infringement was asserted. What was unspecified was the kind of "intellectual property" that was meant. Not even the category of "intellectual property" was clear after IBM's first reply.

  • IBM then provided the patent list and repeated its infringement assertion with the letter I published on Tuesday.

IBM now tries to downplay its action of threatening by waving patents, saying that the patent list was only provided on request. Even if the thing about the request were true (which it is not), it would be a hostile, threatening act against a FOSS project. But as the exchange of letters undoubtedly shows, IBM brought up a patent infringement (without even saying clearly it was about patents, but that's what was meant) before TurboHercules ever inquired about it.

IBM told TurboHercules in its first letter: "Don't mess with us because we're armed and can kill you any moment." Of course, "armed" can mean lots of thing: a gun? a knife? a dagger? two or more different weapons? And TurboHercules was wondering: after a decade without any conflict, what's going on now? IBM then drew its gun. When asked, it said: "We didn't threaten with it. We were asked to show what we had."

Give me a break.

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.