TurboHercules and Florian Mueller accuses IBM on going back on its pledge not to assert patent infringement on open source software by claiming TurboHercules infringes two patents in the list of patents included in the pledge.
Brian Profitt, my favorite Tech writer, has the best article on an overview of the event and his conclusion that IBM had not done what TurboHercules and Mueller had accused it of doing is shared by me. PJ, as usual, has the best links in her article about the subject. However, I must note that her article concentrate too heavily on ‘holiness’ of GPL and give the impression that IBM is free to sue any company and other open source software as long as it leaves Free Software alone, which in my opinion irrelevant (The GPL bit) and wrong (free to sue any company and non-Free software).
Now that the storm had passed and the smoke had cleared, we see that there is no substance in the claim. Ironically, the best evidence is from TurboHercules itself, its publication of the correspondence trail between it and IBM shows that the mention of the two patents were taken out-of-context. Excerpt of IBM’s 11 March letter:
“According to your own statements, your product emulates significant portions of IBM’s proprietary instruction set architecture and IBM has many patents that would, therefore, be infringed. For illustration, I enclose with this letter a non-exhaustive list of IBM u.S. patents that protect innovative elements of IBM’s mainframe architecture and that IBM believes will be infringed by an emulator covering those elements. For your information, the enclosed list also includes a non-exhaustive list of relevant IBM u .s. published patent applications. Apart from concerns about unauthorized use of proprietary IBM information by one or more TurboHercules contributors, IBM therefore has substantial concerns about infringement of patented IBM technology.”
Excluding the last sentence, this is the most quoted passage in the letter. I do not believe this IBM letter has not been seen by its legal department. Therefore, any omission/inclusion is carefully crafted by wordsmiths in the legal department. I included the last sentence because IBM said the infringement might be made by TurboHercules contributors, not Hercules (the open source software) contributor. While we can expect overlaps in contributor pool of TurboHercules and Hercules, they are not one and the same. Therefore, any infringement allegation is not targeted at open source, but the company TurboHercules. Moreover, the letter mentions emulator in general and is not referring to the Hercules emulator specifically. One can argue that Hercules does infringe the two patents, but is cured by the open source pledge.
Furthermore, if you dig into the history of the correspondence, TurboHercules is asking IBM for licenses to run IBM’s mainframe OS on Intel hardware which IBM declined. The correspondence is always about TurboHercules and IBM. The Hercules project is only mentioned in passing as background to the request in the first letter. Discussions between the two parties had moved on from that to what TurboHercules wants to do. Therefore I do not think it is fair to drag Hercules project back to the fray because when IBM drafted the letter that mentioned the two patents, it is likely that it is no longer refers to the Hercules project.
It is, however, true that IBM could had avoided the whole thing by not referring to the two patents. Since the list is ‘non-exhaustive’ after all, omitting the two from the list of 170+ will not make any difference.
It is clear to me that TurboHercules has been asking for something that IBM refuses to give. As a result, it was out on a campaign to get IBM to change it mind. This is part of the negative part of the campaign to achieve this. I am giving TurboHercules the benefit of doubt in deciding that the EC complain is not part of the negative campaigning.