One of the many functions that standards (and open source) play is that they provide a forum for cutting through patent thickets so that an industry can grow. But it’s hard. Players that stay out of the standards making process, but who are active in the industry, are well positioned to poison the well. Of course you do that after the standard is widely deployed. The only defense is to acquire patents in the spaces around the standard. That can be done defensively by the supporters of the standard, or offensively. There should be a term of art for this tragedy – where in the attempt to cut a path through the thicket triggers the sudden growth in the surrounding thicket.
Recently I’ve been quite interested in an emerging standard, AMQP (a high speed enterprise message bus). AMQP lacks, as far as I can tell, any association with a mature standard’s body. There is an AMQP project at the Apache Software Foundation, but that is not a formal standard’s body. The AMQP working group’s governance is a bit ad hoc; but they do have an stated IP policy[1].
Today I learn that RedHat has at least one patent application in this space (patent #20090063418). This has got the makings of a shit storm. For quite a few reasons. RedHat is a key member of the adhoc standards group working on AMQP. RedHat is a members on the Apache Software project tied to the AMQP ‘standard.’ RedHat donated much of the code to that project.
I assume their donation to the ASF is part of a very typical open source busines model. One encouraged by the Apache licence; i.e. there is a Apache licenced open source product that lots of people adopt for zero licensing costs. That creates a market and then vendors offer enhanced versions of on top of that. Redhat has a very extensive offering in the AMQP space.
Standards create opportunities to do stuff. These opportunities may well be patent worthy. So if you want to grow out the thicket around the emerging standard you just lock some smart guys in a room and start them brain storming. Some of what they come up with will be obvious, but that hardly means you won’t be able to capture a patent for it. Just to add to fire to the shit storm it appears that Redhat’s patent is for the mind bogglingly obvious idea of transfering XML data over AMQP. Of course any patent worth it’s lawyering starts with some broad claim and then get’s more focused.
Some other links:
- Good posting, more pissed than mine.
- Red Hat – a statement on this issue. This serious bull: “Although there have been some recent questions about one of our patent applications relating to the AMQP specification, they appear to originate in an attempt to spread FUD. There’s no reasonable, objective basis for controversy.”
- Well written note from one of the other firms implementing AMQP. “… We are however very annoyed about the Fear, Uncertainty and Doubt that actions like this cause. We are astonished…”
[1]: I’m happy to report that the AMQP working group’s IP Policy is not as ad hoc as I earlier guessed. See here: svn link, but you may need to poke your way thru some ssl cert complaints.
Pingback: Notional Slurry » links for 2009-03-18
But… um… RH has been seeking defensive patents for years, and there are not one but two redundant, binding pledges ensuring that anyone who wants to use this code is free to do so, and that RH would be estopped from pursuing any patent claims. Given that, I’m not seeing what the problem is, other than kneejerk anti-patentism?
(Well, knee-jerk anti-RHism is another plausible explanation, I suppose?)