I used to work in wireless R&D (at a startup). It's frightfully expensive, and we would have never be able to do a whole stack with original technology, so we licensed and built on top of WiMAX. I'm not sure why we should've been allowed to build on all that technology for free.

Patents are a huge boon to standardization in capital-intensive industries like that. Getting your patent in the spec is what creates the incentive to contribute to this joint venture that's going to be opened up. Otherwise you're doing R&D for your competitors for free.

"R&D for your competitors for free" sounds like the open source movement to me.

In the brief amount of time I worked in telecom, I was shocked at how paranoid and backstabby it appeared to be.

The GnuPG maintainer is begging for nickles, while most commercial investment into open source is incidental to some other activity that generates the real money. Those aren't good things. When you make it difficult to monetize a product, you create a huge incentive for companies to vertically integrate until they control part of the stack that can be monetized. E.g. Android can exist as open source because it's an on-ramp to Google's proprietary web services.

If companies like Qualcomm couldn't directly monetize their R&D, the likely end result would be their getting sucked up into companies like Apple and Samsung that build end-user products.

A lot of open source comes from companies who make their money not by selling technology but by selling services, often by leveraging data, which they guard jealously. They have little to lose and a lot of geek cred to gain by open sourcing useful projects, not to mention the bonus of "commoditizing your complements" a la Spolsky.

This does not apply to companies that sell the technology itself. Not to mention the fact that telecomm R&D is not like software R&D, which, let's face it, is mostly just re-doing D for R that happened decades ago.

There is a paper by a Qualcomm engineer that talks about how standards are set and how patents play a role in it, specifically in the context of 3GPP. It is a pretty complex process involving attending meetings and casting votes, and many participants do not contribute any technology. Surprisingly, patents seem to be an afterthought.

But then Qualcomm makes a billion or so every year in royalties, so that's something to keep in mind. Will link the article if I can find it.

I would love to see that paper.

Found it!

http://www.law.northwestern.edu/research-faculty/searlecente...