Theo is famous for being an abrasive ass; his redeeming feature is that he's intolerant of bad code. Thus, the security focus of his project (since security exploits are but bugs in either coding or in the design of the coded protocols).
The bitching about Intel's open-source efforts are based on Theo's beef against Intel (and others; there was a high-profile, for OpenBSD, beef against Adaptec that ended in the unsupport of Adaptec SCSI cards, which Adaptec may or may not have noticed yet, other than as a source of hate mail) for not providing documentation for writing device drivers for the hardware they sell. The lack of documentation gives Theo's projects a draconian choice: fail to support popular hardware, or spend a terrible price in time and in performance reverse-engineering (perhaps poorly) the methods to access the device properly. The "blobs" are Binary Large Objects (BLOBs) that constitute a binary digestible to the Intel hardware, and fully capable of being delivered by Theo's device drivers (as they are in some Linux drivers), if only Theo weren't philosophically opposed to delivering code that either (a) doesn't contain the license he prefers, or (b) he has no chance to inspect. With regard to (b), Theo never got to inspect firmware on SCSI cards, video cards, and even in network interface cards, so delivering what amounts to firmware within the device driver every time the device driver is loaded (instead of storing it in expensive, separate, on-device storage) is really no different in some ways than what Theo used to accept before BLOBs became more common.
Apple ships some drivers written by third parties, to which it does not own the code, and to which only binaries are available. The political bent of those who complain about that should be pretty clear reading who they're from: GNU-Darwin. Darwin wasn't released with a GNU license for a reason, and it's unsurprising a release that is trying to be a GNU release would have some unhappy folks associated with it.
As long as Intel thinks it's got a competitive advantage in its interface with its hardware, it'll be keeping its firmware secret. People who want to use the hardware will have to get used to this, though of course like Theo they're entitled to bitch about it in the hope that Intel changes its mind. Sometimes, though, even the vendor can't just wish and make it so: vendors may owe obligations to third-party software authors:
As to the reason Intel refused to update their licensing, Theo explained that they referenced obligations to outside parties.http://kerneltrap.org/node/4202
What Theo wants is, I suppose, either to be contracted to be the outside author and to release the code free (being paid for it), or to have the outside contractors agree to allow vendors to open-source property that helps keeps the outside software vendors in future business if kept secret (hmm, yeah).
News 
Market Data 
Discover 
