Thanks for the response. I'll keep an eye on the wiki.

In the OHL, I don't understand the legal basis for section 4, "Manufacture and distribution of Products". What gives this section any force beyond a "gentleman's agreement" or a legal bluff, which is easily ignored? I can see that copyright is the legal basis for section 3, dealing with documentation, but don't see the same for section 4. Don't get me wrong, I want the OHL to be binding, but currently I don't see how it can be.

What are the current licensing options for open hardware? Has anyone found a "copyleft" equivalent?

About a decade ago, this issue was discussed at length on the OpenCores mailing lists. At the time, the best we (engineers) could come up with was that the design documents/files could be copyrighted and so GPLd, but there was no way to oblige that a physical device be distributed with design data.

It seemed to be okay for someone to take a design, make secret modifications, build it and distributed a physical product that could not be replicated. The obligation to share modifications only kicked in when the GPLd design data was distributed, not when the physical product was distributed. Is this the case, or has a real legal mind figured out that we were wrong?

It's bad for the integrity of policing. Good for the integrity of the police.

I'd like to know what number patents Innovatio are using. The CSIRO patent (5,487,069, filed in 1992) was a pretty complete description of the 802.11 OFDM physical layer. Surely anything else has to be a minor and obvious increment? The oldest patent I can see, with inventors Meier and Mahany, is 5,394,436, filed in 1994, and it does not refer to the CSIRO patent (meaning it is open to challenge from the CSIRO patent?) 5,394,436 might apply to the MAC layer, rather than the physical layer, and it is quite vague. Defenders might want to refer to the PARROT project, which the CSIRO was running as part of its WLAN work, predating 5,394,436. PARROT was a complete WLAN MAC layer (google: csiro parrot mac layer).

There was also a PhD thesis that came out of Macquarie University in the early 1990's. The name of the author escapes me, but the supervisor was David Skellern. The thesis described a MAC layer for mobile IP, and fed heavily into the standards at the time.

> Step by step !!

No harm in dreaming! :-)

Be careful to distinguish between synthesis and ppr. Synthesis is doable. PPR requires knowledge of the FPGA's structure as well as complete timing info. I agree that clues can be gleaned from the FPGA editor tools, but I don't think it's enough to write a PPR. (I could be wrong though, since I haven't tried it!)

I'm keen to contribute, though I'm constrained by other things that take my time. One thing I do have is a complete MIMO capable reconfigurable radio platform. It would be suitable for use with GNU Radio. I did it as part of a Master's by Research. My thesis and all the designs are under the GPL, the licensing being written into the thesis. I plan on putting it on my website, but I need to time to clean a few things up first. Everything is already on the 'net, in the form of the online thesis. Know any people who would be interested in such a beast?

is the highly proprietary FPGA technology used to implement the CPU. FPGA partition, place and route (ppr) is some of most proprietary software on the planet, slathered in trade secrets and patents. The chips themselves are worse. Think of them as a type of processor (after all an FPGA is just a bit cruncher) with a secret instruction set and compiler (ppr). Xlinix (major FPGA company) want potential customers to sign an NDA simply to have their salespeople say more than "we sell FPGAs".

If the Free Software community is to use FPGA's, as more than just a curiosity, first task is to design/build its own silicon and write its own toolchain. Then they come up against the proprietary nature of semiconductor manufacturing.

I'm not belittling the Milkymist project, as what I describe above is a separate project. It's a huge project, essentially a reimplementation of 50 years of semiconductor progress, ultimately linked to the (seminal) desktop manufacturing projects that some have started. Imagine RepRap mk42 with semiconducting, conducting and insulating inks, printing circuits at the micro-scale.

The disruption you mention almost happened in the early 90's. NeoCAD produced a compete competing tool chain for Xilinx FPGAs, including the place and route, for the then state-of-the-art 4000 series. Their software was better than Xilinx's, including things like a graphical layout editor. Xilinx was having none of it and bought NeoCAD. Quite a few NeoCAD features made it into the Xilinx software, eventually. Soon after that Xlininx started publishing less information on their FPGA's interconnect networks, and there has never been another attempt at writing such software.

Personally, I think writing a clone of the Xilinx software, today, is the wrong thing to do. It would be less effort to design and manufacture an "open source" FPGA, and write the necessary software from scratch, than to reverse engineer Xilinx's place and route.

A critical question is: Who owns the result? A prize should be about promoting development, and NOT about acquiring ownership. Any prize recipient who transfers ownership is a bunny. The sequence should be:

1. Win prize (prize posted to promote development.)
2. Use prize money as seed funding for business
3. PROFIT! Prize giver has a solution and developer has money.

The reward for the 2nd, 3rd, .. place getters is the opportunity to still develop a business, albeit without the benefit of the prize money as funding.