What I gonna say is maybe wrong or incorrect, but sometime, it´s useful to miss the point. > > > Now, going out on a limb and at risk of being called a troll: from > > > *my* point of view, Traversal is welcome to require very expensive > > > licenses/very constrictive NDA's in order to use/view the RTL code, > > > *provided* *that* they make a covenant with the community to always > > > release*full* interface & behavioural data & specifications for the > > > hardware[1]. I agree with Peter, the community will focus on OGA1 not on the RTL code, so TT can require a restrictive license for the latter. Depending of course who would want the RTL, the license can not be the same as well as the committement to provide a full spec in return could be badly perceived. An university could agree, but a *concurrent* would want a counterpart, especially about IP issues, that is, he wouldn´t be liable.
Certificate is a good idea, though, if as a manufacturer I pay a license, I think that that will be enough. Why should I bother with your sticker or else, not mentioning the involved cost on packaging (as Peter notices)? In most case there will be a bargain. > > > > I agree. I think it should mainly be a business decision by Traversal. > > How much revenue does TT miss when opening up the RTL versus how much > > does TT gain from having outside developers helping out. That's hard to > > say, and it will probably change over time as well. > > One of the real experiments is going to be to see if OGA1's release > results in much change by the community. If there's much enhancement, > we might find it worth-while to fab a 1.x revision based on that. But > more likely, people will be preoccupied by the design of OGA2. That seems to be very probable, but that confirms the focus on OGA, not on the RTL code. Though, what can be the utility of some spec about an hardware not yet on the market ? > > A similar experiment will happen sooner with OGD1. We need libraries > of logic blocks built up around that. SPI is done, video is > essentially done, I'll be doing a memory controller, PCI is 80% there, > etc. But people will want more blocks. If the community can help > work on these pieces, it'll show the kind of give and take that will > go on in the future. OGD1 is a totally open project from the > beginning, so this will be very telling. [It's also therefore vital > for the OHF to get donations so we can get OGD1 boards into the hands > of the developers.] > > One thing we should have in mind for the Ohio LinuxFest is seeing if > we can get the hardware (snip technical details) > The slowness (if any is observed) will be "proof" that we're > really using this hardware. IMHO, one better has to come with something in the hands, even if that doesn´t look like a graphics card yet. Otherwise, it will be just about a project and nothing else. It would be good as well to show some spec of OGA. The both could convince that there´s something to do, and a lot to do (enhancement, drivers...) and lets Traversal doing the *dirty job*. > _______________________________________________ > Open-graphics mailing list > [email protected] > http://lists.duskglow.com/mailman/listinfo/open-graphics > List service provided by Duskglow Consulting, LLC (www.duskglow.com) _______________________________________________ Open-graphics mailing list [email protected] http://lists.duskglow.com/mailman/listinfo/open-graphics List service provided by Duskglow Consulting, LLC (www.duskglow.com)
