And Windows, whose APIs are from OS/2's Presentation Management layer. And Linux/OS X, many of whose APIs are from Unix! I don't know where Oracle thinks the madness should end, and I don't think they care.
Go back to the Phoenix BIOS. The original IBM BIOS interface would be covered by this, I expect. So if IBM could prevent BIOS clones, we never would have had the PC-compatible business, and IBM would have been able to prevent PCs from ever being powerful enough to encroach on their mainframe business (apart from Apple, I guess). Since copyright expires after 95 years, we could have had PC clones starting in... 2076.
I prefer the way our industry evolved without interfaces being copyrightable.
Mac OS would probably be safe because it is a Unix [1]. I'd expect that the legal framework that was set up to allow the Open Group to certify things as being officially Unix included having such certification include permission to use the necessary IP to implement a Unix.
Ah, I think you're right about the OS/2 one. (I did some researching, and it seems like it actually might have been developed by MS, which I did not know. I knew there was some collaboration, but I thought it was only for the ability to run Windows apps natively in OS/2.)
> You guys need to find an example where one company offered licensing and the other simply refused to negotiate.
Part of the trouble with this, I think, is that prior to this case, nobody thought there was a need to license this, since it wasn't covered by copyright by any sane reading of the law. (And I still think it isn't, and the appellate court erred in overturning the district court.)
Late reply but it frustrates me when these examples are all (1) public standarda, or (2) two companies in bed with each other.
The best counter-example I can think of is "Microsoft NWLink", which was a Netware 3-compatible file server offered with Windows NT. How long and under which legal threats is unknown to me.
I prefer the way our industry evolved without interfaces being copyrightable.