> On Tue, Aug 28, 2007 at 04:08:02PM +0100, Edd Barrett wrote: > > On 28/08/07, Craig Skinner - Sun Microsystems - Linlithgow - Scotland > > > Yay! Action at last. > > > > Wow! This is great news. > > Better late than never, but damn is it late.
Indeed, that is the correct sentiment regarding Sun's action here. The facts of the industry are simply this: Approximately 95% of machine parts are documented (whether they are documented well or not is a totally seperate question). Starting roughly around 1990, Sun put themselves on the path of supplying only the absolute minimum documentation for their machine parts. Meanwhile, the PC really took off, and all the documentation for PC parts has always been out there (minus a few special cases that we have had to fight for). DEC released pretty much all the documentation for the Alpha right from the start, and later a few people pressured HP to release pretty much all the HPPA documentation. That left the largest straggler in the industry: Sun. And the case is that Sun has always had the documentation in-house; because of solid engineering principles in-house they document everything, perhaps because their hardware and software groups are seperated so much. Apple also has done a poor job of documenting their hardware, but looking at the quality of their hardware (with entirely pointless divergences between models that come out 3 months apart) we can guess that maybe we don't want to see them. Finally, there are a few American chip makers that resist the status quo, like Marvell and (to a lesser degree) Broadcom. Even Intel tries to play the open game now. Then there are a handful of (increasingly irrelevant) American wireless chipset manufacturers. But in general there are fewer and fewer closed vendors. But Sun had no excuse for this behaviour in 1990, and it is incredible that only now they will try to redeem it. So I don't say bravo, but I say "about time". They don't get any points from me, because they are so late. I give the most credit to Craig Skinner who started the conversation at Sun with us (he found the right place to push Sun -- right at the top), and David Gwynne for continuing the soft pressure through the last couple of months. My biggest hope is that Sun's cleanup process does not delete too much information from the pages... like descriptions of hardware bugs and the workarounds needed for "best effort" operation. Because we already know that some revisions of Sun hardware have brutally bad bugs that ... even sometimes cannot be worked around.
