On 03/05/07 08:52 -0700, ron minnich wrote:
let me say it differently. I think if we want to target OLPC, we ought to do it from V3, not V2. I'm still in favor of dropping the current code and forcing a 'rework' of the code.
To me, the current OLPC code is the worst possible example of a Geode port we have. I feel comfortable saying that because a bunch of that hack job is mine.
I have no problem at all with somebody porting V3 to work on the OLPC - I hope they do it for every Geode platform they can get their hands on. But perpetrating iffy code that will never be used in production for a board that will never *be* in production to me is silly. Lets stop giving people the wrong impression and lets stop leaving a poor example of platform support around.
Very few LinuxBIOS developers have platforms, and those that do are too busy trying to get the production image ready to go to worry about any side hobbies. I'm confident that by the time enough people have the platform to make a LinuxBIOS port useful, then V3 will be solid, and we'll have dozens of good Geode examples under our belt to choose from.
Jordan