XPost: comp.sys.mac.advocacy, alt.os.linux.mandrake, comp.os.linux.advocacy   
   XPost: alt.os.linux.redhat   
   From: znu@fake.invalid   
      
   In article <1114654478.789320.227840@l41g2000cwc.googlegroups.com>,   
    imouttahere@mac.com wrote:   
      
   > >Big deal. MacOS has been rumored to run on x86 for decades.   
   >   
   > Just sayin'. If you don't want to believe me that I know what I'm   
   > talking about here, that's fine.   
      
   Well, we know Apple maintains an x86 port of Darwin -- they give it away   
   on their web site. Assuming the upper bits of OS X are actually well   
   written, which I guess we'd all like to think they are, there's not much   
   reason why they shouldn't just compile and run on top of that Darwin   
   port, though possibly without accelerated video support, etc.   
      
   It would be a good idea for Apple to maintain cross-platform builds of   
   the entire OS, just because writing for two architectures provides a   
   sort of sanity check, that you're not doing things that could reduce   
   portability at some future date, and that you're not introducing bugs   
   that you don't notice because they're hidden the the eccentricities of   
   one specific chip.   
      
   None of this should be taken as an indication that Apple has any desire   
   to move the Mac to x86 or (reaching even further) to release OS X for   
   generic x86 hardware. Honestly, discussion of such a thing at the   
   present time just seems totally wacky.   
      
   Back in the days when Motorola couldn't manage to get over 500 MHz,   
   while Intel and AMD were racing for 1 GHz, the idea might have made some   
   sense. Now, it makes no sense at all. The PPC is very competitive now.   
   At the same time, development seems to be slowing down dramatically   
   industry-wide. In this environment, an architecture switch makes no   
   sense at all. Any kind of transition, even if perfectly executed, would   
   take years and a lot of people would be running slow, emulated apps   
   along the way. And there's no way at all to know what architecture will   
   have the edge in three years or whatever. Everyone seems to he having   
   the same fundamental problems right now, and who knows who's going to   
   crack them first? I don't see anyone making much progress, honestly.   
      
   Dual core isn't the solution. I mean, it'll happen, and it'll be good   
   for people who need the performance, but what's the followup, if the   
   more fundamental problems don't get solved? Four cores next year? Eight   
   cores the year after? Good luck pulling that off, with die shrinks   
   getting harder every year. And good luck writing tools that allow   
   developers to leverage that much parallelism effectively (because   
   they're sure not going to manually code everything to use eight   
   threads). Adding cores is just not a viable long-term strategy for   
   general-purpose processors.   
      
   --   
   "This notion that the United States is getting ready to attack Iran is simply   
   ridiculous. And having said that, all options are on the table."   
      -- George W. Bush in Brussels, Belgium, Feb. 22, 2005   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)