From: cross@spitfire.i.gajendra.net   
      
   In article <107768m$17rul$1@dont-email.me>,   
   Thomas Koenig   wrote:   
   >Dan Cross  schrieb:   
   >> [snip]   
   >> If you're just talking about RISC design techniques generically,   
   >> then I dunno, maybe, sure, why not,   
   >   
   >Absolutely.  The 801 demonstrated that it was a feasible   
   >development _at the time_.   
      
   Ok.  Sure.   
      
   >>but that's a LOT of   
   >> speculation with hindsight-colored glasses.   
   >   
   >Graph-colored glasses, for the register allocation, please :-)   
      
   Heh. :-)   
      
   >>Furthermore, that   
   >> speculation focuses solely on technology, and ignores the   
   >> business realities that VAX was born into.  Maybe you're right,   
   >> maybe you're wrong, we can never _really_ say, but there was a   
   >> lot more that went into the decisions around the VAX design than   
   >> just technology.   
   >   
   >I'm not sure what you mean here.  Do you include the ISA design   
   >in "technology" or not?   
      
   Absolutely.   
      
   >[...]   
   >   
   >> While it's always fun to speculate about alternate timelines, if   
   >> all you are talking about is a hypothetical that someone at DEC   
   >> could have independently used the same techniques, producing a   
   >> more performance RISC-y VAX with better compilers, then sure, I   
   >> guess, why not.   
   >   
   >Yep, that would have been possible, either as an alternate   
   >VAX or a competitor.   
   >   
   >> But as with all alternate history, this is   
   >> completely unknowable.   
      
   Sure.   
      
   >We know it was feasible, we know that there were a large   
   >number of minicomputer companies at the time.  We cannot   
   >predict what a succesfull minicomputer implementation with   
   >two or three times the performance of the VAX could have   
   >done.  We do know that this was the performance advantage   
   >that Fountainhead from DG aimed for via programmable microcode   
   >(which failed to deliver on time due to complexity), and   
   >we can safely assume that DG would have given DEC a run   
   >for its money if they had system which significantly   
   >outperformed the VAX.   
      
   My contention is that while it was _feasible_ to build a   
   RISC-style machine for what became the VAX, that by itself is   
   only a part of the puzzle.  One must also take into account   
   market and business contexts; perhaps such a machine would have   
   been faster, but I don't think anyone _really_ knew that to be   
   the case in 1975 when design work on the VAX started, and even   
   fewer would have believed it absent a working prototype, which   
   wouldn't arrive with the 801 for several years after the VAX had   
   shipped commercially.  Furthermore, Digital would have   
   understood that many customers would have expected to be able to   
   program their new machine in macro assembler.   
      
   Similarly for other minicomputer companies.   
      
   >So, "completely unknownable" isn't true, "quite plausible"   
   >would be a more accurate description.   
      
   Plausiblity is orthogonal to whether a thing is knowable.   
      
   	- Dan C.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)