From: user5857@newsgrouper.org.invalid   
      
   Robert Finch  posted:   
   >   
   > Is the need for backwards compatibility killing things as technology has   
   > improved?   
      
   No with respect to::   
   Little Endian   
   IEEE 754 floating point   
   Byte addressable memory   
   Misaligned memory   
   PCIe peripheral access   
   CXL  interconnect   
   CXL  added memory   
   CXL  added cache   
   access to Linux   
   access to gnu   
   access to LLVM   
   access to qemu   
   access to gem5   
   numerical libraries/packages   
      
   yes with respect to::   
   x86 condition codes   
   x86 shift by 0   
   x86 descriptor tables   
   4096 byte pages   
   long latency exception/interrupt control transfer   
   need source to port application   
   SIMD considered harmful   
   ATOMIC activities   
   Exception walk-back across block structure   
   Signal/exception delivery   
   language evolution   
   environment evolution   
      
   >  There seems to be a lot more known good/bad approaches making   
   > me think that the lifetime of newer designs could be longer.   
      
   Yes, but the people making the decisions are still to young to have   
   the history needed to make better decisions.   
      
   The graduates of major universities go right out and start designing   
   without being exposed to "enough" of the disease of computer architecture   
   to be in a position to understand why feature.X of arch.Y was bad overall,   
   or why feature.X of architecture.Y was not enough to save it.   
      
   Each generation reaches employment after university at about the same   
   level as we did when we invented RISC.   
      
   Architecture is only 1/3rd ISA--and it is the other 2/3rds where the   
   {trouble or success} lies (85% confidence level).   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)