[CS-FSLUG] Query: 64-bit vs 32-bit

[Frank Bax]

At 10:36 AM 8/30/06, Ed Hurst wrote:
>It's true there are currently few advantages apparent to the average
>user. I know so little, so I have to ask: Is it possible coding with
>32-bit is a mindset, a basic design structure in itself? Does taking
>full advantage of 64-bit -- the extra registers, the SSE2/SSE3 stuff,
>etc. -- require a completely different way of looking at code? If
>someone started a new project from scratch, and were thinking in 64-bit
>from the start, would it be very different?


Perhaps its a support issue.  Before a developer can produce a 64-bit 
binary, they need a 64-bit compiler that runs on a stable 64-bit 
OS.  Presumably there are some coding difference, if only in system API 
calls.  They can then start coding for 64-bit, but will there be any uses 
of their software on that platform yet?  Why invest the effort before there 
is a real market?  During the coding/release process, developer also needs 
to decide if they have resources to support both 32 and 64 bit versions for 
a period of time.

Does anyone remember the change from 16-bit to 32-bit Windows and the long 
delay before 32-bit apps actually appeared (in some case several 
years)?  Was there a similar transition with Linux?

Frank