🚀 go-pugleaf

In article <2848@omepd> pcm@iwarpo3.UUCP (Phil C. Miller) writes:
 [...]
>
>   I obviously didn't make my point clear enough.  I think that computer
>   science will reach such an advanced state (at some point in the future)
>   that applications programmers and users will have no need for such 
>   knowledge.  
>
>   At that point, architectural and algorithmic details will hopefully
>   become background details which are handled transparently.  An example
>   of where this has already occurred is in the use of overlays: virtual 
                                                                  ^^^^^^^
>   memory has made one troublesome programming detail an obsolete 
    ^^^^^^
>   consideration on most scientific computers.
                     ^^^^^^^^^^^^^^^^^^^^^^^^^^
 [...]
>Phil Miller


I think you're right when you say most, but ...

Crays at least do *NOT* have virtual memory.

I once tried to argue that this was a real shortcoming
with some of the hardcore coders around here. (see organization line)
Of course being that most of my experience and information
came from my udergraduate CS studies and their experience
and information came from *real* experience, I lost.

The fact is that until some sort of amazingly sophisticated
automatic analysis system is created programmers (real ones
anyway) will know much more about their progams amd their 
memeory requirements etc. than any compiler/os/runtime envirionment
will ever know.  So when real perfomance is a must there will
be no substitute for do it yourself, preprogrammed management like
overlays.  (yes, people here use overlays often for a variety of reasons)

In the past I've seen people suggest that ease of expression and understanding
should always take precidence over trying to write efficient code and
faster hardware will solve the inefficiency problem.  Well we've got
a lot of pretty fast hardware and nobody seems satisfied that it runs
fast enough that programming efficiency can be ignored.  

However, as you've said maybe some day computer science will advance to
the point where no obscure code or special hardware understanding will
be needed to write optimal code.  Let's hope so.  But I don't think
we should all start holding our breath quite yet.

	David Huelsbeck
	dph@lanl.gov
	{ihnp4,cmcl2}!lanl!dph

	#include <std/disclaimers.h>