🚀 go-pugleaf

In article <2844@garth.UUCP> smryan@garth.UUCP (s m ryan) writes:
[quoting someone from watdragon who says:]
>>Hear ye!  Segments are evil.  If my usable address space is N, it is a priori
>>*wrong* to assume that I won't mind dealing with segments of any size < N.

  No, Sir Dragon, silly restrictions are evil, and tend to disguise
the elegance of good ideas, as well as merely fair ones.
  If my address space with current processor speeds/densities/etc is
N, where N is about half a terabyte, then I am likely enough to accept
intel-like segments of about 4 gigabytes.  As we are supposedly using
up address space at the rate of 1 bit every eight months, this will
hold us to the end of the century.  I wouldn't bet on it doing much
better (:-)).

  I'd really like a multi-level naming scheme, so I can say
fred->mary[j] and not be terribly concerned whether fred is
outrageously large or not. If its not, it should be in the "data" or
"bss" segment.  If it is, it should be in a segment all its own. 
  I certainly would **not** want to have to worry about its size in
the general case. 

  If it's in some way special (eg, if it's a file), I'll accept the fact
that it's special and declare/initiate/open it in the required special
way.
  

I hope to post a series of articles (a sort of mini-review) of different
ways of approaching the multiple -vs- sparse address space issues this
weekend, if the weather doesn't stay nice.

 --dave (only four more days to monomania!) c-b