Re: following up on speed

[Sundar Narasimhan <address@hidden>]

   (Pause.) Sorry for the catharsis but I needed to get that out of my 
   system.  I hope
   that this might provide some good research ideas for some good 
   researchers...
Dan -- no need to apologize. You have experience w/ both the Lisp and
then C/C++ side on large systems, which mirrors mine and consequently
I read all your notes (cathartic or otherwise) with interest (as I'm
sure do others).

It's not clear to me that what kind of research you envision
though. It would seem that demand paging is here to stay and OS level
open-implementations aren't going to happen anytime soon, and that
language designers (even at ll1) seem quite merrily divorced from many
of the concerns that you and I share, and starting from scratch is
oh-so-expensive so it's a better strategy academically to noodle at
the edges (or call names at older solutions that seemed to work for a
while :). Commercially it's wiser to leverage 'other people's' code or
keep redefining the problem. Perhaps why, as an industry, we haven't
made much progress. Empirical work bores theoreticians, and good
empiricists don't necessarily want to divide and conquer problems
where once divided you end up with things like W2k or Solaris paging
that are viewed by most as opaque or too dynamic between releases to
be of much use.