[FoRK] The rise of the stupid?

zuzu sean.zuzu at gmail.com
Sun Feb 27 12:47:42 PST 2005


On Sun, 27 Feb 2005 20:23:53 +0000, Andy Armstrong <andy at hexten.net> wrote:
> On 27 Feb 2005, at 19:57, John D. Mitchell wrote:
> > Ah, well, I ignored that just like most of the comments on /.
> 
> Yeah, but he's saying it ON HERE :)


because i prefer to write code in a language (ruby) where the current
interpreter / virtual machine for it runs something like 30x slower
than its compiled C equivalent.  (though i would argue "equivalency",
that in many respects the ruby version *does more*, which define the
reasons why compiled C appears to run faster.)

but i'm simply not in the business of writing code which is time
critical, such as scientific computation, maintaining coolant pressure
in a nuclear reactor, or the firmware for radar collision detection
for commercial aircraft.

i'm curious to know what software you write where a compiler such as
gcc just isn't good enough that you have to try out-wit it with arcane
syntax.  i admit there's plenty i don't confront w/r/t IO and
databases as well.

-z

p.s.  however, my current research project _is_ writing a better ruby
virtual machine in forth for power4 architecture processors;
particularly the multiple processor architecture of the IBM PPC 970
(aka apple G5).  also, that same research throws into question the
very nature of IO (ken thompson's fatal flaw?) and likewise the
niche-minded development of databases.  (maybe hans reiser has
something to say here...  as well as anyone who really _programs_ in
SQL...)


More information about the FoRK mailing list