Linus vs. NVidia

[Christopher Browne]

On Wed, Jun 20, 2012 at 12:27 PM,  <sciguy-Ja3L+HSX0kI at public.gmane.org> wrote:
>> Someone had asked him if there was any other language other than C
>> that can fit operating system development. He said no and he showed a
>> couple of reasons why that is the case.  One of that case was the
>> ability to optimize, as he can see the assembly lines that will be
>> generated by just looking at the C code.
>
> He means that, having coded raw machine code without an assembler since
> about age 10, he can now look at C code and visualize in his mind how
> efficient the machine code will be, allowing him to micro-optimize the
> code effectively. That is a rare skill.

Yep.

The main other language I can think of that has *reasonably* entered
such debates is BLISS, which was used heavily for systems programming
on VMS.
http://en.wikipedia.org/wiki/BLISS

If you step back in time, the main operating system implementation
languages have been:
a) C, for quite a large set of systems
b) Assembly language, for a large set of now-forgotten systems
c) PL/I derivatives, for MULTICS, OS-360, AIX
d) BLISS, for VMS

There have been experimental systems using other things, but with not
much success.

I wish that Go were a plausible answer to add to the list, but it is a
bit too keen to have garbage collection pre-provided, and I don't
think it tries terribly hard to provide ways to jump out to assembler
for the "pointiest bits."
-- 
When confronted by a difficult problem, solve it by reducing it to the
question, "How would the Lone Ranger handle this?"
--
The Toronto Linux Users Group.      Meetings: http://gtalug.org/
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://gtalug.org/wiki/Mailing_lists