Version française
Home     About     Download     Resources     Contact us    
Browse thread
[Caml-list] Rewriting UNIX in Caml and getting rid of the C disease
[ Home ] [ Index: by date | by threads ]
[ Search: ]

[ Message by date: previous | next ] [ Message in thread: previous | next ] [ Thread: previous | next ]
Date: -- (:)
From: Tom <tom7ca@y...>
Subject: Re: [Caml-list] Rewriting UNIX in Caml and getting rid of the C disease
There are is probably a lot more non-C/C++, safe software in Linux and
UNIX than in Windows NT.  Linux and UNIX use shell, Perl, Python, and
awk extensively--safe, high-level programming languages (if dynamically
typed).  Linux installations often also use cfengine, a kind of expert 
system shell for system configuration.  And this is nothing new: 15 years
ago, statistics at Bell Labs showed that people spent a lot more compute
cycles in awk than in native C programs.

Still, dynamic languages have their limitations, and it would be 
nice to rethink and build larger parts of the Linux
environment in ML, in part to achieve more safety, and in part to make
it easier to modify and extend system software.  Both UNIX and ML 
would benefit from the experience and results.  Utilities 
like "unison" and "bibtex2html" (both written in OCAML)
are a good indication that ML can be used for such applications.

I think most people don't spend much time on it because there isn't
really a big itch that needs to be scratched for them.  Apache, or "GNU 
ls", may both be messy, hard-to-maintain C programs, but they get the
job done and someone else is doing the testing, so why bother 
rewriting them?  Still, all it takes is volunteers, so go ahead 
and pick a project.

Tom

--- Eric Newhuis <enew@bigfoot.com> wrote:
> <seriousness>
> Is there a market for such a thing?  Defense?  Secure operating systems?
> What are all the tangible benefits?
> </seriousness>
> 
> <humor>
> What will the name fo the OS be?  SafeLix?  FreeBSafeD?
> </humor>
> 
> <opinion>
> Windows NT might already be part of the way there depending on how one views
> Prolog as being closer to Caml.  I guess they're both functional.  Has
> Prolog, used in the following manner, added the safety you hope to achieve?
> And could Caml have also achieved it?
> </opinion>
> 
> <fact>
> Microsoft used Prolog in NT's network stack.
> http://research.microsoft.com/research/dtg/davidhov/pap.htm
> </fact>
> 
> -------------------
> Bug reports: http://caml.inria.fr/bin/caml-bugs  FAQ:
> http://caml.inria.fr/FAQ/
> To unsubscribe, mail caml-list-request@inria.fr  Archives:
http://caml.inria.fr


__________________________________________________
Do You Yahoo!?
Find a job, post your resume.
http://careers.yahoo.com
-------------------
Bug reports: http://caml.inria.fr/bin/caml-bugs  FAQ: http://caml.inria.fr/FAQ/
To unsubscribe, mail caml-list-request@inria.fr  Archives: http://caml.inria.fr