Re: Coyotos : A restatement

[Jonathan S. Shapiro]

On Fri, 2006-08-11 at 01:07 +0200, Guillaume FORTAINE wrote:
> Let's go for war :-) :

Let's not. Some of us have *productive* work to do.

All:

Matt Kaufmann has just replied to the private thread, indicating that he
doesn't have any problem with my statement that Kit should be viewed as
a prototype system or my statements about ACL2.

> 5 months are not considered as "high robustness". Ok, so you want to 
> test 10 years an OS and say 10 years later, it's production ready ... 

Funny. That's definitely not what I said. I don't remember anything in
my post about waiting 10 years, but hang on and let me check... Nope.
That seems to be your idea.

There are ways to test and validate systems to much stronger standards
of uptime. I think that most of the people on the l4-hurd list have a
general sense of where the Coyotos project is going with this, so I
won't waste time repeating myself.

> All right about security :-)
> 
> Nmap ( http://insecure.org/nmap/ ) this :
> 
> http://www.lanl.gov/
> 
> ( Linux + Kerberos : http://web.mit.edu/Kerberos/ )
> 
> Did you speak about high security ?

If you feel that it meets your needs, then Linux is a perfectly fine
operating system and you should certainly use it.

Some of the customers for Coyotos need a system that successfully
resists well-funded, professional attacks by Organized Crime and/or
national governments. Linux just isn't there, and I don't think that it
ever will be. I could be wrong, and that would be a very nice thing.

Meanwhile, I have customers to serve.


shap