From mboxrd@z Thu Jan  1 00:00:00 1970
Message-ID: <4cb10edc5f742a2850beec63559426f7@coraid.com>
From: erik quanstrom <quanstro@coraid.com>
Date: Tue,  8 May 2007 09:27:34 -0400
To: 9fans@cse.psu.edu
Subject: Re: [9fans] Re: Novice
In-Reply-To: <1178628858.111317.26280@e65g2000hsc.googlegroups.com>
MIME-Version: 1.0
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Topicbox-Message-UUID: 5f863e42-ead2-11e9-9d60-3106f5b1d025

On Tue May  8 08:59:23 EDT 2007, isliguezze@gmail.com wrote:
> Thanks for the attention. I actually didn't mean it's simple, but
> It is elegant, because there is small amount of code, for example
> take 8 1/2 and X window systems - they both are graphical, though
> Plan9 graphics are a little bit strange, it is many times smaller in
> source...

the current system is "rio" --- a successor to 8=C2=BD.

>=20
> Ken Thompson and Rob Pike say that UNIX(R) and all clones of it
> are ancient systems, Plan9 is modern; C language is bad, though
> Limbo is very good.=20

(prefaced with "as i see it".)

the basic problem with unix-type operating systems is they treat the
network in a fundamentally different way than they treat the filesystem.
network resources are named differently, use special-case file-descriptor=
s, etc.

in linux that differences goes all the way down to the slab allocator.
there is no way to hand the vfs a socket buffer.

i don't think you can say "c is good, limbo is bad".  they're different.
c is self-hosting. limbo is not.

- erik