[prev in list] [next in list] [prev in thread] [next in thread] 

List:       9fans
Subject:    Re: [9fans] Plan9 advancements over Unix
From:       erik quanstrom <quanstro () coraid ! com>
Date:       2007-07-27 15:56:36
Message-ID: 1e314e347f4ab316668ef8d6e187c01b () coraid ! com
[Download RAW message or body]

> - Why was 8 1/2 abandoned? Is it because the graphics is much advanced
> over bitmap operations? Does plan9 use Xorg now? I think at least the
> idea of having a single private console sounds good.

plan 9 uses rio, which is a decendent of 8 ½.  plan b, a close relative of plan 9,
uses omero.

> 
> - Having private namespaces means maintaining a namespace *per-
> process*. Would this not be more complicated than having a single one?
>

in principle, each process may have its own namespace.  in practice large
numbers of processes do share the same namespace.

this is simplier than having one global namespace.  one prototypical use
of this in plan 9 is the handling of network stacks.  on plan 9 i can
import the network stack from a gatway that has the access i need and
bind it on /net (the place where ip networking stuff lives).  processes
started in this namespace will do their networking through the gateway,
while other processes in the system are oblivious.

> Is there a paper that compares or evaluates how well Plan9's new
> features performed over Unix?

i don't think plan 9 has new "features".  it's quite different.  "devices"
are implemented by fileservers, for example.  thus it makes not difference
(to the client) if the "device" is in the kernel, in userland or on another
machine.

- erik
[prev in list] [next in list] [prev in thread] [next in thread]