[coyotos-dev] Status and roadmap

[Jonathan S. Shapiro]

On Tue, 2007-01-16 at 07:26 +0100, Pierre THIERRY wrote:
> Scribit Jonathan S. Shapiro dies 15/01/2007 hora 12:40:
> > 1. Get CapIDL working, so we can document interfaces and
> > specifications and generate some critical header files that the kernel
> > needs.
> > 
> > 2. Get the system image builder working, so that we can start building
> > test images for the kernel.
> > 
> > 3. Implement kernel, core utilities for embedded.
> 
> What languages do you plan to use to write them?

At the moment, these will be done in C (kernel) and C++ (CapIDL,
MkImage).


>  Is it already planned
> who will work on what, when, which somewhat means, where is it possible
> to help without getting on someone's feet?

Let me say first: we would really *welcome* the help!

CapIDL and MkImage are going to be done too soon to really be able to
help with them. The kernel is something that is mostly specified, and I
think it would slow us down to take on more people there. There are a
few places that I will do, but already I am trying to stay out of it so
that Jonathan Adams has room to work unmolested.

Above that layer, there do start to be things that others could do. As
soon as CapIDL is working we can start writing the interface
specifications for those things, and it will become possible for others
to implement them.

> Parcularily, separately for each one, would you accept a working
> implementation in Lisp (because it's the language I'm probably the most
> efficient with, currently)?

No. Once we get past CapIDL and MkImage, everything we are building
needs to run natively (on Coyotos), and we have no lisp environment for
Coyotos yet. If support for LISP-like languages is the place where you
want to help, and you know LISP, one obvious candidate is to work on
getting BitC re-implemented in BitC to the point where it is
self-bootstrapped and has a stand-alone runtime. The BitC project is
still hosted at Hopkins, but all of this stuff will eventually help us.

> Would some these parts of the system also need specifically
> documentation and analysis to be published (like papers that were
> written on the window and network systems in EROS)?

Yes.

> >   HARD deadline for all of above is to be complete by October 7, 2007
> 
> Definitely too long for my limited patience in this field. I *need* to
> help.

Oh, don't worry. We'll find something for you to do. :-)
-- 
Jonathan S. Shapiro, Ph.D.
Managing Director
The EROS Group, LLC
+1 443 927 1719 x5100