KUT, etc

[Francois-Rene Rideau]

Dear LLL Tunesers,

On Thu, Jan 07, 1999 at 08:14:54PM -0500, Tom Novelli wrote:
> I just found out about KUT today, and I've been working on something similar
> for a few weeks. (How's that for disorganization :-)
Yeah, and at the same time, Emmanuel Marty is on the verge of releasing
clementine, his own GPL kernel-less system...
A few of us have CVS access (I've sent a few patches already),
and you may nag core@suntech.fr about giving you an account,
without waiting for public CVS access to be available
(which will be read-only, anyway!).

Not surprisingly, the keyboard driver has problems, that vary greatly
depending on the machine on which the system is run (KUT had similar
problems, last time I tried). Down with crappy "bug-compatible" PC
hardware! When there is source-level compatibility, clear-cut
technological transition will be possible instead of that stuff!

> Actually, I wrote a
> bootable Forth interpreter (8086 real mode only) back in August, before I
> joined Tunes.
I invite you to port eforth, oof, or another 32-bit FORTH system to
clementine!

> 	* I'm writing a V86 handler to support BIOS calls from protected
> mode. I have example code to look at, to make this easier.
This is good. Examples include SOLO, OOF, Tran's pmode, too. See
	http://www.tunes.org/Review/OSes.html#Code

> Anyway, we both need an object format, and also a program to read & write
> objects from Linux.
If we write as much as possible in a HLL (at least FORTH),
then we can live independently in Linux and over the raw hardware.
Actually, the oskit may help you do that, too!

> Another problem: The 80x86 doesn't support
> position-independent code, like the 680x0 does. Bummer. More work for us.
Nope. PIC or not, we need linking for inter-module calls, anyway!
So that's not *more* work. But the 80x86 stuff is ugly indeed.

BTW, a very important feature will be the ability to unlink then relink
an object, for purposes of migration, gc, persistence, etc.
This requires access to information usually not revealed by the
linker/assembler.

> Rough sketch of my object format:
I recommend you get direct inspiration from CMUCL or RScheme or OCAML,
rather than invent something new.

Regards,

[ "Faré" | VN: Уng-Vû Bân | Join the TUNES project!   http://www.tunes.org/  ]
[ FR: François-René Rideau | TUNES is a Useful, Nevertheless Expedient System ]
[ Reflection&Cybernethics  | Project for  a Free Reflective  Computing System ]
Some people will argue that since there's no evidence either way whether
the smurf fboinks or not, it's ok to firmly believe that indeed it does.
Such people are insane, often as the result of a life-long endoctrination.