I haven't been playing around with CMU/CL lately. I was wondering what
people though of porting CML/CL to the upcoming Rhapsody. Difficulties?
Potential? Work in Progress?
Rhapsody will be the first NeXTSTEP (Unix) based Mac OS. Presumably running
the new (now faster cheaper and cooler than Pentium) PowerPC.
--
William P. Vrotney - ·······@netcom.com
·······@netcom.com (William Paul Vrotney) writes:
> I haven't been playing around with CMU/CL lately. I was wondering what
> people though of porting CML/CL to the upcoming Rhapsody. Difficulties?
> Potential? Work in Progress?
>
> Rhapsody will be the first NeXTSTEP (Unix) based Mac OS. Presumably running
> the new (now faster cheaper and cooler than Pentium) PowerPC.
CMU CL was developed initially under Mach, so making it run under
NextStep should be pretty straightforward. If Rhapsody retains Mach's
copy-on-write capability, you could even run a gang of Lisp process on
the same machine, sharing all the pure pages.
However, interfacing CMU CL to a user interface written in Objective C
might be bloody. You could either cobble an interface together by
hand or try to build an automatic wrapper-generator. The latter would
be cooler, but harder.
Producing a compiler back-end for the PowerPC would be a big job. We
never started that, and I don't know of anyone else working on it. I
don't know of any special problems that this architecture would cause.
In fact, CMU CL was originally developed on the IBM RT PC, which was a
direct ancestor of the PowerPC RISC architecture. But doing a new
back-end is still a big job. Just how big depends on how much you
already know about our compiler.
Personally, I think that Apple/Next will probably be out of business
before you could finish this task.
-- Scott
===========================================================================
Scott E. Fahlman Internet: ···@cs.cmu.edu
Principal Research Scientist Phone: 412 268-2575
Department of Computer Science Fax: 412 268-5576
Carnegie Mellon University Latitude: 40:26:46 N
5000 Forbes Avenue Longitude: 79:56:55 W
Pittsburgh, PA 15213 Mood: :-)
===========================================================================
Scott Fahlman <···@clyde.boltz.cs.cmu.edu> writes:
[...]
>However, interfacing CMU CL to a user interface written in Objective C
>might be bloody. You could either cobble an interface together by
>hand or try to build an automatic wrapper-generator. The latter would
>be cooler, but harder.
I'd like to point out that Objective-C is a very brave cicitzen when
it comes to integrate it into other languages (compared to C++ and
such). Using an ObjC based GUI framework from languages that interface
to C might be very straightforward.
Most features of Objective-C are implemented as a normal C
library. This is radically different from C++ where you need a lot of
compiler support. ObjC's compiler support is just some syntactic
sugar, text-replacing stuff only.
ObjC's syntactic extensions can easily be left unused and you
can call the ObjC runtime functions from pure C code. Other things
like object creation, object copying and conversion are either
expressable as C function calls also or have been ignored by the ObjC
designers anyway :-)
And ObjC is quite reflective, the runtime tells you a lot about the
objects, classes and messages you can use.
[...]
>Personally, I think that Apple/Next will probably be out of business
>before you could finish this task.
There's nothing like a good race :-)
Martin
--
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
···············@wavehh.hanse.de http://cracauer.cons.org Fax.: +4940 5228536
"As far as I'm concerned, if something is so complicated that you can't ex-
plain it in 10 seconds, then it's probably not worth knowing anyway"- Calvin
In article <··············@CLYDE.BOLTZ.CS.CMU.EDU>, Scott Fahlman
<···@clyde.boltz.cs.cmu.edu> wrote:
> Producing a compiler back-end for the PowerPC would be a big job. We
> never started that, and I don't know of anyone else working on it. I
> don't know of any special problems that this architecture would cause.
> In fact, CMU CL was originally developed on the IBM RT PC, which was a
> direct ancestor of the PowerPC RISC architecture. But doing a new
> back-end is still a big job. Just how big depends on how much you
> already know about our compiler.
>
> Personally, I think that Apple/Next will probably be out of business
> before you could finish this task.
Another twenty years to write a compiler backend? Is it that difficult?
--
http://www.lavielle.com/~joswig/
·······@netcom.com (William Paul Vrotney) writes:
>I haven't been playing around with CMU/CL lately. I was wondering what
>people though of porting CML/CL to the upcoming Rhapsody. Difficulties?
>Potential? Work in Progress?
>Rhapsody will be the first NeXTSTEP (Unix) based Mac OS. Presumably running
>the new (now faster cheaper and cooler than Pentium) PowerPC.
Porting to a new processor is a major effort, although the PowerPC
might reuse code from the old PC/RT port. I watched the x86 effort and
can honestly say it took a real wizard to glue things together. The
number of people involved was irrelevant, the thing took up on the
base of one individual's restlessness :-)
Porting to a new OS is easier, I heard someone is already approaching
the BSD server on NeXTStep/Intel.
The mailing list archives of the x86 port effort are a must-read for
anyone interestedt in this. See http://www.cons.org/cmucl/
Martin
--
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
···············@wavehh.hanse.de http://cracauer.cons.org Fax.: +4940 5228536
"As far as I'm concerned, if something is so complicated that you can't ex-
plain it in 10 seconds, then it's probably not worth knowing anyway"- Calvin