Defsystem

[Scott.Fahlman@B.GP.CS.CMU.EDU]

Mark Kantrowitz sent me the following.  I'd like Blaine to pick this up,
torture it a bit, and if it's not obviously busted, stick it in the
Lisp library.  Some of us might even want to use this, given the number of
different systems being built.

-- Scott

------- Forwarded Message

replied: Thu, 01 Feb 90 23:20:44 EST
Return-Path: <Mark.Kantrowitz@A.GP.CS.CMU.EDU>
Received: from a.gp.cs.cmu.edu by B.GP.CS.CMU.EDU id aa21512;
          1 Feb 90 19:36:36 EST
Date: Thu, 1 Feb 90 19:23:51 EST
From: Mark.Kantrowitz@A.GP.CS.CMU.EDU
To: sef@B.GP.CS.CMU.EDU
Subject: defsystem
Cc: mkant@A.GP.CS.CMU.EDU

I recently updated my defsystem macro to have a lot more features and
to be Common Lisp portable. [I started writing it for work I was
doing at MIT before I came here, and updated it for use in the OZ
system.] 

It does a topological sort of the DAG, uses nice defaults, and gives
the user lots of options on what gets compiled/loaded.

I've used it in CMU Common Lisp, Allegro CL (PMAX and vax), and 
Allegro/Coral CL (Macintosh). It's at least as powerful as the System 
Construction Tool on Symbolics Lispm's.

My advisor suggested I send you mail about it in case you would be 
interested in including it with CMU Common Lisp.

If you're interested, source is in /afs/cs/user/mkant/Defsystem/defsystem.lisp
and some notes on how to use it in 
/afs/cs/user/mkant/Defsystem/Documentation/defsystem.{mss,ps}

A couple of people in the MIT AI Lab and Lincoln Laboratory are using
it. A good example of a real system that uses it is in
/afs/cs/project/oz/oz2/mark/FrameKit/framekit.system

- --mark

------- End of Forwarded Message