In file behave/behave.lisp, function behave, use *default-pathname-defaults* 
as binding for homedir, replace (user-homedir-pathname) everywhere

In file newcomp/compile.lisp, function subcompile, use *default-pathname-defaults* as binding for homedir

behave/behave.lisp: function "structurep" is not C.L., so define explicitly
(defun structurep (something) (typep something 'structure))

behave/bcmood.lisp: change order of parameters of "direct-activate" into:
(defmacro direct-activate (&whole w behavior percentage) 
(because &whole parameter has to come first)

newcomp/frags.lisp: 2nd arg of format cannot be nil in C.L.,
so you need to change function pp-comment-body, add the test:
 (if (not (null (cdr comment)))

1,,
Summary-line: 24-Jan                 to: brooks  #behavior language on suns
Return-Path: <pattie@ai.mit.edu>
Received: from vub (vub.ai.mit.edu) by life.ai.mit.edu (4.0/AI-4.10) id AA14668; Wed, 24 Jan 90 23:34:21 EST
From: pattie@ai.mit.edu (Pattie Maes)
Received: by vub (4.0/AI-4.10) id AA00295; Wed, 24 Jan 90 23:36:51 EST
Date: Wed, 24 Jan 90 23:36:51 EST
Message-Id: <9001250436.AA00295@vub>
To: brooks
Cc: ariel@ai.mit.edu, pattie
In-Reply-To: Marty Hiller's message of Wed, 24 Jan 90 22:44:05 EST <9001250344.AA01265@lucky-charms>
Subject: behavior language on suns

*** EOOH ***
Return-Path: <pattie@ai.mit.edu>
From: pattie@ai.mit.edu (Pattie Maes)
Date: Wed, 24 Jan 90 23:36:51 EST
To: brooks
Cc: ariel@ai.mit.edu, pattie
In-Reply-To: Marty Hiller's message of Wed, 24 Jan 90 22:44:05 EST <9001250344.AA01265@lucky-charms>
Subject: behavior language on suns


One of the bugs Marty ran into is "function structurep undefined".
There are a bunch of such small differences between Sun Lisp and
Allegro Lisp which I suspect we'll run into.  I suggest that the
necessary changes are made to the Mac source files (i.e. make them
Common Lisp), instead of having all sorts of patches for the Sun
version. OK? So here's an initial list.  Shall I take care of them, or
are you going to?

1) function "structurep" is not C.L., so define explicitly

2) bcmood.lisp: change order of parameters of "direct-activate" into:
(defmacro direct-activate (&whole w behavior percentage) 
(because &whole parameter has to come first)

3) frags.lisp: 2nd arg of format cannot be nil in C.L.,
so you need to change function pp-comment-body, add the test:
 (if (not (null (cdr comment)))

I have a script to compile everything. 



1, filed,,
Summary-line: 24-Jan                 to: brooks  #behavior language on suns
Return-Path: <pattie@ai.mit.edu>
Received: from vub (vub.ai.mit.edu) by life.ai.mit.edu (4.0/AI-4.10) id AA14668; Wed, 24 Jan 90 23:34:21 EST
From: pattie@ai.mit.edu (Pattie Maes)
Received: by vub (4.0/AI-4.10) id AA00295; Wed, 24 Jan 90 23:36:51 EST
Date: Wed, 24 Jan 90 23:36:51 EST
Message-Id: <9001250436.AA00295@vub>
To: brooks
Cc: ariel@ai.mit.edu, pattie
In-Reply-To: Marty Hiller's message of Wed, 24 Jan 90 22:44:05 EST <9001250344.AA01265@lucky-charms>
Subject: behavior language on suns

*** EOOH ***
Return-Path: <pattie@ai.mit.edu>
From: pattie@ai.mit.edu (Pattie Maes)
Date: Wed, 24 Jan 90 23:36:51 EST
To: brooks
Cc: ariel@ai.mit.edu, pattie
In-Reply-To: Marty Hiller's message of Wed, 24 Jan 90 22:44:05 EST <9001250344.AA01265@lucky-charms>
Subject: behavior language on suns


One of the bugs Marty ran into is "function structurep undefined".
There are a bunch of such small differences between Sun Lisp and
Allegro Lisp which I suspect we'll run into.  I suggest that the
necessary changes are made to the Mac source files (i.e. make them
Common Lisp), instead of having all sorts of patches for the Sun
version. OK? So here's an initial list.  Shall I take care of them, or
are you going to?

1) function "structurep" is not C.L., so define explicitly

2) bcmood.lisp: change order of parameters of "direct-activate" into:
(defmacro direct-activate (&whole w behavior percentage) 
(because &whole parameter has to come first)

3) frags.lisp: 2nd arg of format cannot be nil in C.L.,
so you need to change function pp-comment-body, add the test:
 (if (not (null (cdr comment)))

I have a script to compile everything. 



1, answered,,
Summary-line: 24-Jan          brooks@ai.mit.edu  #behavior language on suns
Return-Path: <brooks@ai.mit.edu>
Received: from rice-chex (rice-chex.ai.mit.edu) by life.ai.mit.edu (4.0/AI-4.10) id AA14747; Wed, 24 Jan 90 23:43:21 EST
From: brooks@ai.mit.edu (Rodney A. Brooks)
Received: by rice-chex (4.0/AI-4.10) id AA13293; Wed, 24 Jan 90 23:43:19 EST
Date: Wed, 24 Jan 90 23:43:19 EST
Message-Id: <9001250443.AA13293@rice-chex>
To: pattie@ai.mit.edu
Cc: ariel@ai.mit.edu, pattie
In-Reply-To: Pattie Maes's message of Wed, 24 Jan 90 23:36:51 EST <9001250436.AA00295@vub>
Subject: behavior language on suns

*** EOOH ***
Return-Path: <brooks@ai.mit.edu>
From: brooks@ai.mit.edu (Rodney A. Brooks)
Date: Wed, 24 Jan 90 23:43:19 EST
To: pattie@ai.mit.edu
Cc: ariel@ai.mit.edu, pattie
In-Reply-To: Pattie Maes's message of Wed, 24 Jan 90 23:36:51 EST <9001250436.AA00295@vub>
Subject: behavior language on suns


I agree---lets fix the mac sources.  Why don't you make the changes on
the file server, and let me know which files you change so I can
propagate to my development versions.

If you use loadup then meta-h in emacs sets up a new edit history line
at the top of the file.

