Re: Common Windows

[Daniel A Haug <aihaug@SHRIKE.AUSTIN.LOCKHEED.COM>]

> Date: Thu 9 Mar 89 14:24:09-EST
> From: MAYBURY@RADC-TOPS20.ARPA (Mark Maybury)
> Subject: Common Windows
>  
> Does anyone know about common windows?
We have been using it as our window standard for two years now on the
AirLand Battle Management project.  Intellicorp currently sells a version
of Common Windows, although they plainly admit that they are not in the
window business, but have been using it to facilitate the portability
of KEE.

> 1) Is Common Windows a moving target or has it been standardized.  If so, 
> can one purchase an implementation of Common Windows for any machine running 
> Common Lisp (e.g. the Symbolics, TI, Macintosh, IBM-PC etc.)

Intellicorp sells it for any workstation to which they have ported KEE.
Certainly Symbolics and Suns, and probably the Explorer.  Don't know
about other workstations.  The source is pretty cheap.  I believe we
picked it up for around $300.  About the standard, I believe it is
fairly dead.  It was proposed before any solid oo standard was available
for Common Lisp, and it also came out about the same time that X began
picking up support.  However, I am not really knowledgable about current
window standards in work.

> 2)  How does it work?  Is it a total window system implementation, or 
> does it make calls to the resident windowing system on the machine or is 
> X-Windows involved or something else?

It provides a standardized interface to the underlying window system.
Thus, on a symbolics, it provides a standard set of accessor functions
to a specific flavor of TV windows.  It is a very lack-luster standard,
largely due to the fact that it was developed before CLOS was fully
spec'd.  Instead, it attempts to emulate various aspects of an object-
oriented window system, but falls short.

> 3)  Will it run on unix-based machines running X-windows 
> as well as on Symbolics and TI LISPMs?  Macintoshes?  IBM PC's?

As stated before, it does run on the Sun, but using Sun's native window
system, not X.  I do not know of any common-windows implementation that
uses X.

> 4) Does anyone have any comment on X-windows versus Common Windows?

To suit our purposes, BBN developed a CLOS-based superset of Common
Windows (known satirically as Uncommon Windows), thus giving us the
power of an object-oriented window system and a somewhat portable
window standard.  This strategy has worked out well, and is not terribly
difficult to implement.

For our application, we did not have a need for a network-based window
system, and thus, X was unnecessary functionality.  As a window standard,
I cannot comment on the differences, as I am not an experienced X hacker.

> Thanks in advance.
> 
> Mark


dan haug
haug@austin.lockheed.com