Path: utzoo!attcan!uunet!cs.utexas.edu!tut.cis.ohio-state.edu!bloom-beacon!ATHENA.MIT.EDU!wdc
From: wdc@ATHENA.MIT.EDU (Bill Cattey)
Newsgroups: comp.windows.x
Subject: Re: Toolkits & User Interface Deevolution
Message-ID: <8YbmmboGG0471hEIIO@athena.mit.edu>
Date: 21 Jun 89 05:50:31 GMT
References: <1552@harrier.ukc.ac.uk>
Sender: daemon@bloom-beacon.MIT.EDU
Organization: The Internet
Lines: 64

There are the beginings of a user interface builder and an extensible
toolkit environment right there on your X tape.  It's the Andrew
Toolkit, and it's ADEW subsystem.  The ADEW system was featured in a
talk at January's X conference at MIT.

It has some major problems:

1. It's not compatible with the X toolkit intrinsics.  Interestingly,
it's window server independent graphics layer does some useful stuff
that was missing from the XVT system.

2.  It is not user-interface policy independent.  I like to think of it
the way we used to talk about the Bell System:  One system -- it works.

3. It contains an operating system specific dynamic object loading
system.  This took two person months for Olivetti to port to their UNIX
port, but it _was_ UNIX dependent.

4. Some coding is still required.  But the modules are templated with
helpful comments automatically generated showing where the user has to
add code.

5. It is 400,000 lines of C code that is intimidating to get running. 
This code yeilds a different, simpler intrinsic toolkit layer, the
dynamic object loader, a multi-font text editor, a rich widget set,
including geometry management, and a variety of high level applications,
such as a drawing editor, a spreadsheet, a raster editor, and a
multi-media mail interface.

----

I have been trying to motivate development with this stuff along several
fronts:

1. Getting people to at least look at the stuff.  Maybe compile it and
try it out and think about it.

2. Discussion of what unique features and technology the ATK stuff has
to offer.

2. Re-consideration of the X toolkit intrinsics in the light of the
emerging Motif user interface policy.

3. Development of useful applications with the ATK objects which offer a
much wider diversity of functions than presently existing X intrinsics
based widgets.

4. Development of code that will enable applications written using the
unique ATK abilities to migrate to what eventually becomes the
full-featured industry standard.

----

Here at Athena, the home of X, we also look at other useful and
interesting technologies.

People interested in cooperating on studying, using, evaluating, and
developing using contrib/toolkits/andrew can contact me directly or can
enroll in the CMU based andrew mailing list by sending mail to
info-andrew-request+@andrew.cmu.edu

-wdc

wdc@athena.mit.edu