[Oz] Oz vs. Squeak

Mikael Kindborg mikki at ida.liu.se
Tue Jan 30 11:30:17 CET 2001


Hi again.

While I am at it I thought I should bring up an issue I have been
thinking about for some time. During the last 2-3 months I have
been exploring Oz/Mozart and Squeak in parallel (Squeak is a Smalltalk
implementation, see http://www.squeak.org). I can see strong points and
weak points with both efforts and would like to hear the reactions from the
Oz community on the following.

To begin with, Squeak appears to have a massive user community, and
the creators of the language are very active on the mailing list which
has up to something like 25 messages per day. At this mailing list
there are only a few mesages every month. I am certain that a lot of
activity is going on with Oz at SICS and other places, but as an outsider
it is difficult to take part in these activities.

A related issue is the update frequency of the websites. The Squeak
site uses a wiki and is updated frequently by many people. The most recent
entry of the Oz site news page is dated Feb 29, 2000. I visit the Oz site
quite often, and I have a hard time to detect any updates. Still I have
heard of for instase Oz-Lite, Oz and Java bindings, and I have seen
Oz for .Net mentioned on the Mercury webb (if I recall correctly). It would
be good to have this information first hand, either on this mailing list
or at the Oz website (or both).

Squeak has an extremely appealing user interface and development
environment. To me, who has a background in HCI-research and
graphical user interfaces, Emacs is not the development environment
of choise. Also, in Sqeak, since it is Smalltalk, all the graphics and
all other "library" classes are written in Sqeuak and thus "easy" to
browse, use and modify. The approach with the Tk graphics module
feels "closed" by contrast. I would want an Oz programming environment
similar to Smalltalk/Squeak, written in Oz itself.

By now some of you may wonder why I don't go ahead and use
Squeak rather than Oz. The reseaon is that I percieve the computational
model of Oz as being based on a much more modern approach. The
use of locgic variables and the concurrency support paired with the
distributin support feels to me very superieor to what I have seen in
Smalltalk (where concurrency for instance does not work well, often
one process has exclusive use of the machine and the system "freezes"
until it is finished).

To summarise, Squeak seems strong in the areas of community
support, graphics, its integerated approach and its programming
environment, while Oz seems much stronger in the support the language
has for concurrency and networking (I am aware that Oz has interesting
stuff for constraints etc, but this is not my main area of interests).

What I am asking is if there is any interest in the Oz-community for
pursuing programming environment and user-interface issues?
I would love to use Oz for my research (for buildning end-user
programming tools) but right now I feel that this is not happening.
Another approach that might interest the readers of this forum,
could be to add Oz concurrency and networking constructs
to Squeak. This would certainly requrite a major rewrite of the
Squeak VM and likley require a new language semantics for
Smalltak, but it woul be a way to combine the best of two worlds.

Best, Micke

-
Please send submissions to users at mozart-oz.org
and administriva mail to users-request at mozart-oz.org.
The Mozart Oz web site is at http://www.mozart-oz.org/.





More information about the mozart-users mailing list