Date:         Sat, 20 Nov 1999 10:26:52 +0200
From: Annick Fron 
Subject:      [BSUG] OOpsla report

I saw too much at OOPSLA to detail in any reasonably sized email.
Hopefully others that attended will contribute their observations. Here
is my decidedly high-level OOPSLA '99 review from a Smalltalk
perspective:

Very exciting!

The mood has changed since OOPSLA '98 (in Vancouver). Last year the mood

could be summed up as anxiety and frustration. Anxiety about what
GemStone, Sun, and rest of the market will do (most obvious for our
Smalltalk customers). Frustration because so much of the industry was
under pressure to take two steps backwards (in terms of OO development)
to take one giant step forward (e-commerce). [...]

The mood has also changed since Smalltalk Solutions '99. For the most
part, the mood was much like it was at OOPSLA '98; however, there was
also a feeling of desperation and resolve. Desperation because it was
obvious to Smalltalkers that something had to be done to meet the
challenge of Java. Resolve because some vendors were already rising to
meet the challenge. IBM announced WebSphere and ultra-light-clients.
ObjectShare released VW 5i (with support for web browsers). Applied
Reasoning released a new version of Classic Blend. GemStone introduced
GSNC. There were slight signs of optimism in the air among Smalltalkers.

The mood last week at OOPSLA '99 (in Denver) was chiding and optimistic.

Java seemed to have a large presence at the show, but was consistently
and publicly criticized in very subtle ways by most well known people in

the industry. The most humorous of which was a mock trial of the "gang
of four" in which defendants were attacked for adding usefulness to the
language of "curly brackets". Smalltalkers seemed far less threatened by

Java. Java was largely viewed as less likely and capable of taking over
the industry. On *six* occasions I spoke with Java developers that were
interested in learning some wonderful language that everyone had been
talking about (Smalltalk). [...]

Smalltalk vendors had a great deal of excitement to offer. Cincom showed

dedication, vision, and an emphasis on the Linux market. Both
Smalltalk/X (from eXept) and Mission Software/KSC showed revolutionary
technology to bridge the gap between Smalltalk and Java. Squeak put on a

killer presentation that is bound to generate more Squeak evangelicals.
IBM was rumored to be busy working on their Linux port.

Some apparent trends are:
--Smalltalk attracting new grassroots developers into the industry
through exciting low-cost products.
--Smalltalk starting to be recognized as the best choice for building
maintainable enterprise applications.
--Smalltalk starting to be observed as a non-proprietary language with
performance superior to Java.
--Vendors continuing to develop products that bridge the gap between
Smalltalk and Java.

All is not positive for Smalltalk though. Smalltalk is as fragmented as
Unix while Java is as ubiquitous and defacto of a standard as MS
Windows. Smalltalk's lack of native threads is one of the few advantages

Java has over most Smalltalk dialects. That one issue will likely cause
new projects to choose Java. Smalltalk's diversity between dialects is
likely to be a significant problem that would likely retard long term
acceptance and growth. Lesser known Smalltalk dialects (like Squeak,
Smalltalk/X, and Dolphin) now offer some specialized advantages over the

current market leaders (VisualAge and VisualWorks). This, combined with
lower product costs, means that there will be more challenges and
opportunities over the next couple years.

It may be two years before Smalltalk gets considerable attention in
trade rags. By next OOPSLA, I expect it will be vogue for developers and

consultants to be knowledgeable enough of both Java and Smalltalk to
present both as real options for new projects. By next year, I expect
that most new projects will still be Java based, but that Smalltalk will

have a large and growing acceptance at a rate approaching that of Java.
As Dale Henrichs observed, GemStone/S (and Smalltalk in general) has
traditionally been driven by developers. Java and C++ was driven by
marketing and peer-pressure. Developers are likely to be the driving
force behind future GemStone/S sales--and will be heard more often as
Java hype diminishes or transforms as C++ hype did years earlier.

And the fun continues...


Paul Baumann
Software Engineer
GemStone Systems, Inc.
pbaumann@gemstone.com