 |
|
||
|
|
Camp Smalltalk (OOPSLA 2000) Thursday |
Sunday Oct 15th 2000
Thursday Oct 19th 2000
Mob Software by Richard Gabriel, found at his site:
http://www.dreamsongs.com/NewFiles/MobSoftware.pdf
You had to be there. It's difficult to describe, Dick's costume is even more difficult to describe (Pictures, Pictures, someone needs to get me a picture), I think the software industry should start a fashion fad, robes are fitting for icons in the industry, not that I'm a icon, but most picture us as the computer geeks in jeans. One wishes one was 20 years younger to see how it will turn out. Our industry is younger than we think. Dick mentioned the shuttle software team, in fact I've a link to an article about them on my readings page. I perhaps should have recorded the session, but I decided to listen carefully instead.
In talking to some of the folks after the presentation, they thought many of the concepts were neat. But one of them told me it would be interesting to unleash say someone from the University of Chicago's Economics dept to attempt understand how the gift model works in this technically driven environment. One can understand how in the past patrons would give money to artists to have them and their possessions painted to carry forward their memory into eternity. But can you do the same for Software? Goodness, we can't even remember who invented the computer languages we use today. Oh and yes Smalltalk and of course Lisp was mentioned. However this was more about who we are and where we need to go, a vision versus the implementation.
Today at Camp, it's not fair to say things were slow. Various people (20+) were sitting around, somewhat suffering from post-conference burnout, that made it slow. Some were still working, more on that later. In general the opinion was the location in the exhibit hall and the chance to show "NO! Smalltalk is not dead", it's doing the tough stuff behind the curtain was great. We had exposure that we've not had for a few years. Like for example:
I was listening to Rodger speak to Domink a person interested in Squeak. Domink was at the conference to present his work about a distributed C++ Corba? system. His background was C++, he had heard about Smalltalk, and here at Camp was an option to see what Smalltalk was, sure the web has lots of information but here you see the man speak. He felt that Java was too restrictive, he struggles programming with it everyday. We did some explaining, pointing to web URLs then showed Domink a Swiki. Later with Michael's support we showed Domink the Disney submarine game. This of course quickly gathered a crowd of onlookers, with some debating about how feasible it would be to do outside of Morphic and Flash.
Rodger then showed me and Domink a presentation about why Rodger stopped using Java. He had put it together with Squeak, not PowerPoint. Its active! Then we had about five people crowd around to get a better view. Squeak as a presentation tool is neat. Powerpoint is so restrictive. Sell Squeak as a replacement, just say you've got this cool open source presentation tool with a scripting language build it, look at the cool slides it can make. Later you can admit it's a programming language. I think he should make the image available, later we convinced him to post the information to the Squeak list about how to get a copy.
To: squeak@cs.uiuc.edu
From: Roger Whitney <whitney@cs.sdsu.edu>
Subject: Why I stopped using Java talk
Earlier this semester I gave a talk in the SDSU CS colloquia series titled
Squeak: Inventing the future again, or why I stopped using Java. One goal
of the talk was to show how on paper Squeak and Java might seem similar
(VM, byte codes, class libraries, OO language, cross-platform, IDE, etc)
while demonstrating the differences between them. The talk was done using
Squeak with PowerPoint slides. (Try doing that in Java). I demonstrated
things like flash, Alice, workspaces, browsers, and morphic. While this is
all very routine stuff in Squeak, it made a big impression on the audience.
I showed parts of the talk to a few people at Camp Smalltalk@OOPSLA. They
were interested in getting a copy of the talk. The Squeak image containing
the talk is available via anonymous ftp at:
ftp://rohan.sdsu.edu/faculty/whitney/squeak/talkImage/Squeaktalk.tar.gz
Roger Whitney Mathematical & Computer Sciences Department
whitney@cs.sdsu.edu San Diego State University
http://www.eli.sdsu.edu/ San Diego, CA 92182-7720
(619) 583-1978
(619) 594-3535 (office)
(619) 594-6746 (fax)
In my last few hours I wandered about and polled the campers about their thoughts. Joseph had been working with Rodger on porting XML into the Gemstone server some of the tests (at least 3) run now. Joseph said Rodger had cleared up a lot of fog about what the issues were. Note of course SUnits 3.0 works for GemStone! Joseph said some people asked him about Smalltalk what is it? Show me! He was pleased with the ability to show off Gemstone and talk to people which he's had no previous exposure to.
I talked to another fellow who had dropped by to see what was up and the concept of greeters came up. Next year signs for tables? (XML here), (Refactory Browser kingdom here). This would make things easier for people who come in to get access to the things they have interest. Ya and since most of us are introverts more roles are required to get people interested. I think Camp Smalltalk at OOPSLA is slightly different that your normal camp. More people are needed to draw people into the mass.
Stefan Aust dropped by with some comments about the difficulties of getting the source code that is contributed to the image into the image and tracked for the community. John Sarkela talked about material available from www.smalltalksystems.com about Smalltalk as Modules, if by magic somehow Allen Wirfs-Brock some how managed to appear. In John Sarkela opinion this should be a reference for the community to use to build Squeak from scratch. In some people's opinion the 2.8 image has become very complex and not understandable from an individual's viewpoint, credit to the SqC for being able to build it, but we can't continue in the same manner.
Micheal Hewner hewner@uiuc.edu was quietly writing a front end to the SmallLint test from the Refactor browser. These tests are valuable and Micheal decided to build a UI for them as the entire Browser interface is too complex and pending completion. Hopefully the basic framework will be done by the weekend. In some respects I've missing running smallLint against my systems and having it whine at me about how my code has problems. Please get me a copy…
Shaun dropped by and asked me about variablebytesublassing and some Objectworks code that creates a subclass but the subclass is of a different data type. This works in Objectworks but is forbidden in Squeak. Having a sequenceable collection with instance variables could be interesting but it looks like a VM change. Mmm it's time to drag the language spec out and see what's legal.
Plane Plane got a plane to catch.
First thanks for all the notes of encouragement.
Second: I've had another Smalltalker volunteer to take notes at the next Camp Smalltalk, this will be great since it's hard to write code and do history at the same time.
Three: I had someone commit to porting the MPEG code (unless of course some Georgia Tech student beats them to it).
And lastly, as Gabriel pointed out it's the community that makes a difference.
So ends another OOPSLA and even if I only got 20 hour sleep for the entire week I still had a great time.