We had a grande olde timee. With three different speakers and plenty of beer I think this was one of the best uSmalltalk User Group meetings I've never been to. I started my day of by presenting at Cincom Australia which gave me a chance to polish for the STUG meeting that night.
Then when we got to the venue, we had some dinner and people started to gather around. We had Smalltalkers and Rubyers and we kicked off the presentations with myself, presenting Seaside followed by Web Velocity. This was the first audience to see Web Velocity in action and I got lots of useful feedback from them. The reception to it was quite warm so I was very pleased.
Next up wite had Myles Byrne who presented his forst attempt at a Seaside application. It took the rapproach of 'don't give me tree views, i don't like clicking lots to read my code' -- it was a code browser in the web browser and used lots of neat Ajax, such as an autocomplete when you wanted to navigate to a different class, etc.
It turns out this was an experiment to see what programming in Seaside was really like. Myles liked it but decided that the experiment itself was a bit of a failure. The crowd disagreed, giving him lots of encouragement to keep going as it was an interesting thing to see done.
Next up was Bruce Badger, all the way from London town. He gave a talk on SLAPS, a Smalltalk PDA interface that can act as both client and server. You can actually query the Smalltalk image through the LDAP protocol. This is quite cool as it might mean we can query a running Smalltalk image from Jabber. Right now, they can use Smalltalk as an authentication service for PAM used by Postgres.
Bruce then moved on to talk about the Smalltalk ANSI process and how what the community wanted - an open democratic debate on how to evolve Smalltalk - is at significant odds with what the ANSI process wants - a money paying insular set of overlords who dictate the standard. As a compromise Bruce hopes to setup a Python PEP like process called STEPs (Smalltalk instead of Python) which will drive community discussion and change, pushing the approved STEPs up to the ANSI committee who will then produce the revised ANSI specification.
Next up was myself again and I presented Searchlight, as there were people intereseted in seeing better ways to jump around the image as quickly as possible. We jumped about the image, looked at the different ways you could explore and nagivate and there was some interesting feedback from that too. After a bit of jumping around we moved on to OpenGL in VisualWorks. Many there had seen OpenGL before and even knew how to do it, so while they were excited to see that the speed of OpenGL in VW was exceptionally fast and every appreciates pretty pictures, it was decided we'd move on to the last and final topic of the evening - Xtreams.
Xtreams is a rewrite of Smalltalk-80 Streams APIs with a reference implementation in VisualWorks. It lets you stack streams on top of streams, it has a core 'transform' concept and buffering and it undoes a few of the mistakes that were made in the original design that don't match up with how operating systems actually work any more. There was a great deal of interest in this and the further we got in to the discussion, the more pleased I was to see that the various requests for functionality already existed in our code base and the various concerns were already addressed too.
That wrapped up our evening and we had a few more beers before heading to our respective homes/hotels. It was a very fun night and the energy level from the crowd was quite high. I really liked the way the evening turned out - having multiple talks that were all shorter than the norm was a good format.
Thanks for having me Sydney Smalltalk Users Group!