I'm writing this nearly 5 years into SAIL.
I first became interested in a common system to support education technology research during a CILT conference in 2002
I had just completed a software framework that took content and turned it into html/pdf/ and (the hard one) interactive handheld activities. The handheld version of the activities recorded student work so a student could answer questions and then come back and continue the work later. The framework was modular so new components could be added to the activities without modifying the framework itself. This framework was called LabBook. It was used by both the TEEMSS project and Data and Models project.
This CILT conference was focused on "Ubiquitous, Mobile Computers in Education". I realized that there were several groups doing very similar things to Concord. And we were all wrestling with similar problems. At the end of the meeting there was a chance for participants to make wrap up comments. Many people talked about what they wanted to build to support future development. I stood up and said, "Hey I've got a framework which does a lot of these things, if we all worked together it would be good for all of us". I probably didn't say exactly that, but that was my intention. There was no response to this. At the time I thought the other groups were dumb for not pursuing this. I now think they were just being rationally cynical about the success of such a venture.
Soon after this conference the TEEMSS project ended which was funding my work on the LabBook framework. I now needed to work on the MAC project which was well underway using Edward Burke's Pedagogica framework. I started learned this new framework and found several parts of it that I wanted to change. I then spent several months not want to just throw out the work I had done on the LabBook.
I spent several months trying to make the LabBook code portable so it could be used both in its initial environment "Waba" and the more common environment Java. Eventually, I had to stop this work as there were more pressing needs in the MAC project, and the LabBook code would not solve them. In the beginning there were 2 needs for MAC, the deployment system in use was not working well, and there wasn't a good system for doing the logging. I started worked on both problems, but ended up only being able to really work on the deployment system. This turned into the Workspace framework. It was a dependency management system, a client for launching activities, which including a login system so different sets of activities could be presented to different users. It also supported multiple caching and logging configurations so it would work in various network configurations found in schools. It was modular so new "activities" could be added by creating new Java classes that extended particular interfaces. It was modular enough so it could have been extended to define the activities themselves. In end it was approaching the functionality of LabBook framework, but it didn't have a way to save user data. The user data in the MAC project was stored as log files which were saved locally, posted to the server, and deleted if the post was successful. These files were not used in the activities themselves. The Workspace framework used CVS as it content management system. Activity authors would check in their changes to CVS which would make them available to testers in a special testing workspace. And then the jars and pedagogica files would be "released" to a production workspace so the schools would get the changes.
During this time I also worked on build environment to support various jar files of the MAC project. This build system ended up being used. By MAC project and the TEEMSS2 project.
Around Nov. 2003, the Seeing Math project started up and needed my attention. I started by trying to make the workspace code more flexible so it could be used to deploy the interactives needed by the project. This was not used because the project opted for using applets which were easier to integrate in their web content.
We also tried to find a way to make the TEEMSS authoring system more user friendly. The goal was to author the Seeing Math content in a presentation independent way so it could be shown as web pages, built into Blackboard course imports, pedagogica activities, LabBook activities, or just nice looking pdfs. We hoped this would be easy to do using a wysiwyg xml editor like XML Mind. We contracted a consultant to customized XML Mind to do this work. But eventually we realized there were too many parts of this design that would not work well. And the project resorted to just authoring using Dreamweaver and saving the html files in CVS. And then manually importing them into Blackboard.
Soon after (Jan. 2004) I started working part time on the TELS project. Initially trying to work on the workspace code so it could be used for the TELS project. And at the same time I was working on designing the technology for the TEEMSS II project.
In March 2004 I attended the first SAIL retreat. It was called OWL at that time. We spent much of the retreat debating about the pluses and minuses of a web based framework versus a Java client framework. In the end we settled on a Java client framework. Based on my work in TEEMSS and vision for the TEEMSSII project I suggested we try to make the activities work both as web based and Java client based. The group decided this was too much work so it would be kept in mind but not actively pursued. I also suggested we base the OWL framework off of my Workspace framework, but I wasn't able to describe it well enough to convince anyone.
At this same time a collaboration between myself and 2 other developers at Concord began to make a common graph component. I had made a graphing component for the TEEMSS project, and was going to need to use one for the TEEMSS II project. So this seemed like a great place to collaborate and share a single component. Instead of modifying one of our existing graphs we decided to design it from scratch.
The technology for the TEEMSSII project was in the planning stage. And I was working on a design for talking to multiple Sensor interfaces.
From March 2004 - NOW we had regular SAIL meetings designing and discussion what would be called SAIL.
July, 2004
I started working on a distributed object store to be used in TEEMSSII this was called CC Portfolio at the time.
Worked on the Sensor library so we could demo this at the August TELS retreat.
August 2004
Had the TELS retreat, I had to fill in for Edward to support a new version of Pedagogica that provided a new kind of authoring, where a scripter could create an activity that was a interative authoring environment to create similar scripts.
Continued work on distributed object store.
Sept.
TELS - Added simple html component into sail framework
TEEMSSII - started work on loading portfolio components from xml. (beginning of OTrunk)
Oct.
TELS - researched use of IoC containers for SAIL with Tony
TEEMSSII - first reference to OTrunk - Refactor portfolio into two packages OTrunk and Portfolio.
Nov. 2004
TELS - SAIL retreat [fill in more about the retreat here]
TEEMSSII - designed sensor device framework