I just read the latest column of "I, Cringely", which talks about the different view of M$ and Linus concerning the coding abilities of a company vs. many distributed developers. At the peak of it's time AD had I'd say around 100 coders working on ACS TCL in client projects. They did not care that much anymore about what the outside world was doing, but drove their truck at a speed defined by marketing and project management (aka. business guys). Though a momentum like this is worthwhile it bears the risk of releasing code when it is not ready to be released. Which, in effect, alienates other developers because of the crappy maturity of the code, regardless how easy it is to get going with it. So I do not think that the existance of aD contributed to the groove of ACS. The existance of Philip, the marketing machinery and the fun environment of people with a huge ego, but willingness to help (as part of their ego) did attract some people to ACS, more than the existance of AD.
Now, looking at this community, I see we are releasing well tested code that has put a lot of thinking behind it. What we lack though, is to release the knowledge, how to use the code, as the methods got considerably more sophisticated. Furthermore, what happened to the concept of Bootcamps. I know, there is no AD anymore, that could run a bootcamp on it's own (well, okay, we do it occasionally, but that does not really count, as we did not open them up to the public so far), but why did this stop. Furthermore, how many universities are using dotLRN? How many are teaching how to use it?
Getting fresh people into the toolkit through a university course or a bootcamp will give you a lot of feedback, how to make it easier to develop with OpenACS. Having a teaching forum for exchange between bootcampers and university students across bootcamps or courses, would be a good place to collect all this feedback, which in return could make it back into the documentation effort and maybe even into improvements for the toolkit.
And would it be too much to ask for developers in the community, to commit one week of their time to teach a bootcamp? Or give a class at a university next to them.
One of the strengthes as Walter pointed out, is dotLRN. But not necessarily only because of the (monetary) lifeforce (though it looks like that most developement nowadays is dotLRN driven, at least if you look at hours spend on development), but because of it's potential to attract new developers and become a standard in the world. I found it very intersting that the language with the most translations in the latest CVS checkout is Arabic. I doubt Blackboard or WebCT would have an easy sell in the Arabic markets nowadays.
