I'm not sure I buy these ad hoc, arbitrary marketing insights about Open Source. "Itch scratching", "consolidation" etc are just buzzwords. I'm sure that emacs and GCC would be surprised about not existing because their not at Oreilly. They don't show up because they realize that Oreilly is pretty much a clique that needs an excuse to spend $1500 a year.
The truth of why people are leaving the community is not because technical inflexibility and obscurity - that's been true since the beginning of the project. We've always been dealing with the fact we're not OO, we don't use a P-lang, we don't use Apache and we don't use MySQL. Nothing new with that.
The reason that people are leaving the community is because the promise that it held in the form of .LRN has been totally unfulfilled.
There are many ways to show how .LRN has botched the lead it had on Sakai but we're all familiar with that. The question is whether anything can be regained either in attractive new developers (which I think is unlikely) or attracting new customers.
Focusing on attracting new developers is a waste of time. We have a technology that no one is terribly excited about putting on their resume (tcl and AOLserver). It doesn't play well in heterogeneous environments. And the learning curve is such that very little previous knowledge can be applied other than general database and programming experience.
These things can be improved in order to attract new developers, but it requires that things like the RAD facilities are improved, the object systems, merged, etc. That's apparently not going to happen anytime soon given the resources we currently have. Expecting that we can attract new resources that will take up these hard problems is wishful thinking.
Attracting new customers, though, may be possible and that is only through marketing. Marketing, though, is the major competence that is lacking in the OpenACS and .LRN communities. While OpenACS doesn't need marketing competence - we are a developer community, if we had it, great, if we don't we'll survive - .LRN's incompetence in this area is both discouraging and mind boggling.
It certainly doesn't need to be that way. .LRN has been around for 4 or 5 years or so and has a fairly impressive roster of users. But it's leadership has been disorganized in gathering momentum and shown a lack of creativity in managing to do the most important thing to make the project effective - raise money.
That is fundamentally the important issue here. If .LRN doesn't raise at least $1M in short time, it is doomed. That money should go to marketing itself better and getting out of the obscurity in which it languishes. Once that happens, .LRN should hire developers to document the system and and make RAD more accessible.
But the chances of that happening are increasingly slim. So why should anyone want to work with a group that can't pull itself together after 5 years of existence?
talli
