A year or two before Sakai began, I tried experimenting at UC Berkeley with easy commercial-web-hosting-style deployment of LAMP tools on an instructor-by-instructor, site-by-site basis. But even with scripts in place to reduce the cost, it couldn’t scale to meet our mandate. Hereabouts, anyway, most instructors and GSIs were simply too pressed for time to deal with the extra cognitive and administrative overhead of tracking updates, maintaining backups, taking care of memberships, and so on.

I’m still hoping, though, to break out of the in-or-out mindset and reach the sweet point of supporting LAMP-style (or nowadays Google-Apps-style) environments alongside (or in place of) the all-in-the-LMS environments. A lot of the political barriers have fallen over the past few years, and I think the underlying technical issues are better understood nowadays.

Especially if you decouple memberships and permissions from class enrollments and teaching assignments, you get some exciting possibilities for learning communities working across disciplines, across semester, across institutions.

(But I agree that you don’t achieve the traditional LMS functions of automatic and easy account creation and maintenance).

It would be good to imagine a way (and then invent it) to get that integration when you want it, without the lockdown when you don’t want it. I totally agree about that, and would love to see it!

What I’d like most is to support that style of integration _without_ forcing teachers, learners, and collaborators into the traditional Swiss-Army-knife-cum-penitentiary LMS mindset.

I don’t think there’s anything necessarily or inherently bad about management, or assessment, or gradebooks. They’re all fine things to have, and not evil.

But they’re also not interesting, or exciting, or innovative–and they don’t have a lot to do with learning. They’re the necessary substrate (thanks for that term, Michael), but they’re not sufficient, and they don’t provide the exciting and creative added value that is what I’m looking for in using new tools.

And as Luke says, the substrate shouldn’t suck all the oxygen out of the discussion or the classroom.

I think, too, that part of the problem with Blackboard (as it’s used at CUNY) or any big system is that it’s big. Centralization brings economies of scale. And it brings convenience for students moving among different units. But it also brings single points of failure, and it can stifle the independence and context-specificity that really is important.