A new month, a new file. I’ll try to get back to a non-prehistoric system over winter break, although other things may take precedence.

Outsourcing to India in Business Week and at MIT… – now this is some harsh practical education…

The more sophisticated portion of ocw.mit.edu is a 100 percent Microsoft show. A student asks the speakers why they chose Microsoft Content Management Server, expecting to hear a story about careful in-house technical evaluation done by people sort of like them. The answer: “We read a Gartner Group report that said the Microsoft system was the simplest to use among the commercial vendors and that open-source toolkits weren’t worth considering.”

Students began to wake up.

A PowerPoint slide contained the magic word “Delhi”. It turns out that most of the content editing and all of the programming work for OpenCourseware was done in India, either by Sapient, MIT’s main contractor for the project, or by a handful of Microsoft India employees who helped set up the Content Management Server.

Thus did students who are within months of graduating with their $160,000 computer science degrees learn how modern information systems are actually built, even by institutions that earn much of their revenue from educating American software developers.

I’ve been working on and off over the past couple months (hmm, since SXSW – wow, I’ve been slow on this) on what I’m labeling ‘distributed social software‘, sort of redundant one might think, except that ironically, besides the original social-software catylyst (blogs) no current generation social networking systems are distributed, and even with blogs, not formally so (not necessarily a bad thing, perhaps, but that’s another argument).

Anyway, sort of surprising that a Google search turns up exactly zero matches, but I was talking to a friend who had simultaneously invented the same term for a CS paper he’s writing, so at least we’re making up terms on the same wavelength. 🙂

Issues I’m currently tackling:

  • Defining complex (typed, n-way) trust relationships
  • Dealing w/ trust segmentation; trust transitivity across fragmented identity
  • pseudonymity and/or double-blind transactions
  • not re-inventing any wheels I don’t need to
  • maximum simplicity, extensibility, compatibility
  • surviving last week of classes (this may be the toughest thing)

Trying to figure out making QuickTime Playlists right now. There’s a free program called QT Xlist which will play text lists, however one can’t navigate sections… There’s also a program called QT Playlist Maker, but it’s $15. It looks like it’s just generating SMIL, so I’m putting together something to do just that.