one of the reasons why i posted what i did was because matt mentioned that "if this was a bug in a linux program, I could have googled for the fix and been back to balancing my bank account in a matter of minutes."
reliance on this sort of community may work for some things, but the community didn't do a lot of bug finding or testing when it came to chandler. maybe that is because people, perceiving it as a funded project, refused to do so. maybe it's because, as roseberg hints, iirc, that chandler needed to have been put out there, bad shit and all, and develop it in "a patchy" way. ha ha. :)
also, the problems rosenburg identifies are not with poor project management. rather, he notes that the 'bazaar' model (remember! there was an entire book published online advocating the 'bazaar' model of open source development, as opposed to the bad, 'cathedral' model) doesn't work the way it was advertised. for another, rosenburg's argument is that it may just be in the limited way we conceptualized what we do. i remember woodchuck writing about these years ago on dc-stuff, pointing out that 'file' was a really bad way to thinking about computing.
there's more but i was up at 1:30 for a disaster recovery test. brain's fried.
shag