Another little train of thought, whose conclusion will probably be obvious to anyone:
I wonder how hard it would be for me to make a little personal idea processor like Eastgate Tinderbox using RDF? Very likely much harder than it was for its author to create the original - I'm by no means smarter than Mark Bernstein, and he's got years on me in developing tools like Tinderbox. So why would I even consider rolling my own in this case? Certainly not the price - for a tool like that, the price is a steal.
No, I think it's because I don't have full access to hack the thing, and it has a few itchy spots for me. At least, that's the way it looked when I last tried a demo on MacOSX. I wish I could fix them, and rolling my own is the best way I know for that. And besides, I'm in a real RDF love fest lately.
But... Is it really so bad as it is? Bad enough to try to play catch up with what someone else has already devoted so much to? Nope. Bad idea. Best to promote recombinant growth, and rephrase the question: I wonder how long it would take me to get used to the tool's quirks as I percieve them, make suggestions to the author, and then use the extraordinary hackability already present in the tool to get it soaking in RDF?
This is a lesson that's taken me awhile to resign to learn, so I figured this would be a good exercise to document it. And to think, I once spent a year or two trying to re-implement Zope, by myself, mostly in quiet, and with not much to show in the end.shortname=ooocec