I doubt that facebook or twitter programmed their FIRST iteration with
that kind of scaling in mind. Once they approached about 3/4 million
users - and were able to plot a bankable growth curve - they could
easily leverage those numbers for an investment in a more scalable system.
Sometimes, I find myself dwelling on the technical minutiae of MY
personal projects out of fear: That as long as I continue to develop my
concept to ultimate perfection, I can justify delaying the actual
release... and the -hopeful acceptance- or -dreaded rejection- that follows.
I'm not saying with any certainty that's what you're doing, it's just
what I often catch myself doing. Matthew's comment just nudged me into
putting it out there at this moment. So I'm sorry for polluting your
thread, Jason.
But to everyone - and nobody in particular: If you've got a cool idea,
then whip up a working model and just THROW IT OUT THERE! If it really
sticks, you'll have ample opportunity and 'agility' to whip it into
shape. Apps get changed no matter what. So face your fears... fellow geeks!
Al
Oh yeah, and consider using GAE's datastore or CouchDb - if you're
thinking of each user having something like "their own table".
--
official tag/function reference: http://openbd.org/manual/
mailing list - http://groups.google.com/group/openbd?hl=en
