I would probably make Wikipedia a front end to a Subversion repository. Then I would get all the data handling for free. I would of course add caching, to limi

I would probably make Wikipedia a front end to a Subversion repository. Then I would get all the data handling for free. I would of course add caching, to limit the number of reads to the Subversion server.