By running this on a devicemap server (or any other server -
*.apache.org maybe), we can start capturing information when people to
visit a page that includes the logcapture.js script found in the same
folder (that script can be improved of course).
There's no privacy issue IMO if we collect just the above parameters,
and throw away client IP, referrer and other potentially sensitive
We can then parse those logs and process to generate devicemap repository data.