Ow, ow, ow, ow...

First you want to rebuild the indexes, then you want to break the log chain (no more point in time recovery), shrink the log file, ensuring that there will be slow downs and log fragmentation from the growths and restart the server ensuring that it'll be slow when the users start using it.

Well, to be honest I could think of worse things to do, but it might take work. 😉

For the log, please read through this - Managing Transaction Logs[/url]

Why not to shrink a DB, see - http://sqlinthewild.co.za/index.php/2007/09/08/shrinking-databases/, much the same goes for the log

Why restart daily? Some servers I work with haven't been restarted in months and there's seldom a need to restart SQL regularly, doing so means more work for the DB engine to repopulate all its caches.