On Tue, 7 Jan 2003, Stefano Mazzocchi wrote: > It can hardly be simpler than what it already is. Introducing an > explicit version control cycle (because commits are exactly that) might > seem relatively simple (or even totally natural) for somebody (all of us > if we are committers) that is used to it but might not be for users that > simply want to give some help. > > As Jeff brilliantly put it on the forrest-dev list: users don't rant, > they quietly leave. > > Here, you could say: users don't complain about how hard the > documentation system is, simply they don't use it to contribute. > > Between having a smarter diff-sending algorithm transparently do the job > on the users behalf and an explicit commit cycle, I'd go for the first > even if it removes only *one* step, it might be a difference between 100 > people using the wiki and 150, 50 of which are scared by the concept > of 'committing'. > > Paint me PITA, but I think it's worth playing devil's advocate on this > muddy ground.
We use JSPWiki in our company together with the Hula server (can't remember the URL but Google will know it) which checks a Wikipage (NotificationList in our case) that users can put their address and notification time into it and get a mail with the diffs of the last 24 hours (if there are any). That's IMO the Wiki way of notification for interested users. Giacomo
