Here's an idea to get around Twitter's admittedly crap api: people install a bookmarklet, browser extension etc. that detects when you are reading Twitter, and extracts the tweets from the DOM and sends them to a server. If enough people use it then you can start to build up a relatively good alternative to Twitter's unreliable and heavily rate-limited API. Does anyone know if this would work effectively, and more importantly would it breach Twitter's TOS?
