New project: streaming irc concordance cloud.
Dealing with this carpaltunnelwhatever flareup is a drag in many ways, perhaps none more so than that I've basically quit all chat apps nearly all the time. It's a huge source of extra typing, and keyboard time, and I have work stuff due now anyway.
The worst part that is #code4lib withdrawal. It's not like I'm envisioning little men running up and down my arms arguing about whether whitespace is syntax, or anything. But, I'd been hooked up to it for two years running, and now the IVAX is going NEEP NEEP NEEP and I know the nurse isn't coming to refill it.
So this is the app I want to help keep track of what's going on in there without being logged in.
- Connect to channel (probably use twisted)
- Log all msgs privately and to db records
- Every minute, produce a concordance of tokens from the last 30 minutes of conversation
- Rank tokens gathered by frequency
- Eliminate stop words
- Rank msg poster usernames by frequency
- Create a private three-column web view of what's going on in the chatroom:
- Col1: running log. colorized, linkified, etc.
- Col2: concordance token cloud, log-scaled-text-size like tag clouds
- Col3: username cloud, log-scaled-text-size per numposts
- Refresh the msg log several times a minute
- Recalculate the username/token corcordances as a 30 minute moving avg every minute
- Refresh the usernames/tokens with animation to shrink/grow their size and move their position with every recalc
You could even specify differing intervals per "peak time" and "off time" so the window grows/shrinks accordingly depending on a channel's usage patterns.
Okay, it's, um, a little, well, obsessive. But, it would look cool!
Does anything like this exist already? It shouldn't be too hard to hack starting from some of the irc logger bots and colorizers.