Nkrs Today at work we set up a link to a camera so people will be able to see from home if there's a long line at the reception desk. Technically it is sort of impressive (we have a server that streams the RealTime feed from the camera and transcodes it as MPEG, and a WebSocket streams that MPEG into a canvas element), but I wonder if it is really needed or useful (line could grow during the 15-20 minutes that someone would use to get there).
Eric It's a nice idea but your concern about it changing by the time the client arrives is very valid. Maybe a "queue forecast" or estimated wait time for next X hours could work better in simple plain text?
Login or register your account to reply
Nkrs Perhaps it would be better as plain text. There could be a system where every person gets issued a ticket when he enters, and when they get their turn, the ticket gets destroyed. The system could count time between ticket being issued and destroyed, and by accumulating that data, there could be forecasts. But that's a lot of work, and cameras were there already.
Martijn I would volunteer for the work. Sounds pretty cool. I know there is at least one paper published on queueing wrt picking a lane at checkout. Not sure about anything on time forecasts.
Martijn That would be an interesting forecast to work on, I wonder if their are any documented algorithms for it. Hmmmmm...