My thought on this is that we would pull the real-time data down on the server and transform/cache it there. For two reasons:
Cut down on the amount of data that the browser has to download and process (it's already doing plenty of processing for the visualization, and we don't want to overburden it with things that the server can do)
Cut down on the number of requests that the client has to make (since there's no way to get all the real-time data across all lines at once -- to my knowledge -- we'd have to cycle through the lines, collecting them each individually)
For visualization, we could use the number of vehicles on each line as a metric to determine the frequency of the crawlers along that line.
My thought on this is that we would pull the real-time data down on the server and transform/cache it there. For two reasons:
For visualization, we could use the number of vehicles on each line as a metric to determine the frequency of the crawlers along that line.