While I entirely agree and preach latency often, there are two major points I would like to make:

1) Remote computing doesn't have to use X11. The local machine can do some of the lifting. Wolfram Alpha on mobile phones is a good example of this- a server does the crunching, but neither is the client side a dumb terminal.

2) I cross my fingers that one day latency will be prioritized more... IMO the majority of latency problems are not caused by the limits of physics, but rather the simple fact that not many people care about it.

The theoretical peak ping speed across the USA & back is something like 20ms, an entirely reasonable figure for remote computing- if only it was ever anywhere close to realized.

> IMO the majority of latency problems are not caused by the limits of physics

If you look at network QoS research in the past 20 years, you'll see that a lot of people do care. It's just not an easy problem to solve.

And with modern browsers and AJAX we already have clients in place that can deal with latency better than X11 servers.

True, but is the real solution to remove the latency in the first place and just run it on your local machine without a network connection? (Like ordinary desktop software that we have at the moment.....; seems the terminal approach to software is attempting to fix a problem that has already been fixed - just run it locally)

its not really the same in this case it updates in the background and stuff runs locally using NaCl