"OS-independent remote display (e.g. show a GUI on a Windows machine or a Mac from your *NIX netbook)."
remote x need high bandwidth and low latency networking. Especially round trip issues. So it works OK on LAN but suck on the internet.
Most users use vnc and or nx anyway. Anything is faster than remote x on internet. If X did it right the first time there is no reason anyone would ever need nx.
(NX is basically a roundabout way to get around X's problems/bugs/limitations to make it works in real live situation. It's written by people really use X and know all the low-level problems. Amount other thing X protocol is very verbose, often send redundant and repetitive parameters over the net. NX try to cache/compress/delete those parameters to save bandwidth)
BTW, Google just release Neatx, an Open Source NX Server
http://google-opensource.blogspot.com/2009/07/releasing-neatx-open-source-nx-servier.html
remote X's #1 problem is round trip delay. To solve that problem, browser use client-side javascript. So X need some sort of server-side scripting. I'm sure Google Chrome solve that problem already!
"The only serious improvement I've seen suggested over the X model is to provide a vector scene-graph API so that you can store the entire sequence of drawing commands in things like OpenGL vertex arrays in the GPU's memory. While this is a nice idea, it would require a radical redesign of all existing GUI toolkits and applications to be used to its full capability."
Bingo. OSX did it.
overall X's #1 problem is toolkit. Anyone still use Xt? On X everyone write their own toolkits. "radical redesign of all existing GUI toolkits" is easy to do if there is only one standard toolkit.
Similarly X have multiple window manager, so any real improvement require radical redesign of all existing window managers.
pH
