#Compsci people:

if you were to compare #distributed computing #architectures, what are the features along which you'd compare them?

Some ideas:
1. Where on the CAP tradeoffs something is.
2. Tolerance against hostile nodes.
3. Supporting localized innovation/extension.
4. Statistical properties such as max latency.
5. Ease of recovery from node failures.
...

Anybody have a good list so I don't need to make one up on my own?