On 01/26/2012 12:50 AM, Henrik Nordström wrote: > > ----- Ursprungsmeddelande ----- >> If they have sync problems, they may violate HTTP. I am just doing my >> best trying to stay focused on the [local] cache architecture topic; I >> do not want to get into discussion about distributed hierarchies.
> Exactly. HTTP spec is written for a consistent single-path forwarding path, > not a mesh. >> It may be a good idea to polish HTTPbis if it indeed allows both yours >> and mine points of view to coexist even though they contradict each >> other. > It's a little late now for such adjustments. And not needed i think. I do not think it is too late because we do not need to add or change any normative text. Just add a paragraph clarifying the specs scope: "This specification defines behavior of a fixed chain of HTTP agents. Content Distribution Networks, cache hierarchies, load balancers, and other complications often result in variable forwarding paths, where two requests with the same Request-URI from the same client may arrive at different agents. While all HTTP requirements ought to apply to collections of cooperating agents, treating each collection as a single agent, such compliance is usually not guaranteed (and, hence, should not be relied upon) in complex distributed environments." Or something of that kind. Cheers, Alex.
