[PATCH] Remove obscure "low-latency" parts in the introduction of spec

[Justin Lee]

On Fri, May 31, 2013 at 6:00 AM, Andrey Sidorov
<andrey.sidorov at gmail.com> wrote:
> It is related to two way transfer. Imagine high latency (say, 0.1
> second) network. With d-bus you can send 100 messages in 0.1 second
> and receive reply for all 100 on 0.1 s. With protocols wich don't
> support two-way communication (like, for example MySQL protocol) this
> can't be done in less than 10 seconds. "Low-latency" thus makes sense
> to me in D-Bus description.

The latency is 0.1 second whether we are sending 1 message, 100
messages or a billion messages synchronously/asynchronously, because
the sentence "Imagine high latency (say, 0.1 second) network." tells
us the latency is 0.1 second.

Latency is a measure of response time of a network for *single*
message/packet AFAIK (I could be wrong though). In general, we don't
sum up time taken by multiple messages when calculating latency. For
instance, if we send 3 messages and they take 0.05s, 0.2s, 0.1s
respectively. The latency is 0.05s for the best case and 0.2s for the
worst case, but we don't usually calculate latency as being 0.05s +
0.2s + 0.1s = 0.35s (well, I haven't seen the analogy before).

Justin