Xiaofan Chen wrote:

> I do not quite understand this. If the jitter is quantifiable, it would
> have a maximum value. So it has determinism. Right? Kind of confusing.

Not necessarily. Jitter may be quantified in statistic terms like average,
mean, distribution -- without quantifying a maximum value.

> I know there are many applications which do not need hard real time and
> allow packet loss (you just wait for the next packet which contains the
> same data).

There's no form of determinism that can guarantee "no packet loss" -- just
cut the cable (or all cables), and there /will/ be packet loss :)

I don't think this ("deterministic") is a "deterministic" term, so to speak
:) It has a meaning that depends on the context and the author. The text
you cited in your last message contained the term "better determinism",
which implies that determinism is not an is/is not characteristic but a
more/less characteristic. I don't think this is how many other people use
this term.

I also think that some authors confuse determinism with reliability. IMO
the former is a characteristic of the network protocol (and it either is or
is not, given a set of assumptions), while the latter is a probabilistic
characteristic of the whole setup, which includes such things as
connectors, cabling (you can reduce the reliability by cabling all over the
factory floor with standard connectors :) and so on.

Gerhard

-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist