25 Jan
2016
25 Jan
'16
7:47 p.m.
On 2016-01-25 19:40, Robert Armstrong wrote:
And I'm not sure I understand that sentence. :-)
All links are the same, I guess. They all have a cost associated. The
cost might be different. Which would suggest they are not the same. So
what do you mean?
I feel that it
is impossible to setup some some systematic approach to
costs when one sort of direct link is considered a general problem.
I don't understand what that means. Aren't all direct links the same, at
least as far as DECnet cost is concerned? Unless you're saying that you consider the
bridge to be a "direct link" despite its extra hop thru psilo. If we go that
way then any and all links are equivalent and you've rendered my only goal
superfluous.
Yes, I consider the bridge to be a direct link as well. All our traffic
is carried over IP, meaning it is running through a number of hops in
between that are invisible. The host running the bridge is not, in my
view, any different than any router, or the extra layers that a Multinet
link is running through inside the VMS box, in order to carry over the
DECnet traffic into an encapsulated UDP or TCP stream.
If the encapsulation happens inside the machine, or externally, how much
do that matter? Especially when the external encapsulation might in fact
be way faster?
Johnny
--
Johnny Billquist || "I'm on a bus
|| on a psychedelic trip
email: bqt at softjar.se || Reading murder books
pdp is alive! || tryin' to stay hip" - B. Idol