[aprssig] NOSaprs update - cross port digi with callsign substitution

[Robert Bruninga]

>>> mckeehan at mckeehan.homeip.net 06/22/05 3:29 PM >>>
>I can not think of examples where the no source 
>routing approach would cause problems. 

The NSR routing says that the sysop of each digi
will decide how far your packets can go, and/or
who you can talk to and how.  Clearly this is to
limit congestion which is very laudible.  But there
is a cost.   Based on my experience in HAM radio
here is why I think it wont work:

1) Get 4 hams in a room and you will never get
agreement on one way of doing something.  In
fact, you cant even get agreement on what
ham radio is for an why.  DX, rag chew? Experiment?
Mobile? VHF, HF?

2) When repeater owners or clubs make draconian
rules, people often just leave, or QSY elsewhere.

3) There are very different needs on APRS for
   a) simply operating mobile for commuting
   b) Just trying to get to an Igate
   c) SAR anywhere in the state
   d) WX for locals
   e) WX for region to the NWS cener
   f) Local digital chats
   g) Statewide EOC nets and checkins
   h) 24/7 monitoring and availability
    i) Club communications and bulletins updates
    j) Rapid DFing jammers  interference

All of these usually have different needs.
Under the NSR proposal, the sysops may even 
pre-define different NSR plans for each purpose,
but this reeks of inflexibility when in fact, the
thing about Ham radio is supposed to be its
extreme flexibility.

4) My experience is that most digis and nodes
are installed and forgotten.  It takes MONTHS
to get a sysop to make even the simplest change.
Under NSR, the SYSOP has to be asked to re-configure
the network for EACH need and each application.

Sure this CAN work, but my exlerience in HAM
radio is that it never does.

Again, NSR *is* the ideal way to design a
reliable network designed to do exactly what
the sysops set it up to do.  No questino about
that.  My experience is that  such beautifully
designed systems requiring every node to be
fine tuned to exactly the settings of each
sysop will fail in HAM radio in most cases.

People will just give up and go elsewhere.
I dont want to see APRS fall apart like that.

Just witness how long it is taking the digis
to change to the New-N paradigm which
really only takes a sysop 10 minutes to do
by remote sysop!  Its been several months
and we still only have less than 20% 
changed.   Imagine if every digi in your
area had to have sysop changes everytime
you or your club or another club wanted to
do something different than what the sysops
pre-thought of.

>Can you give some example situations where some 
>"flexibility" would be needed.

Sure!  We want to get everyone to cut back to
WIDE2-2 in congested areas.  Baltimre is such
an area that is saturated.    Yet, I have absolutely
no problem with someone running WIDE4-4 during
a cllub demo, or school demo.  In fact, he has to
run 4 hops just to find a live human to demonstrate
APRS messaging with...

Its simple "gaming theory".  If everyone tries
to step on everyone else to get the most for
their packets all the time, then everyone loses.
But if everyone doesnt try to "win" all the time
and let someone else use the asset, then it
will be there when he needs it.

That is why I am opposed to the network and
a few sysops deciding what one can do on APRS.
That is not what the network is for.  It is for
anyone to use with whatever path they need
for LIMITED immediate needs that are flexible 
eno0ugh to let communications needs be met.

Again, technically there is absolutely nothing wrong
with NSR.  It is the ideal design of such a network
(as long as you can get agreement on exactly what
that network is supposed to do).    And my only
warning is that you will never get that in HAM radio
with something as flexible as APRS.

And if you try to just force a local solution on people,
they will just go elsewhere.  This will kill either the
flexibility of APRS or kill participation......

Bob,

On Wed, June 22, 2005 3:05 pm, Robert Bruninga said:
<snip>
> But that approach is based on one thing:
> *having the network KNOW what you want
>   done with the packet*
>
> There are many sysops who are quite happy to
> define where your packets should go, and will
> happily make a network do it.  But that kind of
> approach on an ad-hoc come-as-you-are
> kind of system that APRS is supposed to fullfil
> is full of pitfalls and kills the kind of flexibility
> that APRS was intended to provide, and that is
> to use it for what you need at the time.
<snip>
> Ah, but Do what?  Many people have very different needs
> for their APRS packets, and very different ideas about
> how they want their packets to be used.  Trying to force
> ham radio operators into a one idea, one-app system has
> the potential to kill the system.  HAM interests
> are too varried to try to take those draconian steps.
-- 
William McKeehan
KI4HDU
Internet: mckeehan at mckeehan.homeip.net 
http://mckeehan.homeip.net 

_______________________________________________
aprssig mailing list
aprssig at lists.tapr.org 
https://lists.tapr.org/cgi-bin/mailman/listinfo/aprssig