[aprssig] New n-N success in North Carolina
aprs at kd4rdb.com
Fri Feb 11 22:22:29 CST 2005
The problem with the "trap out" method is that even if you trap W7-7,W6-6, and
W5-5, you still let the W7-6's and W7-2's thru....
There is another way and that is to support the alias of WIDE2-2,W2-1,W3-3,W3-2
under the UIDIGI setting. It'll give ONE more hope to each elligible packet.
Eric, I agree with you... SSn-n is the way to go. Ok, fine, we can support ONE
hop and one hop only for W2-2 people by putting W2-2 in the UIDIGI setting.
Ideally, I'd really like to see is a hop count limit. If this existed, it would
be fine for a fellow to run W7-7. When he wandered in to an area where 2 hops,
his packets would be digipeated until they got to WIDE7-5. Now imagine a guy
approaching a metro area. Some of his packets would go toward the metro area,
other packets would head off in the rural direction. Those packets going into
the metro area's 1 or 2 hop limit digi's would not be digi'ed once those
packets were decremented to W7-5. Meanwhile the copies of packets going toward
the rural area's 5 hop limit digi's would continue to be digipeated until they
got to W7-2. They could feasible route around the city! What I'm getting at is
that each area could filter the same packets to varying degrees... wow.. what a
shame a hop limit wasn't spec'ed when the UIFLOOD was first written.
I kinda get a chuckle out of this next suggestion. If I (or someone) was to
produce a KISS mode controller based on a PIC chip, it would probably cost $35
(around the price of a tiny trak). This device could be added on to the back
of a kpc3 tnc in kiss mode and powered from pin 13 or pin25 of the serial port.
The part I chuckle about is that kantronics wants $60 for firmware, but it
would be a neater solution.
Quoting "Eric H. Christensen" <kf4otn at earthlink.net>:
> > 1) SSn-N is still a perfect way to go for most local ops
> > it remains a great feture of the New n-N Paradigm
> > 3) We found a simpler way to TRAP large N hops without
> > haivng to drop WIDEn-N completely
> This will not work. Yes, you can trap SOME of the WIDEn-N packets
> but that won't stop 4-4, etc. PLUS, in NC we are putting additional
> LANs up to support gating of weather bullitens and objects so they
> don't go outside of the area. (We have someone on the Western part of
> Virginia or NC that is sending out bulletins to a very WIDE area.)
> So we don't have room to trap all these bad n-N packets. By doing
> away with WIDEn-N, we have freed up our network from a lot of packets
> that used to come in from all over. John's assessment is accurate.
> > 4) LNKn-N will never work since it simply takes too much
> > coordination and wont be of any value unless all links
> > in the chain are good. It has been abandoned
> I disagree. As you coordinate the SSn-N you coordinate the LNKn-N.
> We ALREADY have THREE digipeaters up on the 95LNKn-N and that was
> within a month. Not a lot of work there...
> > What you have done is a great positive step. But please
> > re-visit the New n-N Paradigm site and prepare for one more
> > round of changes. Basically, send out a bulletin reminding
> > everyone how much better SSn-N (or Wn-N) is than W,W or R,W.
> > Then you can turn back on the WIDEn-N but with the filters on
> > 55,66,77 or what ever limit you want...
> I think the next best thing is to eliminate the WIDE and TRACE.
> > >This has been the best thing that has happened to APRS
> > >to date. I hope we don't back track to our older ways.
> > ANd thanks for trying it. And dont be upset with this
> > latest change. We have learned a lot from the user
> > feedback and it will make it even better...
> I disagree. Even letting WIDE4-4 through would bring in many packets
> from hundreds of miles around and then we are back at square one.
> Eric KF4OTN
> > de WB4APR, Bob
> aprssig mailing list
> aprssig at lists.tapr.org
More information about the aprssig