[aprssig] Why RELAY,WIDE... is so bad....

[Wes Johnston]

If it comes out thru a home station, it'll come out like
WIDE2-2*,WIDE2-2, and the first digi that catches it will digi it as
WIDE2-2*,WIDE2-1...  So if a home station were to hear it first, it'd
progress like this:
WIDE2-2,WIDE2-2
WIDE2-2*,WIDE2-2
WIDE2-2*,WIDE2-1
WIDE2-2*,WIDE2-0

If it hits a digi first, it'll come out as WIDE2-1,WIDE2-2.  Of course
it'll just keep going...

WIDE2-2,WIDE2-2
WIDE2-1,WIDE2-2
WIDE2-0*,WIDE2-2  ->now it begins to bounce along, using the 2nd widen-n
WIDE2-0*,WIDE2-1 ->and keeps going and going....
WIDE2-0*,WIDE2-0

Looks like it could go a whole lot further that we'd intend.. but it
wouldn't be ping ponging around the locals.  But what's to stop someone
from using a path of WIDE2-2,WIDE2-2,WIDE2-2,WIDE2-2,WIDE2-2,WIDE2-2 or
worse, WIDE7-7,WIDE7-7,WIDE7-7,WIDE7-7,WIDE7-7,WIDE7-7,WIDE7-7

Although I like the idea of WIDE2-2,WIDE2-2, it ultimately goes further
than we want it to...

BUT.... UITRACE is now supposed to do WIDE for us... so if I send a
packet out W2-2,W2-2,W2-2,W2-2, it would grow longer like this:

W2-2,W2-2,W2-2,W2-2
W2-2*,W2-2,W2-2,W2-2   ->home station digi'ed it as if it were RELAY...
W2-2*,digi1,W2-1,W2-2,W2-2
W2-2*,digi1,digi2,W2-0,W2-2,W2-2
W2-2*,digi1,digi2,W2-0,digi3,W2-1,W2-2
W2-2*,digi1,digi2,W2-0,digi3,digi4,W2-0,W2-2  -> what happens here when
we exceed 7 path elements?

This thread has also showed there is a need to have smarter
"pre-emptive" digipeaters that can look for multiple WIDEn-n's in a path
- and cut them off.

Pete's Javaaprs won't be fooled by my bad example of 7 W7-7's above...
that program counts the total hops in a packet and won't digipeat if
above a threshold. 

Digi_ned can do the same thing, a slightly different way.

What about ui-view?

I know the KPC3 digis would be tricked.  But how do they behave when a
UITRACE path grow longer than 7 elements?

What about UIDIGI on the tnc2 clones?

Wes

Stephen H. Smith wrote:

> On the issue of how to combine home digi first-hop-assist with true
> WIDEn-N systems without using first-hop RELAY:
>
>
> How about a path of    "WIDE1-1, WIDE1-1"    or    "WIDE1-1,
> WIDE2-2"    ??
>
>
> Set the alias of the former home "RELAY" digi to WIDE1-1 instead.   If
> a "dumb" home former-RELAY-type digi does the first digipeat, it marks
> WIDE1-1 as used.
> Then the true WIDEn-N digi(s) get a shot at the second part of the
> path.   I assume that in the absence of a nearby home digi, a "real"
> WIDEn-N would digi the first hop, then the WIDE2-2 part would cause
> two more digipeats.   
> The path is compatible with either    first-hop-via-home-station,   or
> with all hops via    n-N-only    true wides and preserves exclusively
> dupe-supressing WIDEn-N type paths.
>
>
> [ BTW,   Here in L.A. the monster mountaintop digis are now swallowing
> WIDE2-2 on a single hop.  Before that was implemented,  when I  was
> just driving around the greater L.A. area, I would routinely set
> WIDE1-1 into my mobile UI-View setup to get into the WIDEn-N-only
> digipeaters with an absolute minimum of dupes.   In my experience, ONE
> hop from a high digi will get into an igate from anywhere in the
> greater L.A. area.  ]
>
>
>
> Stephen H. Smith             wa8lmf (at) aol.com
>                                                    Home
> Page:                   http://wa8lmf.com
>
> New/Updated "Rev G" APRS     http://webs.lanset.com/wa8lmf/aprs
> Symbols Set for UI-View,
> UIpoint and APRSplus:
>
>
>
>
> _______________________________________________
> aprssig mailing list
> aprssig at lists.tapr.org
> https://lists.tapr.org/cgi-bin/mailman/listinfo/aprssig