[aprssig] Version 8.2 WIDE1-1 problem

Robert Bruninga bruninga at usna.edu
Fri Nov 9 15:07:41 EST 2007

> > SO yes, the KPC-3 version 8.2 digis have to
> > have WIDE1-1 as a myalias...
> Bleh.  Yes, SANDIA and LAMOSC (the two that 
> ping-pong my packets all the time)
> are KPC-3 8.2 digis.  
> What do they do wrong at the last hop?

They decrement to zeero but do not filp the has-been digipeated
bit.  Works just fine for a WIDEn-N.  But when a WIDE1-1,WIDEn-N
packet hits it, it digipeats it as SANDIA,WIDE1,WIDEn-N which
looks ok, but the WIDE1 never gets marked as used up (there is
no *).  So the packet dies, since there are no "WIDE1" digis.

So we fixed it in the 8.2's by having them use WIDE1-1 as a
callsign substitution alias instead of a real WIDE1-1.  SO these
WIDE1-1,WIDEn-N packets in the above case get digipeated as
SANDIA*,WIDEn-N instead and then the next digi can work on the
WIDEn-N as it is supposed to.

I suppose we could have also fixed it at the other end, by
having all other WIDEn-N modern digis support "WIDE1" as an
alias, but then this puts the fix in the good digis.  It is
better to have the fix only in the digis that have the problem,
so that if they ever go away, the problem and the fix dissappear
at the same time.


More information about the aprssig mailing list