[aprssig] Preemption Digipeating

[Robert Bruninga]

> Is preemptive digipeating defined anywhere? 

Im not sure, but I defined it back in 1995/6 as allowing any
unUSEDup field matching the digi to be digipeated by that
pre-emptive digi, and then marked as used (including all prior).
All prior have to be marked as used up to comply with the basic
AX.25 packet spec.

So your anser of A is correct.  Anything else would open
pandoarah's box of issues.

Further, PREEMPTIVE digipeating was NEVER intended to apply to
generic addresses (such as WIDEn-N or TRACEn-N.  But simply to
be a useful way of listing possible alternative digis
specifically.  Allowing it to work generically can add huge
amounts of unintended QRM to the channel.

But the example below of SAR does make sense.

Bob, Wb4APR


 list to
> maybe I should say other than the T2 source code, which as far
as I
> know is t he only device to implement it.
> 
> 
> > If a station transmits a packet with a path of 
> WIDE1-1,WIDE2-1,SAR and the
> > SAR-1digipeater is configured to preemptively digipeat on
SAR, what
> > will/should the digipeated packet path be A, B, or C?
> >
> > A) - WIDE1-1,WIDE2-1,SAR-1,SAR*
> >
> > B) - SAR-1*,WIDE1-1,WIDE2-1
> >
> > C) - SAR-1,SAR*,WIDE1-1,WIDE2-1
> >
> > I tried this with a T2-135/OT2m and got D) 
> WIDE1-1,WIDE2-1,SAR-1* which is
> > basically A with callsign substitution instead of insertion.
> 
> This is actual preemption.  Preemption as used here is defined
as "a
> taking possession before others".  The SAR preemptive digi
takes
> possession of the PATH and the packet out of turn, i.e. before
the
> WIDE digis use it, which is the normal progression.
> 
> > I personally find B or C more useful as it preserves the 
> original desired
> > path rather than preemptively removing the path by jumping 
> the Used flag to
> > the end of the path.
> 
> B and C _reorder_ the path, not just the _processing_ of the
path.
> That's not preemption, it's reordering, mangling, munging,
etc.
> 
> -Jason
> kg4wsv
> 
> _______________________________________________
> aprssig mailing list
> aprssig at tapr.org
> https://www.tapr.org/cgi-bin/mailman/listinfo/aprssig
>