[aprssig] Digipeaters with local sysop rules...

[Joel Maslak]

On Sep 14, 2007, at 6:42 AM, Robert Bruninga wrote:

>> It would have been cool if the spec allowed
>> for a "live operator" bit.
>
> It does.  It was added many years ago in APRS addendum 1.2.  If
> the case of the N/S byte is UPPERCASE, then there is an operator
> present.  If it is lower case, then no operator is present.
> This is fully backward compatible with all software, since
> attended stations will all show normally.


What about things like a D700 which have (albeit severely limited)  
messaging capability?  And wouldn't lower case have been the better  
choice, since a KPC3+ connected to a NEMA source will send upper  
case, but likely have no operator?

I probably need to mess with my IGate (although it may be going off  
line soon due to a move).

>> I do path experimentation and it would be irritating
>> to find that someone was NSR'ing my packets especially
>> if that digi was the only digi I could reliably get into.
>
> My thoughts exactly.


I'd love to see a path like "IGATE" work to get someone into the  
nearest IGate.  That would get rid of 95% of QRM probably (most digis  
could do this with no work if they have a free digi-alias slot - one  
hop is probably all *MOST* would need; Some [some home digis] would  
need zero hops).

That would be better than the current recommendation for long- 
distance travelers of WIDE1-1,WIDE2-1 or WIDE1-1,WIDE2-2 (note not  
WIDE3-2 at the end because that's trapped by some digis - which is  
essentially modifying the path and keeping a knowledgeable operator  
from doing what he expects the network to allow him to do), *IF* the  
travelers don't have a way of sending/receiving messages (if they do,  
use WIDEN-n).

To make it safe, it's easy.  Make the substitution like:

IGATE becomes PATHSUB,N7XUC-1,KC7AAA-2

Where N7XUC-1 is the digi that did the substitution, KC7AAA-2 is the  
next hop (explictly named!).  Because PATHSUB does *NOT* set the  
repeated bit, nor does N7XUC-1, KC7AAA-2 won't repeat the packet  
unless it's a new-generation "smart" digi that does good dupe  
checking (part of implementing the smarts that looks for PATHSUB in  
the packet).

N7XUC-1 would never repeat the packet again, because the rule is  
"don't repeat a packet if you see 'pathsub' and your own call  
immediately following."

Then those who want to make an actual contact can use a path that  
uses the radios more.

But I know this is too radical, probably would need minor tuning (so  
throw it out completely!), etc...