[aprssig] Tracker Smart Pathing

scott at opentrac.org scott at opentrac.org
Tue Mar 21 10:02:22 EST 2006


> Surely though, this then should be a function of the 'Network' i.e.  
> the Digi and not the tracker.  It only means changing the code in  
> every Digi (not trying to start an argument now or anything), but it  
> would mean that all trackers could avail of the facility as long as  
> their UI-PATH wasn't complete nonsense in the first place.

I agree.  I really can't see the justification for blind path rotation.

> Basically, I should be able to set my path to AUTO or something  
> (DISCVR?), and the network then ensures that my packets make it as  
> far as the nearest I-Gate, (or nearest Big High Digi, or 
> whatever the  
> sysops set the upper bound to be) and don't cause any more QRM.  Any  
> misconfigured trackers could be 'controlled' at this point as well.

I could easily add a rewrite option to my digi code that'd let you
substitute something for AUTO (or any other keyword) when it's received by
the digi.  It'd be a little more difficult to expand it to a full path, but
I don't know how necessary that'd be.  I think most places could get by with
a WIDEn-n substitution, or a single named hop.  Also, with preemption
enabled, you could keep a fallback (e.g. WIDE3-3) path, and have AUTO take
precedence if it hits as supported digi.  I think digi_ned can do this with
the 'swap' command - not sure about others.

Scott
N1VG







More information about the aprssig mailing list