[aprssig] Throttleing EchoLink Objects

[Pete Loveall AE5PL Lists]

One principle guides this over all others: ALL packets seen on APRS-IS are available to ALL.  What you are asking is hundreds of "regional" subnets where data is only injected for that region.  Great idea; completely undoable in an ad-hoc network which is APRS-IS.  And, there is nothing that prevents someone from botching their configuration up causing all of those "regional" packets to show back up to everyone.

Also, there are not 20,000 nodes we are looking at but 2,000.  And only a fraction of those will set up usage for APRS.  Hence, my statement that this should be left to the individual Echolink sysops.  I don't believe a central, non-real-time server (or servers dispersed regionally) provides anything other than excessive overhead (read "no benefit").  The idea to subdivide APRS-IS into regional subnets is wrought with the potential for improper configurations breaking APRS-IS even further.

My recommendation: follow Lynn's lead and let it drop.  If you want to pursue it further, contact Jonathan directly (not through a group blast) and see if APRS-IS server connectivity could be added to Echolink for links and repeaters only (similar to what is available today for TNCs).  I will be happy to work with him off-list on the particulars of the APRS-IS idiosyncrasies.

73,

Pete Loveall AE5PL
pete at ae5pl dot net

> -----Original Message-----
> From: Robert Bruninga
> Sent: Thursday, July 23, 2009 10:46 AM
> 
> But I think that even if we had each echolink node source the
> data, we would still have the same problem if this became
> popular, 20,000 nodes putting out their object globally just so
> that it could be seen locally via one local Igate.  I don't
> think this solves the long range problem as APRS grows into an
> info resource for the mobile.
> 
> Help me here.  Since the APRS-IS has achieved great progress by
> the filltered feeds, and there are even regional feeds, and the
> 2-tier system, can you help me understand why we cannot have
> regional filtered feeds.  I assume this will require new code,
> since these filtered feeds still have to drink from the global
> APRS-IS fire-hose, but why cannot these servers be modified so
> that they can also SIP from only a subset of the central
> Echolink location data and only serve that out on their regional
> feed?
> 
> 1) Lynn rewrites his code to run at each APRS-IS server that
> implements a regional feed.
> It downloads periodically the Echolink data and parses for nodes
> in its region.
> 
> 2) The server then only spews this subset oout on its regional
> feed.
> 
> What would it take to do something like this?