On Apr 10, 2009, at 12:02 AM, john_ke5c wrote:

> > -The One-Touch reply works only for those JA stations who are  
> populated
> > in one of G2 gateways (this is his/her default gateway). Most likely
> > those JA stations got registered to use DVDongles.
>
> Kay shares a key point - this only works for 20 or so of the J- 
> Landers:
>


Yep, it is in interesting "side-effect" of two different things... non- 
US-Trust-Server stations being registered on the US-Trust-Server  
because they own a DV-Dongle...

And the code added to the local Gateways to "cache" local port changes  
and make them faster, if a user switches modules/bands on the same  
system.

What's interesting about this is the "caching" was reported not to be  
able to "know" which port a user came in from, thus it doesn't help a  
"roaming" user too much who jumps from Gateway to Gateway, even if  
they key up and talk back to the other person.

If you recall, without any data to back it up, I was surmising a while  
back that if a Gateway could cache the information received from OTHER  
Gateway during a transmission (including which port the person was on)  
and compare that timestamp to the last known update from US Root, the  
whole "roaming" thing could be made to work MUCH faster, as long as  
you were continuing a conversation with someone who wasn't switching  
Gateways.

So... I'm starting to think this leads to something that could be  
helpful for some... this JA thing helped me "see" something else here.

Let's say I'm talking to John and he's mobile, and we're callsign  
routing.  I'm "parked" for all intents and purposes on W0CDS Port B.   
John is on Gateway-A, we'll call it.  John "roams" out of the coverage  
area of Gateway-A.  We all know that this means today that me routing  
to UR: KE5C... is going to stop working at that point.  John is now on  
Gateway-B, we'll call it... but the Trust Server and my local Gateway  
think he's still on Gateway-A.

But, if John continues to callsign route to ME, continuing the QSO, we  
know that part will continue to work.  I haven't moved Gateways.

Now, if John says into his mic, "Hey Nate, I just switched systems...  
I'll stay keyed here for a moment while you hit your ONE TOUCH  
button..."  That... should work.  My rig -- not the local Gateway  
caching -- copies his new "routing" information into temporary memory,  
and now routes to the repeater he's on at each key-up, completely  
bypassing the lookup that normally has to take place when I'm routing  
to UR: KE5C.   The rig is now going to route directly to whatever it  
"copied" from John's last transmission.

This works as long as John stays on Gateway-B.  If John travels out of  
coverage of that system and into a third, he just has to tell me  
again, so I can hit the one-touch button again.

Effectively we're doing our own "routing", and you don't need the  
callsign roaming at all anymore.  You can use it to "find" someone for  
the initial call-up, but if they move to another port, gateway, etc...  
and just TELL you to hit your one-touch... it's all going to work  
normally until they move again.

So... "mobile roaming" with callsign routes DOES work, but it requires  
one more step at each "handoff" today.   It's not really using the  
network to handle the roaming, it's just a knowledge of how the system  
works and an extra manual step to tell your rig "copy down where John  
is right now and route there".

Do I have anything wrong there?  Is that scenario correct?  It sure  
sounds like with a little preparation/training on how it all works,  
users can easily hit their one-touch and "follow" any mobile user,  
moving from system to system... and it might not have been an  
"obvious" solution to those who are miffed that callsign "roaming"  
doesn't keep up with people changing repeaters "mid-flight".

--
Nate Duehr, WY0X
[email protected]




Reply via email to