[aprssig] OpenAPRS has assumed "EMAIL" alias(was UniversalMessaging)

[Robert Bruninga]

Thanks Pete,

> Be advised that... many APRS-IS servers and 
> APRS clients consider callsigns that are less 
> than 3 characters to be invalid.  

Yes, even APRSdos ignores AX.25 packets with less than 3 letter
calls. 

> Also, a problem seen with the Sprouls' email 
> server early on (I think they fixed this) was 
> that it would ack back using a callsign 
> instead of EMAIL causing confusion with many 
> clients.  

Lemme think about this... Help me remember...

Sending:
WB4APR>APRS::EMAIL....:message...{line#

Expected response:
WU2Z>APRS::WB4APR....:ackline#

Many clients would not recognize this as an ACK because the
SENDING callsign (WU2Z) did not match the outgoing callsign
(EMAIL).  So, there were 3 possible fixes...:

1) If the ACK was from WU2Z, then accept the ack
2) If the line number matched ANY outgoing message, then accept
it?
3) Have WU2Z fake a "EMAIL>APRS::WB4APR...:ackline#"

I think we ended up loosening up the line-number match algorithm
and adopted #2.  If #1 is hardcoded in any systems, then that is
a problem.
If the fix was #3, then that is a problem at the AX.25 level as
Pete raises.  Pete, can you confirm that if the fix was #2, that
then nothing in in the APRS-IS would care about a "message"
addressee (though they do care about AX.25 addresses)?.

Thanks
Bob, WB4APR

> Things to consider when looking at doing short callsigns 
> and/or callsign substitution with server applications.
> 
> 73,
> 
> Pete Loveall AE5PL
> pete at ae5pl dot net