[aprssig] APRS protocol replacement ideas: protobuf

Heikki Hannikainen hessu at hes.iki.fi
Sun Feb 13 03:35:02 EST 2022

On Sun, 13 Feb 2022, Andrew Pavlin wrote:

> You know, if we are even going to consider a totally different protocol, 
> we probably should look at the lower levels of the protocol stack, too.

Yes, that would be a good idea, but I think that could be perhaps be a 
different discussion thread, and the APRS payload be considered somewhat 
separately from the low levels.  m17project.org is one interesting 
development on the lower level of the protocol stack.

> so I'd have forward error correction at the modem level, plus checksums 
> in the frames sent over the modem (in case the error exceeded what FEC 
> could handle).

Yep, +1 for checksums in the frames, end-to-end over the whole ecosystem. 
In practice I've been mourning over missing end-to-end checksums in APRS 
quite a lot.

There's only error correction in AX.25. We have these PASSALL issues, some 
buggy igate software corrupting packets in various ways, and even packets 
truncating/concatenating on APRS-IS sometimes (another packet sometimes 
appears within the comment field of another packet). If the payload would 
have an end-to-end checksum, these corruptions would be easy to detect, 
and likely much less common as it would then be immediately obvious where 
exactly they happen - the APRS-IS server for example could say which 
client is doing it, and complain to the client.

Even if a modem has error correction, an end-to-end checksum in the data 
payload may be quite useful if the payload is later transferred through a 
larger network or a stack of diverse software.

> Now, what about getting back to maintaining what we already have and 
> love/hate, which is APRS As She Is Written?

Yes, this is also an important point to discuss. Are you saying we should 
just stick with it, do minimal maintenance, and *not* discuss based on our 
collective experience how a better eventual replacement could look like?

Should such discussions preferably happen elsewhere than APRSSIG, and 
APRSSIG be dedicated to maintaining the contents of APRS101.PDF and the 
existing addendums?

   - Hessu, OH7LZB/AF5QT

More information about the aprssig mailing list