[aprssig] KISS Protocol Question

[Heikki Hannikainen]

It can also be useful to browse existing open source implementations of 
the protocols. For example, here are the ax25 and kiss implementations of 
aprx, in C:

http://repo.ham.fi/websvn/wsvn/aprx/trunk/ax25.c
http://repo.ham.fi/websvn/wsvn/aprx/trunk/kiss.c

There are lots more of those around the internets.

On Wed, 17 Oct 2012, Lynn W. Deffenbaugh (Mr) wrote:

> The AX.25 header portion is bit-shifted ASCII along with some bit flags in 7 
> octet "chunks".  Once you get through those (as flagged by one of the bits), 
> the packet payload for APRS packets is straight ASCII, hence what you are 
> seeing.
>
> KISS is just a wrapper (and escaper if the boundary character appears inside) 
> for a raw AX.25 packet.  You'll need to grok the AX.25 spec to be able to 
> interpret the actual packet headers.  And then you'll need to grok 
> aprs101.pdf to interpret the APRS packets in the payloads.
>
> http://www.tapr.org/pub_ax25.html links to the PDF for AX.25. Because you're 
> doing KISS, you can ignore the descriptions of "bit-stuffing" and the FCS 
> (checksum) as those are handled by the TNC.  If you're only interested in 
> APRS packets, you can ignore everything about connected mode packets and 
> focus solely on UI packets, but the interpretation of the header portion is 
> the same.
>
> Just remember that every packet starts with a destination address followed by 
> the source address optionally followed by path components and finally 
> followed by some sort of payload.  The first two addresses are opposite of 
> what we're used to seeing in humanly-readable packet interpretations.
>
> Lynn (D) - KJ4ERJ - Author of APRSISCE for Windows Mobile and Win32
>
> On 10/17/2012 4:20 PM, Brett Friermood wrote:
>> Hi all,
>> 
>> A little off topic, but figured this might be the best list to try
>> first. Hopefully someone else can benefit also.
>> 
>> I decided I wanted to try writing my own interface to a KISS TNC. More
>> for fun than anything else at this point, but I do have an idea for a
>> couple little projects down the road.
>> 
>> I did a little online research and have read the document written by
>> KA9Q. It seems pretty straight forward, but I'm having a little
>> difficulty. For a data stream I am using off the air APRS packets.
>> 
>> My understanding of the protocol is data is sent in 8 bit binary. Any
>> frames sent from the TNC to the host will be of the format
>> "FEND,Port,---packet data---,FEND" where FEND is hex C0 and the Port
>> will be hex 00 in my case.
>> 
>> I have written a simple script in Python that reads bytes from the
>> serial port. By doing so I get a hex value I can further use. I do
>> some processing to use FEND as a delimiter and input the remainder
>> into a list. I then print the "raw" strings in that list. Below is one
>> such output:
>> 
>> '\x00' '\x82' '\xa0' '\xaa' 'd' 'j' '\x9c' '\xe0' '\x9c' 'r' '\xaa'
>> '\x88' '\x9e' '@' '`' '\x9c' 'r' '\xa2' '\x92' '\xa0' '@' '\xf4'
>> '\xae' 'r' '\x98' '\xa8' '\x82' '@' '\xf4' '\x9c' 'r' '\x9a' '\x8a'
>> '\x82' '@' '\xfe' '\xae' 'r' '\x9c' '\x9c' '\xa6' '@' '\xf0' '\xae'
>> '\x92' '\x88' '\x8a' 'f' '@' '\xe1' '\x03' '\xf0' '=' '4' '3' '2' '9'
>> '.' '3' '3' 'N' '/' '0' '8' '9' '4' '3' '.' '4' '5' 'W' '-' 'S' 't'
>> 'e' 'v' 'e' ' ' 'i' 'n' ' ' 'B' 'a' 'r' 'a' 'b' 'o' 'o' ' ' 'o' 'n''
>> ' 'U' 'I' '-' 'V' 'i' 'e' 'w' '\r'
>> 
>> By printing the strings in raw format I was expecting to see all hex
>> characters since they had not been converted to ASCII yet, but as
>> shown roughly half of the packet is usually displayed as ASCII. I hope
>> this is only a phenomenon of being printed and somehow being
>> interpreted in ASCII.
>> 
>> The issue, though, is with the remaining hex characters. The first
>> element ('\x00') is of course the port part of the KISS frame. However
>> the remainder are unprintable hex characters. My understanding is that
>> the data portion of the KISS frame would be the ASCII text of the APRS
>> packet, in hex.
>> 
>> Can anyone offer any insight? Am I misunderstanding the protocol, or
>> just missing some little step?
>> 
>> Thanks,
>> Brett KQ9N
>> 
>> _______________________________________________
>> aprssig mailing list
>> aprssig at tapr.org
>> https://www.tapr.org/cgi-bin/mailman/listinfo/aprssig
>> 
>
>
> _______________________________________________
> aprssig mailing list
> aprssig at tapr.org
> https://www.tapr.org/cgi-bin/mailman/listinfo/aprssig
>

   - Hessu