<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
</head>
<body>
<div class="moz-cite-prefix">John Gorkos wrote:<br>
</div>
<blockquote type="cite"
cite="mid:CAJcxM3z2gzT_yEfd=s4KoA9YKe1_4M3X4u3H58mEZRUxi7kehQ@mail.gmail.com">
<meta http-equiv="content-type" content="text/html; charset=UTF-8">
<div dir="auto">Please review this treatise on packet lengths
inside APRS:
<div dir="auto"><a
href="http://lists.tapr.org/pipermail/aprssig_lists.tapr.org/2016-November/046312.html"
moz-do-not-send="true">http://lists.tapr.org/pipermail/aprssig_lists.tapr.org/2016-November/046312.html</a><br>
</div>
</div>
</blockquote>
<p>Reviewed, however, some technicalities are explained in the
linked article. An APRS-IS line is:<br>
</p>
<p>[SRCCALL]>[TNCID][,VIA PATH]*:[APRS PACKET][CARRIAGE
RETURN][NEW LINE]</p>
<p>The "APRS PACKET" part might be limited to 255, but the line sent
to APRS-IS includes some of the outer header fields (the path etc)
so can be longer. If I understand the doc correctly, it thinks
317B</p>
<p>PLUS, as you yourself pointed out, when submitting by UDP you
need...</p>
<p>user CALL pass 00000<linefeed>[above line]<linefeed></p>
<p>Adding presumably another... 24B? Possibly more for suffixed
CALLs?</p>
<p>IN ANY CASE, we're not talking about using this ESP library to
send ANY APRS data, you're hand-crafting some WX format packets,
your total UDP packet would be:<br>
</p>
<pre>26B? user CALL pass 00000<linefeed>
21B? CALL>APRS,TCPIP*:
???? [insert actual WX/data here]
2B <CRLF>
</pre>
<p>If you're limited to 256B UDP by your ESP library, you have
approx 207B remaining for your telemetry/WX, however if I'm
reading APRS101 right, most WX packets are, at most:
</p>
<pre>1B Pkt Type
7B Time
8+1+9+1B Lat+Sym+Lon+Sym
7B WindDir/Speed
4B Gust
4B Temp
4B RainLastHr
4B RainLast24Hrs
4B RainSinceMidnight
3B Humidity
5B BarometricPressure
<span style="left: 144.644px; top: 251.866px; font-size: 10.8645px; font-family: sans-serif; transform: scaleX(1.09066);"></span>1B AprsSoftware
2-4B WxUnit
67B TOTAL ... and you still have 140B left for comments?
</pre>
<p>The longest (legal) telemetry looks like
5+7+7+6+6+5+6+5+4+4+4+3+3+3 = 68B, still leaving TONNES inside
your 256B UDP packet<br>
</p>
<p>Look at MIC-E formats if you wanna pack even more into a smaller
space.</p>
<p><br>
</p>
<p>Out of curiosity I also reviewed yesterday's APRS-IS captures,
looking for the longest lines, and...<br>
</p>
<blockquote type="cite"
cite="mid:CAJcxM3z2gzT_yEfd=s4KoA9YKe1_4M3X4u3H58mEZRUxi7kehQ@mail.gmail.com">
<div dir="auto">
<div dir="auto">255 is better than "the best" you can do</div>
</div>
</blockquote>
... except:<br>
<p>6390 lines were longer than 260B yesterday<br>
3745 were longer than 280B<br>
2366 were longer than 300B<br>
1896 were longer than 320B<br>
330 were bigger than 330B<br>
55 were longer than 340B</p>
<p>Some of the FAIRLY long ones seem to be wx forecasts, but not in
any of the legitimate WX formats:</p>
<p>FK8HM-13>APSFWX,TCPIP*,qAC,T2SYDNEY:)Temp-8HM!2212.85S/16628.14EiNOUMEA:
CHAUD/SATURE HUMIDITE/VAR.:
Temp.:27.50°C/HR:98%/P.atm.:1012.65mbar // JOUR:
Tmin:26.60°C/Tmax:28.00°C/Tmoy:27.16°C/dT:1.40°C // MOIS:
Tmin:25.10°C/Tmax:31.80°C/Tmoy:27.81°C/dT:6.70°C /<br>
</p>
<p>All of the stupid-long ones are one weirdo self-declaring a
"STATE OF EMERGENCY AND HUMANITARIAN CRISIS"<br>
</p>
<p>Nick VA3NNW<br>
</p>
<pre class="moz-signature" cols="72">--
"Nosey" Nick Waterman, VA3NNW/G7RZQ, K2 #5209.
use Std::Disclaimer; <a class="moz-txt-link-abbreviated" href="mailto:sig@noseynick.net">sig@noseynick.net</a>
IBM: It's Bullshit Mommery
</pre>
</body>
</html>