<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<font face="Arial">I think the extended AX25 that Linux supports
addresses those issues. It can certainly handle Maxframe 127
instead of just 7.<br>
<br>
I played with it over a wire link between a YAM 9k6 modem and the
Baycom USCC>4 card 9k6 modem some years back. I remember being
impressed by the performance compared with 1k2 packet.<br>
<br>
Ray vk2tv<br>
</font><br>
<div class="moz-cite-prefix">On 7/11/18 2:43 pm, Carlos wrote:<br>
</div>
<blockquote type="cite"
cite="mid:CAJqR=erGyP42Oskj_C4-wD50K2+KtHJOFxBLV5XXeHh-yH-GTg@mail.gmail.com">
<pre class="moz-quote-pre" wrap="">Hi Tom,
I was reading a bit this morning and I understand that transporting
TCP/IP traffic over AX.25 might not be a wise idea due to the small
frame count.
I don't understand these things well, but with a radio capable of 64
kbps (or more) constrained by a protocol with a short frame length and
count (the number of unconfirmed packets that can exist at a give
time), the transmissions will be constantly halted waiting for
acknowledgements ???
Have you guys explored the use of SLIP/PP over radio ?
73,
Carlos VK1EA
On 11/7/18, Thomas Karpiniec <a class="moz-txt-link-rfc2396E" href="mailto:tk@1.21jiggawatts.net"><tk@1.21jiggawatts.net></a> wrote:
</pre>
<blockquote type="cite">
<pre class="moz-quote-pre" wrap="">On Tue, 6 Nov 2018, at 11:41 PM, Cary Dreelan wrote:
</pre>
<blockquote type="cite">
<pre class="moz-quote-pre" wrap="">You’ll find that your Linux distribution has compiled some defaults that
may not be suitable to this application, but they are *not* hard coded.
if you run ‘sysctl -a’ you’ll find more parameters that can be tweaked
in realtime than you’ll want to ever know about.
</pre>
</blockquote>
<pre class="moz-quote-pre" wrap="">
Oh definitely - I probably gave the wrong impression. There's lots of sysctl
stuff that vastly improves matters, particularly concerning ARP, and it's
usable. However there was at least one fundamental constant where optimising
it for packet required a kernel rebuild. I've forgotten what precisely it
was but I could do a deeper search later if anyone particularly wanted to
know.
Cheers & 73
Tom VK7NTK
_______________________________________________
OZAPRS mailing list
<a class="moz-txt-link-abbreviated" href="mailto:OZAPRS@aprs.net.au">OZAPRS@aprs.net.au</a>
<a class="moz-txt-link-freetext" href="http://lists.aprs.net.au/mailman/listinfo/ozaprs">http://lists.aprs.net.au/mailman/listinfo/ozaprs</a>
</pre>
</blockquote>
<pre class="moz-quote-pre" wrap="">_______________________________________________
OZAPRS mailing list
<a class="moz-txt-link-abbreviated" href="mailto:OZAPRS@aprs.net.au">OZAPRS@aprs.net.au</a>
<a class="moz-txt-link-freetext" href="http://lists.aprs.net.au/mailman/listinfo/ozaprs">http://lists.aprs.net.au/mailman/listinfo/ozaprs</a>
</pre>
</blockquote>
<br>
</body>
</html>