| |
ZL3AI > APRDIG 06.06.03 15:01l 269 Lines 9207 Bytes #999 (0) @ WW
BID : 1433-ZL3AI
Read: GUEST
Subj: TAPR Digest, May 31, 7/8
Path: DB0FHN<DB0RGB<OK0PPL<RZ6HXA<SP7MGD<ZL2TZE<ZL3VML
Sent: 030605/1358Z @:ZL3VML.#80.NZL.OC #:1729 [Chch-NZ] FBB7.00i $:1433-ZL3AI
From: ZL3AI@ZL3VML.#80.NZL.OC
To : APRDIG@WW
Subject: [aprssig] Re: APRS Internet System Design Philosophy
>But how do you differentiate?
I really don't think there's a need to - IGATEs can use TCP_NODELAY, and
everything else can just buffer it.
>If you flush the buffer, then you lose those packets. Now you have
>introduced random packet loss. This has been seen with some software
Not what I'm saying at all. You just time out the buffer and push it on
down the protocol stack after the specified interval, full or not.
>Because of the issues I pointed out before, the latency that can be
>introduced could defeat the dupe checking algorithms and therefore
>destabilize the network (yes, we were able to find 45+ second loops
>before). The gains in bandwidth utilization (what you call efficiency
>and capacity) are minimal and are strongly offset by the potential
>damage to the integrity of the APRS-IS subnet.
Ok, let's look at some real numbers. I grabbed a random chunk of traffic
from first.aprs.net port 23 and ran it through snort. In a bit over a
second, I got 8 TCP segments (I'm not going to debate the terminology here)
carrying 15 lines of CR/LF delimited APRS traffic. The regional feed from
aprsca didn't show any multi-line segments, so I'm assuming this is a result
of the higher data rate of the full feed.
Here's a summary of what I captured. Each packet has 52 bytes of overhead,
not counting link layer - 20 bytes for the IP header and 32 for TCP. ACKs
are 52 bytes as well.
05/31-17:07:38.686033 89 bytes
05/31-17:07:38.686371 Ack
05/31-17:07:38.790591 192 bytes (2 lines)
05/31-17:07:38.790981 Ack
05/31-17:07:38.902229 572 bytes (6 lines)
05/31-17:07:38.902626 Ack
05/31-17:07:39.055256 66 bytes
05/31-17:07:39.055555 Ack
05/31-17:07:39.160335 222 bytes (2 lines)
05/31-17:07:39.160757 Ack
05/31-17:07:39.480247 109 bytes
05/31-17:07:39.480597 Ack
05/31-17:07:39.695418 52 bytes
05/31-17:07:39.695813 Ack
05/31-17:07:39.798163 67 bytes
05/31-17:07:39.798539 Ack
That's 2201 bytes total, with 1369 bytes of payload and 832 of headers.
This could all bit consolidated into a single TCP segment of 1369 bytes,
with 104 bytes of total overhead. That's a savings of 728 bytes, or 33%.
Not what I'd consider minimal.
The maximum latency imposed in this example is 1112 ms, and average latency
is 556 ms.
I really don't see how an extra second of latency is going to destabilize
the network. Seems to me like a cheap way to increase the servers' usable
bandwidth by 1/3.
Scott
N1VG
----------------------------------------------------------------------
Subject: Re: D700 *& APRS
From: WA8LMF2@aol.com
Date: Sat, 31 May 2003 21:30:38 EDT
X-Message-Number: 36
In a message dated 5/31/2003 5:11:42 PM Pacific Daylight Time,
rwwood47@hotmail.com writes:
>I would like to track myself on the screen
>along with other statios. I can go to the TNC screen, and see the info
>coming into the screen from the GPS, but I cannot get the softwart to use
>the info to plot my position on the map...
1) On the SETUP, MAIN Screen, set the MODE in the "Port1 - TNC" window to
"MODE 7". This is a special Kenwood D700/THD7 mode that expects the TNC and
GPS data to be coming in as a mixed stream on the same PC serial port .
2) From the MAPS tab of the main screen, check the box "Track Me Enabled"
near the bottom of the screen (or click the "Track Me" button for a one-time
update).
Stephen H. Smith wa8lmf @ aol.com
Home Page: http://wa8lmf.com
Ham Radio/Mobile SSTV page: http://members.aol.com/wa8lmf/ham
APRS Stuff
http://members.aol.com/wa8lmf/aprs
----------------------------------------------------------------------
Subject: Re: APRSdos problem
From: Bob Bruninga <bruninga@usna.edu>
Date: Sat, 31 May 2003 21:33:52 -0400 (EDT)
X-Message-Number: 37
Error 68 is NO COMM port.
On Sat, 31 May 2003, Earl Needham wrote:
>OK, got an interesting problem with APRSdos --
>
>Using a Zenith laptop with DOS on it, no Windows, Linux, or anything
>else. Booting into APRS826, we get the following:
>
> A$:Seattle
> C$:
> ES$in Yval
> ERROR 68 See TROUBLE.TXT. Hit ENTER to resume or Q to quit?
>
>Trouble.txt indicates some sort of comm port problem -- but I'll
>be darned if I can figure out how to fix it.
Could a MOUSE be using, it, It not turned on, or you have COM1 or 2
reversed?
To recoenfigure APRSdos to use the other port, start it on the comand line
with "APRS826/new" Then you can tell it the other com port or whatever...
bob
----------------------------------------------------------------------
Subject: Re: Email Line Length
<LYR30523-135098-2003.05.31-08.08.49--kb2scs#optonline.net@lists.tapr.org>
From: kb2scs@optonline.net
Date: Sat, 31 May 2003 21:58:14 -0400
X-Message-Number: 38
Hi Bob
Your E-mail displayed on my E-Mail client just fine.
Using Pegasus Mail
Let us hope we never witness the "Silence Of The Hams"
73 DE John KB2SCS
Mail To: kb2scs@wa2pnu.#nli.ny.usa.noam
E-Mail: kb2scs@arrl.net
----------------------------------------------------------------------
Subject: Re: APRSdos problem
From: Earl Needham <needhame@yucca.net>
Date: Sat, 31 May 2003 20:01:53 -0600
X-Message-Number: 39
At 09:33 PM 5/31/2003 -0400, Bob Bruninga wrote:
>Error 68 is NO COMM port.
Understood.
><snip>
>Could a MOUSE be using, it, It not turned on,
>or you have COM1 or 2 reversed?
>
>To recoenfigure APRSdos to use the other port, start it on the comand line
>with "APRS826/new" Then you can tell it the other com port or whatever...
>bob
Thanks for the help, Bob. We finally just changed from port 2 to
port 1 (try saying THAT real fast!) and everything is well. I guess port 2
is fried on that particular antique laptop.
7 3
Earl
----------------------------------------------------------------------
Subject: Re: & TinyTrak3
From: Rich Garcia <k4gps@bellsouth.net>
Date: 31 May 2003 22:23:41 -0400
X-Message-Number: 40
Some others have mentioned advertising... It's the same as when the
programs send out their version numbers, as far as I know all APRS clients
do so.
I think it is even more important on a tracker than on a client. If I see
it is a dumb tracker like a tiny track I don't bother sending a message to
it. If it appears to be a Kenwood or a Tracker with a laptop then I will if
I need to get a message out.
I don't know how many times I see the msg field in findu for K4GPS and look
at the messages a few months later. I have not had a laptop connected to my
KPC3+ in darn close to a year but people still send messages to it. I got
to see if I still have Tracker Only in the text.
--
Rich Garcia K4GPS
Jupiter Farms FL.
www.richgarcia.net
Jupiter Farms Community Emergency Response Team
----------------------------------------------------------------------
Subject: Re: PAPRS 1.94 problem update
From: "Coe" <k4coe@arrl.net>
Date: Sat, 31 May 2003 21:47:14
X-Message-Number: 41
I have reloaded the software and as can be seen the findu listing the
problem persists. I was even in a different state and this time the 'echo'
was 5 minutes after receiving the message.
k4coe
On 05/29/03, ""Coe" <k4coe@hiwaay.net>" wrote:
>I watched the packet traffic and I am sending out the ack.
After the ack has been negotiated, it then echos the message just received
to "blank".
Samples seen at:
http://www.findu.com/cgi-bin/msg.cgi?call=K4COE
Would having the D7 DIGI on do this? I do not recall turning it on.
Still (easily) puzzled...
K4COE
----------------------------------------------------------------------
Subject: Re: & TinyTrak3
From: Robbie <mwrobertson@attbi.com>
Date: Sat, 31 May 2003 21:53:02 -0500
X-Message-Number: 42
Hi Rich,
I think it was suggested a long while back that if we wanted to indicate a
message capability, we could use the SSID and I do since then.. Kenwoods
could use a -7 for a D7, a -1 for D700. For dumb trackers, the conventional
-9 and as you mentioned, it is obvious, a TT?? does not..
I also see where most truckers use -14 and also have message capabilities..
I also notice that this SSID use is dependent on your area.
Robbie, wa9inf
----------------------------------------------------------------------
Subject: D7 not capable of full Voice Alert?
From: "Coe" <k4coe@arrl.net>
Date: Sat, 31 May 2003 21:59:5
X-Message-Number: 43
Try as I might I cannot get Tone and Tone Squelch (CTSS) to operate
simultaneously. If you turn Tone one to transmit the 100 hz PL it turns
CTSS off and vice versa. And by golly, 2 places in the manual state that
they are mutually exclusive on the THD7A/G. So maybe it is only the D700
that this works for?
According to the manual, I can either transmit the PL to break other
peoples CTSS or I can have CTSS set. But if we all have CTSS set, no one
is transmitting the 100 hz with their packets to let some know they are
around.
Is there some setting that the /G has that the add-on manual does not
document?
de K4COE
----------------------------------------------------------------------
Read previous mail | Read next mail
| |
