| |
ZL3AI > APRDIG 30.05.04 22:40l 236 Lines 9055 Bytes #999 (0) @ WW
BID : 3386-ZL3AI
Read: GUEST
Subj: TAPR Digest, May 28, 2/3
Path: DB0FHN<DB0RGB<DB0MRW<DB0ERF<DB0FBB<DB0GOS<DB0ACC<DB0EA<DB0RES<ON0AR<
ZL2BAU<ZL2BAU<ZL3VML
Sent: 040530/1923Z @:ZL3VML.#80.NZL.OC #:25024 [Chch-NZ] FBB7.00i $:3386-ZL3AI
From: ZL3AI@ZL3VML.#80.NZL.OC
To : APRDIG@WW
Subject: Help with homebrew tracker
From: "Andrew Rich" <vk4tec@hotmail.com>
Date: Sat, 29 May 2004 06:17:33 +1000
X-Message-Number: 7
Hi,
I have a PIC 16F628 parsing the GPS Stream no problems. I am sourcing a
MX614 chip to gen audio.
My question is what "extra" data do I have to send in the stream to keep
ax25 happy ?
What actual "BYTES" need to be sent ?
What do I need to send for the txdelay ?
VK4TEC>APU25N,RELAY:=2734.44S/15306.20E-
Andrew Rich (VK4TEC)
www.tech-software.net
vk4tec@hotmail.com
space - electronics - radio - aviation
----------------------------------------------------------------------
Subject: Re: Help with homebrew tracker
From: "Scott Miller" <scott@opentrac.org>
Date: Fri, 28 May 2004 13:46:52 -0700
X-Message-Number: 8
Ah, that's the fun part. And really poorly documented, too.
You probably already know that 1200 hz is a 'mark' and 2200 hz is a 'space'.
But the encoding is NRZI (non return to zero inverted), meaning that you
send a zero as a change and a one as no change in the tone being sent.
Since that would mean losing clock sync if you sent a lot of ones, you have
to do bit stuffing. After every five consecutive ones, you have to insert a
zero. This gets removed automatically at the receiving end. The only
exceptions are for flag and reset sequences.
AX.25 (and all HDLC) frames start and end with a flag sequence of 01111110.
Everything is sent least-significant bit first. The details of what you
need to send for the AX.25 fields like the from/to and digi addresses you
can find in the AX.25 spec.
As for the TX delay, the spec is going to tell you to send a continuous
stream of flags when the channel is idle. Don't. At the beginning of a
transmission, flags don't give the receiver enough transitions to lock on
fast. Just send zeros while you wait for the receiver to sync up, then send
a flag or two. Makes a BIG difference.
The only other tricky part about AX.25 is generating the FCS checksum. It's
a 16-bit CRC-CCITT, sent low-byte first. Most of the implementations you'll
find use a lookup table to calculate it a byte at a time, but you can also
do it very easily one bit at a time while you're sending the packet.
The file fsk.c in the OpenTracker source has some comments on this. The
code's not the most readable, but it's all there and I'll be happy to answer
any questions you've got. See main.c for packet assembly.. it's basically
zeros, flag, tocall, fromcall, path, control byte, pid, data, FCS, and flag.
Scott
N1VG
----------------------------------------------------------------------
Subject: findU backup server
From: Steve Dimse <k4hg@tapr.org>
Date: Fri, 28 May 2004 19:50:49 -0400
X-Message-Number: 9
findU has been working quite well at its co-lo site, which is probably why
it appears no one has noticed that the findU backup server has been down
for some time.
Many of you remember a couple years ago when a collection was made by a
group for a backup server for findU. My involvement in this was limited to
the promise that I would provide the code and support to allow this to be a
full backup for my server. My goals were to provide hardware and internet
redundancy for findU, and more importantly, to assure that findU could be
supported into the future without my involvement. The difficult part of
findU is not writing the software, but rather installing and maintaining
the services used by findU and developing an understanding of how all the
pieces fit together.
The people that took the donations and bought the server promised to
install and maintain the findU software. However, for many months after the
system had arrived there was no movement. Finally, when findU was forced to
move from my home DSL line when DirecTV left the DSL business, I could not
wait any longer and configured the server myself to be sure there could be
service while my server moved to its new home. I did what I could to keep
the backup running after that, but I really don't have time to maintain two
servers, as I made very clear when this backup was proposed.
A few months ago the backup server developed a hardware problem in the RAID
controller which Dell eventually fixed, but not before the drives were
trashed, requiring a complete reinstall. This time the admins again
promised to stay on top of the installation, I sent a large chunk of the
code to them (enough to get the parser up and running) six weeks ago, but
it still has not been installed. Yes, I could do it myself, but that
defeats the purpose of the server, if I wanted two servers I would have
bought another myself. This server belongs to the APRS community, not to
me, and this is not my problem to solve, it is your problem. Perhaps it is
time for you all to apply pressure, and either force these people to
deliver the effort they promised, or hand the hardware off to someone that
will. Or ignore the issue if you feel that is best.
I will continue to do all I can to keep findU working, but I wanted to be
sure everyone is aware that the backup you all paid $5000 for is not
working, and has not been working for the majority of the time since the
money was collected.
Steve K4HG
----------------------------------------------------------------------
Subject: Re: findU backup server
From: "Phil Pacier, AD6NH" <ad6nh@arrl.net>
Date: Fri, 28 May 2004 18:56:8
X-Message-Number: 10
How much in/out internet pipeline would be required to host the backup
server?
73
Phil - AD6NH
----------------------------------------------------------------------
Subject: Re: findU backup server
From: Steve Dimse <k4hg@tapr.org>
Date: Fri, 28 May 2004 20:14:09 -0400
X-Message-Number: 11
On 5/28/04 at 6:56 PM Phil Pacier, AD6NH <ad6nh@arrl.net>sent:
>How much in/out internet pipeline would be required to host the backup
>server?
input is fairly constant at 5 times the internet stream, plus a few
GB/month for requests.
out depends on how it is being used. If the primary is down and the
secondary is taking the full load, the web side is 9.5 million hits, 800 GB
per month. Add to that a few more GB for name server, javAPRS connects, and
other miscellany.
if run load sharing, it would be half the above numbers, if standby just a
few GB/month.
Steve K4HG
----------------------------------------------------------------------
Subject: Re: findU backup server
From: Jeff King <jeff@aerodata.net>
Date: Fri, 28 May 2004 20:42:51 -0400
X-Message-Number: 12
>I wanted
>to be sure everyone is aware that the backup you all paid $5000 for
>is not working, and has not been working for the majority of the
>time since the money was collected.
APRSWORLD I'm sure could put it to good use, as I mentioned on this SIG when
it was first discovered extra funds had been collected. Or if you want to
take out the middleman, donate directly:
http://aprsworld.net/info/sponsors/
as many on this SIG have already done.
----------------------------------------------------------------------
[off topic]
----------------------------------------------------------------------
Subject: Re: findU backup server
From: "Phil Pacier, AD6NH" <ad6nh@arrl.net>
Date: Fri, 28 May 2004 19:56:41
X-Message-Number: 14
Sorry, I wasn't clear on my question - my fault. I'm asking in terms of
how much of an upstream/downstream internet connection would need to be
dedicated for findU to use. I'm not so concerned with hits per month or
data/month. I could probably do the calculations and figure this out, but
can you just tell me basically what kind of upstream/downstream numbers
would be necessary.
Thanks!
Phil - AD6NH
----------------------------------------------------------------------
Subject: Re: APRS Cross Country Flight
From: "Keith - VE7GDH" <ve7gdh@rac.ca>
Date: Fri, 28 May 2004 17:40:01 -0700
X-Message-Number: 15
Bob WB4APR said May 07, 2004 6:25 PM
>APRS tracker on cross country flight:
>We will be flying a tracker (W3ADO-11) on a small plane
>all around the USA in about 2 weeks.
<snip>
>-up the coast of Oregon to Quillayute, WA
>-Pt. Roberts, WA
>-then East through northern Washington, Idaho, and Montana
>-to Port of Del Bonita, WA.
<snip>
Hi to Allen Rothernberg in his 1948 Ercoupe. I and many others have watched
you make your way all the way down the east coast and over to San Diego
(did you see the stealth DEA planes following you along the Mexican border?
hi) and then up the west coast. I had an alarm set in UI-View32 to alert me
when you were in range. This afternoon I watched you make your way from
Port Angeles up past the San Juan Islands and down up to just before you
got to Point Roberts on the BC-Washington border. Enjoy your stay at the
northern-most part of your trip. Have your sun glasses ready because you
will be heading into the sunrise for the next week or so! Have a good
flight back east when you head out.
73 es cul - Keith VE7GDH
--
"I may be lost, but I know exactly where I am."
----------------------------------------------------------------------
Read previous mail | Read next mail
| |
