| |
ZL3AI > APRDIG 04.10.03 11:22l 264 Lines 10849 Bytes #999 (0) @ WW
BID : 2084-ZL3AI
Read: GUEST
Subj: TAPR Digest, Oct 01, 9/11
Path: DB0FHN<DB0RGB<OK0PPL<DB0RES<ON0AR<VK6HGR<ZL2BAU<ZL3VML
Sent: 031003/1412Z @:ZL3VML.#80.NZL.OC #:9644 [Chch-NZ] FBB7.00i $:2084-ZL3AI
From: ZL3AI@ZL3VML.#80.NZL.OC
To : APRDIG@WW
Subject: This is not SPAM
From: Stan Horzepa <stanzepa@earthlink.net>
Date: Wed, 01 Oct 2003 18:42:33 -0400
X-Message-Number: 60
This is Stan, not Spam, announcing that the TAPR editorial board is now
soliciting articles for the Autumn issue of our quarterly newsletter, Packet
Status Register (PSR). If you would like to submit anything for PSR, please
e-mail it to me (wa1lou@tapr.org), your PSR editor, as soon as possible.
(The PSR deadline is October 26.)
[commercial content deleted]
73,
Stan Horzepa, WA1LOU, PSR Editor
----------------------------------------------------------------------
Subject: But... But...
From: Bob Bruninga <bruninga@usna.edu>
Date: Wed, 1 Oct 2003 19:26:40 -0400 (EDT)
X-Message-Number: 61
Somone said my analysis of 196 copies of a WIDE7-7 was flawed in a WIDEn-N
network where every digi could hear 4 others because there would be so
many conclusions that probably only 30 or so would actually go anywhere.....
Well DUH! Where do you think the other 160 disappeared to???
*** By COLLIDING with 160 other peoples packets, that's where ***
We MUST stop this crazy belief that WIDEn-N flooding everywhere and
resultant total vunnerabliity to user errors is the final stage of APRS.
If it is the final stage, then it is terminal meltdown.
Time to move on to the new n-N paradigm...
de WB4APR, Bob
----------------------------------------------------------------------
Subject: RIP-STOP networks
From: Bob Bruninga <bruninga@usna.edu>
Date: Wed, 1 Oct 2003 19:33:47 -0400 (EDT)
X-Message-Number: 62
Even if you dont want to give up WIDEn-N, you can still have universal
WIDEn-N operation in vast areas but still contain the damage of abuse by
some SYSOP planning:
Imagine the New Mexico Example with 30 digis throughout the state and
beyond.
IF, the dozen digis that completely encircle the state simply turned off
WIDEn-N in that one circle of digis only, that would act like RIP-STOP
nylon, at the borders yet still allow a WIDE3-3 Mobile to get to a 3 hop
IGate no matter where he was.. And still all of NM could still use
WIDEn-N if that is their choice...
If the WIDE3-3 user was just OUTSIDE of New mexico, his packets would go 3
hops in other directions to find an IGate. If he was just INSIDE new
mexico, then his packet would go INWARD 3 hops to find an IGate.
But, If some WIDE7-7 jerk drives through or around the state, then his
packets only go one way, they dont go BOTH ways. And any WIDE7-7 user
outside of the state cannot penetrate the thin RING of "NO"n-N digis to
FLOOD the entire state.
Notice that even the routine out-west RELAY,WIDE3-3 mobile can still work
perfectly well anywhere even if he is in the immediate range of one of
those "NO"n-N digis. This is because his initial RELAY packets STILL get
out through the "NO"n-N digis because EVERY digi in an APRS network still
supports RELAY and WIDE. Yet, those same "NO"n-N digis RIP-STOP
any accidental FLOODING from outside, OR from INSIDE.
de WB4APR@amsat.org, Bob
----------------------------------------------------------------------
Subject: Re: APRS Linux digi, was: Re: DigiNed, was: Re: Take Control of your
DIGI part 4
From: "Sean \\The RIMBoy\\" <sean@rimboy.com>
Date: Wed, 1 Oct 2003 16:44:08 -0500 (CDT)
X-Message-Number: 63
On Wed, 1 Oct 2003, Scott Miller wrote:
>I just ordered a Soekris net4501 - essentially a 486-133 with 64 megs of
>RAM.
Are they using the AMD chip? I thought they were ending production on
those. It should not make too much difference, interesting that there are
still people producing 486 based systems. Not that there's anything wrong
with that.
>It's got a CF socket built in, and I'll probably by testing it with
>the 16 meg card I've already got on hand, but 128's are pretty cheap these
>days.
Excellent. Nice to see a company forwarding thinking a little bit.
>Any suggestions for a distro to start with? I was looking at Bering,
>but it's probably too small. I'd rather have something that doesn't cut so
>many corners trying to squeeze onto a single floppy.
Debian is about as good as it's going to get if you don't want to cut
corners. I think you might be able to squeeze a base install into 128
megs. I dunno though, I don't use Debian.
Actually I've not had much of a problem corner cutting. I've learned
quite a bit about the kernel and shells in the process.
That said, it's handy to have bash, but it really holds your hand. That's
one corner I wish I did not have to cut but learning ash was not too bad.
>For those just tuning in, the purpose of this excercise is to develop a
>Linux mini-distro with DIGI_NED and maybe aprsd preloaded, all in a form
>that can be deployed on readily available, fairly cheap hardware.
Well, I've already figured out the aprsd onto a floppy. Anything larger
than a floppy is just gravy.
>So far I've only spent $220, and that includes a spiffy metal case. That
>still leaves a comfortable margin before the setup even matches the price of
>a new 9612+.
Yeah, but more than a new MFJ 1270. Hrm...
The main thing you'll need to figure out is:
Kernel. What do you need and what don't you need. Some stuff is fairly
obvious to trim, others not. Likewise, pick a baseline. Support 486
procs, anything newer will work without the added features. That's not
the end of the world. This project is built to meet a certain need and
processor support is not an issue (read, proc / chipset speed).
Decide on libraries. In my case, everything is built against a RH6.2 set
of libs. There is stuff like uClib (micro c-lib) that can be found not
too far from the busybox stuff.
Busybox. Makes changes very easy. Basically you have one binary compiled
with all of the options you want. From there, you symlink the program
names to it and via some crazy mojo those programs work as expected. Some
of them are fairly stripped, but I've only found one occasion where the
busybox util did not work and I needed to snag the full code (mount IIRC
did not work in the version of busybox).
I've learned quite a bit about how linux works in the process. If
anything you'll get a free education on the boot process and what makes a
linux system tick.
Sean...
--
Believing I had supernatural powers, I slammed into a brick wall.
--Paul Simon
_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_
KG4NRC http://www.rimboy.com Your source for the crap you know you need.
----------------------------------------------------------------------
Subject: Re: The bottom line, simply [was:***New*** TRACEn-N...]
From: "Ev Tupis (W2EV)" <w2ev@arrl.net>
Date: Wed, 01 Oct 2003 19:51:14 -0400
X-Message-Number: 64
This note plays to "the bottom line", users who are uneducated as being "the
problem", not that the APRS network is broken and needs fixing.
============================================================================
Bob Bruninga wrote:
>
>On Tue, 30 Sep 2003, Ev Tupis (W2EV) wrote:
>
>>===== Tier 1 ----- Everyone agrees to use RELAY,TRACE2-2. No figuring
>>if you're a car, a house a digi, etc. This alone will reduce traffic
>>overall...hands down.
>
>Wow, BAD idea. FIVE BIG problems with this:
There are only problems perceived because you took Tier 1 out of context.
It is a GREAT idea with the inclusion of Tier 2 and Tier 3 as well as the
"icing on the cake" of APRSniffer (at the end).
----- The advantages of the legacy KISS proposal -----
Everyone TX's RELAY,TRACE2-2.
Home stations will TX decay to 1 TX/hr, making overall channel load
minimal, even with that RELAY in there. However...the simplicity of R,T2-2
is profound.
The power of defining a path stays with the originator, who doesn't have to
know anyting special about how to communicate. Standard aliases apply, but
now they are fully tracable, allowing for better messaging.
With the introduction of APRSniffer, people are reminded what the area
standard is on a regular basis...automatically. No one to get mad at.
Keeps the network "TRACE7-7 enabled" in case it is important to do so at a
moment's notice.
On the other hand...
>My proposal solves all these problems:
>1) FIXED stations use simply LAN3-3 in their local LAN. One copy per digi
>2) It is network self-limiting. It only goes to X digis in the named
> "LAN"n-N set of digis
>3) The limit is imposed by the network design and the sysops, not
> vulnerable to the users choice of N.
>4) It is fully tracable, the average number of digis is less than 2!
>5) ONLY Mobiles use RELAY,WIDE2-2 because they typically have fewer
> dupes because their lower fluttering signals hit fewer digis on the
> first hop
Item 1: What if a fixed station isn't in range of one of the "firewall" digis?
Item 2: This break APRS. Your philosophy has always been to put the destiny
of a frame in the hands of the originator. An emergency comes up, and
I need to "break in" to a LANn-N protected area and I need to know
what their "LAN" designation is. Alternatively, I can always launch
via W,W,W,W,W ---- which is still allowed in the proposed system but
very ugly.
Item 3: See #2 above
Item 4: So is RELAY,TRACE2-2 (and alternatively TRACE3-3 if one must no
longer keep it simple). Tracability is shared by both proposals.
Item 5: FIXED stations may use RELAY,TRACE2-2 with little overall effect
*IF* decaying TX interval were implemented. Think about this:
even R,W,W once every 60-minutes is trivial overall.
>>APRS "Official Observers" remind those that choose to not play nice
>>about their obligation to be good network neighbors.
<snip>
>This would create animosity.
Not if it were automatic, a'la: APRSniffer -- an automated lurker that
analyzes paths locally and sends a screen message to those stations that
don't play nice (according to the local rules of the local lan). Each LAN
area may define it's own rules as to what is "nice play" and setup
accordingly.
This last item is potentially a powerful enough tool as to singlehandedly
fix the network congestoin problems by itself! Imagine a system that YOU
set the pain level for, according to the needs of YOUR local network. The
level is exceeded by a station and it is messaged with a gentle nudge to
change unproto path. One message every three days, for "4 cycles". Then
one message a day for 5-cycles. Then a message to the local digi operator
for more severe action if necessary.
You have gone to such pain to eliminate R,W,W with Wn-N. I'm not certain
that it is in the best interest of the RF network to force people who want
to abuse the system to reopen the W,W,W,W,W can of worms, just to get
around LANn-N.
Regards,
Ev, W2EV
----------------------------------------------------------------------
Read previous mail | Read next mail
| |
