| |
ZL3AI > APRDIG 28.10.03 15:18l 86 Lines 3776 Bytes #999 (0) @ WW
BID : 2226-ZL3AI
Read: GUEST
Subj: TAPR Digest, Oct 26, 5/5
Path: DB0FHN<DB0FOR<DB0MRW<OK0PKL<OK0PCC<OK0NAG<OK0PPL<DB0RES<ON0AR<IK1ZNW<
ZL2TZE<ZL3VML
Sent: 031028/1252Z @:ZL3VML.#80.NZL.OC #:11460 [Chch-NZ] FBB7.00i $:2226-ZL3AI
From: ZL3AI@ZL3VML.#80.NZL.OC
To : APRDIG@WW
Subject: Re: A Low-Tech APRSteward Proposal [was: SMART digi paths]
From: David VanHorn <dvanhorn@cedar.net>
Date: Sun, 26 Oct 2003 22:15:52 -0500
X-Message-Number: 30
>But, I think that we are getting to the point in network
>use that the digipeater system must be in control of the transmitters in the
>network, or we'll never be able to adequately control the network use.
From an industrial design point of view, (some of you have heard this
before), we are making the system complicated, then depending on people to
keep up with that complication. Not a stellar idea. Note that we don't
type in http://204.95.210.1/... we use mnemonic URLs. There's a tremendous
amount of stuff going on in the IP net system, between you hitting <CR> on
www.ebay.com, and you getting any of that data. Do they make you, or even
let you adjust any of that? Absolutely as little as possible.
I fully agree that it should be possible to manually set things up, and get
going, but to the degree that we can automate things, and make it easy to
do the right thing, we will make the system more robust.
----------------------------------------------------------------------
Subject: Re: A Low-Tech APRSteward Proposal [was: SMART digi paths]
From: Neil Johnson <njohnsn@njohnsn.com>
Date: Sun, 26 Oct 2003 22:28:28 -0600
X-Message-Number: 31
On Sunday 26 October 2003 09:15 pm, David VanHorn wrote:
>There's a tremendous
>amount of stuff going on in the IP net system, between you hitting <CR> on
>www.ebay.com, and you getting any of that data. Do they make you, or even
>let you adjust any of that? Absolutely as little as possible.
Well you can, if you know what you are doing, but by the time you've
learned how to do it, you understand why it's done that way, and end up
leaving it alone ;)
>I fully agree that it should be possible to manually set things up, and get
>going, but to the degree that we can automate things, and make it easy to
>do the right thing, we will make the system more robust.
Bob is big on the "KISS" principle, but I think a better principle is:
Make it as simple as possible, but not any simpler.
Cripes, I'm even for desiging a system similiar to M$ browser service
(sheesh, I can't believe I'm saying that) where the digi's exchange
information about their capabilties and then have an "election" to decide
who gets to be the IGATE, the WIDE, or RELAY (or some other optiimal
configuration), if a digi becomes unavailable, another election occurs and
the next best candidate for a position wins. This information could also be
sent to "smart" trackers while still allowing "dumb" trackers to work.
Heck, the digi's know their location, why couldn't they "agree" on coverage
areas. ("Since we both can contact the IGATE and I'm west of you, I'll
digipeat those packets, you digipeat packets you hear from the east. We'll
divide the space between us in half. There's a digi to the North of me you
can't hear and sees a tracker just wanting to get to the APRS-IS, so I will
digi his packets directly to the IGATE and you don't have to digipeat
them".)
I know, I already see a ton of holes in this scenario, It's just an
example.
The trick is we have two interests to balance: "Tactical" users who want to
utilize APRS according Bob's original vision of APRS (30 minute cycle time,
yada, yada) and "Strategic" (the best antonym I could think of) users who
utilize APRS according to Steve's vision (want to get to an Igate, APRS-IS,
Findu, yada, yada). I don't think these interests are mutually exclusive,
but each interest needs to take the other into account when designing the
network.
--
Neil Johnson
http://www.njohnsn.com
PGP key available on request.
---
END OF DIGEST
Read previous mail | Read next mail
| |
