OpenBCM V1.13 (Linux)

Packet Radio Mailbox

DB0FHN

[JN59NK Nuernberg]

 Login: GUEST





  
ZL3AI  > APRDIG   04.02.03 10:07l 226 Lines 9927 Bytes #999 (0) @ WW
BID : 830-ZL3AI
Read: GUEST
Subj: TAPR Digest, Jan 30, 4/8
Path: DB0FHN<DB0RGB<DB0FSG<DB0ZKA<DB0GPP<DB0LX<DB0RBS<DB0SEL<DB0ZDF<DB0GE<
      LX0PAC<LX0HST<HA3PG<HB9OK<IR0NET<VK3TE<VK4TUB<ZL2BAU<ZL3VML
Sent: 030203/0913Z @:ZL3VML.#80.NZL.OC #:58723 [Chch-NZ] FBB7.00i $:830-ZL3AI
From: ZL3AI@ZL3VML.#80.NZL.OC
To  : APRDIG@WW

Subject: Re: Words of Warning
From: Alex <kr1st@amsat.org>
Date: Thu, 30 Jan 2003 15:28:37 -0500
X-Message-Number: 25

Steve Dimse wrote:

>When talking about a binary protocol, think carefully about why most
>interoperating protocols on the internet are text based. FTP, HTTP, HTTPS,
>POP, SMTP, etc. If binary is such a good thing, why were these designed in
>text, and why, after decades of use for some, do they remain in text?

I honestly don't know and would appreciate at least a hint to why that
is. We do many things a certain way for years on end without questioning
its efficiency. There's probably a good reason why the above mentioned
protocols are in text, but I wouldn't know why that is. I also don't
believe in changes just because we feel there's a need to shake up
things a bit, or because we're simply dissatisfied with the current
system, without a very clear goal in mind.

>Finally, take to heart a central tenet of my profession, expressed
>elegantly by Hippocrates: "Primum Non Nocere", or first, do no harm.

So far I don't see any harm in folks attempting to develop a second
specification, standard, protocol or whatever name it will be given.
From the posted proposals I have seen  so far it's pretty much the same
as the current spec, coded differently, with some minor changes. I think
the real interesting part is yet to come which could have a more
profound impact. I'm talking about things like routing etc.

It's hard to judge an effort without having seen the product of that
effort. So I will not.

73,
--Alex KR1ST

----------------------------------------------------------------------

Subject: RE: New spec - working notes
From: Bill Diaz <william.diaz@attbi.com>
Date: Thu, 30 Jan 2003 14:30:30 -0600
X-Message-Number: 26

Scott,
  I am concerned about any conversions that may be performed.  We have seen 
what can happen when conversions are performed on Mic-E data.

Converting NMEA locations to DD.DDDD may lead to rounding errors / 
inaccuracies.  I assume the locations must be converted back to NMEA to 
support legacy applications.  Would locations for Mic-E data get converted 
as well?

I am also concerned about mangling of binary data.  This is a continuing 
problem on the existing network.

Bill KC9XG

----------------------------------------------------------------------

Subject: RE: New spec - working notes
From: Miller Scott Contr 30SCS/FTI <Scott.Miller@vandenberg.af.mil>
Date: Thu, 30 Jan 2003 20:35:13 -0000
X-Message-Number: 27

I'm aware of the problems with translation and gating of messages.  That's
why I don't forsee it being used on the same frequencies anytime in the near
future.  The protocol also maintains a heard-by list to avoid gateway loops.

Scott

----------------------------------------------------------------------

Subject: RE: New spec - working notes
From: Alex <kr1st@amsat.org>
Date: Thu, 30 Jan 2003 15:43:59 -0500
X-Message-Number: 28

Steve Dimse wrote:

>Even worse, can you guarantee that the new protocol you propose, when used
>on the air, will not cause crashes in the present applications?

I agree mostly with what you said, Steve. However, this is still
hamradio. We do not need such guarantees. Not just a new protocol, not
even a new application or a change in an old application can not give
you such a guarantee. Who knows if my application can generate a string
that will result in a division by zero in some other MIC-E
implementation (this is merely a silly example). A big part of hamradio
is to explore, it's just too bad that things have become so vulnerable
in this digital era. :) If a new protocol would be set free on the
current system, I'd consider it a test of the robustness of the current
system. Maybe we can learn something from it.

73,
--Alex KR1ST

----------------------------------------------------------------------

Subject: Re: Words of Warning
From: Steve Dimse <k4hg@tapr.org>
Date: Thu, 30 Jan 2003 15:46:29 -0500
X-Message-Number: 29

On 1/30/03 at 3:28 PM kr1st@amsat.org (Alex) sent:

>>When talking about a binary protocol, think carefully about why most
>>interoperating protocols on the internet are text based. FTP, HTTP, HTTPS,
>POP,
>>SMTP, etc. If binary is such a good thing, why were these designed in text,
>and
>>why, after decades of use for some, do they remain in text?
>
>I honestly don't know and would appreciate at least a hint to why that
>is. We do many things a certain way for years on end without questioning
>its efficiency. There's probably a good reason why the above mentioned
>protocols are in text, but I wouldn't know why that is. I also don't
>believe in changes just because we feel there's a need to shake up
>things a bit, or because we're simply dissatisfied with the current
>system, without a very clear goal in mind.

They started as text because it is the easiest way to assure compatibility
(no concerns about byte and bit order, compiler data structures, and so
forth), the easiest to observe and debug, and because many of the transport
mechanisms were limited to 7 bits. Much the same as why the APRS RF and
Internet networks use text.

The Internet protocols stay that way, IMO, because while there are gains in
efficiency and elegance that might be made, converting, for example, the
global email system would be an incredible undertaking, resulting in large
amounts of efforts and disruptions in a working an essential communication
system. It simply isn't worth it.

Compare this to IPv6...this is just as big an undertaking as converting
email to binary, yet it is happening (albeit slowly). Why? It is because
there is a very sigificant gain that accompanies IPv6. The world will run
out of IP addresses sooner or later, if the pain of the conversion is not
accepted, we will run into big problems later.

So, is there a big problem looming in the future of APRS to justify this
sort of effort? I don't see it, I haven't heard anyone express it, and I
don't see anything in the design that seem to address one. This is the
lesson I'm hoping the APRS vdesigners will take to heart.

>>Finally, take to heart a central tenet of my profession, expressed
>>elegantly by Hippocrates: "Primum Non Nocere", or first, do no harm.
>
>So far I don't see any harm in folks attempting to develop a second
>specification, standard, protocol or whatever name it will be given.
>From the posted proposals I have seen  so far it's pretty much the same
>as the current spec, coded differently, with some minor changes. I think
>the real interesting part is yet to come which could have a more
>profound impact. I'm talking about things like routing etc.
>
>It's hard to judge an effort without having seen the product of that
>effort. So I will not.

Agreed, but with talk already of using this on 144.39, I'm concerned about
the effect this will have on our functioning communication system. I hope
the designers take the "do no harm" concept seriously, and thoroughly test
any prototypes on a channel other than 144.39, with all legacy applications
and hardware devices, tbefore risking integrety of the APRS RF and Internet
networks.

Steve K4HG

----------------------------------------------------------------------

Subject: RE: Words of Warning
From: Miller Scott Contr 30SCS/FTI <Scott.Miller@vandenberg.af.mil>
Date: Thu, 30 Jan 2003 20:48:47 -0000
X-Message-Number: 30

Steve, see my last post to Bob for answers on the problems I'm hoping to
solve.  I'm adding material to the document to make it more clear what the
intent is.

Analysis of the original protocol and cataloging what is right and wrong is
why I'm here.  I've read the spec, I've seen it in action, and I know what
*I* think is right and wrong.  But there are a lot of other smart folks out
there, and I appreciate their input.

As for text-based protocols, we could spend all day naming protocols that
use text and protocols that use binary.  There are compromises either way,
and both have their strong points.  For efficiency and ease of use with
processor and memory constrained clients, binary makes the most sense to me.

I'm quite familliar with the second-system effect.  For the benefit of those
who aren't, here's a quick definition from the Jargon File:

second-system effect

n.

(sometimes, more euphoniously, `second-system syndrome') When one is
designing the successor to a relatively small, elegant, and successful
system, there is a tendency to become grandiose in one's success and design
an elephantine feature-laden monstrosity. The term was first used by Fred
Brooks in his classic "The Mythical Man-Month: Essays on Software
Engineering" (Addison-Wesley, 1975; ISBN 0-201-00650-2). It described the
jump from a set of nice, simple operating systems on the IBM 70xx series to
OS/360 on the 360 series. A similar effect can also happen in an evolving
system; see Brooks's Law, creeping elegance, creeping featurism. See also
Multics, OS/2, X, software bloat. 

APRS is certainly successful, but I'm not entirely sure that 'elegant' and
'small' really apply here.  =]  What I'm trying to accomplish is almost the
opposite of this - reducing an elephantine monstrosity to a simple, elegant
system.  Yes, it'll almost certainly result in more supported message
elements, but a central point of the spec is that clients DON'T HAVE TO
SUPPORT THEM.  A PIC-based anti-tracker type device can parse and extract
position data without having to deal with half a dozen different formats.

Everyone can stop freaking out about introducing a new protocol on the air.
As you pointed out a few weeks ago, Steve, the only way anything is ever
going to happen is if someone just DOES it, runs it as an overlay network,
and proves that it's superior.  With some help, I'm hoping that we can start
proving by next week that it's possible.

Scott / N1VG

----------------------------------------------------------------------



Read previous mail | Read next mail


 16.11.2026 06:40:44lGo back Go up