[vox-tech] Why I'll Never Be a Network Admin Guy

[vox-tech@lists.lugod.org]

On Tue, Oct 22, 2002 at 07:08:23PM -0700, Harry S wrote:
> On Tue, Oct 22, 2002 at 12:35:38PM -0700, Rick Moen wrote:
> > Quoting Richard Crawford (rscrawford@mossroot.com):
> > 
> > > So what does "Time to live exceeded" mean?
> > 
> > It means that the ICMP packets (comprising the ping attempt) were taking
> > too long to reach the destination, such that they exceeded their allowed
> > lifetime before getting there.
> 
> The above answer is incorrect.

Though it is not technically correct, it is a method to convey the
functional traits to a casual user. If you are looking to avoid explain
hops/nodes and how packets are modified by routers (TTL decremented by 1
for each hop) and be ready to exaplain even more about the topic, then
analogies work.

It sort of comes down to that whole analogy thing. I meet people who are
very strongly and religiously opposed to comparing clear/plain-text
e-mail to postcards sent through the U.S. mail. However, many like the
analogy quite well and continue to use it.

Then there is the analogy of computer crime, and trespassing to illegal
use of resources. These analogies break down rather quickly when you
think about most of them. (Wireless theft was discussed on bugtraq for
example.) We are permitted to "listen" to AM/FM radio broadcasts without
fear of litigation, but what about passivly listening to an
802.11[a,b,g,?] network that is not yours? How about cordless phones
that have no encryption, is there "an expectation of privacy"?, If WEP
is used, is that sufficient for expectation of privacy? Since WEP is
basicly "insecure", is even use of WEP sufficient for expectation of
privacy? Does Expectation of privacy even apply to cases with computer
crime?

Yeah, I will admit analogies often will fail to be perfect, but the
trade-offs include:
Get:
 time savings to convey a thought or idea in a general form (conceptual)
Lose:
 time to explain the item and
 quality of information

Sometimes, I will know the answer to a question posed on a newsgroup or
discussion list, but wont answer it because an explanation that I would
consider acceptable would take too long. (Look at my posts, and you can
see they are often lengthly.) When in a pinch, I'll sometimes use
analogy and use disclaimers like, "It is *kind of* like this, but not
really." ;-) But what does that do? In some cases it can confuse the
reader who does not have any genuine technical interest in the "answer"
but instead want to really know, "Is this something I should worry about
and fix and if so, how?"

Which is better in most cases, a non-answer to a question, or an
incomplete/"conceptual analogy" answer? (everyone has their own opinion
on what they would prefer.)

> When I first started learning about TCP/IP I took answers like the above
> as correct. The answer above certainly sounds right, but it isn't. 
> If you take it literally your understanding of TCP/IP ICMP packets will
> be as warped as my early understanding of TCP/IP.

This is one of the costs of anaology. :-(

If I were the ex-president of the U.S., I could instead redirect and
say, "that depends on how you define 'time' and 'too long'." :-D (tee
hee)

> Several people gave correct answers. I would also suggest checking some
> of the links provided to learn a bit more about ICMP packets (used with
> Ping and Traceroute).
> 
> NOTE:
> Earlier today (10/22/02) I experienced problems with my pacbell DSL
> connection. In particular, I received "Time to live exceeded" messages
> too. Things cleared up sometime in the afternoon.

One place I almost *always* use analogy too much, is when I deal with my
parents and "tech-know-nothings". How would you explain SMP to such a
person? A simple analogy, "it is kind of like doubling the speed of your
machine because there are 2 processors instead of 1." Um. well, that is
not right, and breaks down as an analogy *real* fast. :-/ (Yes there are
better analogies, I just chose one that was really bad on purpose.)

Some will argue that since this is a "technical mailing list" there is
no need to simplify/over-simplify an explanation. It is a tough
thing to solve. Requiring "answers to be complete" is normative
and subjective, not to mention a bit extreme. 

What to do? *sigh* :-(

-ME

-- 
-----BEGIN GEEK CODE BLOCK-----
Version: 3.12
GCS/CM$/IT$/LS$/S/O$ !d--(++) !s !a+++(-----) C++$(++++) U++++$(+$) P+$>+++ 
L+++$(++) E W+++$(+) N+ o K w+$>++>+++ O-@ M+$ V-$>- !PS !PE Y+ PGP++
t@-(++) 5+@ X@ R- tv- b++ DI+++ D+ G--@ e+>++>++++ h(++)>+ r*>? z?
------END GEEK CODE BLOCK------
decode: http://www.ebb.org/ungeek/ about: http://www.geekcode.com/geek.html