How not to use NANOG -- FastEther vs. FDDI

Mon Feb 9 07:13:52 UTC 1998

So, as an exercise in self-flagellation, I went through the messages
in the thread that began with:

Date: Thu, 5 Feb 1998 19:42:02 -0500 (EST)
From: Charles Sprickman <spork at inch.com>
To: nanog at merit.edu
Subject: FDDI or 100Mb Ethernet

My principal observation was that collectively, we did a pretty poor
job of utilizing the resources of this list.

The original question asked seems relatively reasonable (though it's
really related to LAN design/engineering and not the operational
internet per se, but it's at least a question a number of small
operators are asking, and is even similar to some that a number of big
operators may be asking themselves).

Our first response, from Kent, less-than-helpfully notes the existance
of gigabit ethernet (way outside the bounds of the initial question)
and suggests that the concurrent thread about MTUs is not to be payed
attention to. Failure to address the initial query and failure to say
*WHY* the MTU thread isn't worth listening to make this a poor
response.

Next, we get a reply from the oringal querrant, indicating he received
30 replies in favor of Fast Ethernet, and one in favor of FDDI, and
thanking the list. While summarization to the list is appreciated,
this level of summary is *NOT* useful. As should be grossly apparent
to anyone who reads this list for even a short period of time, the
quantity of a response has no discernable relationship to it's
quality.

If you ask a question and get a significant number of personal
replies, summarization to the list is always good. But what's
important to summarize is THE THINGS THAT PEOPLE HAVE SAID, not how
many people have said them.  The reasons why people indicated that
Fast Ethernet might be preferable to FDDI in this particular scenario
are what is important, not the number of people who may have had the
Right Answer for the Wrong Reasons, or the Wrong Answer for the Right
Reasons (or even those rare minority with the Right Answer for the
Right Reasons)!

Then we have a few replies that attempt to provide answers and provide
small technical reasons why one of the choices is better. These are
ok, except it seems likely these people didn't read the message from
the original querrant indicating he'd already gotten "his
answer". Either that, or they somehow felt that the answers he'd come
up with, or the reasons given, were inadequate. Not that anyone
bothered to say that, of course.

I have difficulty seeing how the "Beowulf Project"'s choice of a LAN
medium has much to do with the question at hand, or even closely related
ones. But of course, I didn't slog through the entire web page that was
cited. I skimmed it briefly. Presumably if one is going to cite a web page
that's not directly relevent to the question asked, one should bother to
be specific about what one is referring to and the relevence, if it's
not completely obvious. Most service providers aren't building hypercubes...

Not to mention that most web pages are ephermeral and the nanog mail archive
is not, so a quotation rather than a citation is always nice.

Some more commentary that fits the above classifications, and then a
short anecdote about the nature of cascading layer two hardware in an
enterprise network environment. Surely irrelevent to NANOG, but
probably amusing to some. Personally I think it worsens the
signal-to-noise ratio, especially when a thread has gone to pieces
like this one has.  But perhaps my sense of humor was removed at too
early an age.

Oh yes, then we have Paul and Sean off talking about migration paths,
with cisco BFRs and OC3 connections. Proabably not very relevent to the
initial query, but I suppose someone is curious, so why not re-iterate
a discussion that everyone could find in the archives.

But wait, then a segue into MPLS, the favored nanog discussion topic
of the year, I'm sure.

All-in-all, certainly not the worst thread we've had in a while.
But well-managed, structured, something we can point to as a community
and say "Yeah, we did a good job answering this question and producing
thought-provoking discussion"? I think not.

I suppose my cynicism must have reached new heights if this is what
I'm bothering to waste my time on. Still, perhaps this missive will
make people think just a little bit harder about the things they're going
to say; if so, even if it wounds characters, it will have served its
purpose.

--jhawk