He's fighting against seagull managers. He can outlast them, or take them out. He CAN win.

It's everywhere.

You either play or you lose, you can't ignore it.

The PHD will not save you.

People without technical awareness making technical
decisions, screwing up your life. Forever.

Who's the next step up the ladder? What faculty
gatherings are there that you have been avoiding?

Create an elevator pitch. Get every point you
require to open up a conversation in under 30
seconds. You have to state a problem HE cares
about, and offer enough possibility for a solution
that he wants to hear more.

Practice. Practice. and Practice some more.

Then for each possible question, map out a response.
A simple, low tech, in words that tickle his preferences
response.

What is his educational background? Where did he work
before? What successes did he have that you can refer
to in awe and admiration?

Is this the Lucent exec? Research. What projects did
he work on? In the case of his losses, find out enough
to know what NOT to talk about unless he brings them
up later. Wins? Go for it. People love to talk about
their wins. And maybe it'll affect your viewpoint and
you will learn a bit more.

It might be months before you have your chance. Use them
well. Practice. Research. Be prepared.

Your goal is not to get an answer that satisfies your
issue. Your goal is to impress upon him that you have
a valuable viewpoint. An appreciation for his problems.
A respect of his opinions.

And then, and ONLY then, you can lay the groundwork for
change in a direction that you prefer. That you are SURE
is a good thing. That you can discuss in his terms. For
all you know, there is a position open that he is trying to
fill, that you fit in, that you can have a larger ability
for direction.

People at our level have valuable knowledge which is almost
never tapped unless we do this type of networking. Which
means the people without the knowledge get in dangerous
positions and end up harming us and the company/institutions
we work for. You need a seat at the table before you are
heard. These steps are required to get there.

-----------------------------------------------------------
-----------------------------------------------------------
The following is a tech writeup of my current "under the
radar" implementation of a few Linux boxes. It was written
while you were writing the above depressing post. It may
be helpful on the tech level, but I feel you really need
the above political viewpoint before you can apply the tech
portion.
-----------------------------------------------------------
-----------------------------------------------------------

I'ved just moved into my MIS department.
It is split between IBM mainframe, Coldfusion, and
SQL/Server / desktop programming.

I was told to produce a filter program that must be
executable on the mainframe, processes a bunch of data,
moves a bunch of files around to multiple systems, mainframe,
FTP server, etc.

This program is the definition of mission critical. Without
it in operation, we would be unable to put mail in the
mailstream.

I was told it would be good (wink, wink, nudge, nudge)
if the program ran either on the mainframe or NT to limit
the number of system dependencies. My response is that
I could easily do it on Linux, but painfully / slowly on the
mainframe and in a way I would not be able to support on
NT. If they want, I could code it for NT, but I'm not willing
to stake my job on it's continuing operation.

Told to do it on Linux.

It has been in operation for about a month now. No burps,
everyone happy. Since I had stolen resources to whip it up,
ie: no development investment, and now it is proven, we are
about to buy 2 identical Linux servers to dedicate to it. Not
that it needs the CPU, just we want that level of isolation in
the event of failure.


Our West coast operation has the same issue, and needs to run
the same code with some modifications for their environment.
They are a 100% NT based. They were purchased by us a few years
back, but the operations was never integrated with our MF based
system for the printing. But they have the same business problem.

Their internal IT is fiercely independant. Why shouldn't they
be? They make money so should be left alone.

We can't move their data back and forth to run it "here", would
swamp our comm lines with them. They asked to see my code and
documents concerning the development. No problem, I sent it to
them. About 1,200 line of code that handles the core business
problem and 900 lines of glue code that moves files around, send
emails, etc. And I'd say that 1,200 lines is about 30% comments.

The only programmers they have on their side are VB and Foxpro
coders. They have a good reputation. And I heard nothing back.

So we offered to install a couple of Linux "black boxes" on their
site. I would modify the code for their particular environment,
our IT department would support the hardware / OS, and the
application would be supported by MIS.

This will be the 1st Linux installed on that site, and is solves a
major business problem. This is the type of install that has a major
impact on the people following the $$ and there are no OS religious
issues involved. It just works.

I think that should be the Linux battle cry.
It just works.
[link|http://c2.com/cgi/wiki?ItJustWorks|http://c2.com/cgi/wiki?ItJustWorks]