On Wed, 8 Dec 2004, Herb Peyerl wrote:

> So, if I understand you correctly... You can understand that a company should 
> do what everyone else is doing because it's safe and well known. There's no 
> competitive benefit for a company to perhaps innovate and try branching out 
> in a slightly different direction to see if there's some hidden value there 
> that offers them something that their competition doesn't have?

You're implying too much here. The company I used to work for wanted that 
their systems are up an running. The leadership actually did not care if 
their servers were running Redmond, NetBSD, VMS, AmigaOS, whatever - as 
long as they were running and in case of trouble they could be brought 
online again ASAP. We used to be two full-time systems administrators: one 
of them familiar with NetBSD, the other one not. Since we were both on the 
company's payroll regardless of what we were doing, going with Redmond 
seemed to them as the "safest" way, because it involved no extra cost for 
education and at least basic administrative tasks could be accomplished by 
the software developers people as well in case of some emergency.

There was no "hidden value" associated with switching servers to NetBSD. 
In the long term, they would of course save the money which they (by that 
time) had already sent to Redmond. So the decision was to bring in 
something new which could only be handled by one person versus keeping 
what everybody knows.

As I said, as a systems administrator I think one should look at the 
problem and then find the tool to solve the problem rather than trying to 
do everything with the one tool you already have and know.

But I can understand that the people who run a company probably don't want 
to try something out which might not work and in case of trouble 
probably costs much more than it would save if it worked.

> It's not like administering a Unix machine is alchemy ... It's a fairly well 
> known and commonly available skillset ... There are "Dummies" books about 
> it...

Yeah, of course there are! But the situation I was in and referring to 
would involve to buy such books and have a few people read them only in 
order to realize that they could not take over administration 
responsibilites because of their lack of practical experience. Don't 
underestimate the impact of how safe someone feels in using something! You 
and I and probably most other people here would not be afraid of being 
responsible for a NetBSD server in production use. We know about NetBSD's 
rc.d scripts and how they depend upon each other, we know about PkgSrc and 
how mk.conf affects the build process of packages, we know maybe a lot 
more. And if we run into trouble with any of these, we at least know where 
to look at. My former co-workers, although being smart, simply were afraid 
of things they don't know, because they do not have the experience I 
gained from running several NetBSD systems for years.

I was always unhappy with that situation, but after a while I realized 
there was very, very little I could do about it.

Bye, K&K,
T-Zee
-- 
thiesi@NetWork23.Sytes.NET ---- NetBSD: Power to the people!
Tel.: ++49-(0)171-416 05 09 -- Fax: ++49-(0)171-134 16 05 09
Mirko Thiesen, Soemmeringstrasse 41, D-10589 Berlin, Germany
              "We're with you all the way, mostly"