[TAG] Value by acquistion cost: a case study
rick at linuxmafia.com
Wed Nov 16 22:13:51 MSK 2005
I've received, so far, no reply to this note to NOC staff at a former
----- Forwarded message from rick -----
Date: Tue, 15 Nov 2005 20:34:36 -0800
Subject: Secondary nameservice for zone linuxmafia.com: was switched off
Folks, I noticed that NS1.[$COMPANY].COM, IP=[$SOME_NUMBER], recently
ceased doing authoritative nameservice for my domain, linuxmafia.com,
which you guys had been providing as a courtesy for many years.
(My primary DNS is at NS1.LINXUMAFIA.COM, IP=126.96.36.199.)
In the event that the shutoff was deliberate, then I guess we're done,
though I'm quite surprised to have had no notice.
In the event that it was accidental: I've always appreciated the favour,
and would be delighted if you could switch it back on.
All the best,
----- End forwarded message -----
Some would call this yet another example of "You get what you pay for",
using the societally near-ubiquitous assumption that things should be
valued (to a first approximation) at acquisition cost. Favours you pay
money for are assumed valuable; favours you pay nothing for are assumed
Those assumptions are so ingrained in most people, in most
circumstances, that it's common for audiences to give me the hairy
eyeball when I explain the above: They start worrying that I'm about to
go all Marxist (or some other form of leveller) on them.
However, in the open-source world, and to some degree in the larger
technical community too, the coin (or, at least, one important coin) in
which we hope to get paid for our efforts is _reputation_: When we do
things right in working with others, we enhance our own reputations for
competence, usefulness, and reliability.
Consequently, we tend to apply a _different_ value system -- valuing
things (and people, and companies, and arrangements) at their _use_
value as opposed to acquisition cost.
If that distinction still seems unclear, consider the Linux and *BSD
kernels: Is your kernel worthless because you did (or could) acquire it
free of charge? Don't you, in fact and in contrast, assign value to it
on the basis of what it can be used for?
I won't name the former employer, because I look after the interests of
people I work with (and have worked with), and guard their
confidentiality, even when they've piled up a long record of behaving
like colossal prats, but let's just say that you'd be surprised that a
company that well known would be so inept. (This particular firm also
remains massively important to many open-source projects, which is
somewhat horrifying to contemplate.)
o Prior to this episode, the firm changed the hostnames of all its
authoritative nameservers, when the firm changed its name -- and
neglected to notify the people for who they did secondary DNS
(so that zonefiles could be updated).
o Then, they changed the IPs of all those nameservers, once again
without notifying people for who they did secondary DNS, and
breaking that nameservice.
o Recently, I found out from my own rechecking that they simply
shut off that secondary DNS entirely -- again, without notice.
My three other secondary nameservers are run by individuals,
acquaintances in the technical community. They would never have
committed any of those errors, and neither would I -- because _we_
would not want to be treated that way ourselves, and because
we'd not want to gain a reputation for being screw-ups.
That having been said, I believe there might be an unfilled need
for easily-configured cronjobs to check one's domain for (1) pending
expiration, and (2) various third-party DNS and MX screwups. I need to
research that issue, if only because I need to add information on the
subject to the Linux User Group HOWTO, which I maintain for the Linux
Documentation Project. Any suggestions will be welcomed.
More information about the TAG