On Certification (was Re: [GENERAL] Thought provoking piece on NetBSD)

In the last exciting episode, mdean(at)xn1(dot)com (mdean) wrote:
> Guys, a multiple perspective is important. Your perspective is
> valid, but doesn't address the true purpose of these easy certs.
> They are designed to give the companies involved larger mind space
> among programmers, admins, and companies hiring them. They are a
> self-fulfilling prophecy -- here is our trained army of certified
> blah blahs. Of course the tests are easy. They are meant to suck
> in the maximum number of mediocre technos with large training fees,
> while at the same time getting commitments from these folks to be a
> Microsoft "something" or an Oracle "something" or a Redhat
> something. The cream of the crop are then enticed into tougher
> courses with larger fees. Certification is a Profit Center

Certification is only a profit center if you can get the price tag
down to something reasonable.

It costs on the order of $75K/year to keep an exam in place at Vue,
which is in addition to the cost of initially building an exam, which
means you can only make money if there are thousands of certificants.

It is *expensive* to set up an exam; it only starts to become
profitable if you're giving out tens of thousands of them per year.

My understanding is that the LPI exams for Linux are only *barely*
breaking even, and that's with a fair bit of support from IBM and
Novell.

This represents the compelling part of why PostgreSQL certification
hasn't gone very far thus far; it is *so* expensive to deploy tests
because this has fallen into the hands of the oligopoly of Pearson VUE
and Thompson/Prometric. As competitors pop up, they have the money to
buy them out. (And I probably benefit; I hold shares in Thompson...
:-))

There is a BSD Certification program ongoing; they have run into much
the same issue: to do this easily, they would need a boatload of money
to deploy testing facilities that they simply don't have.

If someone had $1M (of whatever sort of dollar doesn't too much
matter), they might in principle be able to get a new testing network
going; of course, that would probably lead to a buyout from the
oligopoly.

My counterquestion: Do you think that establishing a training network
(only to watch it get snatched up by one of those companies) is a
better use of $1M than any of the other possible uses? It strikes me
as one of the poorer uses of that kind of money...

> And don't mistake their force. a MSCE does get more money, does
> find it easier to get hired in small companies.

Maybe on your planet. The value of an MCSE is discounted pretty
heavily on mine, now that they have come up with curriculum that allow
"certification farms" to spin through unskilled people that can
memorize enough answers in 6 weeks to pass the exam.

The only certs that I hear about that are considered of high value is
CCNA/CCNP (Cisco).
--
output = ("cbbrowne" "@" "gmail.com")
http://linuxdatabases.info/info/slony.html
Rules of the Evil Overlord #132. "Before appointing someone as my
trusted lieutenant, I will conduct a thorough background investigation
and security clearance. <http://www.eviloverlord.com/>