[consulting] Returning to a solution

[Sheryl]

Domenic Santangelo wrote:

> Or, we can mitigate the "study to pass a test" problem by taking control
> of and developing a certification program of our own from the beginning
> to create a test of value (see: Cisco certs)

This really doesn't work, though, because RedHat and Cisco certs are for
system administrators and network administrators respectively, not for
developers.  Therefore they set out a set of tasks to do and grade on the
completion of those tasks.  IMO development certification doesn't lend
itself to that kind of testing.

I'm not familiar with the content of the CISCO tests, but I have studied
for the RedHat exam.  I've spent about half my career as an admin and half
as a developer.  The test is weighted toward an admin who does hardware
and installations primarily, so it's not a good test of an admin who
spends his/her time in other ares (like installing/supporting open source
software).  So if a shop hires someone to be a LAMP/Drupal sysadmin for
their RedHat machines based upon RedHat certifications they're not
necessarily going to be well-served.

> Sure, this is a concept that's been around for a good half-century if
> not longer. Ultimately it's self-interest that drives the market (and
> the aggregate population that composes it). If it's in the clients' best
> interest to reject the concept of a cert, then they will reject it (I
> think your argument at the top, Bill, is a good reason a client might
> choose to reject it for example).

There is a whole industry called "marketing" that tries to get people to
act against their self-interest.  Check out the wikipedia page on FUD
(Fear, Uncertainty and Doubt) some time.  Judging by how much Americans
spend on junk and how little they save for retirement, I'd say that acting
against self-interest is the dominant behavior.

Sheryl