Re: Certification and SRA

On Sat, Aug 27, 2005 at 02:08:48PM -0400, Robert Bernier wrote:
> Hi Chris,
>
> Sorry I didn't get back to you sooner on this, Fridays can be such a drag.

Heh.

> On August 26, 2005 03:26 pm, Chris Travers wrote:
> > >STEP ONE: debate the merits of creating a standards organization
> > >on this mail list
> > >
> > >This I think is job 1. There's going to be respectful
> > >disagreement and I'd like to hear from everybody on this. I'm
> > >pro standards group, especially after seeing how well the BSD
> > >Certification Group have done their work.
> >
> > Ok. I see where you are going. I am pro-standards-group too, in
> > case nobody had guessed :-) However, my question (especially
> > since you work for SRA) is what the perspective is likely to be
> > from the main player in this area? After all we would be lowering
> > the barrier to entry substantially.
>
> I'm not sure I understand what you're asking.
>
> Do you mean, how does SRA feel about how we would react if there was
> such a thing as a standards body? In that cae then I would say that
> a globally agreed set of standards makes it easier to train people
> exactly what they need to know to be able to work with PostgreSQL.
> There is no 'lowering the barrier' in the sense that an organization
> who decides to offer legitimate training obtains credibility by
> demonstrating thay they follow what the industry says they need in
> their DBAs.

Sadly, this is often not true. Organizations quite frequently *buy*
credibility for these kinds of courses through their marketing arms.
I am not saying that SRA would do this, but to ignore this common
practice is disingenuous.

> > The second corrolary question is what the scope of such a
> > standards group should be?
>
> I'd like to tackle this from the practical point of view. Standards
> do two things:
>
> 1. they tell people what the industry says they need to know
>
> 2. they promote good training programs
>
> Fly by night instructors are what normally forces standards to come
> into being. Yet standards don't necesarily exist at the begining of
> a new paradign, especially if the pioneers do it right.

I'm not sure what you mean by this. That an outfit is large and
established does not prevent it from being shot through with rot and
b.s. As far as standards go, it's not the existence of fly-by-night
outfits that makes it important to have them, but the fact that people
quite easily have conflicts of interest. It's *crucial* that the
standards body and the promoting body be clearly separated by a
bright, sharp, uncrossable line.

> > Should it cover curriculum development as well?
>
> Take a look at the BSD Certification Group's work. Their survey
> contains extensive data on this issue.

:)

> > I think curriculum development is more important than
> > certification ... I wonder if it is putting the cart before the
> > horse.
>
> Only if there's no debate in the community about PostgreSQL
> teaching. Consider the following questions:
>
> How do you know if the curriculum is any good?

Partly, you know it's possible that the curriculum is any good because
the people making the curriculum and the people making the test are
not in bed together. It is hard to overemphasize the deleterious
effect that conflicts of interest have on the process.

> How do your peers know what you're about 'without' taking the time
> getting to know you?

What makes you think that this is possible?

> How does an employer know if you have good training?

An employer needs to establish this during the hiring process, and
this process cannot be made easy or cheap, no matter how many hiring
managers wish it were. Taking a bunch of probing interviews and
actually checking references is a *lot* cheaper than hiring somebody
who turns out to be incompetent.

> I've noticed people hire with confidence only when a 3rd party
> agency says you're good.

If by 3rd party, you mean previous employers, educators, and peers,
then I agree. If by 3rd party, you mean some officious certification,
then it signifies only to the foolish.

Cheers,
D
--
David Fetter david(at)fetter(dot)org http://fetter.org/
phone: +1 510 893 6100 mobile: +1 415 235 3778

Remember to vote!