[BSDCert] What's missing?
jpb at sixshooter.v6.thrupoint.net
Sun Jul 31 23:09:13 EDT 2005
* Evan Leibovitch <evan at telly.org> [2005-07-31 01:24]:
> Hi, Dru.
> Dru wrote:
> >Evan, I'd appreciate it if you could elaborate on what you mean by
> >this statement:
> >>group has become extremely proficient at one aspect of a cert program
> >>while _seeming_ to have bypassed the many others.
> I'd hoped for a little more time on the list before wading in on this
> :-). And I beg everyone not to read flamage into what I'm about to say,
> especially since I'm a newcomer to the list. It's all intended to be
> constructive even if it sounds otherwise. REALLY. :-) And, of course,
> these are all my own opinions and I'm not speaking on behalf of anyone
> but me.
Welcome! (a little late, sorry... :-)
> The JTA report was an extremely nice piece of work. This group appears
> to have the actual mechanics of planning and creating exams. What I
> haven't seen on the website is any indication of:
> 1. The business plan and revenue model:
> There _will_ come a time at which volunteer resources will meet their
> limits. LPI has huge expenses in mundane things such as printing and
> certificate delivery. Quality exam development and maintenance takes a
> LOT of person-hours. Professional psychometricians aren't cheap and I
> have yet to find one who will offer any freebies :-). VUE and Prometric
> fees for publishing are quite high, especially in the early days when
> they don't know how much money they'll be making off you. High-quality,
> high-security translations don't come for free. And that's not even
> counting promotional expenses. As a result, LPI didn't even _think_ of
> starting in earnest until we had $300K in sponsorship funding.
Yes, a business plan is needed. We discussed the need for it very early
in our discussions, and while not much has been said about it recently,
it has not been forgotten. To date, we've focused on getting
an understanding of what it is we're trying to produce. Having
an idea of what it is we want folks to pay for is equally important.
> What is the BSD cert's revenue model? How will proctors be paid? (Are
> you willing to live with the risk of corruption of volunteer proctors?
> We weren't.) Will there be official courseware that will augment the
> exam revenue?
Corruption can happen at any time, at any level of service. It is
a serious concern and one we hope to address fully and completely. Many
comments in the Task Analysis Survey declared that responders wanted
the survey to be valuable. With anything of value, comes risk. So, yes-
we intend to protect the value of the certification(s) that we produce.
The methods for doing that will be outlined soon- hopefully in the roadmap.
> 2. Actual demand
> What numbers are envisioned for the certification? In 1999, LPI had
> vendors such as Caldera, SuSE and Linuxcare telling us the demand was
> there (and putting their sponsorship money where their mouths were). The
> number projections will have many consequences in important decisions
> (such as whether to use VUE or Prometric or partner with an existing
> VUE/Prometric client or not to use them at all).
The BSD projects do not have flagship vendors such as those at this time.
But there are organizations that use BSD in the enterprise and have
spoken to us about the desire for BSD certification to proceed.
We don't have numbers yet. But if the survey is any guide, there is
interest on every continent except Antarctica.
> Furthermore, it's important to be aware if BSD admin is the kind of job
> that would _really_ benefit from a certification. There are no big
> vendors such as Red Hat or Novell pushing BSD certification as a way to
> sell software or training. And certifications only succeed if people see
> a _real_ payoff in better job prospects or more pay, especially
> outsiders (existing Windows, Unix and Linux admins) looking to get
> involved in BSD.
One of our goals is to ensure that employers recognize the valuable skillset
that comes with BSD certification. This requires that we construct a
test environment that demands excellence from our candidates. Again,
the roadmap will give more specifics, but the survey comments and our
discussions to date have included both online and hands-on testing.
> Who is the certification for? For an HR person, why would they demand a
> BSD certification? For candidates, what kind of jobs would certification
> get them that they couldn't get without a cert? Are there so many people
> claiming to be skilled in BSD, and/or so many BSD jobs, that
> certification is necessary in the hiring process? Or is it believed that
> the mere existence of a certification will help drive jobs and/or use?
All good questions. Some have already hit this list, and been hot topics.
One way to answer these questions is this- If this certification is
demanding enough to attain, why wouldn't any Unix admin want to attain it?
Yes- of course some won't bother, because they see little value in any
certification or they have too many other concerns. But as the survey
results showed, BSD sysadmins have been wanting a certification for
a long time. We aim to provide it.
> One thing that concerns me is that, judging from the cert's home page,
> it almost appears that the cert is intended to be more of an advocacy
> vehicle than an educational and standards initiative. I suggest caution
> here. Advocacy is important, but IMO it should be seen as an indirect
> benefit of certification. For the cert project, worry more about the
> answer to "why get certified in BSD" than "why use BSD".
The survey results clearly dismissed the idea that the certification
become an advocacy tool. That isn't our focus at all.
> 3. Partnerships
> HR are a pretty lazy lot, and like to cover their behinds. When they're
> told about a BSD certification, they'll want to know who else uses it.
> Who requires it for jobs or contractors, who demands certified people
> for their resellers or instructors, and who endorses it? The BSDcert
> home page lists the four major BSD releases but that doesn't imply
> endorsement or approval. Have people like Jordan or Theo come out to say
> "this is a good thing"? This support could be especially important given
> that the BSD cert might test things that exist in, say, FreeBSD that
> don't exist in OpenBSD.
I really wouldn't expect anyone to endorse something they haven't
had a chance to seriously evaluate. And as of right now, we have
only the survey and results that anyone can evaluate in depth. The
roadmap will provide more content for evaluation as will the first exam.
We do want to work with HR folks regarding the certification. 'Managers'
and 'Hiring Managers' were one of the job task choices. (OK- not true
"HR" in a formal sense- but close.) What we learned from them is that
they also want a cert that provides evidence of skill, not just knowledge.
> Forgive my ignorance of the BSD world. Linux has the Free Standards
> Group, in which the various distributions agree on core components, file
> placements, etc. that are common to all of them. And LPI has a pretty
> close relationship with FSG, considering that FSG defines a software
> standard and LPI creates a skills standard based upon it. Does such a
> body exist to define a standard core that is absolutely common to all
> the BSDs? Remember that certification is a standard, and there are
> plenty of very good standards out there that nobody supports. (Wasn't
> Betamax technically better than VHS?)
There is a lot of close coordination between the projects at the
'core' level. And a good deal of code sharing. But, like the birth
of BSD itself, there are strong differences on some matters (just
ask about vendor driver support).
I tried to find more to say on this topic, but it all degenerated into
advocacy- so I'll let the projects and the code speak for themselves.
That said, I'd encourage you to spend some time on the BSD mailing lists.
It won't be long before you see both coordination and differences.
> So... who supports the skills standard proposed by the BSD cert team?
> The home page mentions Yahoo as a major user. Can they say something
> useful? Once the cert is created will Yahoo require new BSD admins to be
> certified? If people _within_ the BSD community can't be persuaded that
> the cert project is a Good Thing, think how hard it'll be to get
> outsiders interested. Endorsements will be KEY in getting short-term
BSD has survived (and prospered) without much publicity. In fact it's
overcome some major negative publicity. The AT&T lawsuit had many
people on edge for a long time (suit filed in April, 1992). During
the lawsuit, a lot of BSD users saw the rise of Linux and wondered
if BSD was still relevant. The first 386BSD release came out in March,
1992, and the answer proved to be yes, in spite of all the lawsuit
rumors and concern flying around. Consider what might have happened
to Linux if the SCO lawsuit had occurred in 1996 or '97.
Who supports what we're doing? Again- we don't have enough out there
yet for Yahoo or anyone else to really jump on board. Will they? Time
will tell. Will the BSD community jump on board? If the survey
results are any guide, yes- I think they will.
> Yes, this could be derided as crass "marketing". But I personally
> believe strongly that the philosophy of "if you build it they will come"
> is not, and never will be, sufficient to drive the success of a cert.
BSD isn't afraid of marketing. And this group isn't afraid of an initial
failure. That's a risk for any organization. It will take a lot of
hard work to make it succeed. I personally think there's a lot of
untapped energy in the BSD community to make this effort succeed.
> And we haven't even started to talk about the complex relationships with
> book publishers, courseware authors, and training companies. What appeal
> does an accessible certification have if training is expensive and/or
> hard to come by?
True- many good issues here. But, our primary focus is on establishing
the certifications and test methodologies- not on training programs.
Just like LPI, we'll address those concerns when they come up.
> So, Dru, sorry you asked? :-)
> Please believe that I do NOT want to be seen as being mainly negative or
> wanting to put a damper on things, nor that this is a matter of real or
> perceived competition with LPI. I want to do what I can to increase use
> of ALL open source software. But for every success story like LPI there
> are failures such as SAGE, and don't forget that right now the IT
> certification world as a whole is in a tailspin. Part of my task with
> LPI has been distancing it from the growing public dissatisfaction with
> IT certification in general:
> This task is difficult and far from complete. In order to get there, LPI
> has had to be seen primarily as a standards body, secondarily as part of
> a career-building and educational strategy, and distantly behind is the
> broader growth of open source.
> The cruel reality behind all this is that by comparison, the actual
> making of a quality exam appears to be the easiest (and at least the
> best understood) of the tasks at hand :-P.
Evan, I enjoy your piercing insights and welcome your active participation.
All Open Source projects benefit when one is successful.
I see that since I started this email there are now already four other
emails on this topic! I'll have to work harder to keep up with you!
More information about the BSDCert