Linux Adeptus

Posted 8 May 2002 at 13:47 UTC by garym Share This

Here it is 2002 and what, 4 years after the introduction of GNU/Linux Certification Courses? What exactly was got for all that money? Was there anything to be got in the first place, or was it all a grand excersize in missing the point?

Ok, this is old bits, and the fact that what was once oh so hot a topic is now dull old bits is maybe my point already made. This is about certification.

In the past 4 years there have been some very high-profile efforts for creating credible certification (SAIR, Caldera, LPI ...) with prices starting at about $100 and raging up to $1500; $100 is a lot of money to spend on a piece of paper, unless, of course, it is a 500-dollar bill, so what do these desperate hopefulls really get? What can they possibly get?

I'd like to propose an answer: Snake Oil.

Please understand from the start: I oppose all these certifications with a passion, and can rant for hours on the futility of them; those who sell them may have very good and noble intentions, but they prey upon a mis-perceived economics of a scarcity that does not exist and offer nothing more than the "man behind the curtain" giving courage to a lion who has already stood his ground.

The first fallacy: certificates help land a good job.

Job prospects in the GNU/Linux/BSD world are not predicated on certificates, never have been, probably never will. More than any previous technology, everything needed to learn GNU Linux or BSD is free and open to all. We saw this play out in the Microsoft market, and while I have yet to meet even one certified MSCE who had any real wits about computing, I have met many self-motivated and self-taught wizards, and I have seen more than one eBay offer to sell MSCE certificates. By contrast, in the 80's Apple, Microsoft, Oracle and Novell had an iron grip on their proprietary knowledge, wrapped it in expensive licenses and created inside partner agreements with the 'schools'; in theory it probably seemed a good business plan, but in practice, it failed to do what they needed to do to create a competent infrastructure of technical services. Like all royal titles, eventually it becomes a simple black market item with no meaning at all.

The second fallicy: the teachers know something you don't

This current training hysteria repeats what we saw with Websites and HTML. In the early 90's, while myself and others strove to create access points where the everyday people could have their say on the web, I saw collegues charging as much as $1500/day to teach rich execs and sponsored employees how to do the same webpage my daughter learned in an hour at age 8 by just playing. They were teaching basic tags, with no mention of inherent style of the media, no concern for Fitts, Hicks or even net citizenship. They were selling snake oil, bottles of vinegar held out as tonics and elixirs. Today, we see the same again with .Net, courses for which have recently started to crust on my spamfilters and even clutter late night TV -- and let's face it, it is not so much that anyone believes that anyone else can actually teach anything useful about such very new technologies, but far more that people enjoy a day or two expenses paid out of the office!

The third fallacy: proto-hackers can be detected

With my developmental psychologist hat on, I can cite no metric which distinguishes bright people from dull, no more than distinguishing in advance the love of your life from an ex-spouse who will sting you for all you have. There are no magic tests, no magic questions, and certainly no magic examinations. Developmental psychology has laboured hard for over 60 years to find this Holy Grail, and yet none have ever been uncovered. Only the corporate mindset still administers the capabilities tests.

The fourth fallacy: people need to be taught

Here now in my cognitive psychologist and industrial engineer's hat: Regardless what our teaching egos may want to believe, we do not 'teach'; people are learning machines. Ask any human factors engineer. You can barely measure the time resolution of a human in learn mode. I believe this is true from the age of 6 weeks until the age of 112 (I have never met anyone older than 112, so I don't know about after that)

Can we accellerate the learning and create a petri dish to grow new hackers? I do know this much: Play is learning, and play cannot happen without access and community. If we can provide community support and provide a safe-to-explore environment and tools to do it, people learn.

I have worked with local High School children who could barely turn on their computers to play a game (this was years ago) but with only the gift of an old Linux CD, an old 486 to play upon, and an internet connection, three have now landed good jobs as system admins, and one is now his own web design company.

More recently (and I don't know if they will do it) I was asked by a major media corporation about training in web server technologies for their IT people. My detailed reply was to engage their developers in the very open source projects which they use in their day to day work; with the size and scope of their operation, there are frequent problems which arise with a myriad of project products, and that's all the more reason why they should be participating in the communities who build those products. With participation comes everything you need to be considered "trained".

The final fallacy: Paid training misses the point

As the above story relates, there is something intrisically lost when we seek to gain our stature in open source and free software th rough the traditional means of paying for brain bits. Free software offers a community of peer support where, when we need something better, we work together to create it. This is a stark departure from the proprietary mainstream who "get what they pay for". In open source, we do not give out free software as a hook to lure you into later license fees, no sales person will ever call, and there is no unified for-fee service center. There is only 'we'.

Free software brings a global community who want you to succeed because every success is cumulative to the success of the whole of the community. If East Timor students produce even the tiniest fix to extend some software for their special needs, that fix may be instantly grasped and built upon by others. What goes around comes around. This is the way of opensource. With free software, a "bug" is an opportunity to meet new friends, to learn how computers work, and to take control of your own IT destiny. More than this, through opensource, people are not passive consumers of west-coast American software goods and services, they are full peers, equal in every way. It does not matter if they are from Timor or Sudan or Cuba or if they just arrived from an alien planet, everyone is welcome because, frankly, we can use all the help we can get. This is the lesson they miss in certification.

How to grow new hackers

Unlike the proprietary world of secret knowledge, patents and tithes, if a new Linux sysadmin has a question, many people will answer. If a company cannot get their new real-time project to work, they can ask even their so-called competitor; open source does not believe the economics of scarcity apply to knowledge because knowledge is always additive --- your competitor may copy your work, but that only puts you on equal ground; they cannot 'steal' it in the same sense they could steal cattle or gold. The spirit of open communications in open source means your local junior system administrator does not need any basic training apart from knowing how (and where) to look for help.

Thus, if the teachers have nothing to share but the words "go forth and play, make friends and be helpful where ever you can", then the training is already done; we got it in kindergarten! Thereby and to wit, I certify everyone, "GNU Linux Adeptus" -- and you can pick up your certificates just as soon as a few of you help me get a certificate server going.

GNU party time!, posted 8 May 2002 at 14:27 UTC by sye » (Journeyer)

So the certification courses should really be rectified as "GNU party time!". Bring your home brew free beers, bring your free speech, bring your free schemas, bring your free minds, bring any question marks, bring "theorem for free" but don't forget your checkbook with more deposit than the estimated cost for the event. I'll buy that and there are plenty such parties out there, like the UK Linux Developers' Conference, Bristol, 4-7 July 2002. It is always a big deal to bring people together, face to face, fist for fist, isn't it?

On the other hand, some people are more relax in their uptight mind only if they are doing something which is strictly conformed to what the majority is doing. And they are willing to pay $100~$1500 for that piece of paper most people look up to. That doesn't bother me since I've already been pushed into academic rigorous abuse at the age when I didn't know any better. For people who know better, they are doing what they can, for those who didn't know any better.

Is the first fallacy really a fallacy?, posted 8 May 2002 at 15:13 UTC by tk » (Observer)

The concept of "Linux certification" sounds like an oxymoron to me, but I'm curious about the first "fallacy": are there companies which actually decide how much one knows about Linux based on a piece of paper?

(This has nothing to do with whether the certification will create a "competent infrastructure".)

Training or Certification?, posted 9 May 2002 at 03:18 UTC by ncm » (Master)

This essay appears to conflate two barely-related concepts: training classes and verified skills. Maybe the market itself conflates them, but that's no excuse.

It may be that a week-long, or six-week-long, class is worthless, and that any test that can be passed by someone who has only taken such a class is also worthless. If so, that still doesn't make testing itself, not tied to such a class, necessarily worthless.

Demonstrating the ability to solve a variety of problems using the standard tools of the trade, and manifesting awareness of good habits and the reasons behind them, is a good basis for claiming certification. Potential employers can't afford to apply enough testing to ascertain such a basis, and that's really the point (or should be) of getting and displaying a certificate. Do none of the extant certifications depend on such a thorough evaluation?

This may be a case of "Who will certify the certifiers?". In practice, only an employer who does a lot of hiring can afford to establish one certification as better than others, but they can't be counted on to create or publicize it. It's up to the organs of public discourse (e.g. Advogato, or Linux Journal) to promote those certifications that mean something.

there's only one kind of technical certification that matters . . ., posted 9 May 2002 at 05:21 UTC by rjp » (Journeyer)

it comes with 4+ years of study at a university, and it's still useful 5 years later.

industry certifications are all worth much less than they cost.

New Advogato Features

New HTML Parser: The long-awaited libxml2 based HTML parser code is live. It needs further work but already handles most markup better than the original parser.

Keep up with the latest Advogato features by reading the Advogato status blog.

If you're a C programmer with some spare time, take a look at the mod_virgule project page and help us with one of the tasks on the ToDo list!

Share this page