In the Valley of the Blind, the One-Eyed Man is King

I have a reputation as a Linux geek which is only partly based in fact. Fostered mostly by my friends who simply know less about it than I do, there is a grain of truth in it, but mostly it's just hype. This is a common phenomenon in tech and especially in personal computing. The devices we use are simply so complex as to be beyond the ken of those who don't actually study them. I do study them daily, though not in any organized way. My most penetrating self-analysis tells me that I'm not very high on the list, as far as knowledgeable geeks go. I do okay, but I'm nothing special. Not by a long shot!

Why am I writing about this? Well, because I want to avoid giving the wrong impression. Doing so leads to incorrect assumptions regarding my authority when I speak on the show. I try really hard not to give the wrong impression and study to make sure what I say is based on fact or clearly delineated as my opinion.

Still, folks overestimate what I know and can do. I got an email after the show, the other day, asking me to help (not for free) write a set of Linux kernel modules to provide hardware suppot for a small set of new products. I had to tell the company involved "sorry, but I'm not up to the task". Though I have experimented with very, very minor hacks on Linux drivers and had some small success, I frankly do not have the knowledge to manage anything more. In fact; what little I have done in this area has been more of a "cut and try" effort than anything else. I fervently wish I had that level of skill and, especially, understanding.

I did call this company, though, to explain and ask a few questions. They were nice enough to talk with me for a bit and I learned something. That bit of knowledge is that companies who are using Linux or planning to in the near future are having problems finding enough skilled help here in the U.S. (When you think about it; sending a basically blind job offer to me via email is an indication of deperation.) It is this company's assumption that in the next quarter, they will need to hire 3 developers and 8 administrators and technicians skilled in Linux. To this end, they've been running ads and working with headhunters and so far have found very few candidates they consider qualified.

It may well be that the limiting factor in the takeup of Linux will be the availability of skilled people to manage it. This would not be the first time in computing history that something like this has happened. In any event, it's something to think about if you're considering a corporate Linux implementation.

While it is true that there is incredible free support for Linux online and that virtually any situation can be researched there, searching that resource pool takes time; something the business usually does not have when a problem arises. They need a fix now, preferably stored in the brain of someone on-site and at a keyboard.

The pool of Linux talent is expanding rapidly. For the most part, this is not people with certifications, though, as that part of training and testing individuals is lagging behind demand. Prices are high, too. Not outrageous, but the laws of supply and demand apply here and folks of demonstrable talent command a premium. How much of a premium I am unable to ascertain. My best estimate is something near 15%, though I don't put much stock in the accuracy of that figure.

When I am asked about career training for IT, I tell whoever inquires of me not to concentrate on one specific platform to the exclusion of others. Get a broader education. Learn to work and play in several operating systems. The datacenter of the future is likely to be a mix of platforms and operating systems, especially when virtualization becomes more widespread. It is quite conceiveable that you will have to deal with many at any one time on one machine.

Jack