tekra1
Member

Posts: 1
|
 |
« on: August 30, 2010, 09:45:00 pm » |
|
I believe there's an important issue that needs attention. The easiest way to explain it is by example. Some may remember the 'Y2K' scam. What you won't remember are the results of the urgent surveys to determine how many might be at risk. Turns out that most of them were SMEs running legacy DOS apps on 286 boxes! Horrors! Even at that time this was stone-age kit.
There's always been (and probably always will be) a broad disconnect between those who design hitech gear and those who use it. Those ancient 286/DOS boxes were still running for two very good reasons. First, they did their jobs reliably and efficiently, and were easy to maintain and adjust. Second, they offered little scope for 'fiddling about', and as every business manager knows to his cost, an embarrassing amount of supposedly 'technical time' goes into just that. To a businessman, a computer is just a machine that does a job, and is evaluated by most solely on that basis. Now don't get me wrong. My professional career was electronic hardware design, so I'm no tech philistine. Not wanting to surrender yet more of my life to the infernal machine, I avoided software as long as I could; but debugging new hardware with Bill's Bugware is like crap shooting in heavy fog, so I was forced to find an alternative. When I started with RedHat 6.0 it was as reliable as a politician's promise, but I soon became fascinated, not only by the scope and extent of what had been achieved, but even more by the social potential I could envisage, and which still hasn't materialized.
FOSS/Linux is truly one of the great achievements of modern times. Whether considered as a technological achievement, an art-form, an engineering project, a community contribution, or an example of worldwide voluntary cooperation, it's simply outstanding. Yet it receives no recognition, and for very good reasons. It's also the biggest and most successful threat to the corporate philosophy that now dominates everything, including especially the mass media and education, wherefrom it should, by rights, be receiving the highest accolades. Even worse, it proves unequivocally that the profit motive is not supreme, and may not even be viable in the long term. So don't expect this to change anytime soon, certainly not before they've put Dick Cheney up against the wall.
So here's the point of my diatribe. Say I decide to go into business bringing all this technological goodness and nourishment to SMEs, who can surely benefit from breaking their addiction to and dependence on M$, the OS for Dummies. Say I get a contract re-equipping a small company with the latest FOSS goodies, finish the job on time, collect my cheque and get a delighted cheerio from the owner. All well and good. Then, a few months later, the inevitable happens. The owner buys a couple of new machines and invites me back to set them up. I boot up my trusty PCLOS disk, only to find that it doesn't include drivers for the new MB video. No problem! Jump on the net, but find that a new version is out that has the drivers, but the repo for my own trusty version is no longer supported, or even viable. So I take the plunge and install the new version on the new boxes, thus discovering that it's not only a different interface, but has technical problems interacting with the old version. What to do? Reinstall the whole damn lot? Try to fix the technical glitches? Try to persuade the owner and his staff to live with two different interfaces and operational methodologies? Get a gun and shoot the developers?
Even the dimmest can see my point here. What to developers is a superlative art-form that deserves every latitude and indulgence is merely a useful piece of gear to others, but only when it's working. When it's not, it's worse than useless - it can be a frighteningly demanding drain on cash. The solution is simple, obvious, and very inexpensive. Just fork your CVS tree in a couple of places, isolate the most stable versions, and put a geriatric code cutter out to pasture with strict instructions that they're NEVER TO CHANGE, except for keeping up-to-date with the latest driver and library versions. If you're at all interested in grateful, generous voluntary contributions, this is the way to get it, not by flaunting the latest technoporn. Anyone who's set up a string of clients using your eternally stable distro will have a vested interest in ensuring that it remains eternally stable, and if a small percentage of his profits is the cost, it'll surely be thought worth it.
And on this note I must congratulate Vector Linux on retaining KDE 3.5. It is, was, and will always remain a superb achievement, no matter how fashions change. Good code, being an intellectual artefact, does not wear out the way that physical artefacts do, and should not be constrained or discarded due to such ephemeral things as fashion. KDE 4 is the software equivalent of flared trousers, afro hairdo's, the Cadillac El Dorado, or Lady Gaga.
Understand that commercial software vendors do not have this option - they have to 'keep up with the latest', and be seen to do so. One of the biggest advantages of FOSS is that its developers are under no external commercial or deadline pressures - only those that are self-imposed - yet they've never taken strategic advantage of it. It allows them to answer some of the REAL requirements of end users, instead of forcing them into endless upgrade cycles. All except the biggest of them are heartily sick of it - of having to run flat out to stay in the same place, given their limited budgets, technological know-how and enthusiasm.
Tekra1
|