I enjoyed the MP article, well researched and some interesting angles.  However, there was an elephant in the room, namely:

What happens when Windows finally dies?

The business world will still need an OS, and apart from Apple, I can't see any immediate successor.  Now all the Linux guru's will be frothing over their keyboards at the moment, but wait a minute before you get the verbal flamethowers out!

There is a lot of third-party code in Windows that is licenced from other companies.  It's mainly media codec's and drivers, but a "free" OS like Linux can't include it because the GPL and "free as in Beer" distribution system is incompatible with the per-customer licencing structure these Third-Party add-ons require.  I've been pondering this (but not too much, ) and see it falling out one of three ways:

1/.     Everyone starts using MacOS, as it includes versions of most of the required Third Party code.  Unlikely, as there'd need to be a huge investment in hardware, unless Jobs starts selling the OS separately for compatible WinTel systems (Designed for Mac, anyone?)

2/.     Microsoft releases a Linux distro (Windows 8?), include the required code and sell it at a premium - stressing the "value-added" over other Linuxes.  Not a unlikely as it seems, MS did actually have a Unix flavour in the '80s.  But the amount of humble pie that would need to be eaten will definately squash this one

3/.     The bigger (Suse, Red Hat, Mandriva, PCLinuxOS et. al.) Linux distro's include the required Third Party code into their "enterprise class" Distro's and start charging for them.  This is a much more likely proposition, and arguably is beginning to already happen (the relationship between RHEL and Fedora being a case in point).

     So, all is rosy in the garden, no?  Well, I'm getting a shiver down my back at this thought.  Is anyone here old enough and bold enough to remember the "Unix Wars"?  Back in the late '80s and early '90s, there were many vendors of System V Unix (the AT&T version not the BSD version).  Companies like IBM, SUN, HP, Sco, Novell, Microsoft and others were all selling it as an OS.  Note, that this was the *same* OS in all cases. 

     Then, one of them decided that they weren't *differentiated* enough and there was no *vendor lockin* to persuade customers to stick with the vendor's other "great" products for Unix (basically customers would buy Unix from IBM and then go to SUN for an Enterprise DBM).  So, this company (I don't know who started it) started shipping a different Shell from the others, or some library files would be missing, that "just happened" to be required by a competitors application (but not be the originator's) - I'm sure you get the picture.

     Pretty soon *all* the Unix vendors were at it, some more blatantly than others (Solaris was originally Sun Unix, you know), with the result that the Server/Workstation world was filled with versions of the *same OS* that refused to work with applications and utilities from competing companies in any number of small, subtle, but fatal ways.  A real "car crash". 

     Now, I mention all this beacuse if Red Hat, SUSE, Mandriva et. al. start offering paid-for versions of Linux with third-party code in them, I can see that they'll want to maximise the market for their Distro over the others.  And the easiest way to do that?  Yeah, you're smart kids . . . you guessed it.

BTW, if you've ever experienced dll hell on Windows - it's a walk in the park compared to troubleshooting static and dynamic libraries on a *nix system - where even different versions of the same library can cause the whole thing to fatally crash

______________________

Dangerous Dave - Renaissance Man.

Of course, Linux would continue as a community-driven OS under this model as well.  But the $$$ that have persuaded NVidia and Apples QT division to invest in Linux won't be there.  Thus, those iconoclasts amongst us who cleave to a "hobbyist" Linux will have to rely on reverse-engineered versions of drivers and media apps - which take longer to develop, are fraught with legal difficulties and can be voided by a twist in the codec.

Also, most of Microsoft's value is in stock, stock-options and the investment loans raised against them.  Like most companies they keep very little hard cash in the bank.  The end could be swift - but the bankruptcy fire-sale could be veeerrrrry interesting .  Probably a lot like the scramble to get German aeronautics technology by the allies after WWII

Bruce R (21/07/2008)

---

Hi Dangerous Dave
I think that you might have under-estimated the Free and Open Source community's ability to come up with acceptable alternatives, but here's another thought. If the likes of extreme US business model folk like AutoCAD look like like getting strapped for an OS, why not include a 'free OS' in the package ?
(Can I can already hear the likes of Novell and Linspire sales staff scrambling for their address books and revenue calculators ?)
Mind you a behemoth like Microsoft will take a long time dying, and I can even see a future Windows deciding to include 'free hardware' to run it on.

Dangerous Dave - Renaissance Man.

     To paraphrase Churchill said "It has been said that Capitalism is the worst form of government except all the others that have been tried."  Maybe free-market capitalism has had it's day, certainly the strongest proponents would be the first to suffer - Microsoft included.

______________________

Dangerous Dave - Renaissance Man.