One point is mentioned here but not highlighted very much, is the fact that ARM also precluded a price race to the bottom. This is not exclusive for ARM, because it is happening everywhere, but ARM could be a front runner.

In the price race, there’s no room for highly priced software licenses. Image a hardware appliance of $ 50,-: nobody could possibly think of putting hundreds of dollars of software on that thing. In your words: “wishful thinking by the monopolists who need high retail prices to hide the price of their part of the PC, CPUs and licences for software. ”

Just take a look at the new OLPC prototype: http://news.bbc.co.uk/2/hi/technology/8428147.stm
Supercheap ($ 100,-), very innovative, large battery life. It’s probably no coincidence this thing is running on ARM?

Cheap hardware is a friend for free (as in beer AND as in speech) software and will be the enemy of closed software.

• 45 million per year X $60 or so is too big a hit so they are charging what they can for “7″ instead of pushing XP. That makes netbooks with GNU/Linux cheaper to buy by $50 or so which affects quite a few purchases.
• the small OEMs have to compete on price. They cannot do that with that other OS. Margins are incredibly small on each netbook at the low end.
• ARM will be the processor of choice at the low end in 2010. There are a few out there since 2008 but the numbers will be huge in 2010. OEMs, Google, and others are gearing up for a massive sale. I expect by the end of 2010, ARM will begin to compete in the non-netbook segment, especially thin clients, processor in the monitor, and some compact desktops/notebooks. Price/performance/battery life will all be too good to resist.
• That other OS needs huge caches which ARM does not have. That other OS will swap/thresh heavily the caches and drop off the edge in performance so M$ will not be able to run well at all in ARM. This is due to the inherently compact nature of ARM CPUs and instruction sets. M$ could get someone to build ARM with huge caches but who will bother? M$ itself? It is too big to bother. ARM can fly under the radar for a couple of years. Perhaps “8″ will have an ARM chipset and a release for ARM but not “7″. Bloatware and ARM are not a good match.

So, ARM may well be the straw that breaks the back of monopoly. It is too effective to control, too small to swat, and too inexpensive to pay the world not to use it.

ARM has interesting consequences for AMD/Intel/Via. By about 2012 or so, ARM will be kicking butt with multicore/22 nm stuff on clients and servers. AMD totally ignored this space. Via depends on small. Intel is in the game with Atom but is still handicapped with M$ and x86ness.

I would bet that M$ will take years to respond reasonably and they cannot do much more than they are now to screw with the competition without massive intervention by governments and courts. It’s almost unthinkable, but we may see competition arrive/return on the desktop. I doubt we have seen much of that since the early 1980s.

I too have no ARM. This year I bought an AMD64 X4 hair-drier. It is best used as a server. It certainly does not belong on my desktop. Neither does that other OS. I expect within a year, I will have several ARM thin and thick clients. It’s time.

Currently using an Eee 701 Debian Sid KDE 4 as my main workhorse I attached a 19″ LCD, keyboard, mouse and a bunch of peripherals via a 13-port USB hub. Everything runs fine, including watching DVDs at 1680×1050 resolution and occasional video editing. Nevertheless I know x86 is a quite inefficient architecture, so I can’t wait to lay my hands on a device with a Cortex A9 CPU inside (preferably multicore).

I honestly don’t know why they bother unless they believe they can shape the market and point it in any direction they want. I often wondered why Microsoft, with all their money, would lay people off. I’m beginning to think they are using that freed-up cash to create a Zombie Army of journalists.