The choice of distro is not a bad thing in itself, but why spend so much energy with tons of near similar distros with different branding? This does not make Linux better, this doesn't give Linux the better apps it needs. The talent is there, but it's spread so thin between different projects that are trying to achieve the same goal!
I'm not certain they
are trying to achieve precisely the same goal. The variety of software on the Linux/GNU desktop is the result of many different viewpoints and ideologies. The difference between this and the Windows world is fairly simple: projects in the Windows realm compete commercially, with money. The introduction of such a tangible resource and how it is managed, combined with the demands of the user base, tends to put quite a number of Windows programs out of business over time.
Linux/GNU allows old programs (and indeed, old programming methodology) to survive indefinitely because there is no cutthroat competition. Users of free software are unable to
demand growth -- indeed, people who do are usually bashed into the ground. Users of commercial software simply let their purchases do their talking for them.
This doesn't mean free software doesn't evolve; it does (and often more impressively than commercial software). The direction of the evolution, however, is determined mostly by developers rather than users. As the movement has gained more users over time, this state has begun to shift -- how far this will go remains to be seen.
No version of Windows (not even WindowsXP SP2) installs without floppy disk if you have a RAID.
You can boot from the DVD into the installer, right. But that's it.
The installer will offer you "Install onto <<no hard disk found>>" and "Load third party driver from floppy disk". There is no "Load driver from Installation CD". Of course you can "slipstream" your drivers onto the installation DVD, but this is not quite in concordance with the EULA, and certainly not suitable for Joe Normaluser.
My friend recently had this problem in setting up his machine. Why Windows XP lacks native support for most (all?) RAID-capable motherboards is beyond me. It can become a real nuisance; especially if one doesn't have a floppy drive on-hand.
3.5" floppies are rather antiquated now; it amazes me that we still need them (for anything) at all.
Modifying the installation process probably does break the EULA. There's absolutely nothing Microsoft can do about it, however. The binding power of EULAs is quite weak due to the relatively few court cases involving them (of those cases, several have conflicting rulings). Until such a case reaches a federal court, enforcing the terms of any EULA is precarious ground (rarely, if ever, worth the effort of a major corporation).
Of course the prior paragraph applies to the U.S. only -- the legal power of the EULA is probably even more questionable (bordering, I would imagine, on laughable) in many other nations. Considering the piracy rate in China, Southeast Asia, and Eastern Europe, it doesn't seem like the EULA should be a major source of concern or attention for Microsoft.
The drivers that nVidia distributes themselves, well... maybe they work by now, who knows.
They seem to work quite well for many people. I've hardly done any significant gaming on linux/gnu (or any non-Windows platform), so the performance and stability of the drivers is not completely proven to me. However, there are quite a number of reports of success these days. The documentation of the driver has also improved significantly, if you compare the modern README to those provided in the past.
I don't believe drivers is the primary issue preventing broad adoption of free software. Linux has made great (massive, really) strides in improving and expanding the provided hardware drivers over the past decade. A good number of devices work "out of box" just as well in Linux/GNU as they do in Windows. Some third-party devices actually have support I would deem better in the open source world than they do on Windows.
I wish that a few BIG software companies would start SELLING high-end software for Linux, that would really stimulate it's growth.
I chose to address this last because it is probably the widest topic presented.
The question begging to be asked is: "why has commercial software not been successful in the GNU/Linux market?", or perhaps "why don't companies (such as game manufacturers, for instance) release more programs for Linux?".
A common response to such a query usually points to the fact that Linux/GNU users (or indeed, the entirety of all users who are dependent
primarily on free software) make up only a small percent of the market. This does not seem to be a complete response, though.
One issue deterring the introduction of numerous pieces of commercial software on Linux/GNU is simple: ideology. Particularly, it is the ideology of the extreme left (of notable mention, Mr. Stallman) which hampers widespread adoption and embracement of a multitude of proprietary programs. To put it more clearly: some users don't want commercial products, and a few even
fear the introduction of them on a wide scale.
An atmosphere of dislike (or in some cases, contempt) for proprietary programs decreases the liklihood of success for said programs -- a simple conclusion. Yet, one must ask why the introduction of large bodies of commercial software would harm any Linux/GNU user. The truth is that is does not; every person would still have the choice to simply not use commercial solutions. What the introduction does create is new demand among a portion of the user base, and this is a terrible threat to the "purity" of the system.
Ultimately, the deduction to be drawn from these premises is that Linux/GNU is only as commercially viable as the overally user base wants it to be. The greater the demand for proprietary software developments, the less the influence of the extremists will be.
Some people question why open source software itself is not effective in the market. After all, why shouldn't free software work in a free market? The problem is two fold. One, the
freedom of the software
necessitates the
lack of cost for the software. So long as the program source is freely distributable, not only can third parties create new binaries -- they can then turn around and distribute their creations at absolutely no cost to the consumer. This means that any open source project which charges for itself cannot thrive in an environment where there are people willing to give their own time and effort towards cost-free distribution. Two, people are driven by convenience and their wallet's limitations. The average consumer, given the option of purchasing a piece of free software (which will require time to find it, effort to obtain it, and money for the actual purchase)
or downloading that same software at no cost and automatically installing it through their distribution's package manager, will inevitably choose the latter.
So long as we live in a world where time and money are scarce resources, I do not see how Linux/GNU will be able to prosper in the commercial marketplace.