You make some good points. Personally, I use both. It really depends on what your tasks are, and where the software you need is.

If you just want to do e-mail and browse the web, your underlying OS almost does not matter. If you are doing a lot of scripting, customization, and learning of fundamental computer science then you could succeed either way, but I find the open source nature of Linux and the Linux community is very supportive of trying to learn those fundamentals. And for a lot of people, specific applications they need will almost force them to one or the other.

As I said, I use both, and I like both for different things, but I will say that for the majority of my friends that come to me for advice, I would set them up with Windows. For a lot of others, I recommend doing both side by side (whether or recommend virtualization or dual boot depends entirely on how strong their hardware is). But there are very few people I know that I would comfortably recommend a Linux only setup.

While I fully agree that this is a major step forward, I would hesitate before saying this will or should remove the middle man. Remember the journals currently organize much of the peer review and handle vetting and editing functions. Their business model should and must change, but that does not mean they are obsolete just yet.

Yes, but my post was also based on the assumption that all processing would be moved server-side, and in that case, Linux is already as good or better than Windows for that user, since all the apps would be exactly the same. That would make sense, but my suspension is that will never happen. I do believe that the desktop will continue its current trend of declining in significance as server side processing increases, but that is a far cry from saying all processing will move server side.

As I mentioned in my earlier post, there are some things you inherently want to do locally, and even above that currently in many cases processing power is cheaper than bandwidth and that is likely to remain true for the forseeable future. Remember that even many web based apps currently do much of the actual processing on the users local machine with only selected functions actually occurring on the server side.

In fact, assuming the free product works as well as the non-free one, why would anyone pick the non-free one? You are completely correct, but the key phrase there is assuming the free product works as well as the non-free one. For instance, I am a huge fan of SAGE, but it cannot yet replace mathematica.

Also, for some people and some uses, Windows works better than Linux. Personally I use both, for different tasks.

Cloud computing and the client-server architecture in general is definitely decreasing the significance of the desktop and will continue to do so, but there will likely remain some niches where it makes sense to have significant desktop performance.

One example that comes to mind is doing development work, including both traditional programming and CAD work as well as graphics design. To be responsive to the user it seems those would want to keep most of the processing near the end user. Similarly, anything dealing with sensitive information must tread lightly when dealing with the cloud or any other server which is not under direct and immediate control.

I would respectfully beg to differ. For my personal computer, I have two layers of backup, one goes to an external harddrive and works quite well to protect me from harddrive failure or being stupid and overwriting the wrong file. The other is burning a set of DVDs every now and then and store elsewhere, which protects me from physical destruction of my house.

At my last job we had 3 layers of backups; the first was to a large harddrive (actually a RAID away, but still an array of harddrives.) Then from there we made tape backups for short term storage and less frequent tape backups for long term offsite archival. But again, the harddrives where the first layer.

In short, I find harddrives make an excellent first layer of backup protection. They are faster and (depending on setup) easier to use than most other backup solutions, and very affordable.

Of course, I would never recommend relying on harddrives as your only layer of backups when dealing with mission critical data, but then I would never rely on any one layer of backups for truly mission critical data. Harddrives probably provide one of the best first layers in that case.

If you are dealing with less critical data and working on a budget then you may be able to accept just one layer of backups, and in that case then which way you choose to go depends on a number of factors but harddrives are a strong contender even there.

The question is, what were the original advertisements? If this was originally advertised at X speed unlimited usage, then it is immoral (legality is another question and I claim no knowledge there) to throttle in a way that provides less than that without a long transition period which allows plenty of time for customers to look for other options. I understand in this case there may be only one other realistic option, but they should still have abundant time to decide to either accept the change or take that other option or even look at more exotic choices like satellite. On the other hand, if the original advertisements did not promise something like that, then you may have more (moral) room to maneuver. In that case, traffic shaping may be your best option if you can't afford to just get more bandwidth. Also, I agree that billing instead of throttling may be a good idea.

I am a great lover of using python of calculations, but I am a newcomer to the language right now. Also, it does get tedious having to specify decimal all the time.

I partially disagree. MMOs for obvious reasons do not need to have an ending, but there have been good series with full on endings and even games with true endings. Babylon 5 and EFC both had true endings to the series (spinoffs notwithstanding, but a spinoff is different from a sequel), and from what I understand those endings were out least sketched out before the first production began. In games, The Force Unleashed has a true ending (2 actually). Even games with sequels of a sort can have a definitive end to the plot in that game. For instance, KOTOR II was much more of a spinoff than a sequel to KOTOR which had a very conclusive ending. I do think most television is the way you describe, but as you point out most television is ephemera. There is a large corpus of books, movies, and even (non-mmo) games with absolutely conclusive conclusions.

Agreed. It is impossible to have an out of copyright monopoly (at least short of having and directly controlling the only copies in existence). Project Gutenberg illustrates this nicely.

Its worth pointing out that O'reilly is selling many of their books digitally in the epub format without DRM.

You have some good points. Relative to the US market, the Iphone is quite nice. Personally, I have a G1 with Android, but I looked very seriously at the Iphone before making my decision.

As a freshmen in college I enjoyed Innumeracy, and Beyond Numeracy. They are both very easy reads, but at least introduce some very deep ideas. In general, I would recommend almost anything by John Allen Paulos. Another one that provides a more detailed but still very accessible introduction is Chapter 0, but that one is written like a text book.

That's a fair answer to what to recommend, but it does not help the semi-computer literate consumer thinking about trying Linux. It is easy for someone with little knowledge to get overwhelmed in the options.

I do think he is right, having one size fits all would be a major mistake. With that, when speaking about the consumer desktop/laptop market (netbooks are another story) it would be of great benefit to an average consumer with limited computer knowledge to have one clear default choice if they want to try Linux.