I do agree that redundant PSUs are an extremely valuable option, but aren't entirely necessary if you are running clustered application servers. Particularly in the Windows world, downtimes due to patch management and reconfiguration far outpace the agregate downtime of hardware failures.

That said, and acknowledging that it is not mutually exclusive, I'd wager that Apple has more to gain (and probably) wants to sell more servers to achieve redundancy and availability over providing fault-tolerant servers at higher per-server cost where price is an issue for smaller businesses.

If it is anything like IBM, it may void (or at least impact) any vendor support agreement, which can be critical for high-availability environments.

The issue is that the majority of users care more about their transaction committing or their game not stoping in the middle of running more than they care about their CPU utilization. People use poorly written, crappy web applications to do (subjectively) important things. Browsers that interfere with those applications cause unpredictable application behaviors which cause user frustration. It also causes developer and support frustration if the client is making radical choices (particularly if those choices are obscured to the user.)

I would presume that the final ownership of the content is Geocities, or whomever purchased their properties, given the extremely provider-sided nature of most EULA/TOS/AUPs.

In this case, Yahoo, as the owner of Geocities, would likely have the right to distribute it.

The University of Phoenix has an interesting delima. They have a goal of offering as much opportunity as possible (lax admission standards), because it is profitable. I am an MBA student with the University of Phoenix, because I live in the middle of BFE, and drank way too much beer during my undergraduate program several years ago and graduated with a 2.5, which took a lot of schools off-the-table without a stellar GMAT score. Because of their lax standards of admission, they sign on a lot of students who simply cannot handle the program. I was enrolled on Academic Probation in which I had to maintain a 3.0 through the first four classes. During those classes, the quality of my classmates quickly improved as those who were not committed or incapable of the work dropped.

Phoenix gets penalized for giving students like me an opportunity to try to be successful in the program, but having a high failure rate when those students don't cut it in a program that is comparable to a lot of state-school MBA programs.

I moved my entire iTunes library from a dying Windows PC to a new Mac Book Pro a few years ago. Given that the storage is XML, all I had to do was copy the old iTunes Library directory and change the path in the XML file using a big find & replace, it was incredibly easy (IMHO). I also did a similar thing by restoring a playlist I accidentally deleted from a backup by editing the XML file.

I am definitely not a fan of the scope creep that has worked it's way into it, the massive updates for reasons I don't care about, but generally speaking, on my Mac I've found it to work quite well, and at least decently on the PC. Though, I'm generally comparing it to Windows Media Player, which I find very unpleasant to manage playlists.

I know exactly what you are talking about. We had the exact same thing when I was in school. I used a floppy disk in 2008 to create a driver boot disk to perform a Windows Server install (because it wouldn't accept the driver over anything but a floppy) and I needed to get the server back online. At the university we had students 2001-2006 using them all the time for things. I've recovered a handful of thesis and a few dissertations from failed media working in the IT department. I don't know what it was, but we had a Mac OS 9.x machine with one of those "Superdisk" drives which was like a 120MB Zip disk that the drive also accepted floppies. No matter how corrupt, if it was recoverable, that computer would read it. It would practically read a slice of bread.

Except when fake dope dealers sell oregano to the wrong person they end up getting shot in the face. Fake AV companies just end up pissing off nerds on /. who get stuck fixing their mom's computer.

If the customer made a legitimate case that they only purchased the game for the health benefit they may be able to claim that there is implied warranty for a particular purpose, and at that they'd probably only get their money back minus the intrinsic value of playing the game for pure entertainment if the case went in their favor, which is unlikely. However, to prove such a claim would be incredibly difficult and would hinge on what claims the manufacturer made and if it could be clearly claimed to not have such an effect. Most likely, the customer would claim that it didn't make them "smarter" or "have better memory", but the manufacturers would likely assert that it merely provides marginal improvements in areas specifically used by the game (e.g. memorization of math factors) or skills specifically part of the game (e.g. learning to identify the patterns of the game, counting items, etc.) and that it makes no warranty based that that skill will transfer to other areas of thought or mental capacity. In the case of "Brain Age" Nintendo does not make any medical claims and only asserts that the game is "an entertainment product "inspired" by Kawashima's [a neuroscientist] work." (http://www.newscientist.com/article/mg19726381.500-is-it-worth-going-to-the-mind-gym.html?page=2)

That one reason the VM choice is nice. It's not too difficult to throw together a How-To setup the VM host and then allow them to install it at home on their own machines in the VM. Even at that, I expect that students would get the media, follow the directions, and complete an install to build their own environment. Particularly online, the more vanilla and standard something can be made, the easier it is to help students over the wire.

I teach at the community college myself, and find that installing the OS is a really important part of learning to use it (creating partitions, mount points, swap, etc...) and is one of the first part that makes it very different from most Windows installation processes. Doing the install on a USB stick could result in students killing the Windows partition on the disk if they botch the install and accidentally put it on the hard disk. (I've had it happen).

Using a VM host on the lab computers (either MS Virtual PC or VMWare; assuming that your lab PCs are Windows) and then allowing them to create the virtual disk on their 4GB (or larger) flash disks will give them the install experience (without risk of damaging the host system), and allow their install to be fairly hardware independent (assuming they have the same VM host on their home PC.)

This also allows them to use a normal, general purpose distro than a stick-oriented one, that is also likely to have better textbooks available. I know any text should be good enough for derived distributions, but for students having an out-of-the-box or off-the-iso experience can alleviate a lot of first-week frustrations, and gives them a better (vanilla) resource to consult when bad things happen.

You could make the argument that a programming student should be comfortable with the command line, but that is not always the case. Students whom have never used any *nix variant (except maybe OS X unknowingly) have to learn to navigate around a shell, learn the compiler arguments, the language all at once in a very small window in a 101 class.

I teach introductory computer courses and find the vast differences in skill to be a significant challenge. There are some who need to start at "power button" even if the class has no formal pre-requisite and there are some I offer to take the final and if they get an A I'll let them go on to better things. A lot of students had a hard time installing the JDK on Windows, and using the java compiler from the command line. Computer science is a steep learning curve for the uninitiated and I think would benefit well from a pre-entry exam to ensure if you need start somewhere in the series of Computers 101, Basic Application Software 101 or Introduction to UNIX 101 before getting to programming so they don't get smashed the first few weeks and miss important foundational things.

When I was a CS student in the introductory classes, I had a lot of prior programming experience. I had a few peers who wanted to work together on projects (more-or-less do our own work, but they needed more hand-holding than was given by the instructor.) Simply by helping debug their code or explain "how-and-why" I tackled a problem in a specific way naturally caused their work be conceptually like mine. Even discussing the problem colored their solution toward my implementation.

It didn't help that the instructors self-written textbook was for Turbo C++ and we were the first class to switch to Visual C++ either.

What I found disturbing was how many students could pass CS 101 (with A's even) with rote memorization of the instructors programs and samples. They had no clue what was going on but could regurgitate what the instructor had used in their examples and do well (at least for a while). Students with experience (e.g. me) had issues because we used OOP or stdio, or Hungarian notation when the instructor was looking for iostreams and was hell-bent on variables 8 characters or less with no capitals.

To use the literal case of "snake oil"-- There is nothing illegal about selling snake oil, in and of itself (unless it were to be a dangerous product.) The legal issue is that a snake oil salesman implies that it "performs" some feat (implied warranty for a particular purpose). If the snake oil salesman truly believes that it works, but it can be proven that it does not, he has misrepresented the product. If the snake oil salesman knows it doesn't work, and claims that it does, then he is committing fraud. If a person choose to use this product and did not gain the advertised result, then most people would at least agree they are entitled to a refund or, maybe even damages that resulted due to use. In the case of Mr. Trudeau, it is alleged that his claims are false. Because customers are buying his book "for the purpose of implementing his 'treatments'" the book carries an implied warranty that the content of the text is fit for a particular purpose. If you bought the plans to assemble a boom-box from parts at Radio Shack, and were told that the plans worked, and you followed them properly and it did not produce a boom-box, you could claim that the product did not meet its implied warranty duties. If customers were buying his book for entertainment, not any particular purpose, all the book would need to have is "words" in some narrative format. If he said "I am selling an international anthology of alternative medical practices for historical, literary, or critical purposes" [and it's not my fault if you try them, and should they work, good for you] or "I am selling an international anthology of medical research that is the sole opinion of the individual authors" then it would be a different case. However, someone would have to first make a successful claim that the treatment does not work, or that harm was done by not using an alternative treatment, or harm was directly done by the product (which addresses another issue of strict liability.) In this case, the government is making such a claim, right or wrong. If he was not selling the book, but made it freely available, it would be a pure free-speech issue, which is a much more open to interpretation than fraud or misrepresentation in a transaction.

If you do choose FileNet, opt for the Unix/Oracle implementation. They offer the environment on Windows/MS-SQL but in the implementation I supported (inherited), it was very (I mean very) unstable and tended to crash at the slightest hiccup, and left a very poor impression. There was always a fear that a particular MS patch would kill it which was a security and stability issue that would (likely) be avoided on a mature UNIX platform running only the necessary services. (Disclaimer: Most of my work is in the Windows world). They have a decent API, but it is slow (very slow) for mass imports. There are tools that access the undocumented (unsupported) libraries that are very fast, but run the issue of support. At the time I worked with it, the viewers were unsupported on Vista. FileNet also has the issue of lack of community. If anything goes wrong, you're on the phone with IBM, which is not the style I like when it comes to managing a system. On the other systems I've supported, (LAMP servers, IIS, MSSQL, etc.) there is a lot of online vendor support, community and vibrant forums as a first step. With FN, I was able to come across one forum that was "OK" (filesite.org), and other than that it was a call to IBM or pouring though a stack of dead-tree training manuals. Definitely consider the cost of on site-support and consulting from the vendor as a necessity.