Basically, the XAML defines the UI, while the codebehind defines the logic. In the XAML, you can write binding expressions that declaratively bind values from objects to the UI or between objects in UI (if, for instance, you wanted two panes of a window to have the same size). XAML includes a style system that allows you to control rendering, and you can control that through binding expressions.

It's hardly different than other development frameworks that separate the codebehind from the visual definition, and it's nothing like a mass of undocumented XML that defines objects calling each other.

Sure, you can strictly write Model-View-Viewmodel code, which is what I think the GP was talking about, but that has pros (less coupling between the UI and the data model, not repeating yourself) and cons (much more complex, requires some additional classes that aren't in the framework).

I'm not sure how one could look at the chart at the provided link and take away that Mono is missing "many" major components.

With the exceptions of the limited implementation of WCF, which ends up forcing you to use painful (albeit well-documented) approaches for web service access, and WPF, for which you can use the workaround of using any other windowing library with a C# binding, most of what's missing is stuff few people would miss anyway. For example, the Entity Framework is a rather clunky and immature ORM; you can do better by rolling your own, avoiding ORM entirely, or using one of the many available .NET ORM libraries. WF is rather specific to particular business applications and, in my experience, sees relatively little practical use outside of the BizTalk and SharePoint communities. System.EnterpriseServices is of questionable utility due to its reliance on COM+; you can emulate most of what it does by choosing your hosting environment carefully and using System.Transactions.

Mono current implements all of the cool features of .NET (again, WPF and WCF notwithstanding), such as (P)LINQ, ASP.NET MVC, and C# through version 4.0, including lambda expressions (the new hotness).

Really, Mono is pretty compatible and implements most of what a developer needs unless they're doing particular types of deep magic using WCF, such as implementing request interceptors and custom authentication mechanisms. Personally, I think it's unfortunate they won't implement WPF, but I can't really fault their decision; Gtk# is an entirely reasonable alternative and more familiar for the non-Windows crowd at which Mono's aimed.

You mean like this link to the free as in beer development kit, which is linked from a large "Download the free tools" button on the first hit off Google or Bing for "windows phone 7 development kit?"

This is not true.

"The Automatic Updates feature is not affected by the WGA validation check. Therefore, you can use the Automatic Updates feature to make sure that you receive critical Windows updates."

Only some updates are marked as "genuine only," and this doesn't include security updates (which are all critical).

Indeed. An "enterprise PKI," as Microsoft likes to call it, handily solves this issue. Just add the root CA and intermediate CA certificates to the computers via Group Policy -- just as you would if you needed to trust a novel CA (such as, for instance, the DoD CAs). As an added bonus, if you activate auto-enrollment on Windows, your users get access to encrypted and signed e-mail, and you can trivially kick PPTP VPNs to the curb and use IKEv2 or L2TP instead. With a little more work, you can even get IPSec working. From a browser perspective, most if not all Windows browsers rely on the platform's cryptography infrastructure, so there's no need to install the certificates in each browser.

Unfortunately, while the Microsoft CA is relatively easy to use, using it for anything non-trivial requires the Enterprise or Datacenter edition of Windows Server. This is because you can't modify the certificate templates on lesser editions, and you need those to set up specialized certificates for, say, Configuration Manager.

If you're manually distributing certificates in any Windows infrastructure, you're doing it wrong.

One word: Mono.

Put simply, correlation is a necessary, but not sufficient, condition for causation.

You're claiming that the success or failure of an application is a direct condemnation of the infrastructure stack that runs it? On that basis, I could cite any LAMP application that was ditched for a Microsoft stack application and say that Apache, PHP, and MySQL can't compete with (insert name of Microsoft stack application here) running on plain old .NET and an MSSQL database.

Don't confuse the technology platform with the application. One can build garbage -- or, in this case, an unpopular site -- on any stack. In this case, as others have aptly pointed out, Microsoft dropped Live Spaces not because it didn't work or scale, but rather because it wasn't sufficiently profitable to justify the continued expense for its maintenance.

Mm, no. The i7-860 is an LGA 1156 processor, and boards for that are in the same range as the AM3 boards. You can grab an ASRock H55 for $75 on NewEgg, and $109 will get you an ASUS board with 2x PCIe and CrossfireX support. nVidia SLI support, of course, comes at a higher price point because the manufacturer must license the tech from our favorite space heater vendor.

LGA 1366, which is where the i7-9xx processors are, is indeed much more expensive, and probably for questionable benefit; HT isn't that great, though the 920 and 930 have an almost legendary reputation for easy and extensive overclocking (though one must be wary of unbalancing certain voltages that result in the on-die memory controller burning out). However, LGA 1366 is probably more future-proof, as it's currently the high-end non-server socket.

Home Premium OEM is $99 per computer, and this goes down based on the volume purchased.

CrossOver Premium (which you would generally want for this application, especially as it's cheaper than CrossOver Standard plus CrossOver Games) is $69.95 per computer, which also goes down based on the volume purchased. However, you must also renew it every year to continue to get updates. This is "at a discounted price," though I couldn't find what that price is.

So, once you've owned your computer for a year, Windows 7 Home Premium is most likely less expensive than CrossOver, as you only pay for it once. Whether that added cost is worth not giving money to Microsoft, however, is more of a personal decision.

For the same reason why things "just work" on OS X and can get hairy on Windows and Linux -- because your hardware manufacturer violates one or more standards in such a way that Windows can't reliably do this. Linux users have the same battle, though it's worse due to a lack of manufacturer support. Apple, on the other hand, has a much, much smaller space of hardware to support, all of which they built. If an Apple system violates spec and fails to handle ACPI sleep states, it's either a warranty issue or some engineer gets eviscerated by Jobs.

As for me, when I close the lid on my MacBook, Windows 7 puts it to sleep. When I open the lid, it wakes up. Every time. I ripped OS X off it as soon as I got it home, but I imagine OS X would do the same thing. My work Dell M4400 is about 80%, and that's after I installed our corporate base Windows 7 image; the preinstalled Dell image was 0%. On my old desktop that I built myself using Gigabyte, OCZ, and MSI hardware on Intel P55, it was about 98% successful on sleep. My new EVGA-based desktop using X58 varies from 100% on a good week to 0% on a bad one, with about a 60% success on average.

So what's the solution? Microsoft already puts a lot of effort into hardware compatibility. The only way it could get much better is to start moving towards a Jobsian lockdown of the Windows platform, which would alienate developers (who are, let's remember, the reason Windows is successful; an OS relies on application software to survive), anger hardware manufacturers, and raise the ire of technocrats in both the professional developer and free software communities. Microsoft's really stuck in a tough situation here, and Linux is worse off, as they can't throw money at HC testing and use digital certificates and a dominant platform position to at least try to force people to play ball.

The only other solution is to take the Windows 7 approach and try to make the computer more intelligent at dealing with these situations.

(Also, by the by, Windows XP is an ancient OS at this point. Sleep support in Windows has gotten much better in the two subsequent versions. We don't take Linux to task for what it did back in 2.2 -- or, at least, we shouldn't.)

I would argue this says a lot more about the coding capabilities of the subcontractors and their processes than it does about .NET, which has been used in some fairly large scale projects (as another poster noted). One doesn't blame tools for the deficiencies of their wielders.

Also, the fact that any .NET assembly can be disassembled into CIL, traced, and analyzed means that anything that isn't documented either can be or already has been -- for instance, by the Mono project.

Yes, it is. (Though in a sense, it's actually both.)

The mechanism of action of methylphenidate, just like every other drug in the amphetamine analogue group, is to increase dopamine and norepinephrine levels by blocking their reuptake. It also increases the release of these neurotransmitters, but isn't nearly as potent in this regard as, say, methamphetamine or cocaine. Both of these effects are stimulatory, a fact that can be easily discerned by giving it to animal models or humans. Small doses create a feeling of energy and mental clarity, while large doses create a high distinctly different from that of depressants. Overdoses can provoke stimulant-induced schizophrenia, most likely due to the high levels of dopamine.

That said, in the treatment of ADHD, the proposed mechanism of action is that the areas of the brain controlling impulse control (presumably, the frontal lobe) are naturally depressed. Increasing overall levels of stimulation in the brain causes this area to achieve a normal level of stimulation and activity. Obviously, because this area is inhibitory, methylphenidate in this case has an overall depressant effect, but make no mistake -- methylphenidate itself is definitely a stimulant. Just because a drug can exhibit paradoxical psychoactive effects doesn't mean its standard classification is different; it just means that it has idiosyncratic paradoxical effects in a subset of the population. In this case, that happens to be therapeutic (as opposed to, say, fatal in the case of some antipsychotics and antidepressants).

- IHAADIP (I Have An Advanced Degree In Psychology)

Actually, for Windows Mobile 6.x, you can quite literally write whatever you want and your users can install it. Yes, it is possible to restrict installations only to code signed by specific certificates and this is a relatively common practice on Windows Mobile 6.x Standard, but I've yet to encounter a single WM Professional phone that had such restrictions in place out of the box. This is why you can download software such as GSPlayer, GPSTestTool, and the like for free, or hop over to Handango and buy and install whatever you want. If you're up to it, you can even download the SDK and write your own software without having to pay Microsoft a single cent.

Now, Windows Phone 7 is substantially more restrictive in what it can run, but Microsoft doesn't:
* Restrict what ad systems you can use
* Arbitrarily deny specific development languages (the only restriction is that the code run on the Silverlight version of the CLR; this means you can use F#, C#, VB, Python, and even COBOL)
* Ban the use of interpreted code, so you can write emulators in the CLR language of your choice

Microsoft has also said that its final app requirements won't include any wiggle room for random app denials, and they've also strongly implied that the testing process for app approval will be at least partially automated to remove the possibility of an angry or prudish tester zapping your app. They've also said that they're working on parental controls and intend to allow mature content once that's in place.

So, how again is Windows Mobile or Windows Phone development substantially similar to i-device development?

Bear in mind that the certificate store in Windows is shared across multiple applications. I don't have Firefox installed on my fully-patched Windows 7 Professional machine, and I don't have the CNNIC Root certificate in any of my certificate stores. If you have it, you've installed something that's added it or upgraded from a version of the OS that's trusted it. It most definitely isn't something that Windows trusts by default.

My MBP isn't handy, so I can't check and see if OS X has it by default; my MBP has a tiny OS X partition I only use for software and firmware updates, so it's as close to a stock install as you can get.