OMG your post made me laugh so hard.

You also won't find in the news the guy next to him who decided he didn't need health insurance, also got into a terrible accident just like your brother. Of course, he couldn't afford it, landed up declaring bankruptcy, and while bad for him, the hospital and doctors got stuck with the bill, so now they need to pass it on to the rest of us. And welcome to the $15 asprin.

TCP window sizes are on a per-connection basis, and obviously only affects TCP traffic. Most VoIP, gaming, and many common file sharing protocols don't use TCP so any of these would be unaffected by QoS using the above techniques. Also, having many multiple simultaneous TCP connections in the order of hundreds or thousands would also render this technique useless as you would need to shrink each window so small so in case they all filled up that you could respond in a reasonable time frame that latency (vs inflight buffer) would kill the throughput of any TCP link. Which are reasons why this wouldn't work, and why it isn't implemented in anything beyond niche experimental stuff.

Well, you might want to reread the first paper, which talks about live lock, which is a completely different cause than the one you describe, although similar. You seem to understand what the problem is, but I fail to see how you don't understand why your solution is so wrong.

Live lock as described in the paper you linked was solved over a decade ago using the method described in the paper under section 5.1, at least under windows. It's a standard property to rate limit the interrupts under heavy load, and many network cards have many setting that allow you to tweak this from off to fixed to multiple dynamic methods.

Yes, so your argument is that you can, with great difficulty cause a possible security issue in C#, but in order to do so, you have to basically say... I'm about to do something possibly bad, please don't check to make sure what I'm doing is bad. Then modify the compiler from default to allow said code to be compiled, then put it into a fully trusted assembly so it bypasses all security checks, and THEN you might have an issue.

and this is comparision to where in C/C++ where you can write an exploit in 2 lines of code by accident, using nothing but defaults.

Yeah, if you continue to make 50k for 18 years straight, especially as a developer, you've done something very wrong. Living off of 29k pre-tax isn't difficult. Just difficult if you try and live like you have 50k to spend and only really have 29k. Getting ahead isn't easy if you have no self control, but it's really easy if you plan ahead and stick to it.

How does your system handle powering up the softraid when you turn the system on now?

Why did you stop at Windows 7 Ultimate, why didn't you compare it to the price of Windows 7 Datacenter, or a customized version of Windows 7 for Supercomputing clusters? Your copy of Windows XP doesn't have any of the added features in Ultimate, so why choose that?

Here, you go, saved you about half: http://www.dabs.com/products/m...

Of course, you could have upgraded when it first came out and saved yourself a ton of money, but you procrastinated. Maybe you should wait for Windows 9 and see if they have a deal when it launches.

It's obviousness it should be KingMotleyland.

Problem solved. All warrants now specify earth as the location.

Because no one wants to pay the $19.95 for the rental that the movie studios would want to charge for it.

For any particular crime, you can with in reason pin point it to the exact planet on which the criminal is on. I therefore suggest, we just charge everyone on that planet with the crime in the name of security. Or perhaps, just get a search warrant for the planet so we can identify the correct user. If we did this for every crime, it would solve everything.

Lawyer for a trademark issue for under $30,000?

It would likely cheaper to throw them away and make new ones that red with pink stripes.

Get a different job. They aren't hard to find. They don't want a good programmer, they want a code monkey. You'll be much happier in the long run (and the short run, and every run inbetween -- trust me).

It's not a problem. That is my bread and butter.

I fix all the stuff that a team of developers totally botched.