Ctrl-C is what I prefer.

…which is particularly easy to press if you remap Caps Lock to be another Ctrl.

How does Windows work around this?

"Windows" doesn't, the device vendors do.

I don't know anything about this particular bug, but it's no big deal: no doubt fixable for any given device with a quick patch ("If the DMI string is this and the BIOS says this, then it's lying and pretend it said this").

In the PC world, all the vendors have huge teams of engineers doing exactly this: applying hacks to get Windows running correctly on each new device. Then they package those hacks in per-device "driver" packages. This is the reason why, for example, every motherboard has "Windows drivers" for download where in Linux it's all part of the same kernel.

But this bites linux in the face of BIOS bugs: it has to make the assumption that the BIOS follows the relevant standards. It's not that you can't do the work to make this happen on any given device, it's that no one is paying the teams of engineers to actually do the work for every device.

Indeed, this is the crux of Linux compatibility problems: people will break their firmware to ensure it runs smoothly on Windows (e.g., sending ASPM information in an unconventional, broken way) while Linux is left holding the bag. All Linux can do is assume that something works according to spec until proven otherwise, and then in cases like this, it's hard to "undo" it because some things will function correctly and other things won't.

In other words, vendors only test on Windows and only care about Windows compatibility. Compatibility with other systems is almost always accidental. This puts MS at an advantage and everyone else at a keen disadvantage.

More precisely, it's that the team of engineers which is generally willing to do the work for every device, for free, is prevented from doing so by not having access to the internals of the product.

I'm not sure I buy that. If they had real access to the internals, they'd be fixing the BIOS bugs instead of hacking around it in the driver suite.

I think the real problem runs deeper: the culture of "firmware" in the PC world is fundamentally broken. PC BIOS exists more to protect the business models of the important players than it does to abstract the hardware in a meaningful way. And frankly it's hurting those players (like the big one that starts with an "I") badly in the embedded space, where 12-second POST times, duplicated drivers, and thousands of lines of workaround code aren't considered acceptable.

We don't know right now. Someone is cheating.

Can someone help me understand -- is this a CEP for js?

Edit: or rather: CEP + message queue?

Its not a message queue, and although its only for js now, it meant to be for everybody. But i dont know what CEP is. Something similar?

By CEP I mean something like esper: http://esper.codehaus.org/tutorials/solution_patterns/soluti...

That is very cool, thanks for the link. And yes the main idea is very similar to that.

Right, because people from Marketing don't add any value to a company.

I never said they did not. I said I did not feel any sympathy. Though, I would argue, if they added so much value, why are they being cut?

Let's argue that point. I say the Marketing effort has been misguided and not a good return on value. Do you disagree? Or are you saying the blame lies elsewhere. Would you rest it solely with higher management? I certainly wouldn't shield them from blame, but I think the Marketing arm deserves some valid criticism. I point you to this from last year: http://searchengineland.com/yahoo-should-put-ad-budget-into-...