Does this include cussing at a machine that's physically obnoxious to work on, such as one with a case full of razor-sharp edges or one that you have to pull all the cables off the motherboard to change out a card because it's so cramped inside, or with screws in fiddly places that you inevitably end up dropping in such a way that they roll into a tight corner and are nearly impossible to get out? No, I'm not bitter. :-)

"Slashdot Overload" mode was the bane of v1. It was possible for some comments to become unreachable unless you read in linear mode.

That said, I'd prefer to just be given the entire thread^H^H^H^H^H^Hmountain of comments as one page by default, with no paging, clicking, or load-as-you-scroll shenanigans!

Come on people, give it a REST.

... and zero good taste, apparently.

No, it's a forensics bundle from Microsoft. No wonder it tastes awful. ;-)

Maybe so, but dopamine is a very behaviorally significant neurotransmitter, and the gene in question codes for a dopamine receptor, and that specific receptor has well-studied effects on behavior (especially addictive behavior). The link to politics is a new one on me, though.

A broken dopamine receptor D4 (the same gene TFA is about) also has a tendency to make one less vulnerable to addiction, and especially stress addiction. For an example of how painful a runaway stress-driven dopamine feedback loop can be, look no further than Dilbert.

Is there something wrong with the Common User Access interaction design lineage? The very reason the document was put forth was to reduce user ramp-up time learning a new product.

WordPerfect 5, on the other hand, was a shining example of how to confuse the hell out of a new user by not working remotely like anything else out there. I used it for quite some time, even wrote applications with its macro language, and still couldn't get by without a key binding cheat sheet.

On the other hand, you could actually see begin and end tags in its Reveal Codes mode, and if you were willing to sink enough time and brain cells into it, it was wickedly powerful, so in some sense they were optimizing for the dedicated power user at the expense of the casual user.

On the gripping hand, for ~500 USD in the early nineties, perhaps they were right to expect a highly dedicated power user.

1. A write-once serial number doesn't prevent an attacker obtaining a brand-new, unimprinted device and cloning the original device's serial number. Which cells in a block of flash memory fail early is entirely a physical process issue, and falls in the general category of intrinsic physical unclonable functions. The entire point of a PUF is that it can't be duplicated; by definition, this means it can't be backed up, customized, or otherwise controlled, only observed.

2. Using flash memory's bad bits as an IPUF is indeed tied to the technology, but there's no reason an explicit PUF can't be integrated into device design. Yes, these will vary as hardware technology changes, but the principle remains. A hard disk analogue might be reading the remapped-blocks table in a drive's integrated controller -- bad sectors are not generally in the same place on multiple units of the same model.

3. Elimination of process variation can limit the number of bits of uniqueness obtainable from an IPUF but doesn't invalidate the basic technique. More blocks may have to be tested to destruction to uncover enough bits of physical variation, but part density is also increasing so perhaps this is a push. Disk manufacturing leading to correlated failures in disk arrays is indeed an unfortunate side effect of elimination of process variation, though; I guess we're well past the point of needing to configure arrays to tolerate double failure AND include a hot spare if they have to stay up 24/7.

Oh my. So what you're saying is you actually had to cope with colonel panics?

*rimshot*

Hilarity aside, such messy environments are entirely unsurprising for a medical environment. Commercial medical-records software is even worse -- in my experience, much of it dies horribly unless every user has admin rights on every PC they use, typically because the software likes to poo random files everywhere, including such places as the root of the system drive.

Fun with aspect ratios, in'nit?

Putting the taskbar on the left, as wide as the buttons normally are on the bottom, means you can actually see what the heck you've got going on when you have 20+ things open at a time. In that environment, though, what drives me bonkers are modal dialogs and message boxes that exclude themselves from the taskbar while leaving their owning window disabled, so you have to dig through the whole stupid Z-stack on every monitor to find what you did with it. Even worse, sometimes it winds up underneath a disabled window from the same app. (This isn't supposed to happen if the owner window is set correctly, but it still happens.)

Disclaimer: Three 4:3 monitors are required to make sense of that much going on!

I think the OP is talking about total cost of ownership here, not purchase price. TCO is all-inclusive, covering network bandwidth to make use of the space, backup and redundancy, paying someone to keep it running, electricity to keep it spinning, a share of a fileserver box to put it in, etc, etc.

Font hinting is not proprietary to Amazon. It's been a feature of TrueType for a long time, and the patent on bytecode-based hinting just expired in May, so maybe that's why they've enabled it.

|3\_/T d() j00 sp34k l33t?