Windows 10 IS Windows 9. Microsoft engineers are even still calling it Windows 9. The source tree is the same, there have been no major changes.

What has happened is that Windows 9 has been getting very bad press and is still riddled with bugs. Instead of releasing a version number nobody will buy and would only have to patch almost immediately anyway, OR getting slagged off for Yet Another Delayed Release, Microsoft is renaming it version 10 and delaying the release until the bugs are sorted.

You will observe Microsoft has been talking up Windows 9 for some time, but now all talk (and apparently all memory) of it has ceased. Newspapers suffering amnesia is amost acceptable. Slashdotters??? WTF??? I'm sorry, but there is no-one in or around IT that has a single, solitary excuse.

There have been cases of Boeing 777s and modernized 737s developing unexplained system faults. Do not be so sure that RFI was not to blame. These have had much worse reliability than other Boeing models in recent years and as no other faults have been offered by Boeing as explanation, it is illogical to simply dismiss the one fault we know about as unrelated to the unusual number of abnormalities and crashes specific to these two models.

Obviously, Boeing has no interest in being honest about the problems they know about, be they software or hardware. Nor are they likely to Open Source anything, so there is no possibility of scrutiny by an independent party.

Simple logic (and self-preservation) says they have an unattributed defect capable of causing catastrophic failure, and a defect that can potentially cause catastrophic failure, therefore fixing the defect is essential.

The cost? The cost is insignificant. Boeing is hardly poor and is quite capable of covering the airlines' cost as this is a manufacturing defect. The airlines? They're making enough money that they can afford riots on board when seats are tilted. Besides, this is the cost of doing business. There's a price for bad decisions, all other sectors (except, apparently, banks) are expected to take the rough with the smooth. If several go bust because they chose unwisely, that's how life in business goes. You pay your money, you take your choice. Besides, they'd still be doing better than the German in Last Crusade.

If I went into business and made bad choices, would you be telling people to ignore my expenses? No? Good. If I'm not fit for purpose as a businessman, I've no business expecting support. So why should Ryanair, a notoriously incompetent company, deserve better? Because they're too big to fail? Not a good reason.

All you need are some windows that open in flight. Problem solved.

First, guns don't protect, never have, never will. That is not the function of a gun. So anyone on their high horse should look to see if they're suffering altitude sickness.

Second, the design of these specific rifles is a non-issue. The gun market is inherently grey, which means regulation is minimal to non-existent. There's no white hats in weaponry of any kind. And, yes, that includes the re-enactment stuff I work with. I know that, recognize that and accept it*. No shades, just a thick, pea-soup foggy grey.

*That is why I despise "goody two shoes" arguments from both extreme camps. This isn't black, this isn't white, this is murky grey. I own it for my part, I hold nobody to a higher standard than I hold myself, but I refuse to hold them to a lower one either. Own it.

Third, the design of any regular weapon is a non-issue, but nothing stops you from designing an irregular weapon. With modern cheap hardware, a 3D printer and suitable low-cost materials, a person is quite capable of designing a 3-5 mile range sniper rifle that can be controlled via telerobotics from the home. We already know that low-cost cruise missiles with ranges in excess of 100 miles can also be built at home. With 3D printing, the costs become lower. With advances in technology (remember, the $5000 100-mile cruise missile was designed over a decade ago and it wasn't even close to what budget efforts could do), you can expect far greater ranges, far greater precision and far greater payloads today.

This, again, goes back to this being grey hat technology. If a black hat wanted to use such devices, we'd know about. Or, rather, the survivors would. America still exists, so black hats either don't have the courage of their convictions or they don't have the skill. Either way, they're not worthy of consideration. Worthy of being dumped into a deep oceanic trench, bu not worthy of consideration.

White hats? If white hats were building actively guided systems capable of that sort of range, you'd be seeing miniature computer boards running Linux, Squid and Tor relays launched into stable orbits that crossed nations with restricted network access. We don't. We see "peace corps" infiltrators attempting to install such devices directly, along with who knows what malware, causing international incidents and seriously destabilizing international relations, as part of neocon stupidity. White hats putting in a passive alternative with no hostile software and no damage to other nations -- that's an OBVIOUS way to do good for everyone and to minimize harm. But, no, they either don't have the skill or the courage of their convictions.

So it's all grey. That's all there is. Thick, pea-soup fog.

If it's not an abacus, it can't count. Most of the rational people have quit fet due to database failures, update disasters, an incredibly primitive unthreaded discussion format and a contingent of highly abusive individuals. Abusiveness and primitiveness has done for tech forums, too, which is why Kuro5hin has been in death throes for some time.

A community is never stronger than the people who stand behind it and, in sadly far too many cases, the people standing behind the community are crouched down and in hiding.

That's the job of consultants.

Go back to Windows. If you can't cope with people wanting to choose how to run their system, you don't deserve software that's all about choice. If insults are the best reason you have for living, do everyone a favour.

Doesn't matter. It's not tested or validated for every possibility. Hell, given how easily I can break Debian, I wonder if it's tested at all these days. There is no point in using unvalidated setups with a distro, if you're at that point then you should roll your own.

Software that is designed correctly separates out what it does, how it does it, and how it interacts with the outside world.

Ergo, software that is correctly designed is user-agnostic. If the user thinks in a particular way, whatever that way happens to be, it is the job of the software to accommodate that. If it does not, it is not software for users, it is software that has users. Possession is everything.

Software that is correctly designed is configuration-agnostic. If the configuration file states something is enabled, then that is enabled. It is not the job of the software to say the file really means something else. If the configuration is broken, state how and why. Clearly. If the configuration is old, import and update. But don't tell me, or anyone else, what Joe Bloggs thinks would look better. I don't care. And the more other people's preferences get shoved in my face, the less I will care.

Theo clearly has the right idea - the only way to get past the morons is with an attitude of utter contempt. Bugger all else matters, apparently.

It's definitely one of the cleanest. It needs a rewrite, though.

Linux-From-Scratch is easier to use, less user-hostile and less determined to tell me how to think.

ANY software that pretends to know better than me how I want things done is software that deserves to burn. And then sink into the swamp. It is that precise attitude that got me to kick the Windows habit and led me away from the early ix86 BSDs.

I not only think better than a mere machine, I think better than your average distro compiler. I can spec better, I can build better, I can test better. Debian had, up till now, been acceptable, the packages are convenient and it's no great pain to tune. Now, Debian ranks lower than Fedora. I'd recommend the MCC distribution before either and that was last updated during the Ice Age.

Don't Blink.

I've used the site longer and reserve the right to use Doctor Who references where I'm suspicious of technical details, especially as relate to timing vulnerabilities. This is allowed, as per The Hacker's Dictionary. Bonus points for finding the Doctor Who references included.

That was pretty much my interpretation as well. Which would be great for ad-hoc encrypted tunnels - the source and destination can have keys that are valid only until the tunnel's authentication expires (typically hourly) and where the encryption is based on the identity the other side is known by. Ad-hoc tunnels need to generate keys quickly and efficiently, but also don't need to be super-secure. In fact, they can't be.

If RIBE isn't useful in ad-hoc, then you'd end up having to ask when it would be useful.

Anything that depends on a third party, including PGP/GPG with keyservers, is vulnerable to some form of compromise, SSL/TLS certificates all have a third party signer and Kerberos depends on all kinds of behind-the-scenes work being secure. However, although they're imperfect, they're considered adequate for what they do. Well, except for SSL, perhaps.

RIBE presumably therefore also has a niche where it's good. Rapid key turnover is what's wanted for conversation-based protocols with timeouts. That makes RIBE sound promissing for IPSec ad-hoc and SSL, as it makes store and crunch by attackers less likely to work. But is that the right niche?