True, but then multiple guess exams (I will not dignify them with the word "choice") are void of any kind of merit. By simple examination of the answers, you can usually eliminate 50% or more of the options as absurd and guarantee a pass without knowing anything about the subject at all. On the other hand, the old-style O-Levels and A-Levels I took in the UK were predominantly essay-style questions, which the JMB examining board had no problem with marking. (The Joint Matriculation Board was one of a a number of examining boards in the UK, but was regarded as one of the better ones. It was still very easy to determine what the questions would be, you could eliminate most of the syllabus and focus on a handful of topics without much difficulty.)
Now, this is how you circumvent the privatization problem. You have the exams determined from first principles by a varying pool of academics who teach some given number of years over and above the year being examined, with the uppermost years (BSc, Masters, PhD) being developed by the researchers. The consequence of this is that the researchers will shift the questions as the research shifts - essentially randomly - and exploit the pressure to meet the requirements to create a random shift from year to year downwards, steered by the requirements needed to understand the subject as it is now but not controlled by that as the latencies involved mean teachers have to be guided by the subject, because they can't know either what direction research will go in or how this will stir the examination pot.
Because examinations are thus kept entirely within the educational system, it is not "privatized" in the normal sense. Because the examiner pool would vary, there would be less of a propensity for the system to stagnate into a trivial standard pool of questions, which is what most exam systems do. Because the system is iterative and contains feedback loops, you create a chaotic system where you can describe the system as a whole but not how it is going to change from one step to the next.
This does create a problem for employers, true, because this eliminates any form of standardized testing, which means you cannot compare any two years and cannot absolutely quantify what a specific grade means. But, frankly, as most employers don't give a damn what your grades were, it's not a very big problem. I honestly can't see them even noticing, for the most part. HR can't interpret the grade, even if it is standardized, and in-person interviewers will be far more interested in whether you can do their job, if they're any good. If they're not any good, understanding the subject will prove essential.
I agree with you, though this arguably only applies to "standardized" testing. If the exam questions are from a genuinely random selection (with an even distribution) of what can be expected of children of a given age, where the only prediction you can make about what will appear on the exam is that it will be some fixed percentage of the syllabus, then teachers can't be pressured into dropping stuff that is supposed to be taught. There's nothing senior/administrative types can do to determine what should be dropped. This, however, tends to lead towards a restriction on the syllabus itself, thus still eliminating what is important in favour of what is easy to test. This can be beaten if (and only if) the exam board is given incentives to test a wide range of topics and damn what any school district restricts itself to. If it doesn't teach what's needed, too bad.
Now, it's perfectly true that kids can only handle exams of a certain length of time, and you're limited by that. For a single exam. So set multi-part exams. It's also true that school districts will want to ensure their kids get good grades, because that's how they get paid. Which goes back to the whole relative increase in knowledge, rather than absolute scores. Then it doesn't matter how many questions the kid can't answer, it only matters how many questions they CAN answer in comparison to what they knew the previous time they were tested. (Continuous assessment, as tried in the UK, might work better than strict examination, but only if the student/teacher ratio is good enough for quality assessments and quality responses to a student's change in abilities. But if you lack the manpower, the assessments are no good and just serve to distract both teacher and students from the object of the exercise, which is to learn.)
Recognition for passing a standardized test that the good students know is worthless is worthless recognition. Recognition for something that actually requires understanding - ah, now that's something different. The Great Egg Race (as presented by Prof. Heinz Wolff) and the school version (The Granada Power Game), TV shows like "Now Get Out Of That", and open contests like the Micromouse Tournament - these achieved a lot for various branches of engineering and material science at the height of their respective popularity. Maths Olympics do something, but not a whole lot, and not that many schools anywhere field much of a team. Some of Keith Devlin's maths-related puzzles might help too, but you really do need something extraordinary in mathematics that allows people to earn what they regard as both well-deserved and "real" recognition, that can actually stand up to being compared to the top engineering efforts. (No, "battlebots" don't count for engineering, unless they're genuinely hand-crafted rather than COTS plug-and-power-play systems. The idea is to get people to think with their brains, not their wallets.)
Differential test scores. Rating/paying schools by an absolute score just means schools get students who know the end result. Rating/paying schools by how much they've improved, relative to how much you'd expect them to improve given where they were at the start of the year, would tell you how much you've actually taught them versus expectation. Expect the results to be very different.
Teach maths and science as interesting subjects. People can be enthused with these, but not if they're taught as if they're dead.
Stream the kids by subject. I'd suggest 5 or even 7 streams, to prevent over-broad grouping. Also, don't just use absolute rate of learning. If a kid works better with the support of a peer-group, and the peer-group is in a different stream than the one the kid would otherwise be in, put the kid in the other stream or see if there's a workable compromise. Age should not be a factor - if we go by typical UK figures (and the UK has a lousy system too), there should be a Ruth Lawrence-like figure in the US each year, minimum. You can probably assume a properly-tuned system could achieve 3-4 such people a year in a country of the size of the US, and multiply up the graduates from Masters or PhD programs by a comparable factor.
Improve student/teacher ratios. This doesn't necessarily mean over-small classes. A couple of assistant teachers improves the ratio without dividing up the class unnecessarily.
DO NOT teach to the exam, teach the subject. Teaching to the exam just tells you how good students are at tests, and any student who is any good doesn't give a damn about what the exam needs you to know, they want to know what the subject requires you to know. The exam is merely a device to let you progress further or get a better job. The crap students want you to teach to the exam, because it means they don't need to understand anything, they just need to be able to recite the day after they pull an all-night crammer.
Teach the subjects accurately and honestly. If a book is wrong or out-of-date on a topic, don't use that book for that topic. Kids can access the Internet and if they begin to suspect they're being fed bullshit on one thing, they'll regard everything you say as probably bullshit.
DO NOT insist that something is beyond question unless there are sound reasons for contending that it is, and (most importantly) you're willing to present those reasons to any student that asks. Arrogance and ignorance are the hallmarks of a poor lecturer. If you don't know, don't insist.
Students should WANT to spend as much time out of class doing their own research as they spend in class on that subject, above and beyond the time they spend on assignments. This places two additional requirements:
You need to tell them HOW to do research (including how to spot bogus claims and frauds) and suggest places to look
They need to be given a better reason than "because I say so" to do so - such as finding something that might be otherwise utterly trivial that is fascinating to them
This does not guarantee you'll actually get significantly better results, it merely guarantees that the more obvious bugs are fixed and that exceptional minds are not destroyed by tedium and an abusive environment. There are likely many other bugs that will prevent maximal gains.
How interoperable is it with popular e-mail security mechanisms? (SMTP-over-SSL/TLS, SMIME, POP-over-SSL/TLS, senderkey mechanisms, and so on)
What is the uptime rating? (3N's? 5N's? They didn't look?)
Are there known performance, security or reliability issues? (Don't expect an honest answer, it's the uncomfortableness you should be concerned with. Also, check beforehand the better security websites for known issues and ask about any that aren't closed, if they deny any issues exist.)
Is the system highly scalable and at least moderately future-proof? (You want something that can handle both SMP and multicore systems efficiently, is either well-maintained or highly modular, and supports protocols in use. IPv6 is used in Japan, so IPv6 is in use, whether or not the US or that University use it. If it's not being upgraded except "as absolutely needed", then your needs are unlikely to ever qualify. No matter how small the userbase of IPv6 may be, it's bigger than your University, so if their needs aren't interesting to the vendors, your needs aren't interesting to the vendors. Sorry, I don't care if you're paying. Once you've paid, they have your money. So long as they supply something, they have no further obligations - even to answer support phone calls. The EULA the University must agree to will likely also say they're not even obliged to provide software that works, or works even remotely as described. The onus is 100% on you to make sure that if you're given a hot potato you will be able to carry it. When it comes to software, you have no rights whatsoever.)
It is interesting, though, that it has been about a year since the current run on the stock markets and world finances began. (The current credit crunch, if you look at the graphs, is simply a continuation of a trend that began probably about April last year.)
Now, to use the oft-quoted "correlation does not prove causation", it would be totally absurd to say that the coincidence of dates proves the current problem is related to the cyber-attacks. Lots of things probably happened in April of last year. To pick one out, just for the sake of picking something, would be stupid. However, if I were in charge of IT security at the World Bank, I would be wanting to know if sensitive or classified information was continually exposed over that period that would permit someone to destabilize things.
It's almost certain that unencrypted sensitive information would be present on e-mail servers, which is stupid and naive, and members of the World Bank who don't make use of secure methods of communication for sensitive material should be made to walk the plank regardless of whether any harm was done. The IT managers who allowed unencrypted data to be present and who did not properly install suitable intrusion countermeasures should follow shortly thereafter. In the (extremely dubious and unlikely, but arguably possible) circumstance that the crisis is related to the infiltration, then the game changes from a mere fix-things-up and discipline-the-bastards scenario to a more severe lockdown-the-damn-network-now-defcon-1 type of situation.
The former simply means you need to apply suitable patches and/or servers, and maybe hire a pirate ship to escort the former employees to shark-infested waters. Since this is the most likely situation by far, that's all they need to do. But concealing it hasn't helped them apply the measures they needed, or the attacks could not have continued the moment it tripped the first intrusion detector. In this case, the secrecy has caused severe harm to the World Bank, but probably nobody else. Like I said, this is the most likely.
The worst-case is that we're seeing a positive feedback loop. Sensitive/classified information on volatile situations that could cause those situations to get considerably worse being posted, then lifted and used to do exactly that, causing people to post even more such information, and so on. Positive feedback loops are not simply a technological problem but an entire attitude problem and social engineering problem. That requires more than IT security, because IT security can't debug or firewall the brain. Yet. Such a loop might easily require a complete organizational shutdown, because no amount of patching will help. It needs a major attitude shift - not just on the part of internal employees but also on the part of all countries involved - and that takes time. If it's the mind that's the vulnerability -AND- it is causing massive devastation, the World Bank would have to shut down all operations completely. Otherwise, you can't guarantee killing the loop. The chances this would need to happen are extremely slim, but as I said, it is technically possible, and you can't afford to be piecemeal when it comes to such scenarios.
If it's so unlikely, why mention it at all? Because the timing -is- interesting (a crisis is uncommon, so two parallel financial crises should raise eyebrows), along with the fact they even see it is as a crisis is exceptionally interesting, the fact that their response has been one of paralysis (suggesting a non-trivial people problem, rather than an idiotic individual or an unpatched machine), and the fact that everyone else's management of their perceived problem isn't managing it in the least, is suggestive that (a) the wrong problems are being fixed, and (b) that there is a lot of pressure to avoid fixing - or even seeing - the right problems. Suggestive isn't proof, of course, which is why I'm more interested in whether they're even looking to see if this is a possibility.
The 387 will work with the 386, giving you the maths, but the kernel has maths co-processor emulation support if there's no actual co-processor provided.
Re:Did Bill Gates pay Shuttleworth to create Ubunt
on
Linux 2.6.27 Out
·
· Score: 2, Funny
If there were viruses on VMS (well, other than via DCL scripting in e-mail subject lines), I guess we'd be calling them SYS$kits.
Here's a thought. If they're using Gentoo on a 386SX, using the -git kernel tree, and having it auto-rebuild whenever there's a change, they'd never actually get far enough in recompiling to ever be able to boot up a new kernel.
Well, they can't release a 2.7, as SCO has already declared that that's the kernel that has the proprietary code in it. (Y'see, the Master, who cunningly disguised his alien identity by calling himself Darl, made an error in the time calculations and ended up traveling back in time too many years. Now's our chance to really screw up the space/time continuum.)
This is why you have a concept of "absolute" time (which is really time relative to an agreed frame of reference and increments at a uniform rate as seen by the current timeframe), a current deviation from "absolute" time (which allows you to factor in leap-seconds, relativistic effects, the rate of change of the Earth's rotation, and anything else you care to throw in) and a "timezone" (which is a fixed deviation from "absolute" time with respect to position relative to the agreed frame of reference).
Modern high-performance crystals can give you absolute time directly. If you've a variable-speed crystal with low-enough latency and high-enough precision on the variability, you can adjust it by the current deviation directly without having to store that as a distinct value. The fact that it's random, in either case, is irrelevant, as it's an externally-provided or externally-derived delta. The third value is equally random, given that you can relocate at any time and therefore cross timezones at random intervals, but since you can know position with almost absolute certainty (GPS), this delta is not only trivial to add in but also trivial to see how to add it in.
Leap-seconds are inconsequential when they're considered deltas to an absolute time except for real-time work where the precision of real-time required is comparable to the precision with which you can distribute the addition of the delta. (If you evenly distribute a second over a thousand deltas, your hard realtime will work to bounds of around a thousandth of a second. And so on, to whatever accuracy your realtime software requires and/or your hardware supports.)
Deltas also have the benefit that you can express time outside of the bounds of an epoch by windowing. This solves the problem of trying to time-shift before 1970, for example, and eliminates Y2K-type issues, because your timestamp doesn't need to have sufficient width for the full range of dates required, so long as there is a way to point to the correct set of deltas.
Yes, general reputation matters. Cambridge and Oxford in the UK have enormous street cred with employers and it doesn't matter if they're the best in a given field. They'll always be highly regarded. Specific reputation also matters. A university known to have brilliant students and produce top-notch wizards (Hogwarts?) in that specific field will also count highly with an employer. (It'll be the name HR becomes extremely familiar with, if HR is bothering to track such things - and top employers are more likely to than bottom-of-the-barrel types.)
However, once you're past HR, you've the real professionals to contend with. Again, they'll recognize the big-general names and the big-in-field names if they're any good, but they'll also recognize heavily-published places because it's the published stuff that helps the professionals stay on top. HR doesn't read (and in some places I've applied, I wonder if they could read), so major research universities (where students are likely to be more current but not necessarily more educated) won't rank as high with them, but it should matter a great deal more to those in the field.
Since higher degrees are where people get deep into the research, such places should matter more to those going for a masters or PhD than for the first degree, but it is one of those things that does matter to some extent no matter what degree you're going for. After all, the lecturers can't be any better than the information they have available to them, so non-research places may be 5-10 years behind the curve, as they'll rely on secondary sources of information (such as textbooks) which will never be truly current. On the other hand, if you're the one defining what is current, you aught to be able to teach what is current.
Even then, it's no good if the lecturer knows something if they can't present it, so you still need more feedback. Ideally, you'd know how students change in ability from the start to the end of the course. Even a dumb lecturer can teach students who already know everything they need. The rate of change of student ability in relation to the rate of change expected for students of that level of ability is the magic number you need to tell if the lecturers are any good at imparting information. As I don't think any existing tests currently give you the data you'd need to calculate this, since you can't standardize a test that is inherently much more specialized and all tests these days are standardized (fools that they are), such a calculation is impossible. As such, you either have to pick a number out of thin air, or use surveys to get the students to pick the number out of thin air for you. Either way, it's not the number you want and there's no possible way of determining how close or distant it is. Without that number, the rest of the data only tells you what the upper theoretical limit is on quality, not what the practical day-to-day reality is.
That had a decent sound-track that could exploit devices like the Roland LAPC-1. X-Wing could use the LAPC-1 for the music and a Soundblaster for the sound effects, which was even more impressive. But, yes, those did use sound in a very basic way.
Frontier: First Encounters was a bit more sophisticated - the music was selected by type of scenario (so if you suddenly got into a fight, the music would change accordingly, likewise with docking at a space-station, and so on). In fact, First Encounters (although buggy) was one of the best games I've ever encountered (pardon the pun) when it comes to the use of incidental music, because it was genuinely incidental music in the way a TV program was, not chamber music/muzak. (Even Elite, the game that spawned Frontier and First Encounters, had a limited concept of incidental music, playing The Blue Danube when you had an active docking computer provided you were not attacking or being attacked at the time.)
I think that it is to games like First Encounters that game writers should look - genuine incidental music that fits the scene, rather than a one-type-fits-all. This would meet the objection as described. Music that, although not interactive, fitted the interaction. (Though god only knows what you'd use as music for MMORG cybering.)
The third possibility is that it's the presidential candidate (or president) the Top Brass least want to see in power. Don't look so shocked - Harold Wilson's government in the UK is believed to have been brought down by MI5 by quite serious and sane people.
Really, I don't see any advantage in a cyberwarfare division - at least, not any advantage that competing divisions within the military (eg: SPAWAR) and the intelligence community (SIGINT) can't already provide in terms of defense from attack and means to counterattack. If you thought Slashdot was bad for dupes, the US Government is riddled with them. (Those who complain about the cost of big government miss an important point - big government isn't necessarily any worse than anything else. It's duplicated government that's expensive, as it competes within itself and sabotages itself.)
As for the defense side of things, if the US Government insisted that communications be strongly encrypted and that private communications be strongly authenticated and tamper-resistant, the number of vectors that could be used for an attack would be massively reduced. If they also insisted critical infrastructure use those same constraints and in addition use protocols that are non-repudiatable and use multi-path connections with congestion control over both reliable and unreliable protocols, almost nothing of any significance could be impacted.
Instead, we run into this paradox. The need to maximize profits inhibits the desire to maximize commercial security (which includes virtually all critical infrastructure). The need to maximize the ability to monitor communications inhibits the desire to maximize personal security. Everything is a trade-off. We have chosen as a society to ditch security in favour of market forces. (Attacks are rare and aren't here-and-now. Shareholders are common and in the present. Guess who gets the money spent on them.) As is clear from the telecoms immunity, since no major party is going to support something unpopular with the electorate in an election year, the vast majority of Americans are also in favour of ditching personal privacy (which essentially means personal security) in favour of monitoring that the Government has stated cannot be defined or justified even to its own security courts.
In other words, the entire need for a "Cyber Command" (sounds like something they'd have on Mondas) is due to inept decisions that weaken what could passively protect itself by design into something that needs an active, offensive component to provide the same core functionality at added expense. You want to talk Government waste? Secure systems are cheap and essentially a one-off cost, entire military branches are expensive and require continuous feeding with the green stuff.
What's wrong with writing an emulation module that shifts results by the same amount the bugs would have done? The Excel exporter then simply identifies which emulation module is required for a given set of bugs in that version of Excel. The main ODF files then don't need to have bugs added, the stuff that includes workarounds will continue to work, and those who want to clean their spreadsheet of cruft can gradually switch off the bug emulation.
Microsoft could use the same module in Excel to the same effect. They wouldn't need to keep bugs in, they simply need to have the bugs emulated for existing work unless otherwise deselected. Microsoft then becomes the "hero" for solving both cross-compatibility and backwards-compatibility issues and simultaneously reducing the cost of new work (because new work would need fewer workarounds). This would also be a win for ODF, as they then don't need to introduce bugs, they only need to introduce a modifier layer, which could be used to provide all kinds of other benefits and interesting effects. There, everyone wins.
If future systems are going to use 64-bit or 128-bit timestamps that measure down to nanosecond precision, it makes no sense to use a different level of precision that isn't a direct power of two from the system timestamp. Either simplify by not changing the precision or simplify by making things a straight bit shift. Ideally, given resources are cheap, don't simplify at all.
We laugh at AlexH for thinking that because a bug existed in a calculation, it should be specified and mandated that all future calculations contain the same bug, in case people corrected for it?
Or perhaps at Microsoft for creating non-existent dates.
Or at ISO for creating one of the worst backlashes against a standard I think I have ever seen through their inept handling of the crisis and their blatant disregard for their own procedures.
Or at ODF's board for their suicidal willingness to allow the makers of a competing standard dictate their own direction. (Even if ODF survives - and no guarantee of that - AlexH has already made it clear that the bugs present in OOXML are being deliberately introduced into ODF for "backwards-compatibility" reasons. If ODF becomes a re-implementation of OOXML, who is going to use ODF?)
...if (and only if) those principled individuals set up a rival standards organization, have as part of their charter that they refute corruption and automatically negate standards tainted by corruption, re-certify where legal all known-to-be-"safe" standards under their own name, and then lobby research shops and companies hurt by the ISO scandal to work with them. Fork the certification market, but because of rebranding existing standards, no other standards body would ever need to be involved.
Another alternative - standards bodies rely on the income from charging absurd fees for standards, relying more on secrecy than anything. If you pay enough for a standard, you won't just give it away, in theory. If some suitably rich investor with lots of contacts and enough cunning bought up copies of those standards and then just dumped them onto public sites, it could cripple standards organizations for a long time. If it was clearly linked to the ISO debacle, ISO might not be too keen to be seen to complain - most countries deem bribery (even outside of government) a more serious offense than a petty trade secret violation and the press are more into scandals (which ISO is undoubtedly riddled with) than knuckle-rapping.
Slashdot Mirror
True, but then multiple guess exams (I will not dignify them with the word "choice") are void of any kind of merit. By simple examination of the answers, you can usually eliminate 50% or more of the options as absurd and guarantee a pass without knowing anything about the subject at all. On the other hand, the old-style O-Levels and A-Levels I took in the UK were predominantly essay-style questions, which the JMB examining board had no problem with marking. (The Joint Matriculation Board was one of a a number of examining boards in the UK, but was regarded as one of the better ones. It was still very easy to determine what the questions would be, you could eliminate most of the syllabus and focus on a handful of topics without much difficulty.)
Now, this is how you circumvent the privatization problem. You have the exams determined from first principles by a varying pool of academics who teach some given number of years over and above the year being examined, with the uppermost years (BSc, Masters, PhD) being developed by the researchers. The consequence of this is that the researchers will shift the questions as the research shifts - essentially randomly - and exploit the pressure to meet the requirements to create a random shift from year to year downwards, steered by the requirements needed to understand the subject as it is now but not controlled by that as the latencies involved mean teachers have to be guided by the subject, because they can't know either what direction research will go in or how this will stir the examination pot.
Because examinations are thus kept entirely within the educational system, it is not "privatized" in the normal sense. Because the examiner pool would vary, there would be less of a propensity for the system to stagnate into a trivial standard pool of questions, which is what most exam systems do. Because the system is iterative and contains feedback loops, you create a chaotic system where you can describe the system as a whole but not how it is going to change from one step to the next.
This does create a problem for employers, true, because this eliminates any form of standardized testing, which means you cannot compare any two years and cannot absolutely quantify what a specific grade means. But, frankly, as most employers don't give a damn what your grades were, it's not a very big problem. I honestly can't see them even noticing, for the most part. HR can't interpret the grade, even if it is standardized, and in-person interviewers will be far more interested in whether you can do their job, if they're any good. If they're not any good, understanding the subject will prove essential.
I agree with you, though this arguably only applies to "standardized" testing. If the exam questions are from a genuinely random selection (with an even distribution) of what can be expected of children of a given age, where the only prediction you can make about what will appear on the exam is that it will be some fixed percentage of the syllabus, then teachers can't be pressured into dropping stuff that is supposed to be taught. There's nothing senior/administrative types can do to determine what should be dropped. This, however, tends to lead towards a restriction on the syllabus itself, thus still eliminating what is important in favour of what is easy to test. This can be beaten if (and only if) the exam board is given incentives to test a wide range of topics and damn what any school district restricts itself to. If it doesn't teach what's needed, too bad.
Now, it's perfectly true that kids can only handle exams of a certain length of time, and you're limited by that. For a single exam. So set multi-part exams. It's also true that school districts will want to ensure their kids get good grades, because that's how they get paid. Which goes back to the whole relative increase in knowledge, rather than absolute scores. Then it doesn't matter how many questions the kid can't answer, it only matters how many questions they CAN answer in comparison to what they knew the previous time they were tested. (Continuous assessment, as tried in the UK, might work better than strict examination, but only if the student/teacher ratio is good enough for quality assessments and quality responses to a student's change in abilities. But if you lack the manpower, the assessments are no good and just serve to distract both teacher and students from the object of the exercise, which is to learn.)
Recognition for passing a standardized test that the good students know is worthless is worthless recognition. Recognition for something that actually requires understanding - ah, now that's something different. The Great Egg Race (as presented by Prof. Heinz Wolff) and the school version (The Granada Power Game), TV shows like "Now Get Out Of That", and open contests like the Micromouse Tournament - these achieved a lot for various branches of engineering and material science at the height of their respective popularity. Maths Olympics do something, but not a whole lot, and not that many schools anywhere field much of a team. Some of Keith Devlin's maths-related puzzles might help too, but you really do need something extraordinary in mathematics that allows people to earn what they regard as both well-deserved and "real" recognition, that can actually stand up to being compared to the top engineering efforts. (No, "battlebots" don't count for engineering, unless they're genuinely hand-crafted rather than COTS plug-and-power-play systems. The idea is to get people to think with their brains, not their wallets.)
This does not guarantee you'll actually get significantly better results, it merely guarantees that the more obvious bugs are fixed and that exceptional minds are not destroyed by tedium and an abusive environment. There are likely many other bugs that will prevent maximal gains.
...I could give you some really good disreputable ones. In the meantime, researchers have found a home owner unaffected by the crisis so far.
It is interesting, though, that it has been about a year since the current run on the stock markets and world finances began. (The current credit crunch, if you look at the graphs, is simply a continuation of a trend that began probably about April last year.)
Now, to use the oft-quoted "correlation does not prove causation", it would be totally absurd to say that the coincidence of dates proves the current problem is related to the cyber-attacks. Lots of things probably happened in April of last year. To pick one out, just for the sake of picking something, would be stupid. However, if I were in charge of IT security at the World Bank, I would be wanting to know if sensitive or classified information was continually exposed over that period that would permit someone to destabilize things.
It's almost certain that unencrypted sensitive information would be present on e-mail servers, which is stupid and naive, and members of the World Bank who don't make use of secure methods of communication for sensitive material should be made to walk the plank regardless of whether any harm was done. The IT managers who allowed unencrypted data to be present and who did not properly install suitable intrusion countermeasures should follow shortly thereafter. In the (extremely dubious and unlikely, but arguably possible) circumstance that the crisis is related to the infiltration, then the game changes from a mere fix-things-up and discipline-the-bastards scenario to a more severe lockdown-the-damn-network-now-defcon-1 type of situation.
The former simply means you need to apply suitable patches and/or servers, and maybe hire a pirate ship to escort the former employees to shark-infested waters. Since this is the most likely situation by far, that's all they need to do. But concealing it hasn't helped them apply the measures they needed, or the attacks could not have continued the moment it tripped the first intrusion detector. In this case, the secrecy has caused severe harm to the World Bank, but probably nobody else. Like I said, this is the most likely.
The worst-case is that we're seeing a positive feedback loop. Sensitive/classified information on volatile situations that could cause those situations to get considerably worse being posted, then lifted and used to do exactly that, causing people to post even more such information, and so on. Positive feedback loops are not simply a technological problem but an entire attitude problem and social engineering problem. That requires more than IT security, because IT security can't debug or firewall the brain. Yet. Such a loop might easily require a complete organizational shutdown, because no amount of patching will help. It needs a major attitude shift - not just on the part of internal employees but also on the part of all countries involved - and that takes time. If it's the mind that's the vulnerability -AND- it is causing massive devastation, the World Bank would have to shut down all operations completely. Otherwise, you can't guarantee killing the loop. The chances this would need to happen are extremely slim, but as I said, it is technically possible, and you can't afford to be piecemeal when it comes to such scenarios.
If it's so unlikely, why mention it at all? Because the timing -is- interesting (a crisis is uncommon, so two parallel financial crises should raise eyebrows), along with the fact they even see it is as a crisis is exceptionally interesting, the fact that their response has been one of paralysis (suggesting a non-trivial people problem, rather than an idiotic individual or an unpatched machine), and the fact that everyone else's management of their perceived problem isn't managing it in the least, is suggestive that (a) the wrong problems are being fixed, and (b) that there is a lot of pressure to avoid fixing - or even seeing - the right problems. Suggestive isn't proof, of course, which is why I'm more interested in whether they're even looking to see if this is a possibility.
The 387 will work with the 386, giving you the maths, but the kernel has maths co-processor emulation support if there's no actual co-processor provided.
If there were viruses on VMS (well, other than via DCL scripting in e-mail subject lines), I guess we'd be calling them SYS$kits.
Your office is in the Panoptican Library on Gallifrey? Wasn't that destroyed in the Time War?
Here's a thought. If they're using Gentoo on a 386SX, using the -git kernel tree, and having it auto-rebuild whenever there's a change, they'd never actually get far enough in recompiling to ever be able to boot up a new kernel.
Well, they can't release a 2.7, as SCO has already declared that that's the kernel that has the proprietary code in it. (Y'see, the Master, who cunningly disguised his alien identity by calling himself Darl, made an error in the time calculations and ended up traveling back in time too many years. Now's our chance to really screw up the space/time continuum.)
One of these days, the admins should give Anonymous Coward some mod points.
I think it's more a reference to Greek tragedies.
Or got the units wrong. The measurements weren't taken by the guys who designed all those Mars landers that crashed, were they?
This is why you have a concept of "absolute" time (which is really time relative to an agreed frame of reference and increments at a uniform rate as seen by the current timeframe), a current deviation from "absolute" time (which allows you to factor in leap-seconds, relativistic effects, the rate of change of the Earth's rotation, and anything else you care to throw in) and a "timezone" (which is a fixed deviation from "absolute" time with respect to position relative to the agreed frame of reference).
Modern high-performance crystals can give you absolute time directly. If you've a variable-speed crystal with low-enough latency and high-enough precision on the variability, you can adjust it by the current deviation directly without having to store that as a distinct value. The fact that it's random, in either case, is irrelevant, as it's an externally-provided or externally-derived delta. The third value is equally random, given that you can relocate at any time and therefore cross timezones at random intervals, but since you can know position with almost absolute certainty (GPS), this delta is not only trivial to add in but also trivial to see how to add it in.
Leap-seconds are inconsequential when they're considered deltas to an absolute time except for real-time work where the precision of real-time required is comparable to the precision with which you can distribute the addition of the delta. (If you evenly distribute a second over a thousand deltas, your hard realtime will work to bounds of around a thousandth of a second. And so on, to whatever accuracy your realtime software requires and/or your hardware supports.)
Deltas also have the benefit that you can express time outside of the bounds of an epoch by windowing. This solves the problem of trying to time-shift before 1970, for example, and eliminates Y2K-type issues, because your timestamp doesn't need to have sufficient width for the full range of dates required, so long as there is a way to point to the correct set of deltas.
Yes, general reputation matters. Cambridge and Oxford in the UK have enormous street cred with employers and it doesn't matter if they're the best in a given field. They'll always be highly regarded. Specific reputation also matters. A university known to have brilliant students and produce top-notch wizards (Hogwarts?) in that specific field will also count highly with an employer. (It'll be the name HR becomes extremely familiar with, if HR is bothering to track such things - and top employers are more likely to than bottom-of-the-barrel types.)
However, once you're past HR, you've the real professionals to contend with. Again, they'll recognize the big-general names and the big-in-field names if they're any good, but they'll also recognize heavily-published places because it's the published stuff that helps the professionals stay on top. HR doesn't read (and in some places I've applied, I wonder if they could read), so major research universities (where students are likely to be more current but not necessarily more educated) won't rank as high with them, but it should matter a great deal more to those in the field.
Since higher degrees are where people get deep into the research, such places should matter more to those going for a masters or PhD than for the first degree, but it is one of those things that does matter to some extent no matter what degree you're going for. After all, the lecturers can't be any better than the information they have available to them, so non-research places may be 5-10 years behind the curve, as they'll rely on secondary sources of information (such as textbooks) which will never be truly current. On the other hand, if you're the one defining what is current, you aught to be able to teach what is current.
Even then, it's no good if the lecturer knows something if they can't present it, so you still need more feedback. Ideally, you'd know how students change in ability from the start to the end of the course. Even a dumb lecturer can teach students who already know everything they need. The rate of change of student ability in relation to the rate of change expected for students of that level of ability is the magic number you need to tell if the lecturers are any good at imparting information. As I don't think any existing tests currently give you the data you'd need to calculate this, since you can't standardize a test that is inherently much more specialized and all tests these days are standardized (fools that they are), such a calculation is impossible. As such, you either have to pick a number out of thin air, or use surveys to get the students to pick the number out of thin air for you. Either way, it's not the number you want and there's no possible way of determining how close or distant it is. Without that number, the rest of the data only tells you what the upper theoretical limit is on quality, not what the practical day-to-day reality is.
The 80x86 is an antique architecture. The sun is an antique fusion reactor. Hydrogen is an antique element. Hey, this is a cool game!
That had a decent sound-track that could exploit devices like the Roland LAPC-1. X-Wing could use the LAPC-1 for the music and a Soundblaster for the sound effects, which was even more impressive. But, yes, those did use sound in a very basic way.
Frontier: First Encounters was a bit more sophisticated - the music was selected by type of scenario (so if you suddenly got into a fight, the music would change accordingly, likewise with docking at a space-station, and so on). In fact, First Encounters (although buggy) was one of the best games I've ever encountered (pardon the pun) when it comes to the use of incidental music, because it was genuinely incidental music in the way a TV program was, not chamber music/muzak. (Even Elite, the game that spawned Frontier and First Encounters, had a limited concept of incidental music, playing The Blue Danube when you had an active docking computer provided you were not attacking or being attacked at the time.)
I think that it is to games like First Encounters that game writers should look - genuine incidental music that fits the scene, rather than a one-type-fits-all. This would meet the objection as described. Music that, although not interactive, fitted the interaction. (Though god only knows what you'd use as music for MMORG cybering.)
The third possibility is that it's the presidential candidate (or president) the Top Brass least want to see in power. Don't look so shocked - Harold Wilson's government in the UK is believed to have been brought down by MI5 by quite serious and sane people.
Really, I don't see any advantage in a cyberwarfare division - at least, not any advantage that competing divisions within the military (eg: SPAWAR) and the intelligence community (SIGINT) can't already provide in terms of defense from attack and means to counterattack. If you thought Slashdot was bad for dupes, the US Government is riddled with them. (Those who complain about the cost of big government miss an important point - big government isn't necessarily any worse than anything else. It's duplicated government that's expensive, as it competes within itself and sabotages itself.)
As for the defense side of things, if the US Government insisted that communications be strongly encrypted and that private communications be strongly authenticated and tamper-resistant, the number of vectors that could be used for an attack would be massively reduced. If they also insisted critical infrastructure use those same constraints and in addition use protocols that are non-repudiatable and use multi-path connections with congestion control over both reliable and unreliable protocols, almost nothing of any significance could be impacted.
Instead, we run into this paradox. The need to maximize profits inhibits the desire to maximize commercial security (which includes virtually all critical infrastructure). The need to maximize the ability to monitor communications inhibits the desire to maximize personal security. Everything is a trade-off. We have chosen as a society to ditch security in favour of market forces. (Attacks are rare and aren't here-and-now. Shareholders are common and in the present. Guess who gets the money spent on them.) As is clear from the telecoms immunity, since no major party is going to support something unpopular with the electorate in an election year, the vast majority of Americans are also in favour of ditching personal privacy (which essentially means personal security) in favour of monitoring that the Government has stated cannot be defined or justified even to its own security courts.
In other words, the entire need for a "Cyber Command" (sounds like something they'd have on Mondas) is due to inept decisions that weaken what could passively protect itself by design into something that needs an active, offensive component to provide the same core functionality at added expense. You want to talk Government waste? Secure systems are cheap and essentially a one-off cost, entire military branches are expensive and require continuous feeding with the green stuff.
Why is the parent post modded funny?
What's wrong with writing an emulation module that shifts results by the same amount the bugs would have done? The Excel exporter then simply identifies which emulation module is required for a given set of bugs in that version of Excel. The main ODF files then don't need to have bugs added, the stuff that includes workarounds will continue to work, and those who want to clean their spreadsheet of cruft can gradually switch off the bug emulation.
Microsoft could use the same module in Excel to the same effect. They wouldn't need to keep bugs in, they simply need to have the bugs emulated for existing work unless otherwise deselected. Microsoft then becomes the "hero" for solving both cross-compatibility and backwards-compatibility issues and simultaneously reducing the cost of new work (because new work would need fewer workarounds). This would also be a win for ODF, as they then don't need to introduce bugs, they only need to introduce a modifier layer, which could be used to provide all kinds of other benefits and interesting effects. There, everyone wins.
If future systems are going to use 64-bit or 128-bit timestamps that measure down to nanosecond precision, it makes no sense to use a different level of precision that isn't a direct power of two from the system timestamp. Either simplify by not changing the precision or simplify by making things a straight bit shift. Ideally, given resources are cheap, don't simplify at all.
We laugh at AlexH for thinking that because a bug existed in a calculation, it should be specified and mandated that all future calculations contain the same bug, in case people corrected for it?
Or perhaps at Microsoft for creating non-existent dates.
Or at ISO for creating one of the worst backlashes against a standard I think I have ever seen through their inept handling of the crisis and their blatant disregard for their own procedures.
Or at ODF's board for their suicidal willingness to allow the makers of a competing standard dictate their own direction. (Even if ODF survives - and no guarantee of that - AlexH has already made it clear that the bugs present in OOXML are being deliberately introduced into ODF for "backwards-compatibility" reasons. If ODF becomes a re-implementation of OOXML, who is going to use ODF?)
...if (and only if) those principled individuals set up a rival standards organization, have as part of their charter that they refute corruption and automatically negate standards tainted by corruption, re-certify where legal all known-to-be-"safe" standards under their own name, and then lobby research shops and companies hurt by the ISO scandal to work with them. Fork the certification market, but because of rebranding existing standards, no other standards body would ever need to be involved.
Another alternative - standards bodies rely on the income from charging absurd fees for standards, relying more on secrecy than anything. If you pay enough for a standard, you won't just give it away, in theory. If some suitably rich investor with lots of contacts and enough cunning bought up copies of those standards and then just dumped them onto public sites, it could cripple standards organizations for a long time. If it was clearly linked to the ISO debacle, ISO might not be too keen to be seen to complain - most countries deem bribery (even outside of government) a more serious offense than a petty trade secret violation and the press are more into scandals (which ISO is undoubtedly riddled with) than knuckle-rapping.