*Viruses* had nothing to do with the Irish potato famine. While there were many factors for the famine, many of them political, the pathological reason was the *fungus* Phytophthora infestans.
*ahem* [shallow and pedantic] *Fungi* had nothing to do with the Irish potato famine. Phytophtora is an oomycete, not a fungus. [/shallow and pedantic]
This is really just an economic decision. An IT monoculture brings with it certain benefits, such as decreased (non-virus-related) support costs, but also certain costs, such as increased vulnerability to viruses. It's worthwhile for people like Dan Greer to make the IT world aware of those costs, but even once they become aware, a Windows monoculture may still be preferable for some. OTOH, I don't think the costs of supporting multiple OSs are as high as most people think. We have a small network with OS X Server serving files and doing authentication for a bunch of Windows and Linux boxes, and it's really not very hard to keep running.
Breeder reactors are here now only in the sense that there are a few operational reactors--there is not widescale deployment of breeder reactor technology as there is with conventional fission reactors (I was careful to say "widescale" in my original post). Obviously, present-day breeder reactor technology is more advanced than fusion reactor tech, but, on the other hand, the payoff isn't nearly as good as we can hope for with fusion.
Another poster mentioned that the 50-year timeline for running out of uranium was without breeder reactors, with current usage patterns. I was aware that this estimate assumed no breeder reactors, but I thought it was based on a much higher rate of consumption than we have now. Can anyone confirm / deny?
I might approve an order of magnitude increase in funding at that point, but I see no reason to do so now when there are technologies, particularly fission, wind, and solar power that are becoming viable.
Solar and wind are approaching economic viability as supplemental energy sources. What I mean is that they are good at helping meet some of the peak demand, but not so good as a baseline power source, since the wind doesn't always blow and the sun doesn't always shine. In particular, there are several hours each day in which the entire continent is in the dark (wind is also usually calmer at night). In an all-solar/wind energy scenario, you'd also need an enormous amount of energy storage capability (electricity-to-hydrogen-to-electricity in fuel cells, maybe), which rather dramatically raises your costs.
Nuclear, on the other hand, can be more expensive, but are well suited to be run continuously (aside from maintenance every once in a while), since fuel is only a small fraction of the cost per kW/h.
What makes more sense than either source on its own is nuclear fission to supply most of the baseline power, with solar and wind to supply peak power, and a few natural-gas fired plants for backup. Conveniently, peak generating times of solar and wind tend to correspond to peak demand.
Given the above scenario, you'll run out of fuel for your fission reactors in half a century or so (give or take a few decades), unless you start using breeder reactors, which aren't really a widescale-proven technology, and pose some nuclear proliferation issues. If you're going to pour research money into breeder reactors, why not spend it instead on fusion, which is pretty much the ultimate terrestrial power source?
Now I'm the first to admit that his public words and actions suggest someone much closer to (or even slightly below) the mean.
Public speaking is not his forte, apparently. I loved that SNL skit with him and Gore, where Bush's one-word summary of his campaign was "strategery". Did he actually ever say that word, or was that a bit of creativity on the part of the SNL writers?
In fact, Bush's seeming disdain for "book smarts" typifies the result we both expect from a poor educational system (though I'm under no illusion that he ever attended public school).
I think the disdain for "book smarts" comes more from political orientation than anything else. It's more a mistrust of the climate of arrogant intellectualism that pervades the pseudo-meritocratic structure of academia. As an academic myself (a physics grad student, to be specific), I can certainly see the off-putting aspects of the academic world. It's very easy to assume that, because you're the world's foremost expert on X, you understand Y and Z as well (I know I've done this). When that turns into preaching politics from the pulpit of science, it really pisses people off.
An example: recently, a bunch of physicists signed a letter about the dangers of using nuclear weapons against Iran. That kind of thing just makes me cringe. I know some of the physicists who signed the letter, and they've given me no reason to believe they're any more informed or knowledgeable about the Iran situation that your average college graduate. The spin on the letter was something along the lines of the claim, "we invented the nuclear weapon, and we're here to tell you just how bad it would be to use it", but it's not as if any of the signing physicists actually provided any new information. Military analysts, Middle Eastern policy studies professors, Iranian exiles--all of them might have had something valuable to say--but physicists? Hardly. All that kind of stunt does is make the Republicans even more pissed off at academia.
A major backlash from the right is coming--primarily aimed perceived left-wing bias in the social sciences, etc.--but there will be collateral damage to the natural sciences as well. The more reasons we give Republicans to lump the natural sciences in with the "left-wing" crowd, the worse things will be. Read Phi Beta Cons, a higher education blog run by the conservative magazine National Review, to see what I'm talking about.
Not the Einsteins of the world but the George Bushes.
eddeye, I'm surprised you had the courage to say that--even as a joke--on slashdot. From what I've read, GWB's SAT scores suggest he's of average intelligence for a president. Apparently, his SATs correspond to an IQ test score of around 130 (I forgot the exact numbers), which is approximately two sigma above the mean. In other words, your comment is entirely accurate, but nonetheless a serious lightning rod for politically-charged flames.
I agree with your main point; the smartest students will find something to occupy their minds. The moderately intelligent students are left without much, however, and it hurts them when they get to university and are challenged for the first time. As there are far more people in this category than there are geniuses, this is a major loss for society.
How any russians know german and korean and spanish?
No idea, but I know a Norwegian who knows Norwegian, Swedish, German, French, English, Chinese, and some Spanish and some Latin. I also know a Russian who knows German, English, and maybe some French.
It used to be that to get modded up you could read the article leisurely, understand what it's talking about, and then post your comment at any time... letting the merit of what you wrote stand on its own.
We don't do that anymore
Why the hell was this modded as a troll? Granted, nacturation hasn't been around that long (hah! I mock your six-digit user ID), but he does seem to have hit the nail on the head with the extra big hammer.
I know I've been guilty of replying to the first highly-modded comment, even though my reply had nothing to do with that comment, simply because that increased my visibility to moderators. I know I've been lazy as a moderator on occasion, and blown my mod points on the first half-decent posts I found when browsing "Oldest first". I have sinned myself, and so I know there is truth in what nacturation says.
I hit the karma cap many years ago, and they now no longer even display its numeric value, so I can hardly see the point in continuing with such foolishness. Still, the way slashdot is set up encourages such things. What's the point in posting a comment if nobody will read it? Since the number of readers depends on the comment's score, which depends on how appealing it is to moderators and how early it was posted, we get these types of abuses.
We'd probably be better off with a system where moderators were forced to browse at -1, newest first, and where early posts received a karma penalty unless they achieved a sufficiently high score in moderation. I don't see it happening, though.
Would it be that hard to develop a standard (perhaps much like meta-tagging), giving one set of data easily digestible by the bots (and not displayed to the human reader), while retaining an entertaining writing style for human consumption?
As another poster pointed out, something like this is already possible, via CSS and/or meta tags. The problem is that the system gets abused. Scammers will feed "NATALIE PORTMAN NAKED AND PETRIFIED" or some other high-demand content as the headline to Google, while hapless human users get to see Cialis ads and penis enlargement spam. Naturally, search engine designers know about this and use countermeasures to punish sites that send different content to webcrawlers and users, on the assumption that such tricks are usually employed for malicious purposes. The collateral damage is any site that actually has a legitimate reason to serve different content to webcrawlers than to users.
I know from personal experience that designing for Google has had a negative impact on the aesthetics of my wife's website. Some might argue that designing for Google usually results in a "slimmer" design with more text and less unnecessary images, but when your website is about something visual (say, art), that can be counterproductive. Also, making a (visual) art site have better support for screenreaders seems kind of pointless, and maybe even cruel. What would the ALT tags say? "A really nice painting--too bad you can't see it".
I've done tech support, too, and what bothers me more is not the occasional clueless user. After all, that's what we're paid to deal with, and by and large most are polite if you're polite and not condescending. What bothers me is clueless tech support people who know less than what's readily available on the company's website.
Now, I'm a Mac user and big Mac fan, but I've had plenty of experiences with Apple that go something like this:
Me: Hi, I was preparing to do an upgrade from OS X 10.3 to 10.4, and decided to split my Apple software RAID 1 mirror and just upgrade one drive, so that I could use the other to roll back quickly if needed. I did the upgrade, and all is good, and now I want to wipe the "spare" and add it back into the RAID. I've tried this via Disk Utility, but it fails to rebuild the RAID, and gives error message X. Apple tech support guy: Umm... Can you hold? Me: [holding for a few minutes] Apple: You can try rebooting and using diskutil from the command line... Me: I already tried that, and it fails with error message Y, or error message Z if I try this option. Apple: Can you hold again? Me: OK [listening to lovely hold music] Apple: Uh, you'll have to re-initialize the drives, create the RAID, then install everything again. Me: What? What is the point of RAID if you can't rebuild after losing a drive? Apple: Uh, sorry, I can't help you with that.
This is Apple Premium Support, mind you, which we paid a good chunk of change for.
Now, let me forestall the inevitable flames blaming me for using software RAID or bashing Apple or whatever. We wanted hardware RAID, but at the time we ordered our Xserve, the hardware RAID card was on indefinite backorder (now that the card is available and we know how bad Apple software RAID sucks, we have hardware RAID). We are a small research group, and can't afford to have extra xserves lying around for testing purposes. The only reason I'm picking on Apple is because they're who I have the most experience with.
That said, I don't think the rest of the industry is better. In fact, I see the same thing even outside the computer industry, whether it be medical insurance or government agencies or whatever. Phone support is almost always at the level where only the most basic questions can be handled, but the support people get pissed off that they are asked basic questions. If only there was a way we could add an extra layer of phone support menus to sort this all out:
If you are a typical user, press 1 If you are a power user, enter your processor's model number If you are a genius, enter the smallest prime greater than 2^512 If you are a moron, mash the phone keypad, or remain on the line
Mods: parent is not a troll or flamebait, he's just stating an opinion. That said, he's wrong (IMO), but that's no reason to mod him down. Posts like his are useful in that they further informed and relatively civilized debate.
I much preferred Mac OS back in the OS 9 days. OS 9 screamed in comparison to OS X. It had its problems, sure, but at the time it was the only mainstream OS that was not built on technology besides itself.
If you've ever developed for Mac OS 8/9, you'd realize just how serious those problems were. I wrote part of a printer driver for OS 8/9, and it was hell. Memory allocation was an utter mess. Printer drivers *should* just run in userland, and be unable to muck with the kernel, but that's not how things worked. The driver had full access to both the system memory space and the memory space of whatever application called it. It was preferable to allocate from the application's memory space, but we didn't have that freedom. Because of the stupid user-controlled memory allocation system, we had to worry about how much free memory any application might have been given by the user, and make sure we didn't use more than that. We were trying to modernize the UI and make it more flexible, so we used Metrowerks PowerPlant (an application framework). The problem was this increased our memory requirements to the point where we couldn't fit in the 100kB or so of free memory SimpleText would have by default. To work around this, we would allocate from the system heap. This came with its own problems--if you accidentally wrote to a null pointer, you overwrote the debug traps, and crashed hard. It made for wonderful time in debugging, and forced me to very quickly learn to be careful with pointers and memory allocation (this was my first programming internship, BTW).
Then there was all the cruft left over from the Mac OS's Pascal roots... Pascal strings, pascal calling conventions. And the memory management--Handles!--ugh!
OS 8/9 was a pain in the ass to develop for, whereas OS X is much easier. That's why we're seeing so much great new OS X freeware and shareware.
Thank you for the correction. Do you know if the iTMS revenue is comparable to iPod revenue, or are iPods still the main attraction for the shareholders?
I would buy a Creative player, not an Apple player, next time.
Out of curiousity, why? I have a 4GB iPod nano, and, before that, a 20 GB iPod (3rd generation), and it's hard for me to think up ways to improve them (OK, video on the nano would be fun, but not very practical). Are the Creative players cheaper / have more storage, or is there some other reason? I have no experience with Creative, which is why I'm asking.
if anyone knows how to force the 'Reaver to leave my code alone, could you reply, please?
I hate those damn Reavers, too. Upgraded Zerglings are quite effective in getting rid of them, as are Dark Templar if they don't have an Observer in the area.
People have been using Hymn and the like for ages, and if they're stripping the DRM out of bought files for use on other players they are still buying from Apple and giving Apple money for the privilege
Apple doesn't make much money (directly) from the iTunes Music Store--Steve Jobs himself has said this. Primarily, the purpose of the iTMS is to help sell iPods. What Apple doesn't want to happen is for people to buy able to buy music from iTunes for use on third-party players. If the French iTMS stops being a vehicle for selling iPods, it stops being useful for Apple. Moreover, the conversion utility will inevitably spread outside of France, and hurt iPod sales everywhere. It clearly is preferable from Apple's point of view to close the French iTMS, rather than allow such a conversion utility to become widespread.
Oh, sure, people can use Hymn, but Joe User isn't that sophisticated. Also, AFAIK, the Hymn people haven't yet figured out how to crack iTunes v6 encryption, so it's not exactly a fully-functional solution.
Seems to me that they've become almost programmed to think of anything new as necessarily "better" and thus desirable without ever thinking of the old addage "If it ain't broke don't fix it"
They are programmed to think that way because that's how we've programmed them. We (and I mean the collective slashdot "we", not the royal "we", or even necessarily you) are constantly lecturing users about installing all the latest security patches.
Do you really expect the average user to be able to distinguish a necessary security patch from an unnecessary minor functionality fix? Is there even a faint hope that they would be able to accurately assess the risks and benefits of a patch which combines security bug fixes with other changes (as so many patches do)?
Actually, the bigger spending problem is seniors, not the Iraq war or defense spending. Defense is much smaller (at 23% of total government spending even when you include veterans' affairs) than government expenditures on social security, medicare, and other retirement-related programs (which cost 36% of total government spending). My source is the federal government, via a handy pie chart on page 81 of the instructions for IRS tax form 1040.
This problem is only going to get worse as people live longer and have fewer (tax-paying) children.
The obvious solution? We should have fought the Iraq war with senior-soldiers. Social security checks could have been conditional on enlistment.
There's nobody who can compete with the U.S. for magnitude of evil in this millenium.
In the past six years? Even if it were true, don't you think there's some sampling error inherent in such a small amount of time? I mean, if we were to define a 50 milisecond interval and ask who was the most evil person in the world in that interval, it might well be some guy who decided to double-park his SUV. Hardly meaningful.
China? China hasn't spread nuclear waste over the cradle of civilization.
What are you alluding to? The US hasn't done this either. The closest thing I can think of is the Tuwaitha nuclear facility in Iraq, where our troops didn't arrive fast enough to prevent looters from dumping some barrels of radioactive waste.
China hasn't invaded non-beligerent nations in decades
That's because China invaded decades ago and is still there (Tibet), or is gearing up to invade (Taiwan). Since you restricted your argument to the current millennium, I assume you're referring to the US invasion of Iraq. For what it's worth, Iraq was in violation of several UN Security Council resolutions, and also the terms of the ceasefire of the first Gulf War. The ceasefire required them to provide proof of destruction of all WMD within months of the end of that war. The burden of proof was on them, and they didn't meet it, despite having many years to do so. Under the traditional rules and customs of war, if you violate the terms of a ceasefire, the other party has the right to resume hostilities. Yes, we should never have tried to justify the invasion of Iraq on the existence of WMD, but rather on the many other legitimate reasons to do so (not the least of which is bringing democracy to the oppressed people of Iraq). Blame Blair and the Great Mis-Communicator for that....hasn't killed a quarter of a million people in another country in order to keep them from forming a government for their homeland.
Now I really have no idea what you're talking about. When in the past six years did the US do anything like this?
I don't know what the stats are on forced abortions in China, though. Depending on your value system, this might loom large in any comparison.
It should loom large in any value system. Forced abortions are hardly pro-choice. Nice way to subtly imply pro-lifers are to blame for all the US's evils.
A year of C++ from most schools leaves you in a position of being able to solve homework problems and do basic exercises, but it's a long way from being able to make much headway in a significant software engineering situation. Which is a shame, because that's where the best learning takes place. If you can find an internship where you can work with experienced programmers on serious code go for it -- it's got the potential for being an incredibly valuable experience. But if you're only going to be a round for the summer, a lot of companies won't want to invest the effort required to bring a freshman up to speed on their development environment, their code base, and their processes.
I partially agree. I had 3 semesters (1.5 years) of university before I took my first co-op position (like an internship, but fairly well paid, and coordinated by the university), which was with a software company. The first year was all the standard math & science courses, plus two computer science courses. First semester was Scheme, and the second semester was about half and half C++. In the first semester of second year, I took a C++-based data structures course, a theory of computation / discrete logic course, and a hardware course, plus some non-CS stuff.
My first co-op position was really quite in-depth, because it was 8 months long. Initially, I did a fairly easy project, basically re-doing the installer for their software package to make it more user-friendly. My next project was to do a from-the-ground re-design and re-write of the GUI for one of their OEM drivers. In a way, it was made easier by the fact that the GUI could be modular and largely independent from the rest of the codebase. The GUI was fairly simple--maybe 2 or 3 complex windows--so it was a good-sized project for me.
It was quite the experience, as this was during the peak of the dot-com boom and subsequent collapse, and there were many turbulent events at the company. Mass layoffs, sale of the division I was in, etc. I survived all of it, and even scavenged better hardware from those who were "downsized". The share price went from $16 to $69 and back down again (and down some more).
How did I get the job? By being far and away the most qualified. Out of a pool of perhaps 50 or 100 co-op applicants, I was pretty much the only one with extensive experience on their target platform (most of this experience was from the computer consulting I did while in high school). The interview was pretty much a formality. This may sound discouraging if you don't have special skills, but it really is the key to getting the position you want: start early, figure out the job you want, figure out the skills it requires, then get those skills.
Disclaimer: IAAQIS (I Am A Quantum Information Scientist).
The actual journal article that New Scientist is referring to was just published: Nature439 949. I'm not sure if that link will work if you're not at an institution that has a subscription, but you'll probably at least get to see an abstract.
A few bits of background: New Scientist's coverage of quantum information is sometimes horrible. Therefore, it's not surprising that the New Scientist article makes no sense but contains lots of exciting fluff. That said, these guys do have something interesting.
I skimmed through the Nature article, and it looks interesting. It's especially nice that they have an experimental implementation. Nonetheless, the bit about the quantum computer being "off" is just silly.
Here's a summary of how it works, stripped of some hyperbole and converted into something more like plain english (note: qubit means quantum bit).
(1) Create a "control qubit" and some output qubits, with the control qubit initially set to 0, which we will take to mean off.
(2) "Rotate" the control qubit into a superposition of 1 (on) and 0 (off), with most of the "amplitude" being for the 0 state (the qubit is mostly off)
(3) Apply whatever algorithm to the data and the output qubits, conditional on the control bit being on. (Note: we don't actually measure anything here--this is entirely a unitary operation).
(4) Perform a weak measurement on the output qubits, which has the effect of reducing the amplitude of the output qubits being in something other than their initial state (which can only happen if the control qubit was on and the algorithm was applied), since the amplitude for that was small to begin with.
(5) Repeat (2) - (4) N times, such that, if the output bits are unmodified after each algorithm application, you end up with the control qubit in the 1 (on) state. Otherwise, you get the 0 (off) state.
(6) Profit!
This is the simple version, in which you only get to learn whether the application of the algorithm to the data gives you the default output or not. There's a more sophisticated version in which you learn more about the data.
There are a few catches here. One is that N has to be reasonably large, or the probability of an "error" in step 4 becomes an issue (by error, I mean that the weak measurement gives us the wrong outcome). Specifically, the probability of an error is 1 - cos^2N (pi / 2N), which scales as O(N^(1-4N)). Fortunately, that is exponential suppression of error, which is pretty good scaling. Another catch is that their particular experimental implementation used a non-scalable encoding. This isn't a major issue, but it means we should wait for an experiment using a scalable encoding before we really break out the champagne.
Quantum cryptography that does use public key cryptography for the initial authentication can only be broken if the eavesdropper possesses the realtime technology for breaking public key crypto today. If the eavesdropper possesses it in the future after the initial authentication has been done, quantum cryptography is unaffected.
That's a good point, however, realtime factorization is not needed to break public key-authenticated quantum crypto. Any attack on public key crypto that allows one to compromise the trusted party's signing key--even if the attack is slow and laborious--is sufficient to break the authentication and allow a man-in-the-middle attack. The only advantage gained through the hybrid quantum/public key crypto is protection from retroactive attacks. This is an important advantage, but it is partially mitigated by the uncertainty in determining the moment when public key is compromised.
The moment someone gains the ability to break public key--however slowly and painfully--it becomes useless for authentication, because only the trusted signer's key needs to be compromised.
And in public key cryptography, you have to have a trusted party to provide you the public key, and have to authenticate with that party.
The trusted party (Charlie, let's say) has to provide Bob only with a copy of his public key, in a manner that Bob trusts to be authentic and untampered-with. Charlie signs Alice's public key. Alice can then publicly announce her signed public key.
No secure communication is possible without sharing an initial small secret. You cannot in principle say whom you are communicating with if you share no prior secret information with him. This is the way things are in this world. It is common for all cryptography, quantum or not.
Actually, there are no shared secrets in public key cryptography. There are signed public keys, and one or more untampered keys of trusted parties. These things are shared, but they're not secret--no security is lost by publicizing this information. Then there are secrets--the private keys of each party--but they're not shared.
I think what you mean is that public key crypto does require one initial authenticated information exchange, in which one receives the public (signing) key of a trusted party. That's true, but the important difference between public key crypto and quantum crypto is how things scale as more and more parties are introduced. Let's say we have one trusted party, Charlie, and n other people. In public key crypto, Charlie needs to distribute his public key to all parties in a manner that does not allow tampering, but can be public. Subsequently, each party needs to send their respective public key to Charlie for signing. Again, this communication can be public, but must be protected against tampering (a face-to-face conversation is a good example). Now, Charlie is no longer needed, and all parties can communicate securely and privately.
Compare this to quantum key distribution. Initially, each party must privately and securely generate a shared secret with Charlie. Now, we can proceed one of two ways. (1) Charlie generates n (n - 1) / 2 secret keys, and distributes them such that each party has a shared secret with each of the other parties (i.e. we have a secret for each possible pair of parties). Charlie is now no longer needed. Or, (2) whenever a pair of parties wants to communicate, they have Charlie generate a shared secret for them. Assuming each party eventually wants to communicate with most other parties, the workload is about the same.
The bottom line is that the load on the trusted party scales as O(n) for public key crypto, versus O(n^2) for quantum. Also, quantum key distribution requires private initial distribution of secrets, whereas public key crypto only requires that the initial communications not be tampered with.
Disclaimer: IAAQIS (I Am A Quantum Information Scientist).
The parent poster (wwwrench) is completely, 100% correct. Is this really Slashdot, or did I type the wrong URL?
Seriously, though, the parent poster is bang-on. To elaborate a bit, quantum cryptography would be more informatively called quantum key distribution (although both names are common in practice). All it does is allow you to distribute a key for a one-time pad in a secure method, given that the laws of quantum mechanics are at least partially correct (one-time pads are information-theoretic secure, provided the key is not compromised or re-used). If somebody tries to eavesdrop, you can detect it, and respond accordingly. That response could be privacy amplification (if the information the eavesdropper gained was only partial), re-trying the protocol, or bombing the eavesdropper to smithereens. That last possibility is why quantum telecloning might be useful.
One other hitch is that quantum key distribution requires a small shared secret in order to authenticate the two parties trying to generate a key. Thus, quantum key distribution is not a complete replacement for public-key cryptography.
Let's, for the sake of argument, assume the planet is warming, that this warming will have serious negative consequences, and that the warming can't be avoided through reduced human-caused emissions of greenhouse gases. This last assumption is valid under at least three scenarios: (1) humans are not the primary cause of global warming, or (2) reducing CO2 emissions would be prohibitively expensive, or (3) reducing CO2 emissions is politically infeasible.
What then?
I strongly suspect there are alternative means of avoiding global warming, or of abating many of the negative consequences, and that some of these methods may prove cheaper than dramatic immediate reductions in CO2 emissions. I could speculate about why most people haven't considered such measures, but that's not the point. The few ideas I have seen discussed usually involve some method of changing the earth's albedo. This seems like a potentially simple and elegant solution. Thoughts? Comments?
Slashdot Mirror
*Viruses* had nothing to do with the Irish potato famine. While there were many factors for the famine, many of them political, the pathological reason was the *fungus* Phytophthora infestans.
*ahem*
[shallow and pedantic]
*Fungi* had nothing to do with the Irish potato famine. Phytophtora is an oomycete, not a fungus.
[/shallow and pedantic]
This is really just an economic decision. An IT monoculture brings with it certain benefits, such as decreased (non-virus-related) support costs, but also certain costs, such as increased vulnerability to viruses. It's worthwhile for people like Dan Greer to make the IT world aware of those costs, but even once they become aware, a Windows monoculture may still be preferable for some. OTOH, I don't think the costs of supporting multiple OSs are as high as most people think. We have a small network with OS X Server serving files and doing authentication for a bunch of Windows and Linux boxes, and it's really not very hard to keep running.
Breeder reactors are here now only in the sense that there are a few operational reactors--there is not widescale deployment of breeder reactor technology as there is with conventional fission reactors (I was careful to say "widescale" in my original post). Obviously, present-day breeder reactor technology is more advanced than fusion reactor tech, but, on the other hand, the payoff isn't nearly as good as we can hope for with fusion.
Another poster mentioned that the 50-year timeline for running out of uranium was without breeder reactors, with current usage patterns. I was aware that this estimate assumed no breeder reactors, but I thought it was based on a much higher rate of consumption than we have now. Can anyone confirm / deny?
I might approve an order of magnitude increase in funding at that point, but I see no reason to do so now when there are technologies, particularly fission, wind, and solar power that are becoming viable.
Solar and wind are approaching economic viability as supplemental energy sources. What I mean is that they are good at helping meet some of the peak demand, but not so good as a baseline power source, since the wind doesn't always blow and the sun doesn't always shine. In particular, there are several hours each day in which the entire continent is in the dark (wind is also usually calmer at night). In an all-solar/wind energy scenario, you'd also need an enormous amount of energy storage capability (electricity-to-hydrogen-to-electricity in fuel cells, maybe), which rather dramatically raises your costs.
Nuclear, on the other hand, can be more expensive, but are well suited to be run continuously (aside from maintenance every once in a while), since fuel is only a small fraction of the cost per kW/h.
What makes more sense than either source on its own is nuclear fission to supply most of the baseline power, with solar and wind to supply peak power, and a few natural-gas fired plants for backup. Conveniently, peak generating times of solar and wind tend to correspond to peak demand.
Given the above scenario, you'll run out of fuel for your fission reactors in half a century or so (give or take a few decades), unless you start using breeder reactors, which aren't really a widescale-proven technology, and pose some nuclear proliferation issues. If you're going to pour research money into breeder reactors, why not spend it instead on fusion, which is pretty much the ultimate terrestrial power source?
Now I'm the first to admit that his public words and actions suggest someone much closer to (or even slightly below) the mean.
Public speaking is not his forte, apparently. I loved that SNL skit with him and Gore, where Bush's one-word summary of his campaign was "strategery". Did he actually ever say that word, or was that a bit of creativity on the part of the SNL writers?
In fact, Bush's seeming disdain for "book smarts" typifies the result we both expect from a poor educational system (though I'm under no illusion that he ever attended public school).
I think the disdain for "book smarts" comes more from political orientation than anything else. It's more a mistrust of the climate of arrogant intellectualism that pervades the pseudo-meritocratic structure of academia. As an academic myself (a physics grad student, to be specific), I can certainly see the off-putting aspects of the academic world. It's very easy to assume that, because you're the world's foremost expert on X, you understand Y and Z as well (I know I've done this). When that turns into preaching politics from the pulpit of science, it really pisses people off.
An example: recently, a bunch of physicists signed a letter about the dangers of using nuclear weapons against Iran. That kind of thing just makes me cringe. I know some of the physicists who signed the letter, and they've given me no reason to believe they're any more informed or knowledgeable about the Iran situation that your average college graduate. The spin on the letter was something along the lines of the claim, "we invented the nuclear weapon, and we're here to tell you just how bad it would be to use it", but it's not as if any of the signing physicists actually provided any new information. Military analysts, Middle Eastern policy studies professors, Iranian exiles--all of them might have had something valuable to say--but physicists? Hardly. All that kind of stunt does is make the Republicans even more pissed off at academia.
A major backlash from the right is coming--primarily aimed perceived left-wing bias in the social sciences, etc.--but there will be collateral damage to the natural sciences as well. The more reasons we give Republicans to lump the natural sciences in with the "left-wing" crowd, the worse things will be. Read Phi Beta Cons, a higher education blog run by the conservative magazine National Review, to see what I'm talking about.
Not the Einsteins of the world but the George Bushes.
eddeye, I'm surprised you had the courage to say that--even as a joke--on slashdot. From what I've read, GWB's SAT scores suggest he's of average intelligence for a president. Apparently, his SATs correspond to an IQ test score of around 130 (I forgot the exact numbers), which is approximately two sigma above the mean. In other words, your comment is entirely accurate, but nonetheless a serious lightning rod for politically-charged flames.
I agree with your main point; the smartest students will find something to occupy their minds. The moderately intelligent students are left without much, however, and it hurts them when they get to university and are challenged for the first time. As there are far more people in this category than there are geniuses, this is a major loss for society.
How any russians know german and korean and spanish?
No idea, but I know a Norwegian who knows Norwegian, Swedish, German, French, English, Chinese, and some Spanish and some Latin. I also know a Russian who knows German, English, and maybe some French.
It used to be that to get modded up you could read the article leisurely, understand what it's talking about, and then post your comment at any time... letting the merit of what you wrote stand on its own.
We don't do that anymore
Why the hell was this modded as a troll? Granted, nacturation hasn't been around that long (hah! I mock your six-digit user ID), but he does seem to have hit the nail on the head with the extra big hammer.
I know I've been guilty of replying to the first highly-modded comment, even though my reply had nothing to do with that comment, simply because that increased my visibility to moderators. I know I've been lazy as a moderator on occasion, and blown my mod points on the first half-decent posts I found when browsing "Oldest first". I have sinned myself, and so I know there is truth in what nacturation says.
I hit the karma cap many years ago, and they now no longer even display its numeric value, so I can hardly see the point in continuing with such foolishness. Still, the way slashdot is set up encourages such things. What's the point in posting a comment if nobody will read it? Since the number of readers depends on the comment's score, which depends on how appealing it is to moderators and how early it was posted, we get these types of abuses.
We'd probably be better off with a system where moderators were forced to browse at -1, newest first, and where early posts received a karma penalty unless they achieved a sufficiently high score in moderation. I don't see it happening, though.
Would it be that hard to develop a standard (perhaps much like meta-tagging), giving one set of data easily digestible by the bots (and not displayed to the human reader), while retaining an entertaining writing style for human consumption?
As another poster pointed out, something like this is already possible, via CSS and/or meta tags. The problem is that the system gets abused. Scammers will feed "NATALIE PORTMAN NAKED AND PETRIFIED" or some other high-demand content as the headline to Google, while hapless human users get to see Cialis ads and penis enlargement spam. Naturally, search engine designers know about this and use countermeasures to punish sites that send different content to webcrawlers and users, on the assumption that such tricks are usually employed for malicious purposes. The collateral damage is any site that actually has a legitimate reason to serve different content to webcrawlers than to users.
I know from personal experience that designing for Google has had a negative impact on the aesthetics of my wife's website. Some might argue that designing for Google usually results in a "slimmer" design with more text and less unnecessary images, but when your website is about something visual (say, art), that can be counterproductive. Also, making a (visual) art site have better support for screenreaders seems kind of pointless, and maybe even cruel. What would the ALT tags say? "A really nice painting--too bad you can't see it".
Let's hope that the HAL suits don't consider the quadriplegic a threat to the mission.
I think we have bigger problems. The company making HAL is Cyberdyne. We may never know who launches the first strike.
I've done tech support, too, and what bothers me more is not the occasional clueless user. After all, that's what we're paid to deal with, and by and large most are polite if you're polite and not condescending. What bothers me is clueless tech support people who know less than what's readily available on the company's website.
Now, I'm a Mac user and big Mac fan, but I've had plenty of experiences with Apple that go something like this:
Me: Hi, I was preparing to do an upgrade from OS X 10.3 to 10.4, and decided to split my Apple software RAID 1 mirror and just upgrade one drive, so that I could use the other to roll back quickly if needed. I did the upgrade, and all is good, and now I want to wipe the "spare" and add it back into the RAID. I've tried this via Disk Utility, but it fails to rebuild the RAID, and gives error message X.
Apple tech support guy: Umm... Can you hold?
Me: [holding for a few minutes]
Apple: You can try rebooting and using diskutil from the command line...
Me: I already tried that, and it fails with error message Y, or error message Z if I try this option.
Apple: Can you hold again?
Me: OK [listening to lovely hold music]
Apple: Uh, you'll have to re-initialize the drives, create the RAID, then install everything again.
Me: What? What is the point of RAID if you can't rebuild after losing a drive?
Apple: Uh, sorry, I can't help you with that.
This is Apple Premium Support, mind you, which we paid a good chunk of change for.
Now, let me forestall the inevitable flames blaming me for using software RAID or bashing Apple or whatever. We wanted hardware RAID, but at the time we ordered our Xserve, the hardware RAID card was on indefinite backorder (now that the card is available and we know how bad Apple software RAID sucks, we have hardware RAID). We are a small research group, and can't afford to have extra xserves lying around for testing purposes. The only reason I'm picking on Apple is because they're who I have the most experience with.
That said, I don't think the rest of the industry is better. In fact, I see the same thing even outside the computer industry, whether it be medical insurance or government agencies or whatever. Phone support is almost always at the level where only the most basic questions can be handled, but the support people get pissed off that they are asked basic questions. If only there was a way we could add an extra layer of phone support menus to sort this all out:
If you are a typical user, press 1
If you are a power user, enter your processor's model number
If you are a genius, enter the smallest prime greater than 2^512
If you are a moron, mash the phone keypad, or remain on the line
Mods: parent is not a troll or flamebait, he's just stating an opinion. That said, he's wrong (IMO), but that's no reason to mod him down. Posts like his are useful in that they further informed and relatively civilized debate.
I much preferred Mac OS back in the OS 9 days. OS 9 screamed in comparison to OS X. It had its problems, sure, but at the time it was the only mainstream OS that was not built on technology besides itself.
If you've ever developed for Mac OS 8/9, you'd realize just how serious those problems were. I wrote part of a printer driver for OS 8/9, and it was hell. Memory allocation was an utter mess. Printer drivers *should* just run in userland, and be unable to muck with the kernel, but that's not how things worked. The driver had full access to both the system memory space and the memory space of whatever application called it. It was preferable to allocate from the application's memory space, but we didn't have that freedom. Because of the stupid user-controlled memory allocation system, we had to worry about how much free memory any application might have been given by the user, and make sure we didn't use more than that. We were trying to modernize the UI and make it more flexible, so we used Metrowerks PowerPlant (an application framework). The problem was this increased our memory requirements to the point where we couldn't fit in the 100kB or so of free memory SimpleText would have by default. To work around this, we would allocate from the system heap. This came with its own problems--if you accidentally wrote to a null pointer, you overwrote the debug traps, and crashed hard. It made for wonderful time in debugging, and forced me to very quickly learn to be careful with pointers and memory allocation (this was my first programming internship, BTW).
Then there was all the cruft left over from the Mac OS's Pascal roots... Pascal strings, pascal calling conventions. And the memory management--Handles!--ugh!
OS 8/9 was a pain in the ass to develop for, whereas OS X is much easier. That's why we're seeing so much great new OS X freeware and shareware.
Analysts state that it's making money now.
Thank you for the correction. Do you know if the iTMS revenue is comparable to iPod revenue, or are iPods still the main attraction for the shareholders?
I would buy a Creative player, not an Apple player, next time.
Out of curiousity, why? I have a 4GB iPod nano, and, before that, a 20 GB iPod (3rd generation), and it's hard for me to think up ways to improve them (OK, video on the nano would be fun, but not very practical). Are the Creative players cheaper / have more storage, or is there some other reason? I have no experience with Creative, which is why I'm asking.
if anyone knows how to force the 'Reaver to leave my code alone, could you reply, please?
I hate those damn Reavers, too. Upgraded Zerglings are quite effective in getting rid of them, as are Dark Templar if they don't have an Observer in the area.
Perhaps I've been playing too much Starcraft.
People have been using Hymn and the like for ages, and if they're stripping the DRM out of bought files for use on other players they are still buying from Apple and giving Apple money for the privilege
Apple doesn't make much money (directly) from the iTunes Music Store--Steve Jobs himself has said this. Primarily, the purpose of the iTMS is to help sell iPods. What Apple doesn't want to happen is for people to buy able to buy music from iTunes for use on third-party players. If the French iTMS stops being a vehicle for selling iPods, it stops being useful for Apple. Moreover, the conversion utility will inevitably spread outside of France, and hurt iPod sales everywhere. It clearly is preferable from Apple's point of view to close the French iTMS, rather than allow such a conversion utility to become widespread.
Oh, sure, people can use Hymn, but Joe User isn't that sophisticated. Also, AFAIK, the Hymn people haven't yet figured out how to crack iTunes v6 encryption, so it's not exactly a fully-functional solution.
I believe the phrase you're looking for is shallow and pedantic.
Seems to me that they've become almost programmed to think of anything new as necessarily "better" and thus desirable without ever thinking of the old addage "If it ain't broke don't fix it"
They are programmed to think that way because that's how we've programmed them. We (and I mean the collective slashdot "we", not the royal "we", or even necessarily you) are constantly lecturing users about installing all the latest security patches.
Do you really expect the average user to be able to distinguish a necessary security patch from an unnecessary minor functionality fix? Is there even a faint hope that they would be able to accurately assess the risks and benefits of a patch which combines security bug fixes with other changes (as so many patches do)?
Actually, the bigger spending problem is seniors, not the Iraq war or defense spending. Defense is much smaller (at 23% of total government spending even when you include veterans' affairs) than government expenditures on social security, medicare, and other retirement-related programs (which cost 36% of total government spending). My source is the federal government, via a handy pie chart on page 81 of the instructions for IRS tax form 1040.
This problem is only going to get worse as people live longer and have fewer (tax-paying) children.
The obvious solution? We should have fought the Iraq war with senior-soldiers. Social security checks could have been conditional on enlistment.
There's nobody who can compete with the U.S. for magnitude of evil in this millenium.
...hasn't killed a quarter of a million people in another country in order to keep them from forming a government for their homeland.
In the past six years? Even if it were true, don't you think there's some sampling error inherent in such a small amount of time? I mean, if we were to define a 50 milisecond interval and ask who was the most evil person in the world in that interval, it might well be some guy who decided to double-park his SUV. Hardly meaningful.
China? China hasn't spread nuclear waste over the cradle of civilization.
What are you alluding to? The US hasn't done this either. The closest thing I can think of is the Tuwaitha nuclear facility in Iraq, where our troops didn't arrive fast enough to prevent looters from dumping some barrels of radioactive waste.
China hasn't invaded non-beligerent nations in decades
That's because China invaded decades ago and is still there (Tibet), or is gearing up to invade (Taiwan). Since you restricted your argument to the current millennium, I assume you're referring to the US invasion of Iraq. For what it's worth, Iraq was in violation of several UN Security Council resolutions, and also the terms of the ceasefire of the first Gulf War. The ceasefire required them to provide proof of destruction of all WMD within months of the end of that war. The burden of proof was on them, and they didn't meet it, despite having many years to do so. Under the traditional rules and customs of war, if you violate the terms of a ceasefire, the other party has the right to resume hostilities. Yes, we should never have tried to justify the invasion of Iraq on the existence of WMD, but rather on the many other legitimate reasons to do so (not the least of which is bringing democracy to the oppressed people of Iraq). Blame Blair and the Great Mis-Communicator for that.
Now I really have no idea what you're talking about. When in the past six years did the US do anything like this?
I don't know what the stats are on forced abortions in China, though. Depending on your value system, this might loom large in any comparison.
It should loom large in any value system. Forced abortions are hardly pro-choice. Nice way to subtly imply pro-lifers are to blame for all the US's evils.
A year of C++ from most schools leaves you in a position of being able to solve homework problems and do basic exercises, but it's a long way from being able to make much headway in a significant software engineering situation. Which is a shame, because that's where the best learning takes place. If you can find an internship where you can work with experienced programmers on serious code go for it -- it's got the potential for being an incredibly valuable experience. But if you're only going to be a round for the summer, a lot of companies won't want to invest the effort required to bring a freshman up to speed on their development environment, their code base, and their processes.
I partially agree. I had 3 semesters (1.5 years) of university before I took my first co-op position (like an internship, but fairly well paid, and coordinated by the university), which was with a software company. The first year was all the standard math & science courses, plus two computer science courses. First semester was Scheme, and the second semester was about half and half C++. In the first semester of second year, I took a C++-based data structures course, a theory of computation / discrete logic course, and a hardware course, plus some non-CS stuff.
My first co-op position was really quite in-depth, because it was 8 months long. Initially, I did a fairly easy project, basically re-doing the installer for their software package to make it more user-friendly. My next project was to do a from-the-ground re-design and re-write of the GUI for one of their OEM drivers. In a way, it was made easier by the fact that the GUI could be modular and largely independent from the rest of the codebase. The GUI was fairly simple--maybe 2 or 3 complex windows--so it was a good-sized project for me.
It was quite the experience, as this was during the peak of the dot-com boom and subsequent collapse, and there were many turbulent events at the company. Mass layoffs, sale of the division I was in, etc. I survived all of it, and even scavenged better hardware from those who were "downsized". The share price went from $16 to $69 and back down again (and down some more).
How did I get the job? By being far and away the most qualified. Out of a pool of perhaps 50 or 100 co-op applicants, I was pretty much the only one with extensive experience on their target platform (most of this experience was from the computer consulting I did while in high school). The interview was pretty much a formality. This may sound discouraging if you don't have special skills, but it really is the key to getting the position you want: start early, figure out the job you want, figure out the skills it requires, then get those skills.
Disclaimer: IAAQIS (I Am A Quantum Information Scientist).
The actual journal article that New Scientist is referring to was just published: Nature 439 949. I'm not sure if that link will work if you're not at an institution that has a subscription, but you'll probably at least get to see an abstract.
A few bits of background: New Scientist's coverage of quantum information is sometimes horrible. Therefore, it's not surprising that the New Scientist article makes no sense but contains lots of exciting fluff. That said, these guys do have something interesting.
I skimmed through the Nature article, and it looks interesting. It's especially nice that they have an experimental implementation. Nonetheless, the bit about the quantum computer being "off" is just silly.
Here's a summary of how it works, stripped of some hyperbole and converted into something more like plain english (note: qubit means quantum bit).
(1) Create a "control qubit" and some output qubits, with the control qubit initially set to 0, which we will take to mean off.
(2) "Rotate" the control qubit into a superposition of 1 (on) and 0 (off), with most of the "amplitude" being for the 0 state (the qubit is mostly off)
(3) Apply whatever algorithm to the data and the output qubits, conditional on the control bit being on. (Note: we don't actually measure anything here--this is entirely a unitary operation).
(4) Perform a weak measurement on the output qubits, which has the effect of reducing the amplitude of the output qubits being in something other than their initial state (which can only happen if the control qubit was on and the algorithm was applied), since the amplitude for that was small to begin with.
(5) Repeat (2) - (4) N times, such that, if the output bits are unmodified after each algorithm application, you end up with the control qubit in the 1 (on) state. Otherwise, you get the 0 (off) state.
(6) Profit!
This is the simple version, in which you only get to learn whether the application of the algorithm to the data gives you the default output or not. There's a more sophisticated version in which you learn more about the data.
There are a few catches here. One is that N has to be reasonably large, or the probability of an "error" in step 4 becomes an issue (by error, I mean that the weak measurement gives us the wrong outcome). Specifically, the probability of an error is 1 - cos^2N (pi / 2N), which scales as O(N^(1-4N)). Fortunately, that is exponential suppression of error, which is pretty good scaling. Another catch is that their particular experimental implementation used a non-scalable encoding. This isn't a major issue, but it means we should wait for an experiment using a scalable encoding before we really break out the champagne.
Quantum cryptography that does use public key cryptography for the initial authentication can only be broken if the eavesdropper possesses the realtime technology for breaking public key crypto today. If the eavesdropper possesses it in the future after the initial authentication has been done, quantum cryptography is unaffected.
That's a good point, however, realtime factorization is not needed to break public key-authenticated quantum crypto. Any attack on public key crypto that allows one to compromise the trusted party's signing key--even if the attack is slow and laborious--is sufficient to break the authentication and allow a man-in-the-middle attack. The only advantage gained through the hybrid quantum/public key crypto is protection from retroactive attacks. This is an important advantage, but it is partially mitigated by the uncertainty in determining the moment when public key is compromised.
The moment someone gains the ability to break public key--however slowly and painfully--it becomes useless for authentication, because only the trusted signer's key needs to be compromised.
And in public key cryptography, you have to have a trusted party to provide you the public key, and have to authenticate with that party.
The trusted party (Charlie, let's say) has to provide Bob only with a copy of his public key, in a manner that Bob trusts to be authentic and untampered-with. Charlie signs Alice's public key. Alice can then publicly announce her signed public key.
No secure communication is possible without sharing an initial small secret. You cannot in principle say whom you are communicating with if you share no prior secret information with him. This is the way things are in this world. It is common for all cryptography, quantum or not.
Actually, there are no shared secrets in public key cryptography. There are signed public keys, and one or more untampered keys of trusted parties. These things are shared, but they're not secret--no security is lost by publicizing this information. Then there are secrets--the private keys of each party--but they're not shared.
I think what you mean is that public key crypto does require one initial authenticated information exchange, in which one receives the public (signing) key of a trusted party. That's true, but the important difference between public key crypto and quantum crypto is how things scale as more and more parties are introduced. Let's say we have one trusted party, Charlie, and n other people. In public key crypto, Charlie needs to distribute his public key to all parties in a manner that does not allow tampering, but can be public. Subsequently, each party needs to send their respective public key to Charlie for signing. Again, this communication can be public, but must be protected against tampering (a face-to-face conversation is a good example). Now, Charlie is no longer needed, and all parties can communicate securely and privately.
Compare this to quantum key distribution. Initially, each party must privately and securely generate a shared secret with Charlie. Now, we can proceed one of two ways. (1) Charlie generates n (n - 1) / 2 secret keys, and distributes them such that each party has a shared secret with each of the other parties (i.e. we have a secret for each possible pair of parties). Charlie is now no longer needed. Or, (2) whenever a pair of parties wants to communicate, they have Charlie generate a shared secret for them. Assuming each party eventually wants to communicate with most other parties, the workload is about the same.
The bottom line is that the load on the trusted party scales as O(n) for public key crypto, versus O(n^2) for quantum. Also, quantum key distribution requires private initial distribution of secrets, whereas public key crypto only requires that the initial communications not be tampered with.
Disclaimer: IAAQIS (I Am A Quantum Information Scientist).
The parent poster (wwwrench) is completely, 100% correct. Is this really Slashdot, or did I type the wrong URL?
Seriously, though, the parent poster is bang-on. To elaborate a bit, quantum cryptography would be more informatively called quantum key distribution (although both names are common in practice). All it does is allow you to distribute a key for a one-time pad in a secure method, given that the laws of quantum mechanics are at least partially correct (one-time pads are information-theoretic secure, provided the key is not compromised or re-used). If somebody tries to eavesdrop, you can detect it, and respond accordingly. That response could be privacy amplification (if the information the eavesdropper gained was only partial), re-trying the protocol, or bombing the eavesdropper to smithereens. That last possibility is why quantum telecloning might be useful.
One other hitch is that quantum key distribution requires a small shared secret in order to authenticate the two parties trying to generate a key. Thus, quantum key distribution is not a complete replacement for public-key cryptography.
You do see the difference between China and the US, right? I was hoping you were just a troll, but your comment history suggests otherwise.
A repressive, unelected government has no legitimate authority, so one can hardly fault people for rebelling against the Chinese government.
Let's, for the sake of argument, assume the planet is warming, that this warming will have serious negative consequences, and that the warming can't be avoided through reduced human-caused emissions of greenhouse gases. This last assumption is valid under at least three scenarios: (1) humans are not the primary cause of global warming, or (2) reducing CO2 emissions would be prohibitively expensive, or (3) reducing CO2 emissions is politically infeasible.
What then?
I strongly suspect there are alternative means of avoiding global warming, or of abating many of the negative consequences, and that some of these methods may prove cheaper than dramatic immediate reductions in CO2 emissions. I could speculate about why most people haven't considered such measures, but that's not the point. The few ideas I have seen discussed usually involve some method of changing the earth's albedo. This seems like a potentially simple and elegant solution. Thoughts? Comments?