My response: cry me a river, and congrats to the grad students for their innovative work in the field of distributed communications.
I'd pause before calling this innovative. It doesn't really take much to encrypt data, chop it up and stash it on MediaWiki sites -- either in theory or in practice. If you want something "innovative" in the same vein, I'd vote for the guy who wrote the device driver that lets you use GMail as a drive (spawning many copies). Sure it isn't "distributed", but you could set up multiple GMail accounts to handle the contents of your drive. Clogging up other people's wikis is d**k at worst (and possibly a violation of the CFAA), and really not too much of a security threat at best ("oh? my disk is full? hmm...just dump this spammy user account, or restore the last backup, and password protect the whole business.").
What these grad students have done is demonstrate that open mediawiki setups can be spammed. Whee.
And why exactly are we throwing millions of page views at these racist fucks???
Sigh...the dangers of news aggregation, which is probably how the submitter stumbled upon this piece of bigoted tripe. Somewhere, a thousand page ranking algorithms just went mad seeing these two ostensibly different areas of the Net link to each other.
Can we please be careful not to link this sort of website to slashdot? Somebody please think of the page ranking algorithms!! For those who don't want to click through, this is a summary from the site in the second link of the story (amren.com).
*
The Dangers of Diversity, Part II, Editor Jared Taylor continues a multi-part examination of the effects of diversity by taking a look at just what happens when races mix. This segment provides additional examples of the violence and conflict that ensues when different racial groups are forced to mingle in public schools, and also in prisons.
*
In Three Race Murders in Seattle, journalist Nicholas Stix reports on the murders of three white men, Edward Scott McMichael, James Paroline, and Kristopher Klime, by blacks. Mr. Stix shows how the media (and government) consistently downplayed the racial angle in these cases.
*
In A Voice For Our People, Peter Bradley reviews Frank Borzellieri's new book Lynched: A Conservative's Life on a New York City School Board, which documents the author's one-man crusade to keep Western literature and values alive in increasingly "multicultural" New York City. Mr. Bradley holds him up as example of what a difference one determined man can make in his community.
*
Plus, paying the price for insulting Obama supporters, the GOP takes its message to "hip hop settings," Attorney General Holder on the "nation of cowards," the US military becoming another foreign legion, another monkey cartoon controversy, and more!
Suppose you're visiting Slashdot, and your web browser constantly hangs due to the crappy tag formatting code on the front page. If there was a way to modify and replace that code, you could fix it, and distribute the fix to other slashdotters.
Can you say fucking stupid, kids? Humans are not machines (at least not yet), they have bad days and bad weeks and some have bad decades (imagine your child dies). Evaluating them through "rigorous" methodological measures is pure idiocy.
From the company's point of view, if you're having a "bad decade", you're probably not of much use to them. As dehumanizing and cynical as it sounds, companies exist to make a profit, and if you're not serving that function or are not somehow contractually bound to them, then you don't have a place there. Don't jump on a sound bite and delegate the whole process to "fucking stupid". They're not just blindly looking at email patterns. Done sensibly, I would rather have a series of thoughtful quantitative measures (I said thoughtful, not KLOCs or email centrality) to back up my job performance than have my future depend on the whims of a possibly irrational boss. How many people have lost jobs because their boss was having a bad day, or a bad week?
I think any technical writer that isn't scared away by the syntax of LaTeX should be able to master "svn update", and "svn commit". And if that's too much, there are plugins for Windows, Mac, and Linux that integrate Subversion with the normal file browser.
Exactly. Our lab submitted a collaborative paper that involved five people editing the document. SVN was more than enough for our needs, and all you need is an Apache install running somewhere. It literally was painless because of SVN, just make sure everyone types in descriptive log messages. Bonus: the commit logs can help you determine the order of authors:)
On the frontend, the best SVN clients I've used are TortoiseSVN for Windows and RapidSVN for Linux. As I said, couldn't be happier with the setup. IMO, any more functionality is absolutely unnecessary.
Firstly, you've got Goedel incompleteness to worry about (which stems from statements that are fundamentally ambiguous as to their interpretation, such as "this statement is false"). Secondly, languages are there for people to communicate with, and people seem to prefer ambiguity
What exactly does Godel's theorem have to do with what you just said? The incompleteness theorem deals with axiomatized systems. This leads me to think that you might be confusing the popular meaning of "language" with the mathematical definition. People (at least normal people) do not communicate with mathematical languages.
Was that Yamaha pickup any good? I remember seeing it in a music store some years ago and wondering if it actually worked. It seems like a relatively trivial problem to solve if you can attach a pickup, as they did. Transforming the audio into some kind of useful MIDI information in realtime seems horribly difficult but also doable... just not without immense processing power.
I tried it in a music store and it seemed to be capturing basic pitch and velocity pretty well for each string, which is pretty neat. I don't imagine that it would take too much computational power, since the underlying computation is basically just doing an FFT on the audio signal and picking up the fundamental frequency. That translates quite well into a MIDI note on/note off command. FFTs can be done really efficiently in hardware I believe.
Wasn't this idea thrown around when the MIDI interface was created, 20+ years ago?
But of course...we're now coming full circle when people realize that they can do this Guitar Hero stuff on real guitars, and without the buttons either. Turns out, guitars actually make music all by themselves (with possibly just an amp)!!! But I digress--for those interested in music, here are the relevant things that precede this in time and in awesomeness:
Yamaha MIDI pickup for guitars -- turns the note you're playing into a MIDI note that can then control a synth on your computer. And I don't mean the 80's era crappy "synth" sounds, modern sound synthesis engines are INCREDIBLY realistic.
Alternatively, you could skip Guitar Hero completely and do the following:
Download a Guitar Pro file of your favorite song / use TuxGuitar to import a MIDI file -- the modern software synthesizer sounds nothing like what you might remember from the 80s. Bass and drums are very realistic.
Mute the guitar track in Guitar Pro/TuxGuitar, and play along with full drum/bass accompaniment. The software even scrolls as you play along.
If you like eye candy, route your sound card's output into Winamp and run a visualization.
This is the method I use and it's incredibly satisfying if you don't happen to have a band lying around. Plus it also lets you choose which part you want to play along with, speed up/slow down the song. Sure there aren't any vocals, but it's still mucho fun and way better for impressing people with.
Given classic random theory, given a series of 50/50 type decisions, out of 32 people, one person will be completely screwed and one person will win every time. For larger data sets, the lucky runs are only longer.
This is absolutely meaningless, a hodgepodge of statistical fallacies and vagueness. Given a "series of 50/50 type decisions", assuming each decision is independent, then it is absolutely not necessary that "in larger data sets, the lucky runs are only longer". If each decision is a coin flip, then there is no "memory" of prior decisions, so the probability of getting a long "lucky run" is exactly the same regardless of the size of the data set. The length of a "lucky run" is independent of the size of the data set.
In the context of Gladwell, you could say that "genius" is the independent co-incidence of (1) having innate ability and cultural preparation and (2) the right circumstances presenting themselves. One or both of those events have very low probability, so when they coincide and create a Bill Gates, we get an "outlier" relative to the life experience of the vast majority of people.
The point of authentication is to get accountability, not to get instant filtering. If a spammer is using a fake certificate, that certificate can be blacklisted.
You might be confusing business and personal certificates. Individuals, presumably, will have personal certificates that are usually free and therefore plentiful/renewable. If you're talking about business certificates for the e-mail provider, then read on:
If random-mail-provider.com is doing good at stopping fake accounts, I could whitelist them as well. And when you would send your twin mail via a good email provider it would arrive just fine.
Regardless of whether you're whitelisting or blacklisting, the "amount" of spam sent from a provider could be minuscule compared to their total volume. Thus, there really isn't a good way to tell a "good email provider". Even Gmail is used to send spam, although they may be quicker at shutting those accounts than others.
There's a checkbox on the "your spam solution will not work..." form that circulates on/. about how "ideas like yours are easy to come up with, yet none have been shown to be practical".
Let's replace it with something that authentifies sender and receiver properly, and that allows for efficient transmission of binary data.
Sigh...it's so tiring to hear people on/. say things like "it's a technological problem" about spam. Do you know how easy it is to get a personal digital certificate from Thawte? Fill out a few forms, download your PKCS certificate. What's to stop your sooper-dooper anti-spam system if you can authenticate a spammer? Remember, if you can legitimately receive an e-mail message from ME (a stranger to you, presumably), you haven't "solved" spam. If you can't legitimately receive an e-mail message from me, I can't tell you that I'm your long-lost twin brother (i.e. your email system is then useless).
Didn't Apple release System7 about two decades ago? http://en.wikipedia.org/wiki/System_7 [wikipedia.org] So... Microsoft is a bit behind the times, huh?
This is voodoo science. And I don't mean the LHC experiments.
It's not science, it's just probability. It's senseless to try to assess any statistical estimates *themselves* based on Physics, just the probability that they could be wrong based on some very broad assumptions. Specifically, any estimate is arrived at by a chain (rather, DAG) of logic. What you CAN estimate is the probability that any Physics-oriented estimate is based on incorrect assumptions, by (presumably) analyzing that chain of reasoning down to first principles and assuming that a "logic error" might have been made at any point. I hope that the authors aren't taking it further than this, in which case, this is statistical masturbation.
which is a world record for a quad core CPU and they dished out and astonishing 45,474 3DMark05 score!
Watch the video below to see how it was done and how history was made:
Truly PHENOMenal, but I can't help but (cynically, I admit) think about how history inevitably mocks overclockers. Cue back to the 90s and a headline might have read "486 overclocked to 500Mhz -- history has been made!". Like Ozymandias, nothing beside remains...
You know, they could have extremely tame erotic websites to cater to kids who are interested. Probably like softcore Playboy pics or something.
This is the funniest thing I think I've ever read on Slashdot. You, sir, seem to live in some reality where a controversial but possibly reasonable argument about pornography and children will be taken seriously. Anyway, let's assume that such a proposal does make it to the general public. In the "real" world, "tame erotic websites" will have the same connotation as marijuana being a "gateway drug": (a) that it's addictive and harmful (b) it leads to "harder" stuff (in both weed and porn contexts) and (c) it will ruin the children, even though adults enjoy it responsibly everyday.
Nope, there's a lot more that has to do with creative problem modeling and problem solving, and nothing to do with "making things fast". To name a few random areas: statistical machine learning (induce from given data), data mining/pattern analysis, graphics (such as new algorithms for e.g. HDR photography), graph theory (such as new methods for analyzing social networks, chemical compounds, gene regulatory networks), and it's pretty obvious I could go on if I wasn't so sleep deprived. Certainly, a large part of these areas is improving existing algorithms, but there is a significant amount of creative new stuff coming out too.
But I have seen a lot of bad code the follows all the rules to a point where it losses it usefulness, as it has became "too organized" to a point it lost flexibility and readability. I have seen other code that seems to break all the rules but somehow it is rather easy to maintain, and quite workable and performs well and fast.
You seem to be referring to software engineering, not computer science. There is a LOT more to computer science than "data layers" and "UI layers" and "best practices", which belong to the realm of -- you guessed it -- software engineering. There are no "rules" in computer science, just theorems that get translated into (hopefully) better/more efficient algorithms.
So Yes I would say "skips resource-sucking parts of the operating system and focuses on the browser".
Makeing a real product that is useful is different then those thought exercise in Computer Science. Modern business needs and user requirements conflict with intellectual purity.
I wonder if any of us in computer science, with our bastions of "intellectual purity", could possibly address a business problem such as this -- how can we keep only those parts of an OS that are required for specific tasks and still be intellectually pure and stable too??
I agree with the principle of what you're saying, but concluding with a silly troll about computer science makes you seem like someone who flunked his OS class.
Choose a 25 year old topic (for example, a Pacmangame), reinvent it using lots of buzzwords such as swarm, hive, collective, competitive, but secretly just program a system using some generic rules, and a gradient descent algorithm that will force those generic rules to conform to the behaviour we wanted in the first place.
Yes! Thank you! I was going to say that using an EA just means that you're setting up a function space over all possible games of a certain type. In essence, you've already defined the parameters of "possible games" -- the EA is just stumbling through that parameter space -- hardly "creativity". And that goes for a bunch of other approaches "biologically-inspired" search algorithms that all essentially boil down to random search + heuristics.
and to be honest, its not like WPA is uncrackable, and if someone knows how to get through WEP they can probably figure how to get through WPA in the end.
This is incorrect. WEP has a well-known attack that uses statistical properties of captured packets to limit the search space of the brute-force search. With enough captured packets, it takes under a few minutes to crack a WEP key. WPA does not have this vulnerability, although some variants of WPA are still less secure than other. In other words, all you have to do to crack WEP encryption is put most cheap $20 wireless cards into "monitor" mode, capture WEP encrypted packets and with enough packets, crack the key in under a few minutes. Again, WPA is not perfect, but does not have such an easy statistical attack either.
All it takes is for you to watch one long YouTube video or transfer a relatively large attachment to capture enough WEP packets. I had included a link with the original story that was removed by Soulskill, here. You don't even have to be "hacker smart" to crack WEP, just know enough to read a few articles, run linux and download a few pieces of software.
What's worse is that where I live, AT&T ships 2WIRE routers that only have WEP encryption with 40bit keys and no MAC filters. This almost guarantees that AT&T DSL can be easily cracked by someone with a laptop and the (freely downloadable) aircrack suite. That's something to think about the next time you're downloading a torrent... So, monitor your routers access log, set up a MAC filter, use better encryption if possible, or at least longer WEP keys -- nothing's perfect, but it each puts up one more obstacle.
Slashdot Mirror
My response: cry me a river, and congrats to the grad students for their innovative work in the field of distributed communications.
I'd pause before calling this innovative. It doesn't really take much to encrypt data, chop it up and stash it on MediaWiki sites -- either in theory or in practice. If you want something "innovative" in the same vein, I'd vote for the guy who wrote the device driver that lets you use GMail as a drive (spawning many copies). Sure it isn't "distributed", but you could set up multiple GMail accounts to handle the contents of your drive. Clogging up other people's wikis is d**k at worst (and possibly a violation of the CFAA), and really not too much of a security threat at best ("oh? my disk is full? hmm...just dump this spammy user account, or restore the last backup, and password protect the whole business.").
What these grad students have done is demonstrate that open mediawiki setups can be spammed. Whee.
And why exactly are we throwing millions of page views at these racist fucks???
Sigh...the dangers of news aggregation, which is probably how the submitter stumbled upon this piece of bigoted tripe. Somewhere, a thousand page ranking algorithms just went mad seeing these two ostensibly different areas of the Net link to each other.
Can we please be careful not to link this sort of website to slashdot? Somebody please think of the page ranking algorithms!! For those who don't want to click through, this is a summary from the site in the second link of the story (amren.com).
* The Dangers of Diversity, Part II, Editor Jared Taylor continues a multi-part examination of the effects of diversity by taking a look at just what happens when races mix. This segment provides additional examples of the violence and conflict that ensues when different racial groups are forced to mingle in public schools, and also in prisons.
* In Three Race Murders in Seattle, journalist Nicholas Stix reports on the murders of three white men, Edward Scott McMichael, James Paroline, and Kristopher Klime, by blacks. Mr. Stix shows how the media (and government) consistently downplayed the racial angle in these cases.
* In A Voice For Our People, Peter Bradley reviews Frank Borzellieri's new book Lynched: A Conservative's Life on a New York City School Board, which documents the author's one-man crusade to keep Western literature and values alive in increasingly "multicultural" New York City. Mr. Bradley holds him up as example of what a difference one determined man can make in his community.
* Plus, paying the price for insulting Obama supporters, the GOP takes its message to "hip hop settings," Attorney General Holder on the "nation of cowards," the US military becoming another foreign legion, another monkey cartoon controversy, and more!
or PhD CS program, so I'll have another two or more years after that
I'd bet on the "or more" if I were you.
There's something to be said for conventional data centers: They're rather hard to load onto a truck and drive off with.
Yes, but imagine the bandwidth!
Plus 5 to the 'nostalgic games' index of computing power that is cheaply available.
Suppose you're visiting Slashdot, and your web browser constantly hangs due to the crappy tag formatting code on the front page. If there was a way to modify and replace that code, you could fix it, and distribute the fix to other slashdotters.
It's called Greasemonkey.
Can you say fucking stupid, kids? Humans are not machines (at least not yet), they have bad days and bad weeks and some have bad decades (imagine your child dies). Evaluating them through "rigorous" methodological measures is pure idiocy.
From the company's point of view, if you're having a "bad decade", you're probably not of much use to them. As dehumanizing and cynical as it sounds, companies exist to make a profit, and if you're not serving that function or are not somehow contractually bound to them, then you don't have a place there. Don't jump on a sound bite and delegate the whole process to "fucking stupid". They're not just blindly looking at email patterns. Done sensibly, I would rather have a series of thoughtful quantitative measures (I said thoughtful, not KLOCs or email centrality) to back up my job performance than have my future depend on the whims of a possibly irrational boss. How many people have lost jobs because their boss was having a bad day, or a bad week?
I think any technical writer that isn't scared away by the syntax of LaTeX should be able to master "svn update", and "svn commit". And if that's too much, there are plugins for Windows, Mac, and Linux that integrate Subversion with the normal file browser.
Exactly. Our lab submitted a collaborative paper that involved five people editing the document. SVN was more than enough for our needs, and all you need is an Apache install running somewhere. It literally was painless because of SVN, just make sure everyone types in descriptive log messages. Bonus: the commit logs can help you determine the order of authors :)
On the frontend, the best SVN clients I've used are TortoiseSVN for Windows and RapidSVN for Linux. As I said, couldn't be happier with the setup. IMO, any more functionality is absolutely unnecessary.
Firstly, you've got Goedel incompleteness to worry about (which stems from statements that are fundamentally ambiguous as to their interpretation, such as "this statement is false"). Secondly, languages are there for people to communicate with, and people seem to prefer ambiguity
What exactly does Godel's theorem have to do with what you just said? The incompleteness theorem deals with axiomatized systems. This leads me to think that you might be confusing the popular meaning of "language" with the mathematical definition. People (at least normal people) do not communicate with mathematical languages.
Was that Yamaha pickup any good? I remember seeing it in a music store some years ago and wondering if it actually worked. It seems like a relatively trivial problem to solve if you can attach a pickup, as they did. Transforming the audio into some kind of useful MIDI information in realtime seems horribly difficult but also doable... just not without immense processing power.
I tried it in a music store and it seemed to be capturing basic pitch and velocity pretty well for each string, which is pretty neat. I don't imagine that it would take too much computational power, since the underlying computation is basically just doing an FFT on the audio signal and picking up the fundamental frequency. That translates quite well into a MIDI note on/note off command. FFTs can be done really efficiently in hardware I believe.
Wasn't this idea thrown around when the MIDI interface was created, 20+ years ago?
But of course...we're now coming full circle when people realize that they can do this Guitar Hero stuff on real guitars, and without the buttons either. Turns out, guitars actually make music all by themselves (with possibly just an amp)!!! But I digress--for those interested in music, here are the relevant things that precede this in time and in awesomeness:
Yamaha MIDI pickup for guitars -- turns the note you're playing into a MIDI note that can then control a synth on your computer. And I don't mean the 80's era crappy "synth" sounds, modern sound synthesis engines are INCREDIBLY realistic.
Alternatively, you could skip Guitar Hero completely and do the following:
This is the method I use and it's incredibly satisfying if you don't happen to have a band lying around. Plus it also lets you choose which part you want to play along with, speed up/slow down the song. Sure there aren't any vocals, but it's still mucho fun and way better for impressing people with.
Given classic random theory, given a series of 50/50 type decisions, out of 32 people, one person will be completely screwed and one person will win every time. For larger data sets, the lucky runs are only longer.
This is absolutely meaningless, a hodgepodge of statistical fallacies and vagueness. Given a "series of 50/50 type decisions", assuming each decision is independent, then it is absolutely not necessary that "in larger data sets, the lucky runs are only longer". If each decision is a coin flip, then there is no "memory" of prior decisions, so the probability of getting a long "lucky run" is exactly the same regardless of the size of the data set. The length of a "lucky run" is independent of the size of the data set.
In the context of Gladwell, you could say that "genius" is the independent co-incidence of (1) having innate ability and cultural preparation and (2) the right circumstances presenting themselves. One or both of those events have very low probability, so when they coincide and create a Bill Gates, we get an "outlier" relative to the life experience of the vast majority of people.
The point of authentication is to get accountability, not to get instant filtering. If a spammer is using a fake certificate, that certificate can be blacklisted.
You might be confusing business and personal certificates. Individuals, presumably, will have personal certificates that are usually free and therefore plentiful/renewable. If you're talking about business certificates for the e-mail provider, then read on:
If random-mail-provider.com is doing good at stopping fake accounts, I could whitelist them as well. And when you would send your twin mail via a good email provider it would arrive just fine.
Regardless of whether you're whitelisting or blacklisting, the "amount" of spam sent from a provider could be minuscule compared to their total volume. Thus, there really isn't a good way to tell a "good email provider". Even Gmail is used to send spam, although they may be quicker at shutting those accounts than others.
There's a checkbox on the "your spam solution will not work..." form that circulates on /. about how "ideas like yours are easy to come up with, yet none have been shown to be practical".
Let's replace it with something that authentifies sender and receiver properly, and that allows for efficient transmission of binary data.
Sigh...it's so tiring to hear people on /. say things like "it's a technological problem" about spam. Do you know how easy it is to get a personal digital certificate from Thawte? Fill out a few forms, download your PKCS certificate. What's to stop your sooper-dooper anti-spam system if you can authenticate a spammer? Remember, if you can legitimately receive an e-mail message from ME (a stranger to you, presumably), you haven't "solved" spam. If you can't legitimately receive an e-mail message from me, I can't tell you that I'm your long-lost twin brother (i.e. your email system is then useless).
Didn't Apple release System7 about two decades ago? http://en.wikipedia.org/wiki/System_7 [wikipedia.org] So... Microsoft is a bit behind the times, huh?
Congratulations, you're fucking retarded.
Agreed.
This is voodoo science. And I don't mean the LHC experiments.
It's not science, it's just probability. It's senseless to try to assess any statistical estimates *themselves* based on Physics, just the probability that they could be wrong based on some very broad assumptions. Specifically, any estimate is arrived at by a chain (rather, DAG) of logic. What you CAN estimate is the probability that any Physics-oriented estimate is based on incorrect assumptions, by (presumably) analyzing that chain of reasoning down to first principles and assuming that a "logic error" might have been made at any point. I hope that the authors aren't taking it further than this, in which case, this is statistical masturbation.
which is a world record for a quad core CPU and they dished out and astonishing 45,474 3DMark05 score! Watch the video below to see how it was done and how history was made:
Truly PHENOMenal, but I can't help but (cynically, I admit) think about how history inevitably mocks overclockers. Cue back to the 90s and a headline might have read "486 overclocked to 500Mhz -- history has been made!". Like Ozymandias, nothing beside remains...
You know, they could have extremely tame erotic websites to cater to kids who are interested. Probably like softcore Playboy pics or something.
This is the funniest thing I think I've ever read on Slashdot. You, sir, seem to live in some reality where a controversial but possibly reasonable argument about pornography and children will be taken seriously. Anyway, let's assume that such a proposal does make it to the general public. In the "real" world, "tame erotic websites" will have the same connotation as marijuana being a "gateway drug": (a) that it's addictive and harmful (b) it leads to "harder" stuff (in both weed and porn contexts) and (c) it will ruin the children, even though adults enjoy it responsibly everyday.
CS is about making things fast, only that.
Nope, there's a lot more that has to do with creative problem modeling and problem solving, and nothing to do with "making things fast". To name a few random areas: statistical machine learning (induce from given data), data mining/pattern analysis, graphics (such as new algorithms for e.g. HDR photography), graph theory (such as new methods for analyzing social networks, chemical compounds, gene regulatory networks), and it's pretty obvious I could go on if I wasn't so sleep deprived. Certainly, a large part of these areas is improving existing algorithms, but there is a significant amount of creative new stuff coming out too.
But I have seen a lot of bad code the follows all the rules to a point where it losses it usefulness, as it has became "too organized" to a point it lost flexibility and readability. I have seen other code that seems to break all the rules but somehow it is rather easy to maintain, and quite workable and performs well and fast.
You seem to be referring to software engineering, not computer science. There is a LOT more to computer science than "data layers" and "UI layers" and "best practices", which belong to the realm of -- you guessed it -- software engineering. There are no "rules" in computer science, just theorems that get translated into (hopefully) better/more efficient algorithms.
So Yes I would say "skips resource-sucking parts of the operating system and focuses on the browser". Makeing a real product that is useful is different then those thought exercise in Computer Science. Modern business needs and user requirements conflict with intellectual purity.
I wonder if any of us in computer science, with our bastions of "intellectual purity", could possibly address a business problem such as this -- how can we keep only those parts of an OS that are required for specific tasks and still be intellectually pure and stable too??
I agree with the principle of what you're saying, but concluding with a silly troll about computer science makes you seem like someone who flunked his OS class.
Choose a 25 year old topic (for example, a Pacmangame), reinvent it using lots of buzzwords such as swarm, hive, collective, competitive, but secretly just program a system using some generic rules, and a gradient descent algorithm that will force those generic rules to conform to the behaviour we wanted in the first place.
Yes! Thank you! I was going to say that using an EA just means that you're setting up a function space over all possible games of a certain type. In essence, you've already defined the parameters of "possible games" -- the EA is just stumbling through that parameter space -- hardly "creativity". And that goes for a bunch of other approaches "biologically-inspired" search algorithms that all essentially boil down to random search + heuristics.
...where he is hoping to be inspired by the everyday grind of scientists
What, reading papers, crunching numbers, writing papers and browsing Slashdot? Hmm...think I've already read that story.
and to be honest, its not like WPA is uncrackable, and if someone knows how to get through WEP they can probably figure how to get through WPA in the end.
This is incorrect. WEP has a well-known attack that uses statistical properties of captured packets to limit the search space of the brute-force search. With enough captured packets, it takes under a few minutes to crack a WEP key. WPA does not have this vulnerability, although some variants of WPA are still less secure than other. In other words, all you have to do to crack WEP encryption is put most cheap $20 wireless cards into "monitor" mode, capture WEP encrypted packets and with enough packets, crack the key in under a few minutes. Again, WPA is not perfect, but does not have such an easy statistical attack either.
All it takes is for you to watch one long YouTube video or transfer a relatively large attachment to capture enough WEP packets. I had included a link with the original story that was removed by Soulskill, here. You don't even have to be "hacker smart" to crack WEP, just know enough to read a few articles, run linux and download a few pieces of software.
What's worse is that where I live, AT&T ships 2WIRE routers that only have WEP encryption with 40bit keys and no MAC filters. This almost guarantees that AT&T DSL can be easily cracked by someone with a laptop and the (freely downloadable) aircrack suite. That's something to think about the next time you're downloading a torrent... So, monitor your routers access log, set up a MAC filter, use better encryption if possible, or at least longer WEP keys -- nothing's perfect, but it each puts up one more obstacle.
but having a lot of people dislike something as in the case of Vista means it's pretty damn popular.
I think the word you're looking for is notorious, not popular in its common usage.