big trend of people with relatively crappy ML research gussying it up with some sexy applications (usually bio-related) and then publishing it in a general-readership science journal
Mark Newman! PNAS! The list goes on...generally seem to be people from field X trying to stuff from field Y (where Y is often ML/statistics/algorithms, and X != math or CS).
However, I am concerned that putting developers around a table could potentially be distracting consequently diminishing productivity by increasing coding errors.
I agree with parent, and have you considered that developers whose code quality is affected by seating arrangements relative to other developers might not be...um, the best developers? Otherwise, I'd say you might be overthinking the issue.
Great, so they want to redesign the Internet because people don't want to learn how to identify a phishing site and can't understand that giving your account numbers to unverifiable strangers is a bad idea?
Oh please, I think Sony put an end to the delusion that only grandmas and morons are susceptible to phishing or malware. Allow me to give you an example which most people here won't be able to do detect instantaneously: zero-day exploit in Flash + rootkit + trojan. I run a tight ship like the next nerd, but my AV software still flags trojans that somehow make it onto my system from time to time, and those are only the ones that it CAN detect.
And yes, there are zealots who will undoubtedly say things like "Flash is for suckers" or "what do you expect with Windows?", but these people should consider the fact that (a) not everyone lives in caves, and (b) some people just have more important things to worry about, like losing their homes.
One of the company's key areas of expertise are in "data mining technologies".
Do you really *think* they're THAT concerned with your security, given the situation?
Look, we as nerds must STOP treating "data mining" like an epithet, or at least a scarlet letter on one's resume. The term has been abused by the popular media in connection with the NSA's wiretapping, but people tend to overlook the fact that "data mining" is just a bunch of algorithms to find statistical patterns in different kinds of data. When it's referred to as "exploratory data analysis", no one seems to mind. When it's referred to as simply "applied statistics", no one seems to mind. Read the statement by ACM's data mining special interest group, SIGKDD.
That said, I completely agree with you -- of course Facebook is interested in mining the social graph and f***ing it for all its worth. They're a for-profit company whose only asset is detailed information about people and their interactions. Why is anyone shocked that they don't want to make the world a better place, and would rather become very rich instead off their only asset. For a capitalist country, a lot of nerds in the US seem to have rose-colored glasses on.
It's a relatively interesting article, especially the bit about the Micro SIM it supposedly uses, which is not in use anywhere in the US right now. But in any case, one can't help thinking that the reviewer at Gizmodo would pay good money to vigorously and servilely pleasure Steve Jobs. From TFA:
The seams are perhaps the most surprising aspect of the new design. They don't seem to respond to any aesthetic criteria and, in terms of function, we can't adventure any explanation. But they don't look bad. In fact, the whole effect seems good, like something you will find in a Braun product from the 70s.
Only can a true fanboy turn the phrase "like a Braun product from the 70s" into a compliment. Because we all want to show off our new iEpilators.
You could fund a manned Mars mission (pessimistic estimated total cost: $100 billion) with a 3% cut in the US military budget for ten years.
You could pay for massive upgrades to child protective services, social security, medicare, etc. with $100 billion. You could put a million pedophile priests in jail for $100 billion. You could reinvigorate Detroit and create tens of thousands of jobs for $100 billion.
The point is that you could do a LOT of things with "just a small cut in the military budget", but it wouldn't sit well with the electorate. Obama already takes enough shit for being "soft on terrorists" and "elitist". I doubt he'd want to completely botch his re-election with a snooty re-allocation of military funds ("purtecctt amurreriicaa") to the space program ("scieencee and la dee daa").
Currently, fewer than 1% of search queries are affected by the site speed signal in our implementation and the signal for site speed only applies for visitors searching in English on Google.com at this point.
The main site serves visitors from the US. Thus, measuring speeds from multiple locations around the US is probably the best thing to do. They're presumably measuring speed from all their datacenters (their crawlers are likely to be distributed across the country (and world), so recording the average speed over multiple crawls would be a good approximation when you're dealing with the scale of Google and the Web).
Did anyone think about the poor sod whose job it was to burn the mice? I mean, it's difficult to catch mice that have just unsuccessfully rushed into burning buildings, or set themselves on fire by accident.
Then again, I know a guy who works in tissue engineering whose job is to "harvest" mice, as he calls it. Keeps their heads in a jar above his desk. Apparently, they bob around all day with a smiling expression.
You have to get your hands dirty for a lot of science...
Maybe you should actually, you know,...use Linux before you attempt to troll about security.
What's even worse than with Windows is that since 'rm' is just a normal binary the PDF can launch that, and if you run as root privileges, just issue a command like "rm -rf/". If you don't run as root, then for example Ubuntu should give you the sudo box to input password to. This of course being just one of the examples it could do. Remember that most malware doesn't even need root access to function.
Nobody uses the root account in Linux for everyday activity. In Ubuntu, root login is even disabled by default (you have to sudo). So no worries about the system in general. Although it's pretty devastating to issue a "rm -rf ~" to delete the user's home directory, it's on par with Windows. Then you say that most malware doesn't even need root access to function, but on all the millions of XP boxes out there, it's already given root access by default.
Another reason why it would be even more serious on Linux is the way you can pipe commands and how most systems come pre-packaged with a ton of little utility apps. You can create the whole malware with a series of commands, or wget a bash script from the internet and start that to hide even more malware in the system.
Windows has a pipe function too, in addition to being able to zoink your whole file system with a simple "del". It also comes with ftp and telnet, which are handy replacements for wget. In short telnet+response file = download an.exe from the web = any sort of functionality you might want using Unix command line tools.
Not only that, only nerds would get excited about a STOCK response from HR about discrimination, and then post it on slashdot. Not trying to troll here, but HR folks aren't lawyers, and are trained to be extremely careful when it comes to possible litigation. In short, even the bad publicity makes it worthwhile for HR to apologize to this "Jedi" instead of saying something like "we only recognize jedis on active duty, with working light sabers".
Initially, the idea of "code bubbles" sounds intuitive -- isolated, self-contained, easily testable pieces of code, with well-defined inputs and outputs. Then you could build a complex program by stringing these bubbles together (in theory, anyway).
Then TFS mentions something as banal as "wrapping long lines of code"....and my bubble bursts.
This seems to be a non-story, unless this is the first time these financial ties have been revealed between bit torrent researchers and ISPs.
This is not so much about calling the researchers' methods and findings into question as the ISPs motivation for funding the research. As far as I can tell, the research seems to be sound and pretty neat. The question is WHY are ISPs interested in FUNDING this sort of research?
One possibility that the submitter didn't consider is the fact that many researchers list their funding sources on all published papers, regardless of whether the funding was given to fund that specific project. So it could be that ISPs generally fund this particular research group in any case, and they happened to put out a paper that analyzes BT. In other words, there might not be anything sinister going on.
Don't get me wrong, it would be incredibly cost, labor and time expensive, and require real computer scientists, but it is certainly possible.
Speaking as a "real" computer scientist, I think you might have underestimated the time requirement. Most problems in automatic verification are either undecidable, or intractable.
Never mind the use of 'professional orange and purple'. From TFA:
"We're drawn to Light because it denotes both warmth and clarity, and intrigued by the idea that 'light' is a good value in software. Good software is 'light' in the sense that it uses your resources efficiently, runs quickly, and can easily be reshaped as needed," the design documentation says. "Visually, light is beautiful, light is ethereal, light brings clarity and comfort."
Why do "design documents" always have to be so banal? I mean, "visually, light is beautiful"?!!? Seriously?
CNN was actually discussing this in their reporting yesterday. They were very clear about this being done by bad folks, not the web in general, and the things people should look out for.
Given the amount of money and time CNN has invested in gadgetry and web-whizbangery, over any other network by far, they would be fools to (a) draw too much attention to this unless it becomes a serious problem, and (b) NOT cover it calmly, and without resorting to fear-mongering. I suppose that answers GP's question, at least as far as CNN. Now, Murdoch-owned Fox News is a different story altogether....
Seeing how 66.67% of the time I am either sleeping at home or at work it shouldn't be too hard to fill the other 27% with commute/grocery shopping.
You're not too far off. I worked at the research wing of a phone company, and I can tell you that "tracking" a person using a cell tower is pretty coarse, even in urban areas. Given that most people go to work on weekdays, I'd say that a lot of your "movement" could be predicted on this level by just predicting your average movement. Add in a weekday/weekend variation, and 93% is hardly surprising.
This isn't even one of those "well duh, in RETROSPECT everything is obvious" studies -- anyone who has ever worked with CDR (mobile phone) data knows that this is pretty obvious even before running the experiments.
And for the people who bring up the MIT Reality Mining experiment, keep in mind that they tracked about 100 *individuals*, all of whom were MIT students with pretty regular routines.
Isn't it obvious that the fear of something will have an impact even on the simplest things where something relative to that fear is involved ?
Yes, but I think what this study was trying to test was how basic the task has to be for the fear response to have a measurable effect. Turns out, pretty damn basic.
If they achieve VGA resolution, it's a steady road to full vision for the blind. I'm more interested in, at this point, exceeding human abilities. Think of the case of HDR imaging -- we currently don't have monitors (most of us at least) that are high dynamic range themselves, so images have to be "tone-mapped" to the dynamic range of our monitors, which often results in those ridiculously sharp but somewhat "unrealistic" pictures you see on Flickr.
It would be cool if, say, the IR spectrum or just more dynamic range in the visible spectrum could be tone-mapped to human perception in this way, resulting in perceptually sharper images by way of a direct retinal implant.
Try Mendeley. They're still pretty new, but very promising with their desktop client for Linux/Mac/Win in addition to the web interface. They also sync perfectly with Zotero and CiteULike, which makes migration easier. You can annotate PDFs directly in the desktop, but I think only the latest beta build has support for sync'ing the annotations across multiple computers. I'm hopeful for them -- it's definitely one of the most promising Ref manager systems I've seen (oh yes, they also support Bibtex,Endnote,Refworks formats heavily)
Certainly, Apple will reject the app and Opera knows it. Maybe Opera tries to strengthen Apple's "Evil Empire" image and deal with it with the help of EU (just like they did with Microsoft recently).
My first thought on reading the summary was "where's the leverage?" Either Opera is talking right through their own asses, or they have some serious leverage -- certainly more than just bad PR (which Apple seems immune to anyway). They've invested time and manpower in this project, one can only hope that someone's cojones over at Apple are in a vice grip, and that we will all soon enjoy the big red O on our iPhones.
This might be interesting if they manage to get the privacy thing right. If they don't, I see it as a disaster. I use gmail to communicate with a much wider audience than Facebook. If somehow they managed to let me easily and effectively segment users into different groups, with STRONG WALLS between groups, then it might be interesting.
Although it would take quite a few HCI PhDs to figure out how to do it all without cluttering an already cluttery gmail UI.
If it's so freakin' open please tell me why I still need to have apps signed on my Nokia 6220 classic and will do for the foreseeable future unless I'm willing to try risky hacks.
I'll raise you an anecdote. I just bought a Nokia E63, new and unlocked with a full US warranty for $189 from Newegg, and it's one of the best phones I've ever owned. You simply go to the application manager menu, and for the option that says "Install only signed apps", select "No". It's that simple. I just installed an unsigned FTP client, so now I don't even need Nokia's atrocious PC Suite for syncing.
Slashdot Mirror
big trend of people with relatively crappy ML research gussying it up with some sexy applications (usually bio-related) and then publishing it in a general-readership science journal
Mark Newman! PNAS! The list goes on...generally seem to be people from field X trying to stuff from field Y (where Y is often ML/statistics/algorithms, and X != math or CS).
However, I am concerned that putting developers around a table could potentially be distracting consequently diminishing productivity by increasing coding errors.
I agree with parent, and have you considered that developers whose code quality is affected by seating arrangements relative to other developers might not be...um, the best developers? Otherwise, I'd say you might be overthinking the issue.
Great, so they want to redesign the Internet because people don't want to learn how to identify a phishing site and can't understand that giving your account numbers to unverifiable strangers is a bad idea?
Oh please, I think Sony put an end to the delusion that only grandmas and morons are susceptible to phishing or malware. Allow me to give you an example which most people here won't be able to do detect instantaneously: zero-day exploit in Flash + rootkit + trojan. I run a tight ship like the next nerd, but my AV software still flags trojans that somehow make it onto my system from time to time, and those are only the ones that it CAN detect.
And yes, there are zealots who will undoubtedly say things like "Flash is for suckers" or "what do you expect with Windows?", but these people should consider the fact that (a) not everyone lives in caves, and (b) some people just have more important things to worry about, like losing their homes.
One of the company's key areas of expertise are in "data mining technologies". Do you really *think* they're THAT concerned with your security, given the situation?
Look, we as nerds must STOP treating "data mining" like an epithet, or at least a scarlet letter on one's resume. The term has been abused by the popular media in connection with the NSA's wiretapping, but people tend to overlook the fact that "data mining" is just a bunch of algorithms to find statistical patterns in different kinds of data. When it's referred to as "exploratory data analysis", no one seems to mind. When it's referred to as simply "applied statistics", no one seems to mind. Read the statement by ACM's data mining special interest group, SIGKDD.
That said, I completely agree with you -- of course Facebook is interested in mining the social graph and f***ing it for all its worth. They're a for-profit company whose only asset is detailed information about people and their interactions. Why is anyone shocked that they don't want to make the world a better place, and would rather become very rich instead off their only asset. For a capitalist country, a lot of nerds in the US seem to have rose-colored glasses on.
The seams are perhaps the most surprising aspect of the new design. They don't seem to respond to any aesthetic criteria and, in terms of function, we can't adventure any explanation. But they don't look bad. In fact, the whole effect seems good, like something you will find in a Braun product from the 70s.
Only can a true fanboy turn the phrase "like a Braun product from the 70s" into a compliment. Because we all want to show off our new iEpilators.
You could fund a manned Mars mission (pessimistic estimated total cost: $100 billion) with a 3% cut in the US military budget for ten years.
You could pay for massive upgrades to child protective services, social security, medicare, etc. with $100 billion. You could put a million pedophile priests in jail for $100 billion. You could reinvigorate Detroit and create tens of thousands of jobs for $100 billion.
The point is that you could do a LOT of things with "just a small cut in the military budget", but it wouldn't sit well with the electorate. Obama already takes enough shit for being "soft on terrorists" and "elitist". I doubt he'd want to completely botch his re-election with a snooty re-allocation of military funds ("purtecctt amurreriicaa") to the space program ("scieencee and la dee daa").
Currently, fewer than 1% of search queries are affected by the site speed signal in our implementation and the signal for site speed only applies for visitors searching in English on Google.com at this point.
The main site serves visitors from the US. Thus, measuring speeds from multiple locations around the US is probably the best thing to do. They're presumably measuring speed from all their datacenters (their crawlers are likely to be distributed across the country (and world), so recording the average speed over multiple crawls would be a good approximation when you're dealing with the scale of Google and the Web).
Then again, I know a guy who works in tissue engineering whose job is to "harvest" mice, as he calls it. Keeps their heads in a jar above his desk. Apparently, they bob around all day with a smiling expression.
You have to get your hands dirty for a lot of science...
What's even worse than with Windows is that since 'rm' is just a normal binary the PDF can launch that, and if you run as root privileges, just issue a command like "rm -rf /". If you don't run as root, then for example Ubuntu should give you the sudo box to input password to. This of course being just one of the examples it could do. Remember that most malware doesn't even need root access to function.
Nobody uses the root account in Linux for everyday activity. In Ubuntu, root login is even disabled by default (you have to sudo). So no worries about the system in general. Although it's pretty devastating to issue a "rm -rf ~" to delete the user's home directory, it's on par with Windows. Then you say that most malware doesn't even need root access to function, but on all the millions of XP boxes out there, it's already given root access by default.
Another reason why it would be even more serious on Linux is the way you can pipe commands and how most systems come pre-packaged with a ton of little utility apps. You can create the whole malware with a series of commands, or wget a bash script from the internet and start that to hide even more malware in the system.
Windows has a pipe function too, in addition to being able to zoink your whole file system with a simple "del". It also comes with ftp and telnet, which are handy replacements for wget. In short telnet+response file = download an .exe from the web = any sort of functionality you might want using Unix command line tools.
Your comment, sir, is vapid.
"These are not the dorks you are looking for."
Not only that, only nerds would get excited about a STOCK response from HR about discrimination, and then post it on slashdot. Not trying to troll here, but HR folks aren't lawyers, and are trained to be extremely careful when it comes to possible litigation. In short, even the bad publicity makes it worthwhile for HR to apologize to this "Jedi" instead of saying something like "we only recognize jedis on active duty, with working light sabers".
Then TFS mentions something as banal as "wrapping long lines of code"....and my bubble bursts.
This seems to be a non-story, unless this is the first time these financial ties have been revealed between bit torrent researchers and ISPs.
This is not so much about calling the researchers' methods and findings into question as the ISPs motivation for funding the research. As far as I can tell, the research seems to be sound and pretty neat. The question is WHY are ISPs interested in FUNDING this sort of research?
One possibility that the submitter didn't consider is the fact that many researchers list their funding sources on all published papers, regardless of whether the funding was given to fund that specific project. So it could be that ISPs generally fund this particular research group in any case, and they happened to put out a paper that analyzes BT. In other words, there might not be anything sinister going on.
Don't get me wrong, it would be incredibly cost, labor and time expensive, and require real computer scientists, but it is certainly possible.
Speaking as a "real" computer scientist, I think you might have underestimated the time requirement. Most problems in automatic verification are either undecidable, or intractable.
"We're drawn to Light because it denotes both warmth and clarity, and intrigued by the idea that 'light' is a good value in software. Good software is 'light' in the sense that it uses your resources efficiently, runs quickly, and can easily be reshaped as needed," the design documentation says. "Visually, light is beautiful, light is ethereal, light brings clarity and comfort."
Why do "design documents" always have to be so banal? I mean, "visually, light is beautiful"?!!? Seriously?
CNN was actually discussing this in their reporting yesterday. They were very clear about this being done by bad folks, not the web in general, and the things people should look out for.
Given the amount of money and time CNN has invested in gadgetry and web-whizbangery, over any other network by far, they would be fools to (a) draw too much attention to this unless it becomes a serious problem, and (b) NOT cover it calmly, and without resorting to fear-mongering. I suppose that answers GP's question, at least as far as CNN. Now, Murdoch-owned Fox News is a different story altogether....
Except ... medical studies show that 'Deja Vu' is really just glitches in the matrix.
There, corrected that for you (an old and tired meme, yes).
Seeing how 66.67% of the time I am either sleeping at home or at work it shouldn't be too hard to fill the other 27% with commute/grocery shopping.
You're not too far off. I worked at the research wing of a phone company, and I can tell you that "tracking" a person using a cell tower is pretty coarse, even in urban areas. Given that most people go to work on weekdays, I'd say that a lot of your "movement" could be predicted on this level by just predicting your average movement. Add in a weekday/weekend variation, and 93% is hardly surprising.
This isn't even one of those "well duh, in RETROSPECT everything is obvious" studies -- anyone who has ever worked with CDR (mobile phone) data knows that this is pretty obvious even before running the experiments.
And for the people who bring up the MIT Reality Mining experiment, keep in mind that they tracked about 100 *individuals*, all of whom were MIT students with pretty regular routines.
Isn't it obvious that the fear of something will have an impact even on the simplest things where something relative to that fear is involved ?
Yes, but I think what this study was trying to test was how basic the task has to be for the fear response to have a measurable effect. Turns out, pretty damn basic.
Not really -- I haven't felt the squeeze of 500 MB yet, and the central repository is a plus for me, but I'd be happy to pay for more space.
It would be cool if, say, the IR spectrum or just more dynamic range in the visible spectrum could be tone-mapped to human perception in this way, resulting in perceptually sharper images by way of a direct retinal implant.
Try Mendeley. They're still pretty new, but very promising with their desktop client for Linux/Mac/Win in addition to the web interface. They also sync perfectly with Zotero and CiteULike, which makes migration easier. You can annotate PDFs directly in the desktop, but I think only the latest beta build has support for sync'ing the annotations across multiple computers. I'm hopeful for them -- it's definitely one of the most promising Ref manager systems I've seen (oh yes, they also support Bibtex,Endnote,Refworks formats heavily)
Certainly, Apple will reject the app and Opera knows it. Maybe Opera tries to strengthen Apple's "Evil Empire" image and deal with it with the help of EU (just like they did with Microsoft recently).
My first thought on reading the summary was "where's the leverage?" Either Opera is talking right through their own asses, or they have some serious leverage -- certainly more than just bad PR (which Apple seems immune to anyway). They've invested time and manpower in this project, one can only hope that someone's cojones over at Apple are in a vice grip, and that we will all soon enjoy the big red O on our iPhones.
...if they manage to get the privacy thing right.
LOL
Just to be clear, I meant privacy in terms of your friends. In terms of Google, privacy was pretty much given up a long time ago.
Although it would take quite a few HCI PhDs to figure out how to do it all without cluttering an already cluttery gmail UI.
If it's so freakin' open please tell me why I still need to have apps signed on my Nokia 6220 classic and will do for the foreseeable future unless I'm willing to try risky hacks.
I'll raise you an anecdote. I just bought a Nokia E63, new and unlocked with a full US warranty for $189 from Newegg, and it's one of the best phones I've ever owned. You simply go to the application manager menu, and for the option that says "Install only signed apps", select "No". It's that simple. I just installed an unsigned FTP client, so now I don't even need Nokia's atrocious PC Suite for syncing.