You need to read more about Gerrymandering. it's part of the political sausage factory in the US. It's disgusting. It happens. It violates any definition of democratic fairness. Then there's the various attempts by parties to keep groups statistically likely to vote against them from voting. Insufficient voting machines in college towns and minority neighborhoods; tricks with mailing addresses to remove voters; etc.
In a democracy, everyone has an incentive to mess with the vote (that is, if you actually believe in what you're voting for). Widely available personal information gives political parties more information they can use to statistically correlate with voting tendencies. In the past, it was only party registration that was available to them. Nowadays they can also find out that you're a fan of Birkenstocks, like yoga, and bitch about Bush. Suddenly you're not hiding your preferences behind your Republican or Independent party registration. Then there are a large number of legal, but massively unfair tricks they can perform with that information to marginalize your vote. Statistics are powerful.
I would love to see a proposal to fix the vote that could get rid of these kinds of unfair effects. I do not know one.
One man's online asshat is another man's crusader. Question is...do the people who think you're an asshat have guns/thugs/policemen/jails? Doesn't matter if the government requires your identity or a private organization...only if those who think you're an asshat have access to your Facebook profile.
Asshattery is a price we pay for freedom. It's not pretty but it's one I'm willing to accept. Psychologists have long shown the tendency of people to self-censor when their identity is known or they could be held accountable for their actions (hey it's only logical). When important things need to be said, they should be said, anonymously if necessary, rather than everyone self-censoring until the situation blows up in our faces. Look at Turkmenistan or North Korea or Thailand for examples of self-censorship regarding their glorious leaders...
Not only will the revolution not be televised, it also won't be on Facebook.
For a more 1st world example, imagine posting something critical of a certain candidate or party in the US. Imagine then that party data-mining online posts, classifying people into "favorable" or "unfavorable" to their side using some basic NLP. If they have your name, then they correlate that with voter rolls (which parties have access to), so now they know where you live. Then they use that information to gerrymander your district so that your vote is marginalized, and thus, engineer the election result. I'd be surprised if this isn't happening right now in the US.
Then there's VASIMR, which is an electromagnetic engine more powerful than ion engines. A test unit will be flown to the ISS in 2014. According to this Wikipedia article, fuel for station keeping will be cut by a factor of 20, if this works. That, plus possible improvements in the VASIMR design that may come with space testing, could make boosting it to another orbit viable. So, in principle you take up the VASIMR engine, and a couple resupply vessels containing only fuel...this engine is re-usable. So, we've got between 2014 and 2020 to test, propose, and implement this. It only took us 8 years to go to the moon, we can do this, right?
Consider adding padding characters to a dictionary attack, so you extend the english dictionary by a list of "words" that are repeated characters. Let's be generous and say that this doubles the size of your dictionary. Let's further assume that the cracking software also tries merged word combinations.
Your algorithm has only logarithmically increased the cracking time. You're confusing exponential growth in complexity by length with logarithmic growth in complexity by increasing the dictionary. Your algorithm will work for the 5 minutes between the time that you think it up, and someone implements it in a password cracker.
This is utter bullshit. The entropy of such things is low. So I rewrite my dictionary attack to insert long strings of the same character, only marginally increasing the search space over a dictionary attack. Then add 1234567, asdfghjkl, etc as possible padding strings.
Only if the cracker is stupid, and processes passwords by a "shortest length first" algorithm rather than "lowest entropy" algorithm does such a thing work.
If the algorithm to create your low-entropy password can be written down, it can also be coded into a password cracker. Low entropy passwords are simply not secure.
The answer to all such security questions is
perl -le 'print map { ("a".."z","A".."Z",0..9)[rand 62] } 1..64'
and an encrypted password store... A 1e114 search space is probably large enough.
Such stupid security questions often allow longer input than the password itself...as long as you're not stupid enough to answer the question asked.
TFA complains about simple passwords (containing no non-alphanumeric characters). Over the years I found that every single little stupid corner of the internet decided they had a better idea what should be in a password than everyone else. Each of them excludes a random subset of non-alphanumeric characters from being valid. Another subset of stupid little corners of the internet can't code their way out of a paper bag, and can't properly escape non-alphanumeric characters, especially ['"\%&=] which need to be escaped in certain contexts or are contained in urls. Yet another subset of stupid corners of the internet place arbitrary length restrictions on your password (here on slashdot: 20 characters). Working on wiki software for a while, I watched as time and time again, contributors couldn't understand the basics of properly escaping strings, so they invented stupid crazy regexes that always failed. Then they would pile on more hacks to catch corner cases. On web forms it usually takes the form of some javascript that "checks" the password, and other javascript that has to encode it into a URL or POST request.
So I gave up. As you argue, increasing the length increases the complexity exponentially fast, while increasing the character set increases the complexity only logarithmically fast. So it's better to use a long alphanumeric password than to discover that you can't log in, because the password form can't encode what you typed properly. These days I find it's extremely rare to run across a site or application that requires a non-alphanumeric character to be present.
I've been having kernel panics regularly, I recently figured out that it's due to "hardware acceleration" in the binary flashplayer, coupled with the open source video drivers (ati, for me). Right click on a flash animation and turn off hardware acceleration. And/or un-install that steaming pile of dung. Still, it shouldn't be causing kernel panics. I think the open source/drm drivers need work.
It's common that you will be able to keep a computer account of some kind when you leave. Just set up an ssh tunnel to that computer and use FoxyProxy (Firefox) or Proxy Switchy (Chrome) to set up rules to use the proxy when you hit particular journals. I do this now even though I'm at a university, because every library is reducing their library subscriptions due to increasing journal costs, my research is multi-disciplinary, and it's becoming more and more common to hit pay walls from within the university.
Clearly, one should not distribute raw data, but rather processed data after the experimentalists have taken into account detector resolution, triggering, etc. (In itself a hard problem, I know). No one does actual physics analyses on the raw data anyway. Everyone uses skims or otherwise reduced data.
I'm not so sure. Assuming this is a variation on the Voith Schneider Propeller, consider a configuration of the cylinder of propellers with all the airfoils parallel, and pointed in the direction of flight (so the direction of flight is perpendicular to the cylinder's axis). That's essentially just six stacked wings in an odd configuration, kind of like a triplane. If you have enough forward velocity to maintain lift, all you need to do is lock the airfoils in place. By changing the angle of attack on some of them you can emulate flaps, and increase the lift. The compact configuration of wings would have lots of drag, but you could add fixed wings on the outside to help.
What's more important, the profit of the company or the fate of the employees? Any reasonable business should tie their fates. Employees should be fucked if the company goes under, and the company should go under if it throws its employees under the bus. A high(er) firing cost turns the equation around: how can we make money with these employees while the economy recovers? Instead of: How do we get rid of these extra employees so we can show a profit this quarter? The high firing cost would just make the situation you mention come earlier, as managers evaluate the cost of downsizing (and if managers don't realize this situation until it's too late -- then they're idiots and the company should fold anyway). It should cause companies to think longer term. No? Seems like a win-win all around to me.
You know, why the hell don't regular employment contracts for regular employees have "golden parachutes"? It doesn't make a hell of a lot of sense for executives to have them. It seems to me that it does make sense for regular employees to have them. It makes firing everyone expensive, and incentivizes the company to find ways to utilize their employees, and foster a more cooperative work environment, rather than outsourcing, merging, and layoffs at the whim of this week's executive.
Make the golden parachute a year or two of salary, and make it untaxable. Make the managers equation be (profit employee generates) - (golden parachute) in deciding to retain or fire. This would go a long way to solving unemployment problems too, if all employees could afford to spend a year or two looking for a job.
The data volume issue is brought up occasionally, but is a red herring. The SDSS dataset is comparable in size to that of Fermilab and CERN and is available to the public (hundreds of TB). No one said anyone should publish raw data either. A processed, manageable form is preferable. If the datasets are that large, then we should be working on publicly-funded data warehouses, just as we once built libraries across the country.
Astronomy has a history of sharing data, unlike particle physics. Furthermore, the NSF has data sharing requirements, and the NSF funds most astronomy activities in the US. The DOE funds particle physics (mostly). Data is also not available "on request" at all from particle physics experiments. There are some efforts to change this, but at the tail end of an experiment there is generally not a lot of manpower, money, or motivation to make data public. It's a big effort, really, and the culture of particle physics results in jealously guarded data.
So the best situation is that grant agencies require data sharing as a contingency for all publicly funded activities. (As the NSF has done) And I think we should put effort into data warehousing for publicly funded projects. Hell, give us DNA sequences, drug studies, pesticide tests, EPA water quality, everything publicly funded. Generally everyone wants a subset of the data. Giving anyone that asks the entire dataset is like sending them the library of congress because they want to read one book...
There is absolutely no requirement to share data in particle physics. Most of the data from early colliders is irretrievably lost. There's nothing wrong with time embargoes, but that's not what's going on here.
Bullshit. Look to the the astronomy community for counter-examples. WMAP, SDSS, etc.
The only reason particle physics keeps its data closed is history and turf-protection by its members. Astronomy has a longer history, and realized the benefits of sharing star catalogs hundreds of years ago.
Zotero does categorization. You have a familiar hierarchy of folders (topics). A single reference can appear in multiple folders. (It's your organization, not the filesystem -- each thing appears only once on disk). It does not automatically create categories for you, however. But frankly, I don't think any software could do that in a way I would agree with... You can further attach arbitrary files and notes to things you put in your Zotero database.
You seem to have missed Zotero (it's up there in the comments somewhere) which is a FOSS plugin for firefox. It keeps an offline database, and for nearly any site (e.g. journals) you click one button on the URL bar and it downloads the citation including full pdf so you can read it whenever. It will also let you perform full text searches of your database, and can be configured to perform OCR on scanned documents. Best of all, it's trivial to make bibtex (or many other formats) bibliographies.
I use that in combination with TiddlyWiki for personal typed notes not associated with a journal article/textbook, and Xournal for annotating documents and taking notes with my tablet computer. When annotating documents (textbooks, journal articles) just configure xournal as your pdf viewer and you'll be able to save every annotation you make. TiddlyWiki has a ton of plugins to do whatever you need, including a GTD (Getting Things Done -- it's a book) variant that's probably comparable to Emacs Org-mode, LaTeX math (I wrote that one -- use it every day), and many more.
The one drawback to all this is that I have no way to automatically organize my handwritten notes from xournal. Though they're computer files, my organization for them is horrendous. I still fantasize about some kind of hybrid mutant of TiddlyWiki, OCR (that can magically read my handwriting and equations), and xournal that would let me do all this on a pen-based tablet...
I have hover on my tablet pc... if the stylus is within about 1cm of the screen, it moves the mouse cursor. I still contend that the stylus on an active digitizer is a far superior user interface than your fat greasy fingers. Hey tablet manufacturers, WAKE UP and give us active digitizers, styli, in combination with a capacitive touch screen, and high-resolution screens > 150 dpi, so we can replace PAPER!
Here's a suggestion: certify academics as editors, and give them absolute power over the domain of their expertise (supplanting Wikipedia editors). Certify contributors too, make them subordinate to the editor, but superior to non-expert Wikipedia editors. That would solve the idiot-editor problem, and would also give the academic something to put on their CV: "Certified Wikipedia Editor/Contributor responsible for X pages and contributed to Y pages". You won't get senior professors to do it, but I bet this path would get lots of junior Ph.D.'s. Academics anyway often have unpaid positions as journal editors, and always as referees. This is very similar, and we all put that free work on our CV's. I'd also get rid of the "no original research clause". You'll have academics clamoring to write about their latest theory. Take a look at how the arXiv endorsement system works. It's not a full certification but halfway there. Copying the list of endorsed arXiv authors would be a great starting point for certified expert Wikipedia contributors.
I often find myself landing on Wikipedia when looking up research topics. They're under-cited and often poorly explained, but often can be better than navigating 1000 un-ranked results returned on a journal search (the majority of which use a concept, but don't explain it). There's a real need for Wikipedia here, but it desperately needs improvement.
eventhough its one of the cheapest crap routers out there
Ahem. The WNR3500L they're giving away is a linux-based (openwrt) high-end wireless router. It was $150 when new, now can be had for $80. Its successor the WNDR3700 retails for $185 and it's freaking awesome. A customizable linux-based router is precisely what I'd choose if I wanted to do an experiment like this.
Incorrect. If an insurance company has the opportunity to remove unprofitable members from the rolls, they will take it. If they have the opportunity to refuse treatment, they will take it. If they can select which new customers they will take and which ones they won't, they will use that. If they can write long obtuse contracts outlining things they won't pay for, and have their army of lawyers enforce it, they will do it.
It is a general fact about any kind of insurance that the interests of the insurer are misaligned with the interests of the insuree. They're predatory industries who rely upon promising more than they will deliver and tricking their customers wherever possible.
Only in the circumstance that the insurer is required to insure everyone does the profit motive go in the direction of the patient's interests (in the form of preventative care). Preventative care is a long term investment that wall street doesn't see.
And that, in a nutshell is what's wrong with for-profit insurance providers: the profit motive of the company is directly opposed to the health motive of the customer.
Because of that very fundamental fact, the only medical insurance scheme that makes any sense is a socialized one.
So the premise is that because no one person is uncorruptible, let's take the next logical step and assume everyone is uncorruptible?
You just aren't creative enough. There will be new crimes for new circumstances, and the power will fall to those with the time/resources to mine that data, and those with the resources to avoid the surveillance. (You really think it will be impossible for an untracked person to enter, or avoid cameras?) Business adversaries have an interest in who their competition is meeting with and what they're saying. Scientific rivals have an interest in knowing what their rivals are doing. Competition and negotiation would be impossible. What about the ready availability of information about, say abortion doctors? Someone will be hated, by someone else, and having all this info will be valuable to them. People will find new ways to harass and harm their enemies. What about $despised_minority let's say for example homosexuals, or Mexicans. If the majority doesn't like them, the majority will not punish crimes against them made possible by mining their data. But the majority is a bunch of stupid sheeple, and sometimes they're stupid, or bigoted, or just dead wrong. What about minor crimes such as jay-walking and public urination? Don't you think it would be useful to comb through data of an enemy and find such minor infractions?
Lack of creativity concerning unintended consequences does not a reasonable proposal make. Anonymity is important. It enables society to function when group A doesn't like group B. Anonymity enables dissent, the lack of it enforces conformity. Privacy enables competition and negotiation. Privacy renders stupid laws harmless when no one cares about them. I could go on...
Doubtless in this age, people will have to re-learn these lessons. Maybe Scott Adams hasn't noticed Wikileaks yet. While it laid bare the information about dictators, the only reason it works is that the submitter is anonymous.
Slashdot Mirror
You need to read more about Gerrymandering. it's part of the political sausage factory in the US. It's disgusting. It happens. It violates any definition of democratic fairness. Then there's the various attempts by parties to keep groups statistically likely to vote against them from voting. Insufficient voting machines in college towns and minority neighborhoods; tricks with mailing addresses to remove voters; etc.
In a democracy, everyone has an incentive to mess with the vote (that is, if you actually believe in what you're voting for). Widely available personal information gives political parties more information they can use to statistically correlate with voting tendencies. In the past, it was only party registration that was available to them. Nowadays they can also find out that you're a fan of Birkenstocks, like yoga, and bitch about Bush. Suddenly you're not hiding your preferences behind your Republican or Independent party registration. Then there are a large number of legal, but massively unfair tricks they can perform with that information to marginalize your vote. Statistics are powerful.
I would love to see a proposal to fix the vote that could get rid of these kinds of unfair effects. I do not know one.
One man's online asshat is another man's crusader. Question is...do the people who think you're an asshat have guns/thugs/policemen/jails? Doesn't matter if the government requires your identity or a private organization...only if those who think you're an asshat have access to your Facebook profile.
Asshattery is a price we pay for freedom. It's not pretty but it's one I'm willing to accept. Psychologists have long shown the tendency of people to self-censor when their identity is known or they could be held accountable for their actions (hey it's only logical). When important things need to be said, they should be said, anonymously if necessary, rather than everyone self-censoring until the situation blows up in our faces. Look at Turkmenistan or North Korea or Thailand for examples of self-censorship regarding their glorious leaders...
Not only will the revolution not be televised, it also won't be on Facebook.
For a more 1st world example, imagine posting something critical of a certain candidate or party in the US. Imagine then that party data-mining online posts, classifying people into "favorable" or "unfavorable" to their side using some basic NLP. If they have your name, then they correlate that with voter rolls (which parties have access to), so now they know where you live. Then they use that information to gerrymander your district so that your vote is marginalized, and thus, engineer the election result. I'd be surprised if this isn't happening right now in the US.
Then there's VASIMR, which is an electromagnetic engine more powerful than ion engines. A test unit will be flown to the ISS in 2014. According to this Wikipedia article, fuel for station keeping will be cut by a factor of 20, if this works. That, plus possible improvements in the VASIMR design that may come with space testing, could make boosting it to another orbit viable. So, in principle you take up the VASIMR engine, and a couple resupply vessels containing only fuel...this engine is re-usable. So, we've got between 2014 and 2020 to test, propose, and implement this. It only took us 8 years to go to the moon, we can do this, right?
Consider adding padding characters to a dictionary attack, so you extend the english dictionary by a list of "words" that are repeated characters. Let's be generous and say that this doubles the size of your dictionary. Let's further assume that the cracking software also tries merged word combinations.
Your algorithm has only logarithmically increased the cracking time. You're confusing exponential growth in complexity by length with logarithmic growth in complexity by increasing the dictionary. Your algorithm will work for the 5 minutes between the time that you think it up, and someone implements it in a password cracker.
This is utter bullshit. The entropy of such things is low. So I rewrite my dictionary attack to insert long strings of the same character, only marginally increasing the search space over a dictionary attack. Then add 1234567, asdfghjkl, etc as possible padding strings.
Only if the cracker is stupid, and processes passwords by a "shortest length first" algorithm rather than "lowest entropy" algorithm does such a thing work.
If the algorithm to create your low-entropy password can be written down, it can also be coded into a password cracker. Low entropy passwords are simply not secure.
The answer to all such security questions is perl -le 'print map { ("a".."z","A".."Z",0..9)[rand 62] } 1..64' and an encrypted password store... A 1e114 search space is probably large enough.
Such stupid security questions often allow longer input than the password itself...as long as you're not stupid enough to answer the question asked.
TFA complains about simple passwords (containing no non-alphanumeric characters). Over the years I found that every single little stupid corner of the internet decided they had a better idea what should be in a password than everyone else. Each of them excludes a random subset of non-alphanumeric characters from being valid. Another subset of stupid little corners of the internet can't code their way out of a paper bag, and can't properly escape non-alphanumeric characters, especially ['"\%&=] which need to be escaped in certain contexts or are contained in urls. Yet another subset of stupid corners of the internet place arbitrary length restrictions on your password (here on slashdot: 20 characters). Working on wiki software for a while, I watched as time and time again, contributors couldn't understand the basics of properly escaping strings, so they invented stupid crazy regexes that always failed. Then they would pile on more hacks to catch corner cases. On web forms it usually takes the form of some javascript that "checks" the password, and other javascript that has to encode it into a URL or POST request.
So I gave up. As you argue, increasing the length increases the complexity exponentially fast, while increasing the character set increases the complexity only logarithmically fast. So it's better to use a long alphanumeric password than to discover that you can't log in, because the password form can't encode what you typed properly. These days I find it's extremely rare to run across a site or application that requires a non-alphanumeric character to be present.
I've been having kernel panics regularly, I recently figured out that it's due to "hardware acceleration" in the binary flashplayer, coupled with the open source video drivers (ati, for me). Right click on a flash animation and turn off hardware acceleration. And/or un-install that steaming pile of dung. Still, it shouldn't be causing kernel panics. I think the open source/drm drivers need work.
It's common that you will be able to keep a computer account of some kind when you leave. Just set up an ssh tunnel to that computer and use FoxyProxy (Firefox) or Proxy Switchy (Chrome) to set up rules to use the proxy when you hit particular journals. I do this now even though I'm at a university, because every library is reducing their library subscriptions due to increasing journal costs, my research is multi-disciplinary, and it's becoming more and more common to hit pay walls from within the university.
Clearly, one should not distribute raw data, but rather processed data after the experimentalists have taken into account detector resolution, triggering, etc. (In itself a hard problem, I know). No one does actual physics analyses on the raw data anyway. Everyone uses skims or otherwise reduced data.
I'm not so sure. Assuming this is a variation on the Voith Schneider Propeller, consider a configuration of the cylinder of propellers with all the airfoils parallel, and pointed in the direction of flight (so the direction of flight is perpendicular to the cylinder's axis). That's essentially just six stacked wings in an odd configuration, kind of like a triplane. If you have enough forward velocity to maintain lift, all you need to do is lock the airfoils in place. By changing the angle of attack on some of them you can emulate flaps, and increase the lift. The compact configuration of wings would have lots of drag, but you could add fixed wings on the outside to help.
I think this thing can glide.
What's more important, the profit of the company or the fate of the employees? Any reasonable business should tie their fates. Employees should be fucked if the company goes under, and the company should go under if it throws its employees under the bus. A high(er) firing cost turns the equation around: how can we make money with these employees while the economy recovers? Instead of: How do we get rid of these extra employees so we can show a profit this quarter? The high firing cost would just make the situation you mention come earlier, as managers evaluate the cost of downsizing (and if managers don't realize this situation until it's too late -- then they're idiots and the company should fold anyway). It should cause companies to think longer term. No? Seems like a win-win all around to me.
You know, why the hell don't regular employment contracts for regular employees have "golden parachutes"? It doesn't make a hell of a lot of sense for executives to have them. It seems to me that it does make sense for regular employees to have them. It makes firing everyone expensive, and incentivizes the company to find ways to utilize their employees, and foster a more cooperative work environment, rather than outsourcing, merging, and layoffs at the whim of this week's executive.
Make the golden parachute a year or two of salary, and make it untaxable. Make the managers equation be (profit employee generates) - (golden parachute) in deciding to retain or fire. This would go a long way to solving unemployment problems too, if all employees could afford to spend a year or two looking for a job.
The data volume issue is brought up occasionally, but is a red herring. The SDSS dataset is comparable in size to that of Fermilab and CERN and is available to the public (hundreds of TB). No one said anyone should publish raw data either. A processed, manageable form is preferable. If the datasets are that large, then we should be working on publicly-funded data warehouses, just as we once built libraries across the country.
Astronomy has a history of sharing data, unlike particle physics. Furthermore, the NSF has data sharing requirements, and the NSF funds most astronomy activities in the US. The DOE funds particle physics (mostly). Data is also not available "on request" at all from particle physics experiments. There are some efforts to change this, but at the tail end of an experiment there is generally not a lot of manpower, money, or motivation to make data public. It's a big effort, really, and the culture of particle physics results in jealously guarded data.
So the best situation is that grant agencies require data sharing as a contingency for all publicly funded activities. (As the NSF has done) And I think we should put effort into data warehousing for publicly funded projects. Hell, give us DNA sequences, drug studies, pesticide tests, EPA water quality, everything publicly funded. Generally everyone wants a subset of the data. Giving anyone that asks the entire dataset is like sending them the library of congress because they want to read one book...
There is absolutely no requirement to share data in particle physics. Most of the data from early colliders is irretrievably lost. There's nothing wrong with time embargoes, but that's not what's going on here.
Bullshit. Look to the the astronomy community for counter-examples. WMAP, SDSS, etc.
The only reason particle physics keeps its data closed is history and turf-protection by its members. Astronomy has a longer history, and realized the benefits of sharing star catalogs hundreds of years ago.
Zotero does categorization. You have a familiar hierarchy of folders (topics). A single reference can appear in multiple folders. (It's your organization, not the filesystem -- each thing appears only once on disk). It does not automatically create categories for you, however. But frankly, I don't think any software could do that in a way I would agree with... You can further attach arbitrary files and notes to things you put in your Zotero database.
You seem to have missed Zotero (it's up there in the comments somewhere) which is a FOSS plugin for firefox. It keeps an offline database, and for nearly any site (e.g. journals) you click one button on the URL bar and it downloads the citation including full pdf so you can read it whenever. It will also let you perform full text searches of your database, and can be configured to perform OCR on scanned documents. Best of all, it's trivial to make bibtex (or many other formats) bibliographies.
I use that in combination with TiddlyWiki for personal typed notes not associated with a journal article/textbook, and Xournal for annotating documents and taking notes with my tablet computer. When annotating documents (textbooks, journal articles) just configure xournal as your pdf viewer and you'll be able to save every annotation you make. TiddlyWiki has a ton of plugins to do whatever you need, including a GTD (Getting Things Done -- it's a book) variant that's probably comparable to Emacs Org-mode, LaTeX math (I wrote that one -- use it every day), and many more.
The one drawback to all this is that I have no way to automatically organize my handwritten notes from xournal. Though they're computer files, my organization for them is horrendous. I still fantasize about some kind of hybrid mutant of TiddlyWiki, OCR (that can magically read my handwriting and equations), and xournal that would let me do all this on a pen-based tablet...
It's called the beam dump.
I have hover on my tablet pc... if the stylus is within about 1cm of the screen, it moves the mouse cursor. I still contend that the stylus on an active digitizer is a far superior user interface than your fat greasy fingers. Hey tablet manufacturers, WAKE UP and give us active digitizers, styli, in combination with a capacitive touch screen, and high-resolution screens > 150 dpi, so we can replace PAPER!
Here's a suggestion: certify academics as editors, and give them absolute power over the domain of their expertise (supplanting Wikipedia editors). Certify contributors too, make them subordinate to the editor, but superior to non-expert Wikipedia editors. That would solve the idiot-editor problem, and would also give the academic something to put on their CV: "Certified Wikipedia Editor/Contributor responsible for X pages and contributed to Y pages". You won't get senior professors to do it, but I bet this path would get lots of junior Ph.D.'s. Academics anyway often have unpaid positions as journal editors, and always as referees. This is very similar, and we all put that free work on our CV's. I'd also get rid of the "no original research clause". You'll have academics clamoring to write about their latest theory. Take a look at how the arXiv endorsement system works. It's not a full certification but halfway there. Copying the list of endorsed arXiv authors would be a great starting point for certified expert Wikipedia contributors.
I often find myself landing on Wikipedia when looking up research topics. They're under-cited and often poorly explained, but often can be better than navigating 1000 un-ranked results returned on a journal search (the majority of which use a concept, but don't explain it). There's a real need for Wikipedia here, but it desperately needs improvement.
Ahem. The WNR3500L they're giving away is a linux-based (openwrt) high-end wireless router. It was $150 when new, now can be had for $80. Its successor the WNDR3700 retails for $185 and it's freaking awesome. A customizable linux-based router is precisely what I'd choose if I wanted to do an experiment like this.
Incorrect. If an insurance company has the opportunity to remove unprofitable members from the rolls, they will take it. If they have the opportunity to refuse treatment, they will take it. If they can select which new customers they will take and which ones they won't, they will use that. If they can write long obtuse contracts outlining things they won't pay for, and have their army of lawyers enforce it, they will do it.
It is a general fact about any kind of insurance that the interests of the insurer are misaligned with the interests of the insuree. They're predatory industries who rely upon promising more than they will deliver and tricking their customers wherever possible.
Only in the circumstance that the insurer is required to insure everyone does the profit motive go in the direction of the patient's interests (in the form of preventative care). Preventative care is a long term investment that wall street doesn't see.
And that, in a nutshell is what's wrong with for-profit insurance providers: the profit motive of the company is directly opposed to the health motive of the customer.
Because of that very fundamental fact, the only medical insurance scheme that makes any sense is a socialized one.
So the premise is that because no one person is uncorruptible, let's take the next logical step and assume everyone is uncorruptible?
You just aren't creative enough. There will be new crimes for new circumstances, and the power will fall to those with the time/resources to mine that data, and those with the resources to avoid the surveillance. (You really think it will be impossible for an untracked person to enter, or avoid cameras?) Business adversaries have an interest in who their competition is meeting with and what they're saying. Scientific rivals have an interest in knowing what their rivals are doing. Competition and negotiation would be impossible. What about the ready availability of information about, say abortion doctors? Someone will be hated, by someone else, and having all this info will be valuable to them. People will find new ways to harass and harm their enemies. What about $despised_minority let's say for example homosexuals, or Mexicans. If the majority doesn't like them, the majority will not punish crimes against them made possible by mining their data. But the majority is a bunch of stupid sheeple, and sometimes they're stupid, or bigoted, or just dead wrong. What about minor crimes such as jay-walking and public urination? Don't you think it would be useful to comb through data of an enemy and find such minor infractions?
Lack of creativity concerning unintended consequences does not a reasonable proposal make. Anonymity is important. It enables society to function when group A doesn't like group B. Anonymity enables dissent, the lack of it enforces conformity. Privacy enables competition and negotiation. Privacy renders stupid laws harmless when no one cares about them. I could go on...
Doubtless in this age, people will have to re-learn these lessons. Maybe Scott Adams hasn't noticed Wikileaks yet. While it laid bare the information about dictators, the only reason it works is that the submitter is anonymous.