The Chinese system (and now the UK's) does not overly concern itself with ensuring every possible route through the Great Firewall is blocked--after all, geniuses gonna be ingenious. It merely suffices that the overwhelming majority cannot do this easily, and the technically adept live in fear of the consequences of their clearly pre-meditated actions. The censorship, and the chilling effect, is in the general drag; no need to close down every last loophole. If John Allsup is right above, all that is required to get around this is "a LAMP stack somewhere on the internet" and some code. What percentage of the population are capable of actually doing this in practice? Taking say, the desktop share of Linux/BSD as a proxy for this, which is probably an overestimate, given the ease of installation of modern distros, that's fewer 3% of all internet users. A repressible minority? The UK government evidently thinks so.
It would help focus peoples' minds on the harmfulness of speed if, rather than setting arbitrary thresholds, the fines were proportional not only to income, but also to the excess kinetic energy of the vehicle:
Ek = 1/2 . m . v^2
Kuisla was doing 65 in a 50 zone, so his "kinetic tariff" would be (65x65) - (50x50) = 4225 - 2500 = 1725 units. (It's the exponential nature of the velocity squared factor that yields the disincentive, but, by all means introduce the weight of the miscreant's vehicle too)
This would be a rational way of reflecting the a) the risk of injury in the event of a collision; b) the undesirable environmental effects (noise, gaseous emissions), both of which rise exponentially with the kinetic energy of the vehicle.
This kinetic tariff can then be applied to income data to calculate the actual fine for the individual.
The choice about whether or not to encrypt traffic should be left to each website's administrator. Many sites--shock!--use the web to disseminate information they wish to be public, and the site's users have no problem with their access to it being public either. So get out of their faces! Using the browser to deprecate admin's particular choices is contrary to the spirit of the web, which should always do its damndest to serve something, and degrade gracefully when it's in difficulty, not pop annoying dialog boxes in the user's way.
Self-certificates are already a fairly effective denial of service attack when Firefox is used to access many independent sites that try to implement https, but who fail to do so in a way that offers a smooth user experience to J. Random User (I'm thinking particularly of IndyMedia).
Please note: in China, the censorship does not rely on blocking everything; just on blocking enough that all but the very motivated fail to access it. This troublesome minority can presumably be picked off at leisure later.
yeh, the French know that sacking the public sector in times of crisis does not help the economy; quite the reverse in fact. M. Hollande is old school ENA (Ecole Nationale d'Administration) which turns out highly-educated senior French bureaucrats and politicians, who, whatever else they may be, are not daft.
Diigo seems to work quite handily, but I still prefer Delicious and will mourn its passing. It's easy to export your Delicious bookmarks to an html file and import that into Diigo.
A single seed bank like this doesn't make any kind of biological sense. It is remarkably unlikely to be useful in the event of catastrophe: it's a long old road up there to Norway to replenish stocks of some ancient carrot variety from most parts of Europe.
If you actually wanted to guard biodiversity, you would encourage social networks of gardeners to replant varieties each season and share the ensuing seeds. The French organisation Kokopelli does this, but seems to suffer from legal harassment rather than incur the subsidies it would receive in any sane world.
An analogy for the slashdot crowd might be Napster (centralised) vs. BitTorrent (distributed).
that moldy alfafa will have made really nice mulch/compost for growing veg. Your old computers on the other hand are destined to pollute a water table somewhere.
potential for medical applications is exciting
on
Full-Text Audio Search
·
· Score: 3, Interesting
As 90% of the data for diagnosis comes from the history-taking (interviewing) the patient, the potential for automating/supporting diagnosis is exciting.
Imagine a system that listens to a consultation in real time, making helpful suggestions for diagnosis based on analysis of the patient and the doctor's phoneme streams! And no tedious data entry, just an unobtrusive microphone.
It is a good piece of art because it reminds us that the person who builds the system (the road authority) will not necessarily anticipate the information needs of those who use it (joe drivers).
The solution was rather a drastic hack. The road authority should run a website where you could make suggestions for improved signage. These could be implemented in an economic manner when the signs are upgraded for other reasons.
It's the same thing with software: it's good to design it right: but you're still going to have to capture the thoughts of naive users when they encounter it for the first time, and take their expectations into account when you're upgrading the product.
Here in Hackney in East London we had a political sign hack for a while: someone converted a local route sign from miles to kilometres, presumably to make a point about closer integration with Europe. No one seemd to bother about it for ages, though it was quite a kludge-y job (the material on which the kilometer distances was printed was a noticeably different shade from the rest of the sign).
Then some anti-Euro hacks raised Cain in the local newspaper and the council got round to restoring the sign. No one ever claimed responsibility for it. I quite liked it myself: I prefer to cycle kilometres rather than miles because you go further.;-)
I couldn't reproduce the begin exploit in Outlook Express 5.0 running on my Mac.
It's been my default email client for over three years now; it's really nice to use (and was intuitively so from the start). Of course the best mail client is the one you are used to; much of the discussion seems post-hoc justification of ossifying prejudice.
As for rendering your email illegible to potential recipients... most people just want to communicate, not engage in platform wars. Deliberately making your email unreadable to others seems a bizarrely self-denying ordinance, but, hey, it's his life. Me? I just want to communicate.
Many of the comments have implied Slack is not for the newbie, but in fact Linux is not for the newbie; slack is just more honest, in not trying to bolt inadequate GUIs over a system that was built to be adminstered from the command line and text file.
Maybe if you're buying new hardware the more commercial distros are the way to go, but you can't beat slack for installing on old hardware: I'd tried to install SuSE 7.0, RedHat 7.1, and Debian 2.2 for an old Compaq server before I downloaded a pre-compiled kernel kernel off a slackware mirror, which worked like a charm. Short account of my adventures here: http://www.carnall.demon.co.uk/prosignia_linux.htm
And the lean installs are good too. You have to know what you're doing, but then you have to know what you're doing anyway. So far I've not felt any pain because I had a.tgz rather than a.rpm to install. As for Debian, well, ideologically it's the distro I'd most like to use, but the pain of the UI of the package management database they use ("dselect": the 'd' surely stands for 'devil') has always made it impossible for me to do anything useful with it. Instead it makes me feel physically sick. Presumably they don't update this because they want to keep the riff-raff out. Well, they've succeeded in my case.
I'm aware of Debian's theoretically greater prowess: my dream distro would bolt apt-get like updating over a network to slackware's spare elegance. Maybe one day I'll know enough to it myself (don't hold your breath).
I'm curious about how the quirky one managed to demonstrate mastery of the system he was being employed to develop, yet so spectacularly failed when he actually had to produce the goods. Certainly it seems like he didn't have single-minded attention on the job in hand (but even engineers deserve a life).
I'd be interested to hear his side of the story: there could have been plenty of internal organisational reasons or technical reasons why he didn't gel, which the author of the piece may not be so forthcoming about.
And, if his body odour was really a problem to his inter-relationships with colleagues (and it sounds like it was), why did he not obtain medical help? (Or work from home?)
I have no personal experience of the PathLAN software but the principles underlying the story are straightforward.
It is possible to screen to see whether the pregancy will result in a Down's syndrome child using a number of tests.
The most accurate test is amniocentesis, when a sample of the amniotic fluid in which the foetus floats is removed early in the pregnancy. The foetal cells in the fluid can then be analysed (karyotyped) and their chromosomal pattern determined. (Down's is usually caused by an extra chromosome 21).
Unfortunately, amniocentesis, though a definitive test, is not itself without risk. There is a slight chance--somewhere around 1 in 200--that the test itself will lead to abortion (loss of the pregnancy).
So a lot of energy over the years has gone into selecting pregnancies at higher risk of Down's Syndrome. The best strategy is still up for debate, but basically, tests on maternal blood and also foetal ultrasound have been shown in large trials to differentiate successfully between high and low risk pregnancies. And of course, the age of the woman is a significant risk factor (the risk of Down's Syndrome rises sharply over the age of 35).
Presumably the PathLAN software was programmed to produce an overall risk score using data on maternal age, and from maternal blood and ultrasound tests. The age of the women was incorrectly calculated by the software, altering the overall risk score, and the women were subsequently incorrectly advised on the need for amniocentesis.
I don't know whether PathLAN was a proprietary or an open source system, but it is precisely because of occurrences like this that we must have open source software development methods if we are going to get serious about using computers to support clinical practice in medicine. We need to open the code, as well as the medical knowledge it implements, to proper peer review. And once we have done that, we have an ethical duty to share that peer-reviewed code so that no other patient suffers needlessly from a similar error again.
Max Weber--the father of sociology--attributed the inception of capitalism to the protestant need to do good works on earth for reward in heaven, this being a change from the Roman Catholic cycle of sin, absolution, redemption etc.
Ironically, now we anticipate the Flood (global warming) that is the direct consequence of this sequence of thoughts and behaviours.
Clothing tech is fundamental to all humanoids
on
Nano-pants
·
· Score: 2
It is believed that the Cro Magnons superceded the Neanderthals in Northern Europe largely because of the greater thermal efficiency their sewn clothing gave them.
While I agree that Gore-Tex has many fine qualities, it seems to me that it's biggest problem is its shiny nylon-ness, which precludes its wearing in many "civilian" situations.
The prospect of wearing regular fabrics that have been given an invisible nano-shield to our delightful British weather is a very exciting one.
Yes, I cycle for transport year round. No I don't necessarily want to look like "A CYCLIST"
Stallman's take on this is that the founding fathers of the US constituion explicitly rejected the notion that copyright holders have the the right to unlimited copyright on their work; rather, that the public traded their natural right to copy in exchange for the public good of encouraging more copyrightable works.
As someone who cycles on most journeys I can't *wait* for speeding enforcement to be taken as seriously as, say, drink driving. It is equally dangerous.The speed limit in urban areas should be reduced to 20 mph (which would cut road deaths by two thirds) and enforced by pedestrians, cyclists, and other motorists.
Slashdot Mirror
The Chinese system (and now the UK's) does not overly concern itself with ensuring every possible route through the Great Firewall is blocked--after all, geniuses gonna be ingenious. It merely suffices that the overwhelming majority cannot do this easily, and the technically adept live in fear of the consequences of their clearly pre-meditated actions. The censorship, and the chilling effect, is in the general drag; no need to close down every last loophole.
If John Allsup is right above, all that is required to get around this is "a LAMP stack somewhere on the internet" and some code. What percentage of the population are capable of actually doing this in practice? Taking say, the desktop share of Linux/BSD as a proxy for this, which is probably an overestimate, given the ease of installation of modern distros, that's fewer 3% of all internet users. A repressible minority? The UK government evidently thinks so.
It would help focus peoples' minds on the harmfulness of speed if, rather than setting arbitrary thresholds, the fines were proportional not only to income, but also to the excess kinetic energy of the vehicle:
Ek = 1/2 . m . v^2
Kuisla was doing 65 in a 50 zone, so his "kinetic tariff" would be (65x65) - (50x50) = 4225 - 2500 = 1725 units. (It's the exponential nature of the velocity squared factor that yields the disincentive, but, by all means introduce the weight of the miscreant's vehicle too)
This would be a rational way of reflecting the a) the risk of injury in the event of a collision; b) the undesirable environmental effects (noise, gaseous emissions), both of which rise exponentially with the kinetic energy of the vehicle.
This kinetic tariff can then be applied to income data to calculate the actual fine for the individual.
The choice about whether or not to encrypt traffic should be left to each website's administrator. Many sites--shock!--use the web to disseminate information they wish to be public, and the site's users have no problem with their access to it being public either. So get out of their faces! Using the browser to deprecate admin's particular choices is contrary to the spirit of the web, which should always do its damndest to serve something, and degrade gracefully when it's in difficulty, not pop annoying dialog boxes in the user's way.
Self-certificates are already a fairly effective denial of service attack when Firefox is used to access many independent sites that try to implement https, but who fail to do so in a way that offers a smooth user experience to J. Random User (I'm thinking particularly of IndyMedia).
Please note: in China, the censorship does not rely on blocking everything; just on blocking enough that all but the very motivated fail to access it. This troublesome minority can presumably be picked off at leisure later.
Keep it simple, stupid!
yeh, the French know that sacking the public sector in times of crisis does not help the economy; quite the reverse in fact. M. Hollande is old school ENA (Ecole Nationale d'Administration) which turns out highly-educated senior French bureaucrats and politicians, who, whatever else they may be, are not daft.
Diigo seems to work quite handily, but I still prefer Delicious and will mourn its passing. It's easy to export your Delicious bookmarks to an html file and import that into Diigo.
A single seed bank like this doesn't make any kind of biological sense. It is remarkably unlikely to be useful in the event of catastrophe: it's a long old road up there to Norway to replenish stocks of some ancient carrot variety from most parts of Europe.
If you actually wanted to guard biodiversity, you would encourage social networks of gardeners to replant varieties each season and share the ensuing seeds. The French organisation Kokopelli does this, but seems to suffer from legal harassment rather than incur the subsidies it would receive in any sane world.
An analogy for the slashdot crowd might be Napster (centralised) vs. BitTorrent (distributed).
that moldy alfafa will have made really nice mulch/compost for growing veg. Your old computers on the other hand are destined to pollute a water table somewhere.
As 90% of the data for diagnosis comes from the history-taking (interviewing) the patient, the potential for automating/supporting diagnosis is exciting.
Imagine a system that listens to a consultation in real time, making helpful suggestions for diagnosis based on analysis of the patient and the doctor's phoneme streams! And no tedious data entry, just an unobtrusive microphone.
I've been waiting for this.
It is a good piece of art because it reminds us that the person who builds the system (the road authority) will not necessarily anticipate the information needs of those who use it (joe drivers).
;-)
The solution was rather a drastic hack. The road authority should run a website where you could make suggestions for improved signage. These could be implemented in an economic manner when the signs are upgraded for other reasons.
It's the same thing with software: it's good to design it right: but you're still going to have to capture the thoughts of naive users when they encounter it for the first time, and take their expectations into account when you're upgrading the product.
Here in Hackney in East London we had a political sign hack for a while: someone converted a local route sign from miles to kilometres, presumably to make a point about closer integration with Europe. No one seemd to bother about it for ages, though it was quite a kludge-y job (the material on which the kilometer distances was printed was a noticeably different shade from the rest of the sign).
Then some anti-Euro hacks raised Cain in the local newspaper and the council got round to restoring the sign. No one ever claimed responsibility for it. I quite liked it myself: I prefer to cycle kilometres rather than miles because you go further.
I couldn't reproduce the begin exploit in Outlook Express 5.0 running on my Mac.
It's been my default email client for over three years now; it's really nice to use (and was intuitively so from the start). Of course the best mail client is the one you are used to; much of the discussion seems post-hoc justification of ossifying prejudice.
As for rendering your email illegible to potential recipients... most people just want to communicate, not engage in platform wars. Deliberately making your email unreadable to others seems a bizarrely self-denying ordinance, but, hey, it's his life. Me? I just want to communicate.
I had a similar problem last week. Some spammer was claiming to be me. I got about 800 returned mails/24 hrs :-(
I'm more concerned about the effect of the ones that do get through on my reputation.
Contacted my ISP (Demon) who said there was nothing much they could do.
I set up filtering, so it hasn't affected my work too much, but it is an annoying waste of resources.
Fortunately the storm appears to have abated for now, or I would have to devote some of my copious free time to researching a more durable fix.
Many of the comments have implied Slack is not for the newbie, but in fact Linux is not for the newbie; slack is just more honest, in not trying to bolt inadequate GUIs over a system that was built to be adminstered from the command line and text file.
Maybe if you're buying new hardware the more commercial distros are the way to go, but you can't beat slack for installing on old hardware: I'd tried to install SuSE 7.0, RedHat 7.1, and Debian 2.2 for an old Compaq server before I downloaded a pre-compiled kernel kernel off a slackware mirror, which worked like a charm. Short account of my adventures here: http://www.carnall.demon.co.uk/prosignia_linux.htm
And the lean installs are good too. You have to know what you're doing, but then you have to know what you're doing anyway. So far I've not felt any pain because I had a .tgz rather than a .rpm to install. As for Debian, well, ideologically it's the distro I'd most like to use, but the pain of the UI of the package management database they use ("dselect": the 'd' surely stands for 'devil') has always made it impossible for me to do anything useful with it. Instead it makes me feel physically sick. Presumably they don't update this because they want to keep the riff-raff out. Well, they've succeeded in my case.
I'm aware of Debian's theoretically greater prowess: my dream distro would bolt apt-get like updating over a network to slackware's spare elegance. Maybe one day I'll know enough to it myself (don't hold your breath).
This piece raises more questions than it answers.
I'm curious about how the quirky one managed to demonstrate mastery of the system he was being employed to develop, yet so spectacularly failed when he actually had to produce the goods. Certainly it seems like he didn't have single-minded attention on the job in hand (but even engineers deserve a life).
I'd be interested to hear his side of the story: there could have been plenty of internal organisational reasons or technical reasons why he didn't gel, which the author of the piece may not be so forthcoming about.
And, if his body odour was really a problem to his inter-relationships with colleagues (and it sounds like it was), why did he not obtain medical help? (Or work from home?)
It all smells a bit one-sidedly fishy to me.
I have no personal experience of the PathLAN software but the principles underlying the story are straightforward.
It is possible to screen to see whether the pregancy will result in a Down's syndrome child using a number of tests.
The most accurate test is amniocentesis, when a sample of the amniotic fluid in which the foetus floats is removed early in the pregnancy. The foetal cells in the fluid can then be analysed (karyotyped) and their chromosomal pattern determined. (Down's is usually caused by an extra chromosome 21).
Unfortunately, amniocentesis, though a definitive test, is not itself without risk. There is a slight chance--somewhere around 1 in 200--that the test itself will lead to abortion (loss of the pregnancy).
So a lot of energy over the years has gone into selecting pregnancies at higher risk of Down's Syndrome. The best strategy is still up for debate, but basically, tests on maternal blood and also foetal ultrasound have been shown in large trials to differentiate successfully between high and low risk pregnancies. And of course, the age of the woman is a significant risk factor (the risk of Down's Syndrome rises sharply over the age of 35).
Presumably the PathLAN software was programmed to produce an overall risk score using data on maternal age, and from maternal blood and ultrasound tests. The age of the women was incorrectly calculated by the software, altering the overall risk score, and the women were subsequently incorrectly advised on the need for amniocentesis.
I don't know whether PathLAN was a proprietary or an open source system, but it is precisely because of occurrences like this that we must have open source software development methods if we are going to get serious about using computers to support clinical practice in medicine. We need to open the code, as well as the medical knowledge it implements, to proper peer review. And once we have done that, we have an ethical duty to share that peer-reviewed code so that no other patient suffers needlessly from a similar error again.
Support the Open Source Healthcare Alliance: http://www.oshca.org/
His book The protestant ethic and the spirit of capitalism traces the historical antecedents very clearly. Check out Chapter 2 (54k) in particular for an exposition of how maximising business profit became a devout obligation.
Ironically, now we anticipate the Flood (global warming) that is the direct consequence of this sequence of thoughts and behaviours.
It is believed that the Cro Magnons superceded the Neanderthals in Northern Europe largely because of the greater thermal efficiency their sewn clothing gave them.
While I agree that Gore-Tex has many fine qualities, it seems to me that it's biggest problem is its shiny nylon-ness, which precludes its wearing in many "civilian" situations.
The prospect of wearing regular fabrics that have been given an invisible nano-shield to our delightful British weather is a very exciting one.
Yes, I cycle for transport year round. No I don't necessarily want to look like "A CYCLIST"
The logical conclusion: drive a tank. And then we can all swim to work when your CO2 emissions melt the ice-caps.
Hence, the expiration of copyright.
As someone who cycles on most journeys I can't *wait* for speeding enforcement to be taken as seriously as, say, drink driving. It is equally dangerous.The speed limit in urban areas should be reduced to 20 mph (which would cut road deaths by two thirds) and enforced by pedestrians, cyclists, and other motorists.