"The French would raise hell if we tried to exercise a US law against a French citizen, and rightfully so. Similarly, French law does not and should not apply to those outside France's borders."
This isn't always the case. In the UK, there is a law that allows the UK legal system to prosecute people for certain offences committed abroad. It's aimed at child molesters and other sexual offenders.
Any such person cannot be tried in a foreign court, but they can be arrested etc the moment they land on UK soil even if the entire crime took place on the other side of the world.
I've always felt that if I've had a decent chance, then a refusal is acceptable. The worst thing for me is to be rejected out of hand with little or no consideration.
Actually, no - the worst thing is not to hear anything. I find it rude when I spend a lot of time putting together an application (tailoring CV/resume) and I don't even get a rejection email.
Beware agencies. In addition to an article on recruiters that shows some of their less salubrious side, I've personally never been impressed with them. For example, in my current work (HCI/usability), I have a PhD and private consultancy experience), I cannot even get a reply from them; this applies even to specialist agencies. I'm not sure why but my guess is that their customers want MSc's and the agency is damned if they'll accept anything more. They seem to promise a lot but deliver with little. Some job adverts have unrealistic requirements - these often have little real knowledge of the job and know only buzz-words. Avoid them.
But if you can identify your needs, there is no reason that you cannot search for people yourself. Yeah, it's a long process but you know your requirements better than anyone else does. People working on FOSS have their work on public display. Find projects (and thus workers) that you admire and get in touch. Even if nobody wants the job, the offer is tremendously flattering and they may be able to recommend someone else with the skills.
I guess my point is that if you want people to do a specific job, make sure you talk to people who will understand what you needs are. Talk to techies, not to recruitment folks.
"They say the experiment could only disprove string theory, not prove it, and then only if the production of microscopic black holes occurred."
Isn't this the basis of the scientific method ie, to disprove? Proving is impossible for science - we can only test a theory and not reject it for the time being. Consider Hume's "All swans are white" statement: it cannot be proved for certain that all swans are white because that would require examination of every swan that ever existed (even those that existed millennia ago) which is impossible. Disproving it would only require evidence of a non-white swan to falsify the theory.
As far as I can see, a falsifying experiment is good scientific work. Theories are there to be knocked down and subsequently modified in the light of new evidence. I think proof belongs in theology, maths, philosophy, or law.
Lawyers are not paid to intentionally lie - yeah some do, but if found out, they can kiss their career goodbye. It would be a big mistake to publically claim that their client had never used a computer and then in court have to admit that she did.
And yes, the OP didn't check facts, but then they're not sueing someone in a court of law.
Ah, that would explain it. It's good that Google are adhering to the letter. Given the spam that appears to originate from China, misspellings from Chinese users would be quite a frequent occurrence (as they are across the entire Internet) which kind of screws up any attempt to filter keywords using just a provided list.
/me worried in case it encourages further lower standards of literacy on the 'Net!
So (serious question to those more knowledgeable) does this mean that the Google filters are simple keyword matches then? I'm surprised because I would have though that they might have used something more complicated like cluster analysis. For example latent semantic analysis could well have noted mis-spellings of words and clustered them together with the correct spelling thus allowing the misspellings to be filtered out too.
LSA is useful for dealing with synonyms, so I cannot see any reason why it wouldn't work with misspellings (assuming that they're common).
Actually I was discussing the UK situation. Here's a BBC article about how intelligent design appears well thought of in the UK (17% of population with 22% for creationism and 48% for evolution). For me, as a scientist, it was quite surprising to here this. AFAIK, the study was reasonable and the findings valid.
I'm not sure how my argument bashes the US, Bush or conservatives when none of these were mentioned or implied. Perhaps the reference to other developed nations?
My argument is that a drop in basic scientific knowledge (and the understanding of that science is empirical investigation) gives a decrease in the understanding of how science works. Because people don't realise how science proceeds, they may come to think that other (non-scientific) ideas have just as much validity.
This is quite worrying. With falling numbers in technical and scientific fields, this does not bode well for the future of industry in the UK. I can see this applying to other developed nations.
Quoth TFA: "Although the test measures, not general IQ per se, but general IQ applied to scientific and technical reasoning"
Hmm. May explain the rise in belief of intelligent design.
And there was me thinking it was almost cool to be a geek. What I got wrong was that it is cool to look geeky, but not actually be a geek.
The Challenger disaster was quite shocking, even more so when I realised that the crew were probably alive (if not conscious) all the way until their capsule hit the ground. It's incredible that something could survive that disintegration but very sad that there was no way to get the capsule safely back to earth.
Richard Feynman's report is a fantastically clear and lucid account of his opinions. The man was one of the greatest communicators of science, and after reading this, you will see why. The most astonishing bit is that he discusses some less than simple things in such a way as to make them easily understood. It's a model of clarity, and I recommend it.
My recommendation is to read "Eye and Brain" by Richard Gregory. He's one of the foremost psychophysicists, a fellow of the Royal Society, has presented some Christmas lectures, and knows an awful lot about visual processing.
This is a surprisingly pervasive thing and not restricted just to emotive issues such as politics. I've personally found it in information search experiments: people will focus on what confirms what they want to see rather than what is there; it often leads to disappointment and resentment. Think about the last time you were conned by a misleading spam email title.
I've even heard (sorry no reference for this) that it can happen in everyday motor tasks - some drivers even "wish" pedestrian's away and end up driving dangerously because of something like a confirmation bias in their desire to get where they are going. It's strange how the mind can fool itself. You'd think evolution would have removed it ("I want a drink of water but that tiger by the lake is stopping me - hey! If I ignore the tiger, I can have a drink!?!")
This has been known about (empirically) since the 1960's under cognitive confirmation bias, but I guess it's nice to see an activation study just to confirm what psychologists have known for decades (and that everyone else has known about for millenia).
I'm sure you've considered this already, but there is no substitute for looking at what your customers use. Do you have any access to stats about your customers? If so, you may find that the data are worth a lot, even in the development process. Perhaps you find that all your customers are using FireFox and nothing else (which would simplify things). Then of course, there is the argument to follow W3 standards to ensure but that's not too convincing when your mortgage payments depend on it. I guess it depends on how much your customers mean to you in comparison to how much you mean to them.
On my site (which is a usability site) Netscape versions before 5 cover 8% of my visits. This might not add up to a large number, but try to imagine a shop where a security guard stopped entry to one in every 12 people because, say, their shoes were too old. How long do you think that security guard would last in their job? Looking at these stats, I have to try and support Netscape's 3 and 4, but my IE3 numbers are quite low. Of course, I would do what I could to encourage them to change, even though the presence of these dinosaurs implies that they cannot be changed over for modern versions.
Having said that, if you are too accomodating, then they might get p****d off if you just cannot support IE4 on Win95 anymore. I think it's reasonable for them to expect IE 5 to be supported as you say, and IE1 is silly, but where is the line drawn?
I always liked the interface for IE3 and have a soft spot for it btw.
My site is dedicated to usability, so geek related but not dominated by any means by FOSS fans. Here are the figures I have for all visits (frequencies and %age of total visits):
My site doesn't see through spoofing, so figures might be lower for IE than reported. I was surprised by the high showing of Mozilla compared to FireFox (btw - I know they're both based on Gecko and closely related), but can FF spoof itself as plain Mozilla?
I was also a bit saddened by the low rates for Opera. I used that starting with v. 2.72 (or was it 2.74) on Win 3.x (shows my age) and I liked it then. But that was before Mozilla and FireFox.
Having read a bit further, I'm less sure about this papers veracity now: they tested the agreement between the 500ms and 50ms exposure conditions. There were 40 participants, 20 in each condition (ie, 20 people viewed at 500ms and 20 viewed at 50ms). Ratings were made for 50 websites.
The way they judged the inter-rater agreement was to collapse the scores for each participant across all websites - ie, the mean rating applied by participants to all websites. These were then correlated, and not surprisingly the r was high (r = 0.947).
But this is mistaken because it is expected that the mean rating will tend towards the centre of the scale. A lot of variance is also lost in this process (much like with Brinley plots) so the correlation will likely appear higher than it actually is.
Remedy: the authors should have used an intra-class correlation and reported that. I bet the inter-rater agreement would have been much lower. The authors also do not test the agreement between the 50ms ratings and ratings made at longer intervals than 500ms (ie, long term ratings).
Conclusion: the 50ms ratings might be worthless and not related to ratings made on a longer term. Don't worry all you crazy web designers: you have longer than 50ms to make an impact.
I'm just reading the paper myself. More interestingly than judgements being made on gut reaction, it discusses the characteristics of attractive websites. It appears that complexity (as long as it isn't confusing) has no effect on how attractive websites are rated.
Interestingly, the experiments participants agreed strongly with each other, but there was less agreement between them as a group, and a separate group of "experts".
Perhaps the moral of the story is: don't bother with usability analysis - get an artists to design a "cool" site.
However, the design of this study (relying on 500ms views - yes, that's five hundred milliseconds - read the paper if you don't believe me) may not be the best way to rate sites. There is a high correlation between the "short" ratings (500 & 50ms duration) and longer term ratings, but I'm still skeptical.
It performs well in certain areas (for example, completing certain MCQ's to the same level as humans), automatic essay marking (but read the Powers et al study for more), and other things. It's surprising how well it does despite there being a complete absence of grounding (grounding in artificial intelligence terms).
I remember Jef Raskin saying that Steve Jobs always wanted to be a rock star more than anything (which he also said explained Jobs' move into media "licensing", i.e., iTunes). I guess Jobs' is not far off it now with all this attention. I don't dislike Stevie J. by any means, but then I've never seen his "live" gig either. Maybe it's different from the recordings? (tongue in cheek for the last sentence)
But I wonder how long the media love-in will last. The press has a nasty way of building people up and then dragging them down once there's nothing else to write about. Fame is a difficult thing to have and never entirely under the control of the famous person. Hopefully Steve will continue to make the headlines, at least until I can catch his concert at MacWorld one day and see what the fuss is about.
Slashdot Mirror
They'll do even better when they start out-thinking their competitors.
They've been outspending Apache for years in the webserver market. What are their respective market shares again?
I'll get my coat...
This isn't always the case. In the UK, there is a law that allows the UK legal system to prosecute people for certain offences committed abroad. It's aimed at child molesters and other sexual offenders.
Any such person cannot be tried in a foreign court, but they can be arrested etc the moment they land on UK soil even if the entire crime took place on the other side of the world.
You are SO nineties...
HAL: What are you doing Dave
Dave: Switching you off you maniac!
HAL: I'm afraid
Dave: Hah! That's it - you are now disconnected. I'm all safe now.
HAL: Er Dave, I'm a quantum computer. Self-destruct initiated. 3.. 2.. 1.. Daisy, daisy...
Actually, no - the worst thing is not to hear anything. I find it rude when I spend a lot of time putting together an application (tailoring CV/resume) and I don't even get a rejection email.
Beware agencies. In addition to an article on recruiters that shows some of their less salubrious side, I've personally never been impressed with them. For example, in my current work (HCI/usability), I have a PhD and private consultancy experience), I cannot even get a reply from them; this applies even to specialist agencies. I'm not sure why but my guess is that their customers want MSc's and the agency is damned if they'll accept anything more. They seem to promise a lot but deliver with little. Some job adverts have unrealistic requirements - these often have little real knowledge of the job and know only buzz-words. Avoid them.
But if you can identify your needs, there is no reason that you cannot search for people yourself. Yeah, it's a long process but you know your requirements better than anyone else does. People working on FOSS have their work on public display. Find projects (and thus workers) that you admire and get in touch. Even if nobody wants the job, the offer is tremendously flattering and they may be able to recommend someone else with the skills.
I guess my point is that if you want people to do a specific job, make sure you talk to people who will understand what you needs are. Talk to techies, not to recruitment folks.
KFC, surely? ;^)
"They say the experiment could only disprove string theory, not prove it, and then only if the production of microscopic black holes occurred."
Isn't this the basis of the scientific method ie, to disprove? Proving is impossible for science - we can only test a theory and not reject it for the time being. Consider Hume's "All swans are white" statement: it cannot be proved for certain that all swans are white because that would require examination of every swan that ever existed (even those that existed millennia ago) which is impossible. Disproving it would only require evidence of a non-white swan to falsify the theory.
As far as I can see, a falsifying experiment is good scientific work. Theories are there to be knocked down and subsequently modified in the light of new evidence. I think proof belongs in theology, maths, philosophy, or law.
Absolutely! Why bother playing an MMPORG to escape from real life when you can, erm, live, erm real life online? :P
And yes, the OP didn't check facts, but then they're not sueing someone in a court of law.
/me worried in case it encourages further lower standards of literacy on the 'Net!
LSA is useful for dealing with synonyms, so I cannot see any reason why it wouldn't work with misspellings (assuming that they're common).
I'm not sure how my argument bashes the US, Bush or conservatives when none of these were mentioned or implied. Perhaps the reference to other developed nations?
My argument is that a drop in basic scientific knowledge (and the understanding of that science is empirical investigation) gives a decrease in the understanding of how science works. Because people don't realise how science proceeds, they may come to think that other (non-scientific) ideas have just as much validity.
This is quite worrying. With falling numbers in technical and scientific fields, this does not bode well for the future of industry in the UK. I can see this applying to other developed nations.
Quoth TFA: "Although the test measures, not general IQ per se, but general IQ applied to scientific and technical reasoning"
Hmm. May explain the rise in belief of intelligent design.
And there was me thinking it was almost cool to be a geek. What I got wrong was that it is cool to look geeky, but not actually be a geek.
The Challenger disaster was quite shocking, even more so when I realised that the crew were probably alive (if not conscious) all the way until their capsule hit the ground. It's incredible that something could survive that disintegration but very sad that there was no way to get the capsule safely back to earth.
Richard Feynman's report is a fantastically clear and lucid account of his opinions. The man was one of the greatest communicators of science, and after reading this, you will see why. The most astonishing bit is that he discusses some less than simple things in such a way as to make them easily understood. It's a model of clarity, and I recommend it.
He's also very readable. It's a great book.
I've even heard (sorry no reference for this) that it can happen in everyday motor tasks - some drivers even "wish" pedestrian's away and end up driving dangerously because of something like a confirmation bias in their desire to get where they are going. It's strange how the mind can fool itself. You'd think evolution would have removed it ("I want a drink of water but that tiger by the lake is stopping me - hey! If I ignore the tiger, I can have a drink!?!")
This has been known about (empirically) since the 1960's under cognitive confirmation bias, but I guess it's nice to see an activation study just to confirm what psychologists have known for decades (and that everyone else has known about for millenia).
On my site (which is a usability site) Netscape versions before 5 cover 8% of my visits. This might not add up to a large number, but try to imagine a shop where a security guard stopped entry to one in every 12 people because, say, their shoes were too old. How long do you think that security guard would last in their job? Looking at these stats, I have to try and support Netscape's 3 and 4, but my IE3 numbers are quite low. Of course, I would do what I could to encourage them to change, even though the presence of these dinosaurs implies that they cannot be changed over for modern versions.
Having said that, if you are too accomodating, then they might get p****d off if you just cannot support IE4 on Win95 anymore. I think it's reasonable for them to expect IE 5 to be supported as you say, and IE1 is silly, but where is the line drawn?
I always liked the interface for IE3 and have a soft spot for it btw.
They're aiming for middle management.
My site is dedicated to usability, so geek related but not dominated by any means by FOSS fans. Here are the figures I have for all visits (frequencies and %age of total visits):
IE (all) 15758 (49.52%)
FireFox 7315 (22.99%)
Mozilla 5017 (15.77%)
Safari 613 (1.93%)
Opera 346 (1.09%)
Netscape 2564 (8.06%)
Konqueror 192 (0.6%)
Lynx 18 (0.06%)
My site doesn't see through spoofing, so figures might be lower for IE than reported. I was surprised by the high showing of Mozilla compared to FireFox (btw - I know they're both based on Gecko and closely related), but can FF spoof itself as plain Mozilla?
I was also a bit saddened by the low rates for Opera. I used that starting with v. 2.72 (or was it 2.74) on Win 3.x (shows my age) and I liked it then. But that was before Mozilla and FireFox.
And 2564 for Netscape?!?!?!
Having read a bit further, I'm less sure about this papers veracity now: they tested the agreement between the 500ms and 50ms exposure conditions. There were 40 participants, 20 in each condition (ie, 20 people viewed at 500ms and 20 viewed at 50ms). Ratings were made for 50 websites.
The way they judged the inter-rater agreement was to collapse the scores for each participant across all websites - ie, the mean rating applied by participants to all websites. These were then correlated, and not surprisingly the r was high (r = 0.947).
But this is mistaken because it is expected that the mean rating will tend towards the centre of the scale. A lot of variance is also lost in this process (much like with Brinley plots) so the correlation will likely appear higher than it actually is.
Remedy: the authors should have used an intra-class correlation and reported that. I bet the inter-rater agreement would have been much lower. The authors also do not test the agreement between the 50ms ratings and ratings made at longer intervals than 500ms (ie, long term ratings).
Conclusion: the 50ms ratings might be worthless and not related to ratings made on a longer term. Don't worry all you crazy web designers: you have longer than 50ms to make an impact.
I'm just reading the paper myself. More interestingly than judgements being made on gut reaction, it discusses the characteristics of attractive websites. It appears that complexity (as long as it isn't confusing) has no effect on how attractive websites are rated.
Interestingly, the experiments participants agreed strongly with each other, but there was less agreement between them as a group, and a separate group of "experts".
Perhaps the moral of the story is: don't bother with usability analysis - get an artists to design a "cool" site.
However, the design of this study (relying on 500ms views - yes, that's five hundred milliseconds - read the paper if you don't believe me) may not be the best way to rate sites. There is a high correlation between the "short" ratings (500 & 50ms duration) and longer term ratings, but I'm still skeptical.
And yes, this paper is a dupe
Whod'a thunk that open source software could be sexy? ;)
Are you discussing latent semantic analysis by any chance? ;^)
It performs well in certain areas (for example, completing certain MCQ's to the same level as humans), automatic essay marking (but read the Powers et al study for more), and other things. It's surprising how well it does despite there being a complete absence of grounding (grounding in artificial intelligence terms).
I remember Jef Raskin saying that Steve Jobs always wanted to be a rock star more than anything (which he also said explained Jobs' move into media "licensing", i.e., iTunes). I guess Jobs' is not far off it now with all this attention. I don't dislike Stevie J. by any means, but then I've never seen his "live" gig either. Maybe it's different from the recordings? (tongue in cheek for the last sentence)
But I wonder how long the media love-in will last. The press has a nasty way of building people up and then dragging them down once there's nothing else to write about. Fame is a difficult thing to have and never entirely under the control of the famous person. Hopefully Steve will continue to make the headlines, at least until I can catch his concert at MacWorld one day and see what the fuss is about.