I do the same, originally for privacy reasons, but mostly now just because it's handy. (Although I prefer elinks, if only for the "invoke external editor on text area" option. If you ask me, that one ought to be right up there with "give users a place to enter an URL" as a necessary feature to include when designing a browser.)
Best part is that I can leaving it running on a screen session and pick it up to browse or continue a comment whenever there's a free minute near any computer.
>Look: do you own a printer? ever >noticed you can't buy a printer >with "standard non proprietary" >cartridges? If I follow your train >of thought, you should be outraged, >no? Of course not, you keep printing.
There's a big difference. Most people only own one printer, and cartridges are disposable and are used up many times over the lifetime of a product. No one has to worry whether their library of old printer cartridges will work with their new printer, no one buys a dozen different printers and a single cartridge in order to save money or to allow different printers to communicate with eachother.
On the other hand, people often own many devices that use the class of media being discussed here, and the media often outlive the products themselves. And, unlike in the world of printers, in this field most players have chosen to go with well supported standards.
I'm not outraged at Sony, but I'm also not going to buy their products. If they were obviously and consistantly superior to their competitors, they might be able to pull it off. But they haven't been so for a long, long time.
Given a choice between products with indistinguishable features, I'll take the one that uses standard interfaces and media every time.
That's a developmental strategy, not a support strategy. It's quite orthogonal to the subject at hand.
The two issues are far from orthogonal.
The cost of supporting minority browsers depends critically on how much browser-specific material you design into the website.
Likewise, the decision to sacrifice specific examples of standards compliance or graceful degradation can't be made without considering which browsers you expect to work with the site.
High contrast colors, a dark background, and a large font really help me.
At least, they prevent my eyes from feeling tired and irritated after a day on the computer, and I've been headache free for years. (Don't know if it has any bearing on long term vision problems.)
On the other hand, I've occasionally borrowed friend's computers and felt my eyes fizzle and become sore in minutes. Any time I look at small text in black on a white background, it bugs me. (Don't even get me started on semi-transparent terminals. Why on earth anyone would want to turn down the contrast on their terminal and then past a bunch of distracting stuff all over the text is beyond me.)
I generally make the text large enough that i can easily sit back forty inches or more from the monitor and choose high contrast combinations (slightly off-white or bright green on black, white on navy blue or magenta, etc.) I also avoid anti-aliased fonts, but that's mostly an esthetic choice. In general, I find that if my font is so small that anti-aliasing is useful, it's *too* small to be comfortably read anyway.
And, to satisfy curiousity, what on earth have you made your desktop out of that it doesn't have either built in sound or spare pci slots?
Resources
on
Why KDE Rules
·
· Score: 2, Interesting
[amaroK:]eats *lots* of RAM, but when you look at the alternatives you realize that they eat 50% of the resources amaroK uses to implement less than 25% of the equivalent functionality.
I say, give me something that will eat 10% of the resources and provide 5% of the functionality. Then give me 20 other somethings that do the same, but provide different functionality, and we'll have all of amaroK's functionality with only 10% of the resource commitment.
That's more or less my principle complaint against both KDE and Gnome: in order to get the hand full of features that could be useful, one has to add a bunch of stuff that at best isn't used and at worst gets in the way.
But, the good news is those 20 programs *do* exist, and I can use them. So long as no one is forcing me to use amaroK, I'm glad to see that it exists and that someone enjoys it.
In fact, it brings me great joy to know that both Gnome and KDE exist and that there are fans out there talking each other into using them.
First, people find them useful, and anything that improves people's lives (even in a very small way) is a good thing.
Second, they both help attract new users to free unix-like software, and give talented developers something fun to work on. Anything that makes linux and the BSD's more popular is good for those of us who love them.
Third, I'm happy that *both* gnome and kde exist, because so long as both maintain a large following developers will have no choice but to make projects able to run without requiring either. That's good news for those of us who prefer alternative environments. The day one wins is the day third party apps stop simply requiring libraries and start requiring that the winner be actually running.
3-4% volume is the max for your headphones? Does your portable player also work as a rock concert amplifier?
It's pretty standard stuff. Just happened to hit on a very loud combination. An iriver ifp-890 flash player and sony MDR-EX51LP earbuds. (Incidently, they're not the best sounding headphones I've ever heard, but surprisingly close to it for relatively cheap earbuds. I recommend them highly.)
Playing an mp3 with the default equalizer settings, normal listening is 2/40. (For normal listening I cut down all the equalizer channels in order to get some more resolution on the volume control.) For ogg files and radio I have to increase the volume to something more like 10-14.
Thank you for the background information, and I apologize for what was rather a smart-ass response. (One also full of grammatical errors. Looks like I was more tired than I knew at the time.)
I appreciate the article reference. For anyone else interested, the full reference is Ear & Hearing. 25(6):513-527 (Available from their website only with a personal subscription. Try a university library - mine has electronic copies available via a journal aggregator.)
It appears the recommended threshold was chosen based on careful research and thoughtful analysis.
Still, it does seem to me a a legitimate question whether it's better to provide a concrete threshold with limited accuracy or to refuse to provide one at all.
Among the device and headphone combinations tested, the variation in sound level is quite dramatic: something like 25 dB for the two extreme cases, and judging by eye a standard deviation on the order of 7-10 dB. That suggests that even after dropping the recommendation from the measured level of 70% max volume (for 1 hour) to 60%, there are a significant number of consumers for whom the recommendation is still unsafe.
It might be that only influence such a recommendation will have is to cause those who are already listening to music which is much too loud to reevaluate their choice. In that case, providing a concrete number will only lead to a positive outcome. If someone isn't able to judge a safe volume on their own, giving them a reasonable number may be safer than giving them no number at all.
The danger is that someone will decide that an unsafe volume is okay based on that recommendation. There's power in concrete numbers presented by an authority figure, and it's easy to imagine a scenario in which someone who regularly listens to music at dangerous levels reads an article on the subject and determines that they need not worry about hearing damage because their volume setting is below the threshold. It's not clear to me how likely that sort of response is in practice.
It might well be that the benefits of providing a straightforward guideline far outweigh the danger that someone will read too much accuracy into it. I don't know nearly enough about the subject to put forward an answer. But, it does seem like a worthwhile discussion. (Of course it's always possible the discussion was concluded years ago, in which case an appropriate response is, "Everyone already knows the answer except you, you ignorant twit. Now go back to your own field and stop bugging us.")
We're working (Dr. Fligor and I) on a follow-up with mp3 players. Should be done in a couple months. I'm sure it will hit/. when it's published.
The rule of thumb suggested by researchers at Boston Children's Hospital is to hold the volume of a music player no higher than 60 percent of the maximum, and use it for only about an hour a day.
Of course for those of you whose volume controls go all the way to 11, that should be decreased to 54% of full scale.
Seriously, am I the only person who thinks this advice completely insane? 60% converted on an arbitrary scale, converted to sound pressure with a device-dependent conversion that varies wildly? I hope that comment was misquoted or taken out of context.
With portable and headphone I'm currently using, 60% full scale would blow out my ears in seconds, if the headphones didn't first destroy themselves in a valiant attempt to save my ears. It's true I've got a very good pair of heaphones and a pretty high-ampltude device, but anything above 3-4% max is too loud for comfort. On the other hand, I've listened to equipment at which full scale with all the equalizer channels maximized isn't loud enough.
It seems to me the problem hasn't got anything to do with ear-buds or portables. Unfortunately, it's a lot more widespread and hard to deal with than that. From the article:
"I have an audiologist friend at Wichita State University who actually pulls off earphones of students he sees and asks, in the interest of science, if he could measure the output of the signal going into their heads,'' Garstecki said. Often he finds students listening at 110 to 120 decibels.
120 dB? That's completely insane. Anyone who's *able* to stick 120 dB in their ears without wincing and making a desperate grab for the volume switch has already got some pretty serious hearing damage, if you ask me.
But that's not too surprising, since everything in our society is screaming at us constantly. I'm continually amazed by the amount of volume people tolerate on the subway, in movie theaters, in concerts. I now carry earplugs with me everywhere and end up using them several times a day. That sucks, because it's impossible to find earplugs with a nice flat frequency response, so you end up with badly colored audio. Nothing pissed me off more than paying money to listen to live music and then to have no choice but to hear it through muddy earplugs. Why the hell anyone *wants* to hear a jazz concert in a little tiny club at earsplitting volume is another question. I can only assume it's because we've all burned out our hearing elsewhere.
But, this is in danger of becoming a completely unrestrained rant, so I better quit.
After reading both the journal article, and the writeup linked here, I don't see any obvious problems with it.
Pouring a single shot into short 350 ml tumblers and 350 ml highball glasses are both pretty common activities when making drinks. Seems like a prefectly legitimate thing to investigate.
I'm sure that's true. There are a few other recent cases of people producing surprisingly good films in industrialized countries with budgets of a couple tens of thousands: Following, Clerks, to pick two good ones. (Tarnation, for a somewhat eccentric example, although I have no idea how much the distributor spent on music rights and post production before creating the finished product.)
Of course if you were to include a cash value for the wages for the actors and crew, loaned equipment, and production space as "donations," the budgets increase rather dramatically. Tens of thousands of dollars isn't the answer to the question "what did this film really cost" but rather, "with how little cash can someone with talented, generous friends and free time make a decent movie?" That's also an interesting question to ask.
And, if you allow budgets of a few hundred thousand dollars instead, it's certainly true that the number of examples increases dramatically.
That said, it's not obvious to me that "video blogging" or portable video players will have any impact on low budget film making. The huge improvements in consumer and cheapo professional grade video cameras in the last decade and the availability of cheap video editing software and desktop computers able to run it, on the other hand, are a lot more exciting.
>>As long as "CS251: Introduction to >>Funny Walks" remains an elective, >>I'm fine with it.
>I'm not, and you shouldn't be either.
I gather you're a proponent of removing core competency credit from the Introduction to Funny Walks course, on the grounds that it will devalue your degree?
Surely you wouldn't suggest removing the course entirely. I sure as hell wouldn't want to be the dean who has to explain to the Ministry why it's been removed.
Perhaps a compromise is in order?
Tell you what, we'll put CS251 on the non-credit bearing electives list, if you agree to add CS351 (Graduate Seminar on Silly Walks) to the list of 1 unit advanced electives and add BP24-A (Rectal Examination and the Search for a Sense of Humor in Uptight Slashdot Posters who Missed the Point) to the list of recommended interdisciplinary electives.
>Umm, if it were truly random, it would almost >certainly not be uniformly distributed.
Forgive my imprecision. The above should read, "assuming test answer key is truly random and drawn from a uniform probability distribution." (Or colloquially, "assuming the test answer key is random and that all choice labels have an equal probability of being right.")
>However, in practice, most tests are *intended* >to be either uniformly distributed or random, but >because they are generally constructed by humans >it seldom works out that way.
I suspect you're right, with the exception of the big standardized tests. I can't imagine they'd allow anything except an automatically generated random key for those. (And, for what it's worth, those do make up rather a large fraction of the multiple choice exams students take during their lives. More so if tests are weighted by importance.)
>Also, it is well documented that if there >are five choices, a through e, b and d are >statistically better bets than a, c, or e >(though this of course can vary with specific >teachers).
I don't doubt it. But, I'd be curious to find a source for such details. What humans do when you ask them to make something random is an interesting subject.
A great (if unrelated) example, which I came across in a Steven J. Gould essay, involves asking people to draw a star field of randomly placed dots on a piece of paper. They always end up with something much more regularly spaced than a truly random field. (Gould takes up the subject upon noting that glowing insects in a cave do the same thing - creating something that looks nothing like the night sky, despite being made up of thousands of points of light against darkness.)
Doctor Plaster's strategy sounds like a good one. There's still some nonzero random chance score, but it certainly makes it a lot harder to do well without having to include huge numbers of unique answers.
I took a test prep course and one piece of advice they gave was "pick a letter and if you don't know the answer, use it." The idea being that if you're really unsure, picking 1 letter will give you a consitent 25% chance of getting those wrong answers right.
I remember being taught the same thing. But, now that I stop to consider it, it seems to me that strategy is total bunk.
Assuming test answer key is truly random and uniformly distributed, the chance that you'll win by picking any particular answer to any particular question is a fixed value. (1 / the number of choices.) It can't possibly depend on what one answered for other questions.
So, whether you always choose the same answer, or choose an answer at random for each question, or spell our "abcdeabcdeabcde" makes no difference.
To convince myself that I'm not doing something dumb here, I ran a quick monte carlo simulation using random number generators. For a test consisting of 100 questions with 4 choices each, after one million tests the mean score and standard deviation are:
0.249993, 0.0433486 for totally random answers
0.250037, 0.0433170 for always choosing "C"
From that, the error on the mean (1 sigma) should be 4.32998e-05. The answer agree with each other and with the naive prediction. So, either strategy works equally well.
The only caveat is that we're assuming a truly random answer key. That means, for example, that "every answer to the exam is B" and "no answer is ever B" are both allowed answer keys. That's probably a reasonable assumption for an exam whose answer keys are generated automatically, such as any standardized test.
But, for a human-constructed test, it may be that large classes of solutions will be discarded by the test maker. It's not obvious how to model what a human would do in "randomly" laying out answers. Could be great fun to explore if one had access to thousands of scantron answer keys. (Anyone with a key to the exam repository room want to collaborate on actually doing that?) My guess is that since the vast majority of uniformly distributed random keys look qualitatively similar to human generated keys, the difference can't be large.
The obvious question is why test prep folks have been telling people bogus information for years? Either they are just bad at statistics (a scary thought, given the business they're in), or there's some psychological benefit to convincing people that picking the same letter will improve their score. Perhaps it's marginally faster during the exam? Hard to believe that's significant.
Redundant? Second thread in the whole linup, posted 8 minutes after the article, and it's modded redundant?
Anyway, on the topic at hand:
That's a very good point. It's certainly easier to make a professional sounding documentary/interview/polemic than it is to make a believable piece of narrative fiction.
But, it is also true that there has been a conspicuous absence of fiction in radio for a long time. Even among the big, relatively well funded public radio powerhouses, fiction is hard to come by, and what little there is tends to be the same stilted, rehashed stage plays that have been boring audiences for decades. With a few exceptions (Joe Frank, Selected Shorts, several short run BBC series, etc.), entertainment radio has come to mean almost exclusively documentary and interview. While it may be easier to produce good documentary, even if you have funds and access to talent, that in itself doesn't seem quite enough to explain why fiction has disappeared from the radio. Perhaps what the parent poster sees in postcasts is just an extension of the same phenomenon.
Last time I tried to do this (perhaps two years ago on windows XP) there didn't seem to be any way to do it except by messing with registry entries and rebooting, and the change happened for all users.
But, I readily confess to being clueless and all but unable to use windows. So perhaps there is a better solution out there and I simply wasn't able to find it.
A few quick google searches haven't found it this time either. (But I haven't searched exhaustively yet.) You don't happen to remember a source for info about how to set this up, do you?
Uh rights? You don't need any special privileges to run xmodmap.
True. (Well, technically you do need special permissions - but if you're doing anything useful in X you almost certainly have them anyway...)
But on less user-friendly operating systems (like windows), you do need special permissions to remap keyboards. What's more, it's usually impossible to remap them for only a specific user, or to remap them without rebooting. Real pain in the ass for those of us who are occasionally forced to use windows. (I have enough trouble switching back and forth between capslock/ctrl locations. Can't imagine changing all the other buttons too!)
At least, that was the case the last time I looked into it.
Yeah, I understand what it means in this context, and I don't fault the parent poster for using it. . . but, at some level it just doesn't make sense. Not sure what a better description would be, though. Perhaps "tiered 2D" or something of the sort?
Ultimately, the usual windowed computer screen allows the user to make use of a third dimension in almost exactly the same way that a desk with books and papers on it does. The user stacks things up in piles separated in the principle plane, and then brings whichever bit they want to work on to the center of that plane. Occasionally, they'll leave something else open and visible off to the sides of the main work area.
So long as you're using your computer to do the analog of what we do with desks, anything else is either eye-candy or will hinder getting things done.
That's not to say there aren't places in which 3D interfaces could really be interesting. Just they what most people spend most of their time doing on computers can't be made any easier by arranging things in 3D. At best, they'll end up using it in exactly the same way they do now. At worst, they'll have to go through a bunch of extra steps every time they rearrange things.
Yup. Insisting that people use default configs is pointless, except if your goal is to tweak those default configs.
Ideally, you ought to force people to interact with the machines over a long enough period of time to recognize transient effects associated with having previously used similar systems, and to allow you to appropriately characterize systems which are too willing to sacrifice efficiency in favor of transparency. (Or, even worse, those which sacrifice *functionality* in favor of transparency.)
A disturbing number of studies/surveys/reviews focus on the first few hours during which a user interacts with new software. That's just plain crazy. (Unless your only goal is to help someone design software which won't turn off would-be users with a significant learning curve.)
The vast majority of the time people will spend interacting with their computers happens long after they've had time to adapt to the interface. Whether it takes the average person an hour or a week to achieve a reasonable level of familiarity and an intuitive feeling for an interface isn't nearly as important as how well they can interact with it in the months that follow.
Given appropriate time and resources, you really ought to set things up so that participants predominantly use a single interface for many weeks, periodically testing them throughout that period.
If resources are scarce, starting out with participants who span a range of different experience levels with each piece of software is a decent approach.
Quantifying experience can be tough, of course. It's hard to believe self reporting would work well, unless questions are quite precise. I've met plenty of kernel hackers who will tell you they "know a little bit about linux," and people who are shocked by the idea of using a right mouse button who will call themselves "very knowledgeable.")
For years we've dreamed of doing away with those power-hungry and unpleasantly rigid clocks that line our walls.
Remember when we used to throw our hands up in desperation and say, "if only there were some way to make a clock that's flexible, and one which doesn't consume a whole 1/4 watt, the world would be a better place."
Now, thanks to the wonders of modern technology, the time has finally arrived. Seriously, folks, what gives? e-ink is awesome. A wall clock made with an e-ink display, on the other hand, is just silly.
And, while I'm firing meaningless rants into the void, why do so many people seem surprised to learn that Britannica suffers from such inaccuracies? Anyone who has ever read an encyclopedia article in a field about which they know something ought to know better than to expect accuracy from an encyclopedia. They're great for getting a very brief intro to a subject one has never heard of before, and for picking up enough keywords to find more information, but only a lunatic would rely solely on an encyclopedia article for anything.
I certainly recognize what you're describing. Not sure if it's a result of a society tailored to instant gratification, or whether it's simply human nature and the society in which we live is the inevitable outcome if you throw a bunch of technology and capital at humans. The optimist in me wants to blame elementary education which emphasizes fact rather than process. I hope that's true, since it suggests the promise of improvement.
But, the tendency to skip ahead to the answer to the answer is a powerful one. I certainly recognize it in myself. I've tried to work a problem independently while the solution is open on the desk, and decided that it cannot be done. Even if I believe the solution might be wrong and am checking it, having easy access to it corrupts the process. It's easy to "independently" make the same algebraic mistake a dozen times in a row, so long as one has instant access to the original. (Fortunately, I'm lucky enough to have the training/self-dicipline/genetic-luck/whatever to be able to set the solution aside and not return to it. Many people, it seems, can't do so.)
The extension to worked examples in class doesn't seem too distant.
But, what's the solution? Talking the class into actively paying attention and trying to work the problem in real time would obviously be the best solution, but it doesn't usually seem to work.
Perhaps playing "catch the mistake" games, where you deliberately do something wrong a few times during every section and reward those who catch you? I'm a little afraid of trying that myself, just because there seems to be the potential for some really bad interactions. Tricking students isn't exactly the role one wants to play, and they might easily interpret the whole thing as making fun of them. Certainly, if I bought a textbook and the intro said, "there are a hundred mistakes hidden in this text. The alert reader should document and correct them," I'd be pissed.
I like your computer assembly analogy. But, the sharp student might object that he isn't likely to ever want to build a computer, and that in such a situation *not* causing oneself distress by learning the process is a rational decision. He ends up with a built computer, and is altogether happier than he would be having been forced to learn a skill in which he has no interest.
The answer, perhaps, is that he therefore oughtn't be in a computer building class in the first place. To which, I imagine, he would respond that his advisor insisted that he fulfill a computer-building-class requirement and all he needs is a C to graduate. Which, perhaps, is the heart of the problem. Haven't got a clue how to solve it, though.
Slashdot Mirror
I do the same, originally for privacy reasons, but mostly now just because it's handy. (Although I prefer elinks, if only for the "invoke external editor on text area" option. If you ask me, that one ought to be right up there with "give users a place to enter an URL" as a necessary feature to include when designing a browser.)
Best part is that I can leaving it running on a screen session and pick it up to browse or continue a comment whenever there's a free minute near any computer.
>Look: do you own a printer? ever
>noticed you can't buy a printer
>with "standard non proprietary"
>cartridges? If I follow your train
>of thought, you should be outraged,
>no? Of course not, you keep printing.
There's a big difference. Most people only own one printer, and cartridges are disposable and are used up many times over the lifetime of a product. No one has to worry whether their library of old printer cartridges will work with their new printer, no one buys a dozen different printers and a single cartridge in order to save money or to allow different printers to communicate with eachother.
On the other hand, people often own many devices that use the class of media being discussed here, and the media often outlive the products themselves. And, unlike in the world of printers, in this field most players have chosen to go with well supported standards.
I'm not outraged at Sony, but I'm also not going to buy their products. If they were obviously and consistantly superior to their competitors, they might be able to pull it off. But they haven't been so for a long, long time.
Given a choice between products with indistinguishable features, I'll take the one that uses standard interfaces and media every time.
The two issues are far from orthogonal.
The cost of supporting minority browsers depends critically on how much browser-specific material you design into the website.
Likewise, the decision to sacrifice specific examples of standards compliance or graceful degradation can't be made without considering which browsers you expect to work with the site.
It isn't easy to discuss one without the other.
>This also works beautifully
>for drunken debates at parties.
Just remember to restore the page to it's original state after you've collected on the bet.
High contrast colors, a dark background, and a large font really help me.
At least, they prevent my eyes from feeling tired and irritated after a day on the computer, and I've been headache free for years. (Don't know if it has any bearing on long term vision problems.)
On the other hand, I've occasionally borrowed friend's computers and felt my eyes fizzle and become sore in minutes. Any time I look at small text in black on a white background, it bugs me. (Don't even get me started on semi-transparent terminals. Why on earth anyone would want to turn down the contrast on their terminal and then past a bunch of distracting stuff all over the text is beyond me.)
I generally make the text large enough that i can easily sit back forty inches or more from the monitor and choose high contrast combinations (slightly off-white or bright green on black, white on navy blue or magenta, etc.) I also avoid anti-aliased fonts, but that's mostly an esthetic choice. In general, I find that if my font is so small that anti-aliasing is useful, it's *too* small to be comfortably read anyway.
Cutting overhead light helps too.
What OSes do the two devices run?
And, to satisfy curiousity, what on earth have you made your desktop out of that it doesn't have either built in sound or spare pci slots?
I say, give me something that will eat 10% of the resources and provide 5% of the functionality. Then give me 20 other somethings that do the same, but provide different functionality, and we'll have all of amaroK's functionality with only 10% of the resource commitment.
That's more or less my principle complaint against both KDE and Gnome: in order to get the hand full of features that could be useful, one has to add a bunch of stuff that at best isn't used and at worst gets in the way.
But, the good news is those 20 programs *do* exist, and I can use them. So long as no one is forcing me to use amaroK, I'm glad to see that it exists and that someone enjoys it.
In fact, it brings me great joy to know that both Gnome and KDE exist and that there are fans out there talking each other into using them.
First, people find them useful, and anything that improves people's lives (even in a very small way) is a good thing.
Second, they both help attract new users to free unix-like software, and give talented developers something fun to work on. Anything that makes linux and the BSD's more popular is good for those of us who love them.
Third, I'm happy that *both* gnome and kde exist, because so long as both maintain a large following developers will have no choice but to make projects able to run without requiring either. That's good news for those of us who prefer alternative environments. The day one wins is the day third party apps stop simply requiring libraries and start requiring that the winner be actually running.
It's pretty standard stuff. Just happened to hit on a very loud combination. An iriver ifp-890 flash player and sony MDR-EX51LP earbuds. (Incidently, they're not the best sounding headphones I've ever heard, but surprisingly close to it for relatively cheap earbuds. I recommend them highly.)
Playing an mp3 with the default equalizer settings, normal listening is 2/40. (For normal listening I cut down all the equalizer channels in order to get some more resolution on the volume control.) For ogg files and radio I have to increase the volume to something more like 10-14.
Thanks to all who've suggested musicians/hi-fi earplugs of various kinds.
I'll give them a try.
I've been using disposable industrial earplugs for years and grumbling about them. Never bothered to look into a better alternative.
I appreciate the article reference. For anyone else interested, the full reference is Ear & Hearing. 25(6):513-527 (Available from their website only with a personal subscription. Try a university library - mine has electronic copies available via a journal aggregator.)
It appears the recommended threshold was chosen based on careful research and thoughtful analysis.
Still, it does seem to me a a legitimate question whether it's better to provide a concrete threshold with limited accuracy or to refuse to provide one at all.
Among the device and headphone combinations tested, the variation in sound level is quite dramatic: something like 25 dB for the two extreme cases, and judging by eye a standard deviation on the order of 7-10 dB. That suggests that even after dropping the recommendation from the measured level of 70% max volume (for 1 hour) to 60%, there are a significant number of consumers for whom the recommendation is still unsafe.
It might be that only influence such a recommendation will have is to cause those who are already listening to music which is much too loud to reevaluate their choice. In that case, providing a concrete number will only lead to a positive outcome. If someone isn't able to judge a safe volume on their own, giving them a reasonable number may be safer than giving them no number at all.
The danger is that someone will decide that an unsafe volume is okay based on that recommendation. There's power in concrete numbers presented by an authority figure, and it's easy to imagine a scenario in which someone who regularly listens to music at dangerous levels reads an article on the subject and determines that they need not worry about hearing damage because their volume setting is below the threshold. It's not clear to me how likely that sort of response is in practice.
It might well be that the benefits of providing a straightforward guideline far outweigh the danger that someone will read too much accuracy into it. I don't know nearly enough about the subject to put forward an answer. But, it does seem like a worthwhile discussion. (Of course it's always possible the discussion was concluded years ago, in which case an appropriate response is, "Everyone already knows the answer except you, you ignorant twit. Now go back to your own field and stop bugging us.")
I'll look forward to it.
Of course for those of you whose volume controls go all the way to 11, that should be decreased to 54% of full scale.
Seriously, am I the only person who thinks this advice completely insane? 60% converted on an arbitrary scale, converted to sound pressure with a device-dependent conversion that varies wildly? I hope that comment was misquoted or taken out of context.
With portable and headphone I'm currently using, 60% full scale would blow out my ears in seconds, if the headphones didn't first destroy themselves in a valiant attempt to save my ears. It's true I've got a very good pair of heaphones and a pretty high-ampltude device, but anything above 3-4% max is too loud for comfort. On the other hand, I've listened to equipment at which full scale with all the equalizer channels maximized isn't loud enough.
It seems to me the problem hasn't got anything to do with ear-buds or portables. Unfortunately, it's a lot more widespread and hard to deal with than that. From the article:
120 dB? That's completely insane. Anyone who's *able* to stick 120 dB in their ears without wincing and making a desperate grab for the volume switch has already got some pretty serious hearing damage, if you ask me.
But that's not too surprising, since everything in our society is screaming at us constantly. I'm continually amazed by the amount of volume people tolerate on the subway, in movie theaters, in concerts. I now carry earplugs with me everywhere and end up using them several times a day. That sucks, because it's impossible to find earplugs with a nice flat frequency response, so you end up with badly colored audio. Nothing pissed me off more than paying money to listen to live music and then to have no choice but to hear it through muddy earplugs. Why the hell anyone *wants* to hear a jazz concert in a little tiny club at earsplitting volume is another question. I can only assume it's because we've all burned out our hearing elsewhere.
But, this is in danger of becoming a completely unrestrained rant, so I better quit.
What makes you say the study is weak?
After reading both the journal article, and the writeup linked here, I don't see any obvious problems with it.
Pouring a single shot into short 350 ml tumblers and 350 ml highball glasses are both pretty common activities when making drinks. Seems like a prefectly legitimate thing to investigate.
I'm sure that's true. There are a few other recent cases of people producing surprisingly good films in industrialized countries with budgets of a couple tens of thousands: Following, Clerks, to pick two good ones. (Tarnation, for a somewhat eccentric example, although I have no idea how much the distributor spent on music rights and post production before creating the finished product.)
Of course if you were to include a cash value for the wages for the actors and crew, loaned equipment, and production space as "donations," the budgets increase rather dramatically. Tens of thousands of dollars isn't the answer to the question "what did this film really cost" but rather, "with how little cash can someone with talented, generous friends and free time make a decent movie?" That's also an interesting question to ask.
And, if you allow budgets of a few hundred thousand dollars instead, it's certainly true that the number of examples increases dramatically.
That said, it's not obvious to me that "video blogging" or portable video players will have any impact on low budget film making. The huge improvements in consumer and cheapo professional grade video cameras in the last decade and the availability of cheap video editing software and desktop computers able to run it, on the other hand, are a lot more exciting.
>>As long as "CS251: Introduction to
>>Funny Walks" remains an elective,
>>I'm fine with it.
>I'm not, and you shouldn't be either.
I gather you're a proponent of removing core competency credit from the Introduction to Funny Walks course, on the grounds that it will devalue your degree?
Surely you wouldn't suggest removing the course entirely. I sure as hell wouldn't want to be the dean who has to explain to the Ministry why it's been removed.
Perhaps a compromise is in order?
Tell you what, we'll put CS251 on the non-credit bearing electives list, if you agree to add CS351 (Graduate Seminar on Silly Walks) to the list of 1 unit advanced electives and add BP24-A (Rectal Examination and the Search for a Sense of Humor in Uptight Slashdot Posters who Missed the Point) to the list of recommended interdisciplinary electives.
Deal?
That's a mighty long swim.
>Umm, if it were truly random, it would almost
>certainly not be uniformly distributed.
Forgive my imprecision. The above should read, "assuming test answer key is truly random and drawn from a uniform probability distribution." (Or colloquially, "assuming the test answer key is random and that all choice labels have an equal probability of being right.")
>However, in practice, most tests are *intended*
>to be either uniformly distributed or random, but
>because they are generally constructed by humans
>it seldom works out that way.
I suspect you're right, with the exception of the big standardized tests. I can't imagine they'd allow anything except an automatically generated random key for those. (And, for what it's worth, those do make up rather a large fraction of the multiple choice exams students take during their lives. More so if tests are weighted by importance.)
>Also, it is well documented that if there
>are five choices, a through e, b and d are
>statistically better bets than a, c, or e
>(though this of course can vary with specific
>teachers).
I don't doubt it. But, I'd be curious to find a source for such details. What humans do when you ask them to make something random is an interesting subject.
A great (if unrelated) example, which I came across in a Steven J. Gould essay, involves asking people to draw a star field of randomly placed dots on a piece of paper. They always end up with something much more regularly spaced than a truly random field. (Gould takes up the subject upon noting that glowing insects in a cave do the same thing - creating something that looks nothing like the night sky, despite being made up of thousands of points of light against darkness.)
Doctor Plaster's strategy sounds like a good one. There's still some nonzero random chance score, but it certainly makes it a lot harder to do well without having to include huge numbers of unique answers.
I remember being taught the same thing. But, now that I stop to consider it, it seems to me that strategy is total bunk.
Assuming test answer key is truly random and uniformly distributed, the chance that you'll win by picking any particular answer to any particular question is a fixed value. (1 / the number of choices.) It can't possibly depend on what one answered for other questions.
So, whether you always choose the same answer, or choose an answer at random for each question, or spell our "abcdeabcdeabcde" makes no difference.
To convince myself that I'm not doing something dumb here, I ran a quick monte
carlo simulation using random number generators. For a test consisting of 100 questions with 4 choices each, after one million tests the mean score and standard deviation are:
0.249993, 0.0433486 for totally random answers
0.250037, 0.0433170 for always choosing "C"
From that, the error on the mean (1 sigma) should be 4.32998e-05. The answer agree with each other and with the naive prediction. So, either strategy works equally well.
The only caveat is that we're assuming a truly random answer key. That means, for example, that "every answer to the exam is B" and "no answer is ever B" are both allowed answer keys. That's probably a reasonable assumption for an exam whose answer keys are generated automatically, such as any standardized test.
But, for a human-constructed test, it may be that large classes of solutions will be discarded by the test maker. It's not obvious how to model what a human would do in "randomly" laying out answers. Could be great fun to explore if one had access to thousands of scantron answer keys. (Anyone with a key to the exam repository room want to collaborate on actually doing that?) My guess is that since the vast majority of uniformly distributed random keys look qualitatively similar to human generated keys, the difference can't be large.
The obvious question is why test prep folks have been telling people bogus information for years? Either they are just bad at statistics (a scary thought, given the business they're in), or there's some psychological benefit to convincing people that picking the same letter will improve their score. Perhaps it's marginally faster during the exam? Hard to believe that's significant.
Redundant? Second thread in the whole linup, posted 8 minutes after the article, and it's modded redundant?
Anyway, on the topic at hand:
That's a very good point. It's certainly easier to make a professional sounding documentary/interview/polemic than it is to make a believable piece of narrative fiction.
But, it is also true that there has been a conspicuous absence of fiction in radio for a long time. Even among the big, relatively well funded public radio powerhouses, fiction is hard to come by, and what little there is tends to be the same stilted, rehashed stage plays that have been boring audiences for decades. With a few exceptions (Joe Frank, Selected Shorts, several short run BBC series, etc.), entertainment radio has come to mean almost exclusively documentary and interview. While it may be easier to produce good documentary, even if you have funds and access to talent, that in itself doesn't seem quite enough to explain why fiction has disappeared from the radio. Perhaps what the parent poster sees in postcasts is just an extension of the same phenomenon.
. . . they had a specially designed bomb disposal robot on hand.
I'd hate to be around when a cylinder full of radiation get trapped with nothing but one of those generic, off-the-shelf bomb disposal robots on hand.
Wow. That's great news.
Last time I tried to do this (perhaps two years ago on windows XP) there didn't seem to be any way to do it except by messing with registry entries and rebooting, and the change happened for all users.
But, I readily confess to being clueless and all but unable to use windows. So perhaps there is a better solution out there and I simply wasn't able to find it.
A few quick google searches haven't found it this time either. (But I haven't searched exhaustively yet.) You don't happen to remember a source for info about how to set this up, do you?
True. (Well, technically you do need special permissions - but if you're doing anything useful in X you almost certainly have them anyway...)
But on less user-friendly operating systems (like windows), you do need special permissions to remap keyboards. What's more, it's usually impossible to remap them for only a specific user, or to remap them without rebooting. Real pain in the ass for those of us who are occasionally forced to use windows. (I have enough trouble switching back and forth between capslock/ctrl locations. Can't imagine changing all the other buttons too!)
At least, that was the case the last time I looked into it.
What does a half dimension look like?
Yeah, I understand what it means in this context, and I don't fault the parent poster for using it. . . but, at some level it just doesn't make sense. Not sure what a better description would be, though. Perhaps "tiered 2D" or something of the sort?
Ultimately, the usual windowed computer screen allows the user to make use of a third dimension in almost exactly the same way that a desk with books and papers on it does. The user stacks things up in piles separated in the principle plane, and then brings whichever bit they want to work on to the center of that plane. Occasionally, they'll leave something else open and visible off to the sides of the main work area.
So long as you're using your computer to do the analog of what we do with desks, anything else is either eye-candy or will hinder getting things done.
That's not to say there aren't places in which 3D interfaces could really be interesting. Just they what most people spend most of their time doing on computers can't be made any easier by arranging things in 3D. At best, they'll end up using it in exactly the same way they do now. At worst, they'll have to go through a bunch of extra steps every time they rearrange things.
Yup. Insisting that people use default configs is pointless, except if your goal is to tweak those default configs.
Ideally, you ought to force people to interact with the machines over a long enough period of time to recognize transient effects associated with having previously used similar systems, and to allow you to appropriately characterize systems which are too willing to sacrifice efficiency in favor of transparency. (Or, even worse, those which sacrifice *functionality* in favor of transparency.)
A disturbing number of studies/surveys/reviews focus on the first few hours during which a user interacts with new software. That's just plain crazy. (Unless your only goal is to help someone design software which won't turn off would-be users with a significant learning curve.)
The vast majority of the time people will spend interacting with their computers happens long after they've had time to adapt to the interface. Whether it takes the average person an hour or a week to achieve a reasonable level of familiarity and an intuitive feeling for an interface isn't nearly as important as how well they can interact with it in the months that follow.
Given appropriate time and resources, you really ought to set things up so that participants predominantly use a single interface for many weeks, periodically testing them throughout that period.
If resources are scarce, starting out with participants who span a range of different experience levels with each piece of software is a decent approach.
Quantifying experience can be tough, of course. It's hard to believe self reporting would work well, unless questions are quite precise. I've met plenty of kernel hackers who will tell you they "know a little bit about linux," and people who are shocked by the idea of using a right mouse button who will call themselves "very knowledgeable.")
For years we've dreamed of doing away with those power-hungry and unpleasantly rigid clocks that line our walls.
Remember when we used to throw our hands up in desperation and say, "if only there were some way to make a clock that's flexible, and one which doesn't consume a whole 1/4 watt, the world would be a better place."
Now, thanks to the wonders of modern technology, the time has finally arrived.
Seriously, folks, what gives? e-ink is awesome. A wall clock made with an e-ink display, on the other hand, is just silly.
And, while I'm firing meaningless rants into the void, why do so many people seem surprised to learn that Britannica suffers from such inaccuracies? Anyone who has ever read an encyclopedia article in a field about which they know something ought to know better than to expect accuracy from an encyclopedia. They're great for getting a very brief intro to a subject one has never heard of before, and for picking up enough keywords to find more information, but only a lunatic would rely solely on an encyclopedia article for anything.
Ah well, enough ranting.
I certainly recognize what you're describing. Not sure if it's a result of a society tailored to instant gratification, or whether it's simply human nature and the society in which we live is the inevitable outcome if you throw a bunch of technology and capital at humans. The optimist in me wants to blame elementary education which emphasizes fact rather than process. I hope that's true, since it suggests the promise of improvement.
But, the tendency to skip ahead to the answer to the answer is a powerful one. I certainly recognize it in myself. I've tried to work a problem independently while the solution is open on the desk, and decided that it cannot be done. Even if I believe the solution might be wrong and am checking it, having easy access to it corrupts the process. It's easy to "independently" make the same algebraic mistake a dozen times in a row, so long as one has instant access to the original. (Fortunately, I'm lucky enough to have the training/self-dicipline/genetic-luck/whatever to be able to set the solution aside and not return to it. Many people, it seems, can't do so.)
The extension to worked examples in class doesn't seem too distant.
But, what's the solution? Talking the class into actively paying attention and trying to work the problem in real time would obviously be the best solution, but it doesn't usually seem to work.
Perhaps playing "catch the mistake" games, where you deliberately do something wrong a few times during every section and reward those who catch you? I'm a little afraid of trying that myself, just because there seems to be the potential for some really bad interactions. Tricking students isn't exactly the role one wants to play, and they might easily interpret the whole thing as making fun of them. Certainly, if I bought a textbook and the intro said, "there are a hundred mistakes hidden in this text. The alert reader should document and correct them," I'd be pissed.
I like your computer assembly analogy. But, the sharp student might object that he isn't likely to ever want to build a computer, and that in such a situation *not* causing oneself distress by learning the process is a rational decision. He ends up with a built computer, and is altogether happier than he would be having been forced to learn a skill in which he has no interest.
The answer, perhaps, is that he therefore oughtn't be in a computer building class in the first place. To which, I imagine, he would respond that his advisor insisted that he fulfill a computer-building-class requirement and all he needs is a C to graduate. Which, perhaps, is the heart of the problem. Haven't got a clue how to solve it, though.