Good digital cameras don't use standard batteries. They generally don't hold enough juice.
Well, consumer level ones do.
I have a Canon PowerShot A70. Unless I use the flash and LCD constantly, its four AA batteries (mine are 2000mAh each, but 2300mAh ones exist) really are more than enough.
I second that. I just tried accessing one of these which is connected directly into the wall at work, and I can't get anything on ports 80 or 443 with either a web browser (http) or telnet. I know those ports aren't blocked at the main router because we also have a box hooked up serving web pages.
Perhaps there is some other requirement. Maybe this only works if the firewalling capabilities of the router are turned off? (e.g. this stops protecting the WRT54G itself as well as the computers on the WAN side)
And THAT will be the problem with these video players.
Whaddya mean "will be"? These things exist now. No need for speculation! Let's check out the reviews and see how many hours of battery life these video players actually have (when playing video, that is):
So no, you won't get your 5 hours from any f the current players.
I think the real problem here is that the manufacturers are trying to make these players too similar to audio-only players (e.g. the iPod). They're trying to use the same wimpy 800-1000mAh custom battery packs that most mp3 jukeboxes use. This makes sense from a business standpoint -- you only have to make one battery for your various jukebox products, and you get to sell those custom replacement batteries for $50. But from a practical standpoint, this just won't do.
The solution: video players are bound to be larger than mp3 players anyway, if only because of the screen size, and they're all an inch or more thick. It's time to start making these things use standard batteries, just like digital cameras. They'll fit, and the capacity of regular batteries is plenty high. I've seen AA batteries advertised with as much as 2300mAh of capacity. One of those would do the trick for sure, and two would rock! Even two AAA batteries with a 800mAh capacity each would be an improvement.
Wonderful. "Look, instead of paying an extra 50 cents for a higher quality quiet fan, you can use cheap fans and spend $25.00 in additional parts to make the computer quiet!"
Well, if you have a very hot-running computer that really needs a noisy, high-RPM fan, it might be worth it. Yes, bigger (not faster) fans and cooler components would help more, but if you've just gotta have a dual Xeon workstation, and anything powerful enough to keep the thing running is bound to be noisy, then maybe a $25/fan premium to quiet 'em down isn't such a bad thing?
Actually, last I heard (and admittedly, it's been a while since I heard this), counting cards is not illegal.
Yeah, I did 4-6 upstate for card counting once. My cellmate was a mass murderer. It's like the prosecutor said: card counting helps the terrorists win.
(Hint: of course it's not "illegal"!!!)
Seriously though, this is what I hate about casinos. Card counting is one of the main skills involved in playing card games. It's _the_ skill you need for games like Bridge and Gin Rummy. Yet if you do it well enough in a Casino, they will boot you. That's just sad.
The moral: Casinos are not places where skill is rewarded. If you have skill, go elsewhere.
Oh, and one more thing: if the Casino has a change machine, it's got the best payoff of any machine in the house.:)
OK, I guess we should expect this kind of statement from Slashdot (particularly from an Anonymous Coward), but there are folks with M.D.'s and Ph.D.'s here on Slashdot (like me) and some of these folks work in areas like this.
The problem is, M.D.s wouldn't consider this an appropriate forum for a proper discussion of this sort of thing. If the discussion gets at all personal, then there are issues of patient privacy. If not, then why bother? Why not just recommend some good books instead? And why couldn't the poster get book recommendations and other generic, non-personal advice from a more appropriate web forum?
As for the Ph.D.s, I'm not so sure they should say anything. While they may know quite a bit about the current theories about Schizophrenia, I not sure how helpful that would be to someone who's actually dealing with it. Then there are the Ph.D.s who think that they're M.D.s, who are a whole different kind of problem... if you don't spend your time treating patients, seeing them year after year (running someone on the WCST every six months doesn't count!), then IMHO you don't have as much to say as you think you do.
I think the most useful voices would actually be other SlashDot readers who have had similar experiences (e.g. relatives or friends with Schizophrenia). Again, though, a more appropriate web forum is a better bet, if only because you'll find more people in that category. Not to mention that other venues will be more accomodating to personal discussions (e.g they're not cached by Google, people might actually give out their real e-mail addresses to each other, and so on).
So I have to say that on the whole I disagree with you. While I don't agree with the grandparent posts' tone, I think they're basically right. And no, a discussion about how Schizophrenia affects us personally is not "News for Nerds". Maybe a scientific discussion about the differences between the brains of people with Schizophrenia and other people would be "News for Nerds", but the current topic is too personal to even be considered "News" per se.
If you read the article, that's the mileage it gets _on_the_ground_... not in the air!
Which by the way, brings us to another point. In air travel, you have to go up and down a certain amount. Imagine that you wanted to go ten miles in this thing, but were expected to fly at an altitude of one mile. That's twelve miles of travel where a car would have gone ten.
A flyng vehicle might be able to compensate for this by taking a more direct route (going the "as the crow flies" route:) but that would only last for so long. If these things ever got common, some kind sky highways (routes which are allowed for travel) would presumably be developed.
The point is, there's another little hitch in the direct comparison of mpg between this thing and a car.
Buy 10 children from a thrid(sp)-world nation. [snip] For example, I have taught my purchased child to masturbate when he hears the words "clam chowder". It really is a great party trick.
Note this guy's SlashDot ID.
This is what years of reading SlashDot will do to ya'. Consider yourselves warned, kiddos.
However, I think his issue is that he wants to run XP on a different computer, which is not a$WÃed under the OEM license.
I thought we wanted to run XP on the same computer, just under VMWare. He wants the _right_ to be able to run it on another computer, but he brought this whole thing up because he physically _couldn't_ get his copy of Windows XP to install under VMWare.
Many of the series fans indicate that the newest seasons' writing tends to be watered down. I'll admit that I liked the writing better around seasons 3-7.
See, the problem is that "the fans" have been saying that the show was on the downslide since season 7 (Or even season 6 if you're hard-core enough:). They're on, what, season 15 now?
Economists can look at the same set of data and come to different conclusions, and nothing but more time and experience will help us understand which is the correct one.
Yeah, this occurred to me after I wrote my previous post. Almost all of the areas which I suggested should focus on "hard data first" are areas involving prediction and modelling. They can only be so closely tied to their data.:)
most of the slant seems to be definitely doomsday, environmental, and decidely anti-government..
agan, this is just from reading the article.
Yes, just the ones mentioned in the article. You obviously haven't played HippieBlaster 5000!
Seriously, though, remember NARC? How about these Tom-Clancy-novel-based games I keep seeing commercials for? How about America's Army? I'm pretty sure there are such things as right-wing and pro-government slanted video games. Certainly there are patriotic video games. The writer just has a political bone to pick.
It always bothered me that the SimCity manual editorializes that Reaganomics doesn't work. (Somewhere toward the back, in the section on economic srategies I think -- it's been a while.) I think the game is even set up to demonstrate that 'fact' for you under one of the pre-configured scenarios.
I thought it didn't?:)
Either way, that's the sort of thing that economists should figure out. It actually kind of bugs me that this is considered to be a "political" issue, as though peoples' belief or lack of belief in Reaganomics will affect whether or not "it works" (In one particular situation? In general? What does that mean exactly?) The same thing for environmental issues, energy policy, population growth, and so on... the hard data should come first, "politics" second (or later) insofar as possible.
Of course, SimCity might have meant that _within the game_, Reaganomics doesn't work. Remember, it's their world, and they can make the rules as they see fit. Kind of like any other work of fiction. And like any other work of fiction, the limitations of the world (its necessarily relative simplicity compared to the real world) limit the degree to which that fictional world is applicable to the real world. Thus, shooting drug dealers on sight is wrong, despite the game NARC;).
Err? Compile your kernel with supermounting, set up your/etc/fstab to mount any drives you want supermounted. Then 'emerge hotplug'. Am I missing something?
Yes. Mandrake also has some scripts which interact with hotplug to help out.
Basically, Gentoo did not auto-detect my digital camera. Mandrake detects it and put an icon on the desktop (which runs gtkam, I believe).
Also, "compile your kernel" is not something which you should say to a newbie! Having said that, there are some supermount-supporting Gentoo kernels out there. Then again, "pick a kernel" is also not something which you should say to a newbie...
Second point, having a stripped down system after installing is one of the focuses of Gentoo. You have the ability to customize what you want. This would have been very helpful to me when I was a Linux newbie as half my problem was that there were 3-4 programs installed that all did the same thing. (I was confused)
Yes, but the flip side of that is the "What program do you use to do that?" problem, e.g. "What's this 'hotplug' thing? Why doesn't my computer just detect my camera like Windows/Mandrake does?":)
Personally, I think a Linux newbie, and certainly a computer newbie, would be better off having all the programs they need available to start. Even for a more experienced user it can be annoying. I remember trying to think of the name of my preferred Instant Messaging program and thinking "Why isn't it just installed?" After going through this same sort of thing several dozen times, I got tired of it.
I'm not going to say that Gentoo is right for a Linux newbie.
Then we're in agreement.
Hand-holding is a good thing for newbies, and Gentoo just isn't made for that.
Somebody mod this guy "Insightful". There seems to be a huge "I'm a bigger geek than you" factor involved in OS choice, at least on SlashDot. Despite what the BSD guys seem to think, I use Linux because I like it and because I'm familiar with it, not "because I hate MS". Despite what most Linux users seem to think, I use Mandrake because it gives me a powerful, easy to use desktop, not because I'm a "Linux noob" (exactly the opposite: I'd like to leave my Linux admin workload at work, thank you!)
For a more topical example, note how many posts have already suggested Gentoo as "great for newbies". Now I tried Gentoo a few months ago (before the LiveCDs), got it running with a little work, and it was okay. But I went back to Mandrake pretty quickly. Some advanced desktop features were missing, such as an equivalent to Mandrake's combination supermount/hotplug support. I had to specifically emerge almost all of the programs I wanted to use because the base install was so stripped down. And I sure hope the install is now easier than the old "partition by hand using fdisk" install that I had to do (it wasn't so hard for me, but for a newbie...)
Who would recommend a distro like that to a newbie? I think it's mostly people who want to say "my distro is bigger/badder/geekier than yours". If you're willing to put some work into setting up Gentoo it would probably be a fine desktop, and I'm sure that a stripped down, highly optimized Linux would be a great server OS, but that's not what newbies need.
That's fine. People have proven through the widespread acceptance of MP3 that audio quality is not nearly as much of an issue as cost and availability. MP3 has never sounded as good as PCM and never will, that's not what it's for. What it's good at is fitting down a narrow pipe quickly. In my opinion the development of MP3 was driven as much by dial-up modem speeds as anything else.
Although I basically agree with your first point, I stand by my argument that people won't tolerate too many alterations to their mp3s. After all, if one particular copy of a song sounds a bit weird, why not just download someone else's copy? (A copy which is not so modified).
This is especially true given that broadband is now reasonably common. Downloading someone else's copy will take a few seconds. Sure people used to accept low sound quality when it took 10 minutes to download a 64K sample rate mp3 on a 28.8 connection, but will they anymore now that they have a 3Mbit downstream cable connection or 100Mbit ethernet in their dorm?
Hell, every commercial radio station in existance (in America anyway) uses so many exciters, compressors, subharmonic synthesizers and limiters that I'd be willing to wager that this hypothetical auto-music-listener program wouldn't even be able to recognize the song played on the airwaves, even without obfuscation.
This I think is the core of the disagreement. I'm wagering that people were using a certain means of obfuscation, e.g. compressing in the time domain, that a fuzzy matcher could be built with enough accuracy to still pick out close matches. Frankly, I suspect that the only way to settle this is to actually try to write one and see whether it works.
I worry about how computationally intensive it would be as well -- fourier transforming even one song takes a noticeable amount of time on my Athlon 2200+ (a minute or so), and the matcher would have to transform each song it scanned. maybe a 15-second sample is enough? Who knows whether it's a practical idea. I just figured that it might be a possibility.
Adding small amounts of random data is only white "noise" when it's within the audible frequency range.
So of course the fuzzy matcher only looks in the audible range, thus dealing with this problem. If the song is shifted entirely into super- or sub-audible ranges, then who's going to keep that mp3 around? "Must be one of those junk mp3s the RIAA puts on the filesharing networks."
Beat frequencies wouldn't be a significant problem with dynamic obfuscation since the frequencies would change, preferably as rapidly as possible.
Though that would also hurt the sound quality. Rapidly changing notes would be particularly salient, right?
To me the bottom line is that any system of incrimination and accusation that uses only "fuzzy" comparisons is bound to fail in the courts, even in today's environment of corporate feudalism.
Well, fuzzy matching one song that someone downloaded could be an accident. Fuzzy matching a hundred isn't. It's good enough to get a warrant, I bet. Then the cops collect the real evidence by taking the computer with those downloaded songs on it.
I have to say, though, that a smart police department probably would take your advice at this point and have someone listen to the top candidates (the best matches), just to be sure. Of course, this means that the fuzzy matcher made their job far easier.
-----
If you've read this far, I have a question for you. You do seem to know about audio. I agree with that comment about modem speeds pushing the adoption of mp3s, and I always think it's weird to see people encoding mp3s at a 320K sample rate. It seems to me that a less lossy compression scheme, like mp2, might give better sound quality even when sampled at a lower rate (maybe 192K?) with similar file sizes. Any insight into this?
Ever heard of Sound Forge? Any method for identifying a music file based on it's audio content is practically begging to be circumvented. Change anything in the frequency or amplitude domains and you've created a file that will have a different identifier than it did before the change.
If it's digital, it's hackable.
Since different encoders and encoding bit rates will leave different sounds (e.g. frequencies) in the resulting audio file, the matching has to be at least somewhat fuzzy. But changing the frequencies of the sounds in the file means changing the audio coming out. You may need to make lots of changes to make an audio file a weak match to the frequency pattern of the unaltered track, and there's got to be a limit to how much can be changed while still leaving an audio file which people would want to listen to.
Add in lots of extra frequencies in small amounts? That's white noise (more or less). Not to mention that it won't help much, as the comparison will probably look at the frequencies (or frequency narrow bands) with the highest amplitudes, the most unique features of that particular track.
Add in lots of one frequency? That's a whine. At a high frequency it could be very annoying (and is likely to be ignored, like I said). At moderate frequencies, it'll interfere with the music. A low frequency might work (a low rumble can be ignored) but don't turn on that bass boost! And if that's not enough and you end up adding in lots of multiple frequencies, then you have to watch out for beat frequencies. On the other hand, I suppose a clever mp3 distributor could add pre-determined amounts of certain frequencies, which the recipients then know to remove. Then again, if you're doing that, why not just encrypt the whole file?
Computer keyboarding has NOT, I repeat, NOT been shown or linked to cause CTS. I know that this goes against what most of you have heard, but there are million of people using keyboards, and there are certainly not anywhere near that many people with CTS.
Could you point to a specific source for this? I'm not saying that you're wrong, but I'm not sure that I buy this argument, and of course I can't really confirm that you're the expert that you claim to be.
I've always heard that _excessive_ keyboard use can cause carpal tunnel. The secretary who just types in the occasional appointment, the cashier who types in one price every few minutes when the scanner fails, and the grandma who types an e-mail a day all "use keyboards", but they don't use them like I do! Thus I don't buy the argument about fewer people having CTS than use keyboards. Not _every_ smoker gets lung cancer either:). Furthermore, I thought that CTS develops over time, and the use of keyboards day in and day out (at work _and_ for correspondance _and_ for entertainment) is a relatively recent phenomenon. Thus cases of excessive keyboard use were rare until recently.
If you're right, that would be very interesting, but could you point to a respectable third-party source to back up those claims?
Without downloading the song and making an aural comparison (there is technology available that could discern, as AT&T Wireless is apparently using it, as was reported on/. previously)
Someone recently posted a link to a program called "gjay" recently. This program makes nice playlists by, among other things, doing a fourier transform of the audio of a song (.mp3 or.ogg only) and then trying to put match songs which sound similar.
It occurs to me that this technique might be fairly good for uniquely identifying a song. Presumably certain high frequencies (harmonics of the stronger lower frequencies?) would be weighted less in the comparison, as most forms of lossless audio compression save space by eliminating these.
Interestingly, both groups of researchers were working strictly with visual memory. I wonder whether the working memory used by programmers, mathematicians, etc. will be in the same place, or a different area altogether?
Here's a hint: when you get a song stuck in your head, do you really think that visual cortex is involved? There are almost certainly multiple "working memory" areas for different types of data (various sensory data, and also for preparing movements, linguistic information, and so on). I wouldn't know which areas are involved in mathematics (probably many, especially frontal regions) though I seem to recall that part of the parietal cortex was associated with performing simple mathematical tasks.
Anyway, this article is yet another case of the writer drawing the wrong conclusion. And yes, this is an article written by a journalist, not a research article or letter to Nature or anything else written by a researcher.
Slashdot Mirror
Good digital cameras don't use standard batteries. They generally don't hold enough juice.
Well, consumer level ones do.
I have a Canon PowerShot A70. Unless I use the flash and LCD constantly, its four AA batteries (mine are 2000mAh each, but 2300mAh ones exist) really are more than enough.
I second that. I just tried accessing one of these which is connected directly into the wall at work, and I can't get anything on ports 80 or 443 with either a web browser (http) or telnet. I know those ports aren't blocked at the main router because we also have a box hooked up serving web pages.
Perhaps there is some other requirement. Maybe this only works if the firewalling capabilities of the router are turned off? (e.g. this stops protecting the WRT54G itself as well as the computers on the WAN side)
And THAT will be the problem with these video players.
Whaddya mean "will be"? These things exist now. No need for speculation! Let's check out the reviews and see how many hours of battery life these video players actually have (when playing video, that is):
RCA Lyra RD2780: 4 hours max
Archos AV320: about 3 hours
Archos AV340: 3.5 hours
So no, you won't get your 5 hours from any f the current players.
I think the real problem here is that the manufacturers are trying to make these players too similar to audio-only players (e.g. the iPod). They're trying to use the same wimpy 800-1000mAh custom battery packs that most mp3 jukeboxes use. This makes sense from a business standpoint -- you only have to make one battery for your various jukebox products, and you get to sell those custom replacement batteries for $50. But from a practical standpoint, this just won't do.
The solution: video players are bound to be larger than mp3 players anyway, if only because of the screen size, and they're all an inch or more thick. It's time to start making these things use standard batteries, just like digital cameras. They'll fit, and the capacity of regular batteries is plenty high. I've seen AA batteries advertised with as much as 2300mAh of capacity. One of those would do the trick for sure, and two would rock! Even two AAA batteries with a 800mAh capacity each would be an improvement.
Wonderful. "Look, instead of paying an extra 50 cents for a higher quality quiet fan, you can use cheap fans and spend $25.00 in additional parts to make the computer quiet!"
Well, if you have a very hot-running computer that really needs a noisy, high-RPM fan, it might be worth it. Yes, bigger (not faster) fans and cooler components would help more, but if you've just gotta have a dual Xeon workstation, and anything powerful enough to keep the thing running is bound to be noisy, then maybe a $25/fan premium to quiet 'em down isn't such a bad thing?
Actually, last I heard (and admittedly, it's been a while since I heard this), counting cards is not illegal.
:)
Yeah, I did 4-6 upstate for card counting once. My cellmate was a mass murderer. It's like the prosecutor said: card counting helps the terrorists win.
(Hint: of course it's not "illegal"!!!)
Seriously though, this is what I hate about casinos. Card counting is one of the main skills involved in playing card games. It's _the_ skill you need for games like Bridge and Gin Rummy. Yet if you do it well enough in a Casino, they will boot you. That's just sad.
The moral: Casinos are not places where skill is rewarded. If you have skill, go elsewhere.
Oh, and one more thing: if the Casino has a change machine, it's got the best payoff of any machine in the house.
Last I heard, it's schedule for the day after Duke Nukem: Forever.
Well duh, that's the 3-D engine MS is using for Avalon.
OK, I guess we should expect this kind of statement from Slashdot (particularly from an Anonymous Coward), but there are folks with M.D.'s and Ph.D.'s here on Slashdot (like me) and some of these folks work in areas like this.
The problem is, M.D.s wouldn't consider this an appropriate forum for a proper discussion of this sort of thing. If the discussion gets at all personal, then there are issues of patient privacy. If not, then why bother? Why not just recommend some good books instead? And why couldn't the poster get book recommendations and other generic, non-personal advice from a more appropriate web forum?
As for the Ph.D.s, I'm not so sure they should say anything. While they may know quite a bit about the current theories about Schizophrenia, I not sure how helpful that would be to someone who's actually dealing with it. Then there are the Ph.D.s who think that they're M.D.s, who are a whole different kind of problem... if you don't spend your time treating patients, seeing them year after year (running someone on the WCST every six months doesn't count!), then IMHO you don't have as much to say as you think you do.
I think the most useful voices would actually be other SlashDot readers who have had similar experiences (e.g. relatives or friends with Schizophrenia). Again, though, a more appropriate web forum is a better bet, if only because you'll find more people in that category. Not to mention that other venues will be more accomodating to personal discussions (e.g they're not cached by Google, people might actually give out their real e-mail addresses to each other, and so on).
So I have to say that on the whole I disagree with you. While I don't agree with the grandparent posts' tone, I think they're basically right. And no, a discussion about how Schizophrenia affects us personally is not "News for Nerds". Maybe a scientific discussion about the differences between the brains of people with Schizophrenia and other people would be "News for Nerds", but the current topic is too personal to even be considered "News" per se.
Yes, like sendmail.
:)
Uh, I think that when talking about "peer review", it is implied that when all the reviews are negative, you shouldn't use that software.
In other words, the system is solid, but still doesn't work if end users don't pay attention to the reviews.
If you read the article, that's the mileage it gets _on_the_ground_... not in the air!
:) but that would only last for so long. If these things ever got common, some kind sky highways (routes which are allowed for travel) would presumably be developed.
Which by the way, brings us to another point. In air travel, you have to go up and down a certain amount. Imagine that you wanted to go ten miles in this thing, but were expected to fly at an altitude of one mile. That's twelve miles of travel where a car would have gone ten.
A flyng vehicle might be able to compensate for this by taking a more direct route (going the "as the crow flies" route
The point is, there's another little hitch in the direct comparison of mpg between this thing and a car.
65 dBa is quiet.
That's _per_ flying car, of course. Hopefully these don't get too popular.
Then again, I suppose that regular cars are pretty loud as well. I'm not sure how loud exactly. That's probably the most important comparison.
Also, as I believe someone else pointed out, you have to land eventually. 65dba at 500 feet equals what exactly at 10ft off the ground?
Buy 10 children from a thrid(sp)-world nation.
[snip]
For example, I have taught my purchased child to masturbate when he hears the words "clam chowder". It really is a great party trick.
Note this guy's SlashDot ID.
This is what years of reading SlashDot will do to ya'. Consider yourselves warned, kiddos.
However, I think his issue is that he wants to run XP on a different computer, which is not a$WÃed under the OEM license.
:)
I thought we wanted to run XP on the same computer, just under VMWare. He wants the _right_ to be able to run it on another computer, but he brought this whole thing up because he physically _couldn't_ get his copy of Windows XP to install under VMWare.
Boy, this is confusing, isn't it?
Many of the series fans indicate that the newest seasons' writing tends to be watered down. I'll admit that I liked the writing better around seasons 3-7.
:). They're on, what, season 15 now?
See, the problem is that "the fans" have been saying that the show was on the downslide since season 7 (Or even season 6 if you're hard-core enough
Needless to say, fans' opinions mean nothing.
Economists can look at the same set of data and come to different conclusions, and nothing but more time and experience will help us understand which is the correct one.
:)
Yeah, this occurred to me after I wrote my previous post. Almost all of the areas which I suggested should focus on "hard data first" are areas involving prediction and modelling. They can only be so closely tied to their data.
most of the slant seems to be definitely doomsday, environmental, and decidely anti-government..
agan, this is just from reading the article.
Yes, just the ones mentioned in the article. You obviously haven't played HippieBlaster 5000!
Seriously, though, remember NARC? How about these Tom-Clancy-novel-based games I keep seeing commercials for? How about America's Army? I'm pretty sure there are such things as right-wing and pro-government slanted video games. Certainly there are patriotic video games. The writer just has a political bone to pick.
It always bothered me that the SimCity manual editorializes that Reaganomics doesn't work. (Somewhere toward the back, in the section on economic srategies I think -- it's been a while.) I think the game is even set up to demonstrate that 'fact' for you under one of the pre-configured scenarios.
:)
;).
I thought it didn't?
Either way, that's the sort of thing that economists should figure out. It actually kind of bugs me that this is considered to be a "political" issue, as though peoples' belief or lack of belief in Reaganomics will affect whether or not "it works" (In one particular situation? In general? What does that mean exactly?) The same thing for environmental issues, energy policy, population growth, and so on... the hard data should come first, "politics" second (or later) insofar as possible.
Of course, SimCity might have meant that _within the game_, Reaganomics doesn't work. Remember, it's their world, and they can make the rules as they see fit. Kind of like any other work of fiction. And like any other work of fiction, the limitations of the world (its necessarily relative simplicity compared to the real world) limit the degree to which that fictional world is applicable to the real world. Thus, shooting drug dealers on sight is wrong, despite the game NARC
Err? Compile your kernel with supermounting, set up your /etc/fstab to mount any drives you want supermounted. Then 'emerge hotplug'. Am I missing something?
:)
Yes. Mandrake also has some scripts which interact with hotplug to help out.
Basically, Gentoo did not auto-detect my digital camera. Mandrake detects it and put an icon on the desktop (which runs gtkam, I believe).
Also, "compile your kernel" is not something which you should say to a newbie! Having said that, there are some supermount-supporting Gentoo kernels out there. Then again, "pick a kernel" is also not something which you should say to a newbie...
Second point, having a stripped down system after installing is one of the focuses of Gentoo. You have the ability to customize what you want. This would have been very helpful to me when I was a Linux newbie as half my problem was that there were 3-4 programs installed that all did the same thing. (I was confused)
Yes, but the flip side of that is the "What program do you use to do that?" problem, e.g. "What's this 'hotplug' thing? Why doesn't my computer just detect my camera like Windows/Mandrake does?"
Personally, I think a Linux newbie, and certainly a computer newbie, would be better off having all the programs they need available to start. Even for a more experienced user it can be annoying. I remember trying to think of the name of my preferred Instant Messaging program and thinking "Why isn't it just installed?" After going through this same sort of thing several dozen times, I got tired of it.
I'm not going to say that Gentoo is right for a Linux newbie.
Then we're in agreement.
Hand-holding is a good thing for newbies, and Gentoo just isn't made for that.
Somebody mod this guy "Insightful". There seems to be a huge "I'm a bigger geek than you" factor involved in OS choice, at least on SlashDot. Despite what the BSD guys seem to think, I use Linux because I like it and because I'm familiar with it, not "because I hate MS". Despite what most Linux users seem to think, I use Mandrake because it gives me a powerful, easy to use desktop, not because I'm a "Linux noob" (exactly the opposite: I'd like to leave my Linux admin workload at work, thank you!)
For a more topical example, note how many posts have already suggested Gentoo as "great for newbies". Now I tried Gentoo a few months ago (before the LiveCDs), got it running with a little work, and it was okay. But I went back to Mandrake pretty quickly. Some advanced desktop features were missing, such as an equivalent to Mandrake's combination supermount/hotplug support. I had to specifically emerge almost all of the programs I wanted to use because the base install was so stripped down. And I sure hope the install is now easier than the old "partition by hand using fdisk" install that I had to do (it wasn't so hard for me, but for a newbie...)
Who would recommend a distro like that to a newbie? I think it's mostly people who want to say "my distro is bigger/badder/geekier than yours". If you're willing to put some work into setting up Gentoo it would probably be a fine desktop, and I'm sure that a stripped down, highly optimized Linux would be a great server OS, but that's not what newbies need.
49% of people who want a player don't want music.
Think that's bad? Considering that most of these things have the batteries locked inside, 45% of people must be planning to use the thing just once!
From the article:
Other features that matter to consumers are rechargeable batteries (55%), small size (52%) and the ability to connect it up to a PC (49%).
Wasn't that because the format they recorded it on was quite obscure and they couldn't find a player to read back the data ?
What did they use, 8-Track or Betamax?
That's fine. People have proven through the widespread acceptance of MP3 that audio quality is not nearly as much of an issue as cost and availability. MP3 has never sounded as good as PCM and never will, that's not what it's for. What it's good at is fitting down a narrow pipe quickly. In my opinion the development of MP3 was driven as much by dial-up modem speeds as anything else.
Although I basically agree with your first point, I stand by my argument that people won't tolerate too many alterations to their mp3s. After all, if one particular copy of a song sounds a bit weird, why not just download someone else's copy? (A copy which is not so modified).
This is especially true given that broadband is now reasonably common. Downloading someone else's copy will take a few seconds. Sure people used to accept low sound quality when it took 10 minutes to download a 64K sample rate mp3 on a 28.8 connection, but will they anymore now that they have a 3Mbit downstream cable connection or 100Mbit ethernet in their dorm?
Hell, every commercial radio station in existance (in America anyway) uses so many exciters, compressors, subharmonic synthesizers and limiters that I'd be willing to wager that this hypothetical auto-music-listener program wouldn't even be able to recognize the song played on the airwaves, even without obfuscation.
This I think is the core of the disagreement. I'm wagering that people were using a certain means of obfuscation, e.g. compressing in the time domain, that a fuzzy matcher could be built with enough accuracy to still pick out close matches. Frankly, I suspect that the only way to settle this is to actually try to write one and see whether it works.
I worry about how computationally intensive it would be as well -- fourier transforming even one song takes a noticeable amount of time on my Athlon 2200+ (a minute or so), and the matcher would have to transform each song it scanned. maybe a 15-second sample is enough? Who knows whether it's a practical idea. I just figured that it might be a possibility.
Adding small amounts of random data is only white "noise" when it's within the audible frequency range.
So of course the fuzzy matcher only looks in the audible range, thus dealing with this problem. If the song is shifted entirely into super- or sub-audible ranges, then who's going to keep that mp3 around? "Must be one of those junk mp3s the RIAA puts on the filesharing networks."
Beat frequencies wouldn't be a significant problem with dynamic obfuscation since the frequencies would change, preferably as rapidly as possible.
Though that would also hurt the sound quality. Rapidly changing notes would be particularly salient, right?
To me the bottom line is that any system of incrimination and accusation that uses only "fuzzy" comparisons is bound to fail in the courts, even in today's environment of corporate feudalism.
Well, fuzzy matching one song that someone downloaded could be an accident. Fuzzy matching a hundred isn't. It's good enough to get a warrant, I bet. Then the cops collect the real evidence by taking the computer with those downloaded songs on it.
I have to say, though, that a smart police department probably would take your advice at this point and have someone listen to the top candidates (the best matches), just to be sure. Of course, this means that the fuzzy matcher made their job far easier.
-----
If you've read this far, I have a question for you. You do seem to know about audio. I agree with that comment about modem speeds pushing the adoption of mp3s, and I always think it's weird to see people encoding mp3s at a 320K sample rate. It seems to me that a less lossy compression scheme, like mp2, might give better sound quality even when sampled at a lower rate (maybe 192K?) with similar file sizes. Any insight into this?
Ever heard of Sound Forge? Any method for identifying a music file based on it's audio content is practically begging to be circumvented. Change anything in the frequency or amplitude domains and you've created a file that will have a different identifier than it did before the change.
If it's digital, it's hackable.
Since different encoders and encoding bit rates will leave different sounds (e.g. frequencies) in the resulting audio file, the matching has to be at least somewhat fuzzy. But changing the frequencies of the sounds in the file means changing the audio coming out. You may need to make lots of changes to make an audio file a weak match to the frequency pattern of the unaltered track, and there's got to be a limit to how much can be changed while still leaving an audio file which people would want to listen to.
Add in lots of extra frequencies in small amounts? That's white noise (more or less). Not to mention that it won't help much, as the comparison will probably look at the frequencies (or frequency narrow bands) with the highest amplitudes, the most unique features of that particular track.
Add in lots of one frequency? That's a whine. At a high frequency it could be very annoying (and is likely to be ignored, like I said). At moderate frequencies, it'll interfere with the music. A low frequency might work (a low rumble can be ignored) but don't turn on that bass boost! And if that's not enough and you end up adding in lots of multiple frequencies, then you have to watch out for beat frequencies. On the other hand, I suppose a clever mp3 distributor could add pre-determined amounts of certain frequencies, which the recipients then know to remove. Then again, if you're doing that, why not just encrypt the whole file?
Computer keyboarding has NOT, I repeat, NOT been shown or linked to cause CTS. I know that this goes against what most of you have heard, but there are million of people using keyboards, and there are certainly not anywhere near that many people with CTS.
:). Furthermore, I thought that CTS develops over time, and the use of keyboards day in and day out (at work _and_ for correspondance _and_ for entertainment) is a relatively recent phenomenon. Thus cases of excessive keyboard use were rare until recently.
Could you point to a specific source for this? I'm not saying that you're wrong, but I'm not sure that I buy this argument, and of course I can't really confirm that you're the expert that you claim to be.
I've always heard that _excessive_ keyboard use can cause carpal tunnel. The secretary who just types in the occasional appointment, the cashier who types in one price every few minutes when the scanner fails, and the grandma who types an e-mail a day all "use keyboards", but they don't use them like I do! Thus I don't buy the argument about fewer people having CTS than use keyboards. Not _every_ smoker gets lung cancer either
If you're right, that would be very interesting, but could you point to a respectable third-party source to back up those claims?
Without downloading the song and making an aural comparison (there is technology available that could discern, as AT&T Wireless is apparently using it, as was reported on /. previously)
.ogg only) and then trying to put match songs which sound similar.
Someone recently posted a link to a program called "gjay" recently. This program makes nice playlists by, among other things, doing a fourier transform of the audio of a song (.mp3 or
It occurs to me that this technique might be fairly good for uniquely identifying a song. Presumably certain high frequencies (harmonics of the stronger lower frequencies?) would be weighted less in the comparison, as most forms of lossless audio compression save space by eliminating these.
Perhaps this is how it's done?
Interestingly, both groups of researchers were working strictly with visual memory. I wonder whether the working memory used by programmers, mathematicians, etc. will be in the same place, or a different area altogether?
Here's a hint: when you get a song stuck in your head, do you really think that visual cortex is involved? There are almost certainly multiple "working memory" areas for different types of data (various sensory data, and also for preparing movements, linguistic information, and so on). I wouldn't know which areas are involved in mathematics (probably many, especially frontal regions) though I seem to recall that part of the parietal cortex was associated with performing simple mathematical tasks.
Anyway, this article is yet another case of the writer drawing the wrong conclusion. And yes, this is an article written by a journalist, not a research article or letter to Nature or anything else written by a researcher.