For what it's worth, Gmail has been just the opposite for me. It's Yahoo and AOL which randomly decide to block me
Ditto. I'm not saying the gp is wrong about his experience. But in my own case, I've found that both Yahoo and AOL just stopped accepting email from me ca. 2008. I run my own server on my own domain (not via a residential connection). In Yahoo's case, it was fairly easy to fix; I filled out a form, and after a while Yahoo users started receiving my emails again. With AOL, I haven't looked into trying to fix it. I know one person who uses AOL, and she doesn't get email from me@mydomain, or from people who send her email from academic accounts at the college where I work.
One good thing about installing domainkeys/dkim on your mail server is that Yahoo and Google both pay attention to it, and therefore they won't suddenly decide you're a spammer because someone else starts sending spam that's forged to look like it came from you. As far as I can tell, Yahoo, probably just insituted a policy last year of blocking vanity domains by default, but then unblocking them as soon as the owner of the domain fills out the form.
If schools really cared about anything but profits, then we'd have a mandatory open-source textbook market where academia would be free to create and modify textbooks. These textbooks would cost nothing. Certainly, there would still be a need for private market textbooks (on arcane and/or rapidly changing subjects) but I can see a substantial portion of textbook requirements displaced by an open system.
The "mandatory" part doesn't make a lot of sense. You can't force authors to write books for free. And although a lot of free textbooks do exist already (see my sig), you can't guarantee that for a particular subject, the best book will always be a free book rather than a non-free.
But other than that, what you're suggesting seems similar to something California is doing now. Motivated by the California state budget crisis, Governor Schwarzenegger has announced a Free Digital Textbook Initiative,
which has gathered a list of free, online high school math and science textbooks that are aligned with state content standards.
The intention is to have the books used in classrooms in fall 2009.
This article has some
useful background, but it mistakenly suggests that the arduous state adoption process will be an obstacle to the FDTI; statewide adoption
only applies to K-8, but FDTI is doing high-school books.
There was a previous, unsuccessful
effort called COSTP, which tried to produce a history textbook using
Wikibooks. Here
is a BBC article about the present effort, and here is a newspaper opinion piece
by the Governor. This is a transcript of a speech by the Governor, with some interesting Q&A
at the end.
Twenty books were submitted (press release,
links).
The four books from traditional publisher Pearson are consumable workbooks, not actual textbooks.
>>I think you're incorrect about Heinlein. If you look at his books, the closest I think he comes is The Moon Is a Harsh Mistress in 1966
>While I haven't read it, I've seen recent reviews of Friday that comment on how accurately it foretells the Internet as we have it today. It's not a central part of the book, but just part of the background of the world that it's set in. I've also seen suggestions that For Us, The Living (which I also haven't read), while off in some admittedly important details (like the basic way the technology works) has a worldwide information network with a similar function to the modern Internet.
Hmm...interesting, but I don't think it'll wash.
Friday was written in 1982, at which point ARPANET already existed, and the internet was already on its way. Most of the people using that kind of technology were in academia and the military, but it wasn't anything all that futuristic by then.
I've read For Us, The Living, which is the first fiction Heinlein ever wrote, ca. 1938. It wasn't published until 2003, so even if it had been extremely prescient, it wouldn't really be any more relevant than any other private thoughts of any individual that happened to be extremely prescient. I didn't remember anything internettish in it, and reading the link at nielsenhayden.com, it seems pretty clear to me that there wasn't. They've got what we would now refer to as videotelepony, fax machines, and pneumatic tubes. None of that has anything to do with the internet, which is a public network of general-purpose digital computers.
Actually Heinlein was pretty backward in stuff like this. For instance in Starman Jones (1953), the plot centers around the need for human spaceship pilots to navigate a spaceship by reading numbers out of tables printed in books, and then entering them into the ship's computer.
Such global networks featured in the fiction of Heinlein, Asimov and plenty of others before Neuromancer was published. Plenty of authors predicted the growth and utility of world wide computer networks, although none (including Gibson) grasped the full implications of this.
I think you're incorrect about Heinlein.
If you look at his books, the closest I think he comes is The Moon Is a Harsh Mistress in 1966. A big central computer on the moon becomes self-aware, and he can project a synthesized voice and image over a video phone network. He's also networked to a lot of stuff, and can, e.g., make toilets run backwards. However, it's really not depicted as anything at all resembling the internet. All he really did was take existing time-sharing systems (the Dartmouth time-sharing system started in 1964) and extrapolate to the case where the central computer was self-aware, and the network spread across the whole moon. The way humans use the network in the story is always as nothing more than a video phone network. There is only one computer, and nobody ever transfers any digital data other than video telephony. It's true that the network is described as global (meaning global on the moon), but it's really only depicted as a telephone network, and a global telephone network already existed in 1966. A global network of computers would have been an innovation, but Heinlein doesn't depict the existence of any other computers on the network.
Probably "A Logic Named Joe," by Murray Leinster, is the most relevant example that predates the actual internet.
Okay, this is going to be wildly unpopular on slashdot, but I'm going to play devil's advocate and argue that the damages are not necessarily disproportionate to the actual harm.
It's definitely not correct to calculate the actual harm like this: 24 songs, multiplied by a market price of $1 per track on iTunes, giving 24x$1= $24. She wasn't being sued for obtaining single copies of the 24 songs illegally for her own use, she was being sued for giving them to other people -- possibly a very large number of other people.
Another possibility would be to take 24x$1, and then multiply by the number of people who actually downloaded the specific copies of the songs that she made available. There are a couple of problems with this. One is that nobody has any data on how many people downloaded those specific copies. Maybe zero people downloaded them. Maybe 10 people downloaded a particular song, each of them copied that copy 10 times for 10 friends, and so on, so that conceivably there are millions of mp3 files of that song out there that are all descended from Jammie Thomas's original. We just don't know, because the copies don't carry a history that allows us to know how they spread. You could argue that other people may have bought CDs of the same songs, ripped them, and put them online; again, there was no data about this at trial. And in any case, it's kind of morally obtuse to say that what she did isn't really so bad, because lots of other people were doing it as well. If you want to argue that what she did wasn't blameworthy, then you need to do it on other grounds, e.g., US copyright laws are unjust; well, I agree that US copyright laws are unjust, but we're talking about a court case here, so that's irrelevant. Suppose I set fire to the trees behind my neighbor's house, and around the same time 100 other people are also lighting matches and throwing them into the same bushes. My neighbor's house burns down, and he sues me. IANAL, but I doubt that the court is going to reduce the damages they make me pay by a factor of 100, on the theory that there were 99 other people attempting to do the same thing.
So it seems to me there's a pretty strong logical case to be made that you should calculate the damages like this: 24x$1xNxBxLxD. N is the total number of people who downloaded the song (and, as argued above, not just the number who got a copy directly from the ones Thomas put up). B is a factor that takes into account that only a fraction of the blame is hers, but I argued above that setting B=1 is reasonable (or at least not so unreasonable that it's obviously unconstitutional). L is a factor that measures how much the industry really loses when songs of a certain market value are downloaded. The industry usually assumes L=1, i.e., that every illegal download equals one lost sale. We actually don't know what L is. It's possible that 75% of downloaders would never have bought the song (e.g., they're 12-year-olds who have no money), in which case L=.25. It's even possible that L is negative (illegal downloads publicize and popularize the music, thereby driving legal sales). But these things are all impossible to determine, so I don't think it's completely crazy to say L=1, as a ballpark figure. (Again, the issue isn't whether it's exactly correct, the issue is whether it's so crazy that it gives a result that is obviously unconstitutionally disproportionate.) And finally D is the deterrent factor. When you sue someone for copyright violation, you're allowed to sue for more than the actual damages. As an example, if you write GPL'd software, in many cases the market value of the software is vanishingly small because you're intentionally giving it away for free; nevertheless if you satisfy certain technical requirements, you're allowed to sue for more than your actual damages, i.e., D can be much greater than one. (This is one reason that the GNU project encourages authors of GNU software to formally register a copyright, and sign over the copyrights to GNU. If you don't file a formal copyright re
Pity TFA lacks more detail about the tonality. It would be interesting to know what notes it could produce, and what intervals, possibly indicating whether they leaned to minor or major scale for example..
The major/minor system is less than a thousand years old, and only occurs in Western music.
But what's really interesting about this flute is that the harmonics are very close to a modern-day flute - 35,000 years later! There is a sample of the recreated sound right now on the New York Times website
Thanks for the cool link:-) You're misusing the terminology a little -- the original NYT article is more correct.
Every sound can be broken down into a sum of sine waves. Usually, for basic physical reasons, those sine waves have frequencies that are all integer multiples (or nearly integer multiples) of the fundamental frequency. When they have this integer-multiple relationship, they're called "harmonics;" the more general term for the case where they're not integer multiples (anharmonic) is "partials." Any wind instrument that's made out of an air column is going to have integer-multiple harmonics, not anharmonic partials. So when you say that the harmonics are close to a modern flute, that's not really a useful statement; trivially, for physical reasons, any tone played on any wind instrument is going to have the same harmonics as the same note played on any other wind instrument. The only thing that will be different is the strengths of the harmonics.
What the expert quoted in the NYT article says is "The tones are quite harmonic." This is a different statement. It means that if you had two flutes like this one, and you played combinations of notes, they would sound good together. This has to do with how the scale is constructed. He also doesn't say the scale is the same as any particular modern one, just that it's a scale that sounds good in relation to itself.
The only cross-cultural universal we see today is that all cultures have what's called octave identification, meaning that, e.g., middle C and the C an octave above it are perceived as being similar, and able to play the same musical function. Most cultures don't have harmony at all -- that's mainly a function of Western music. Different cultures generally don't use the same scales. E.g., Beethoven, a Javanese gamelan orchestra, and a Delta blues musician use different scales in different ways. It wouldn't even make sense to interpret the expert's quote as saying that the scale is the same as today's scale, there's more than one scale used today.
Unfortunately I couldn't get the sound widget to play in my browser.
I'd tend to agree that this is useless voyeurism, except that there are some ethical issues that come up in transplants when the patient is very rich. The NY Times had an article about this today, and they specifically mentioned this hospital as one that had a very short average wait time of 3.8 months, compared to the national average of 12.3 months. "If you had access to a jet and had six hours to get anywhere in the country, you'd have a wide choice of programs," they quote one doctor as saying.
If you live in a remote rural area that's off the grid, you need a battery so you can have electricity at night, or when it's cloudy. The battery is big and expensive. That's a non-grid-tied system.
I use multiple OS X, Linux, and FreeBSD machines daily. One cannot sync all home directory files, as all the config stuff differs between Gentoo, Debian, FreeBSD, Tiger, and Leopard.
I use unison to sync my files on two desktop machines, a small-form-factor music server, and a web server. At various times I've had a mix of various OSes on the various machines, including Debian, Ubuntu, and FreeBSD. I use unison to sync them all. It's true that you don't want to sync all home directory files if one account is BSD and another is Debian. However, using USB sticks and email is, IMO, much more difficult, complicated, time-consuming, and error prone than using a file synchronizer. Every decent file synchronization program will have some way to control what gets synced and what doesn't. For instance, I have mine set up not to sync my.bashrc, which has to be different on BSD versus Debian. For syncing my home desktop with my music server, it's configured to sync the music files, whereas it won't sync music files between the desktop and the web server.
The OP says he's contemplating putting everything in a version control system. In my experience, version control systems are a hassle, and you don't want to use them for more things than you need to use them for. The most common reasons you might want a version control system are (1) aw, %#@!, I just deleted (or munged) that file, and (2) I want a three-year-old version of a file for some reason. For me, what works for #1 is to sync using unison very frequently, so that if I inadvertently delete a file, I can just get it back from another machine. For #2, that's what backups are for.
The problem with methods like thumb drives and email is that they're a pain, so you're not going to do them more than once a day or so. I probably run unison 10 or 20 times a day, because it's really quick and easy to do. That means that I can never lose more than about half an hour's worth of work, because that's how old my last sync was. It helps that unison is extremely efficient. For instance, if you have a 100 kB text file that you only changed one line of, it's smart enough to detect that, and just transmit a small amount of data over the network.
When I do have a good reason to use version control for a specific project, I use git. Now, git tends to make duplicate copies of large binary files, which you might think would be a problem. However, there's a nice trick, which is to use rzip rather than gzip for compressing your backup archives. Rzip can detect large, identical segments of a file that are far apart, so your backups that include git directories aren't any bigger than they would have been without the git directories.
He is also sheltered to a certain extent from the failure of the power network so if a situation does arise where there are rolling blackouts again, he knows he will a least have some electricity each day.
This is incorrect. Unless you're in an area where grid power is not available, you buy a PV system that's grid-tied, and then if there's a blackout, you have no power. A non-grid-tied system is significantly more expensive and complicated, so you don't want one unless its your only option.
One of the things that people constantly underestimate the price of is certainty.
This is true, but the uncertainty argument isn't all in favor of getting PV. When people ask me how long my system will take to pay for itself, I always answer that I have absolutely no way of knowing, because I don't know what electricity will cost for the next 25 years, which is the design life of the system. If we get a big spike in electricity prices, I'll come out looking like a genius. But there are also other sources of uncertainty that work against PV. The price of the panels is expected to go down, and their efficiency is expected to go up. I don't know how much the price and efficiency will improve in the next 25 years, which is the design lifetime of my system. That's an uncertainty that argues against buying a PV system now.
I live in an area with expensive electricity and lots of sun. I have a south-facing roof with no shade. Given those factors, getting PV is a fairly reasonable investment, but there's no way to pin down exactly whether or not it's really going to look like a winner for me with hindsight in the year 2032 when the system's design lifetime is over.
What I do know is that (a) it's in the right ballpark to be a reasonable investment in purely financial terms, and (b) when my grandkids ask me what the **** I did to try to reduce global warming, I'll have at least one reasonable thing I can point to.
Second, even the better manufacturers put a barely usable Linux on the netbooks that wouldn't allow you to install any software without using the command line, broke the wireless when you installed software updates, etc.
Yep. I bought a eeePC, and the wireless didn't work. Called Asus tech support, and they told me that it had the wrong card installed, and they didn't have a linux driver for that card. My only option was to return it.
To be fair, there are also some other factors that worked against them. It's a low-margin product, so they can't afford to put too much money into producing a beautifully integrated system. The language barrier contributes to the lack of documentation. If you look at the reviews on amazon, it seems that a lot of doofuses bought a eeePC with linux on it, then tried to install windows, and failed.
TFA says that the way sound is implemented in the kernel is basically okay, but there are problems with how the kernel's facilities are used at higher levels by applications, and with the way the whole thing is integrated by distros. I think he's basically correct.
As an example of what's not broke about the kernel, and doesn't need to be fixed, it's a good thing that we still have support for OSS. OSS allows you to do sound I/O in exactly the way you would expect to do sound I/O based on the fundamental design principles of unix. You just do open(), ioctl(), read() or write() on devices like/dev/dsp. If you couldn't do that, it would be a failure to do the obvious, straightforward stuff to handle sound in the Unix Way.
As an example of what is broken at higher levels: I run Ubuntu Jaunty. Sound works fine every time I boot the computer, and I get the bongo sound as the login screen comes up. Then when I log in, master playback is muted, and the volume is down at 1/31. Also, the way the Gnome icon shows me that sound is muted (a tiny red box with a white x in it) is the same as the way the network icon would show me that I'd disconnected my ethernet cable or something; in other words, it makes it look like it's not just muted, but actually broken. Here's my best attempt to characterize the bug: Here's a bug on launchpad that may or may not be the same thing:
16d. Please analyze your own handwriting for us, and supply a full report on whether the results show that you may be predisposed to workplace violence.
16e. Please build your own polygraph machine, administer the test to yourself, and let us know whether it turns up any proclivity for white collar crime.
Start firing congressmen and senators in significant numbers, and things will change. Otherwise, quit the damn whining.
I live in Orange County, California, which is famous as a bastion of Reagan-style conservatism. In the last general election, my congressman, Ed Royce, outdid his Democratic opponent in fundraising by more than 10 to 1, and won with 67% of the vote.
Your prescription is not going to work here in my district. Vote the bum out? If you tell my neighbors that the NSA is reading people's email, they'll probably say that's great, because it's a good way to fight terrorism. My district isn't unusual, either. The reason incumbents in the US almost always get reelected is that we have a two-party system with geographically defined election districts, and party loyalty is highly correlated with geography.
It's a majoritarian fallacy to say that if the minority's rights are violated, the minority should just vote to have them not be violated anymore. The reason we have a constitution is to protect the rights of the minority, even when violating them is a very popular, majority position.
I will give you $100 if you can provide instructions on implementing this that can be understood by all my friends and family... and that includes my elderly relatives and my "but this is how it come when I bought the computer" friends.
There are two problems. One is the one you refer to, that mail clients make encryption way harder than it needs to be. The other is that there's a network effect. I could figure out how to get encryption working with my own mail client, but that would do me absolutely no good, because the people I send mail to don't know how to read my encrypted mail, and don't want to be bothered. The ease of use problem is much easier to solve than the network effect problem.
First off, pretty much every time we get one of these "OMG!" stories on slashdot about a security flaw going unfixed, we find out that it's not nearly as bad as suggested by the slashdot summary. In this case, the description linked to from the slashdot article says: "The Java plug-in does not block applets from launching file:// URLs. Visiting a website containing a maliciously crafted Java applet may allow a remote attacker to launch local files, which may lead to arbitrary code execution." So that's quite a bit less scary than the slashdot summary makes it sound. If I'm understanding correctly, it apparently doesn't let the attacker launch any code the attacker choses. It only lets the attacker launch code that's already present on the user's filesystem. And doesn't the java sandbox model prevent java applets from writing to the filesystem? So the attacker really may have very little opportunity to execute arbitrary code of the attacker's choosing.
Second: the slashdot summary says, "Apple had previously advised users to turn off Java temporarily in their Web browsers." Wow, that sounds really awful. It makes it sound like a really serious problem. But wait, the apple page doesn't say this. According to the tidbits.com article, Rich Mogull is the one who says the fix is to disable applets. The link to Rich Mogull's advice is a link within tidbits.com.
The problem with allowing causality violations, as I understand it, is that it causes problems for quantum mechanics.
No, there's nothing quantum-mechanical about the issue. The problem with causality violations is that they create paradoxes. What happens when you go back in time and kill yourself?
Thanks for the informative post. The thing is, if it's a proposal to build an Alcubierre drive, there are serious problems with that. And they're not just problems as in "that makes it hard to do," they're problems as in "it wouldn't even do what people have in mind when they imagine FTL."
I'll give a couple of examples, one using special relativity and one involving some general relativity, to amplify a little on what Geoffrey Landis said above.
Let's start with a couple of definitions. An "event" in relativity means a combination of time and place. Event B is defined as lying outside event A's light cone if the distance from A to B, in light-years, is greater than the time-difference between A and B, in years.
Example #1:
Suppose that faster-than-light (FTL) were possible. Then it would be possible for event A to cause event B, where B lies outside A's light cone.
You could simply travel in your FTL spaceship, starting at A and ending up at B, where you'd deliver a message. But according to special relativity, the time-ordering of events is not as absolute as in classical physics, because observers in different frames of reference disagree on the flow of time. Suppose the original setup was described according to one observer, O1, and now we have a second observer, O2, who is moving relative to O1 at an appreciable fraction of the speed of light. If the speed of the relative motion is high enough, then you can always get a situation where O2 says B happened before A, rather than after A. (This only happens if B is outside A's light cone.) So O1 says A caused B, but O2 says B caused A.
Example #2:
In general relativity, wormhole is a possible way to travel between different places, but since time and space are treated on the same footing in general relativity, there's every reason to believe that if wormholes exist, they would also go between different times, i.e., they would be time machines. But let's suppose for the sake of argument that you come across a wormhole that only goes between different places, with both mouths being synchronized in time. This would seem like FTL without time travel. But such a wormhole can always be used for time-travel as well. One method is to use gravitational fields to accelerate one mouth of the wormhole in some direction, bring it to a stop, and then use a similar acceleration and deceleration to bring it back to where it started. When you do this, you get something exactly like the twin "paradox" of special relavitity; the wormholes' times are no longer synchronized. So now if your no-time-travel FTL has been turned into FTL with time travel.
There's nothing special about these two examples. The idea that FTL naturally makes time travel possible is tightly bound to the structure of relativity. Since time travel seems to lead to causality paradoxes (e.g., going back in time and killing yourself), the conclusion seems to be that FTL leads to paradoxes, and that makes physicists suspect that FTL isn't actually physically possible.
No, he didn't. All they did was find one very simple toy model in which there was no inconsistency. And even in their toy model, there are too many solutions, with no way to tell which is the one that actually would occur.
There are a lot of very difficult theoretical problems involved in trying to describe the structure of neutron stars. The classic picture of a star made of nothing but neutrons is probably not quite right, and is possibly qualitatively wrong in important ways. There's supposed to be an upper limit on the mass of a neutron star, and the theoretical uncertainties get greater as you get closer to this mass limit. E.g., it's possible that you get quark stars. We just don't know, because we don't know the behavior of the strong and weak nuclear forces with sufficient precision to be able to extrapolate to these extreme conditions.
Given all that uncertainty, which has existed for many decades, it's not at all surprising to me that there's a corresponding uncertainty about the conditions under which a neutron star is or isn't unstable with respect to collapse into a black hole. The paper, which is linked to from the end of the Technology Review article, is pretty heavy going. My field is nuclear physics, not relativistic astrophysics, and I had a hard time understanding it. The author's English is also pretty hard to understand, so it's hard to tell exactly what he's saying his conclusions are. But if you look at the end, he seems to be suggesting that black holes actually do not form.
I wonder to what extent existing observations constrain this idea. For instance, we know that the Sagittarius A* object at the center of our galaxy has a mass of at least 3.7 million solar masses and a radius of less than 6.25 light-hours.
It would be interesting to know what he proposes this object is, if he says it's not a black hole.
This is literally true, but very misleading. Microsoft has ECMA bless.NET from time to time. Java has the Java Community Process. Yeah, sure, ECMA calls itself a standards organization, and the Java Community Process doesn't. If you look back at the history of Java, its big selling point from the beginning was that it was cross-platform, Sun fought intensely to make sure that it didn't get turned into a nonstandardized mess by MS, and Oracle's reference implementation is GPL'd. Microsoft, on the other hand, has demonstrated with OOXML that they see standards bodies as things that they can cynically manipulate.
The issue isn't that the software had a bug that had to be fixed. The issue is that the author of the software has shown himself to be untrustworthy by making his software interfere with other software, for the purpose of increasing his own financial gain from ads.
Slashdot Mirror
Ditto. I'm not saying the gp is wrong about his experience. But in my own case, I've found that both Yahoo and AOL just stopped accepting email from me ca. 2008. I run my own server on my own domain (not via a residential connection). In Yahoo's case, it was fairly easy to fix; I filled out a form, and after a while Yahoo users started receiving my emails again. With AOL, I haven't looked into trying to fix it. I know one person who uses AOL, and she doesn't get email from me@mydomain, or from people who send her email from academic accounts at the college where I work.
One good thing about installing domainkeys/dkim on your mail server is that Yahoo and Google both pay attention to it, and therefore they won't suddenly decide you're a spammer because someone else starts sending spam that's forged to look like it came from you. As far as I can tell, Yahoo, probably just insituted a policy last year of blocking vanity domains by default, but then unblocking them as soon as the owner of the domain fills out the form.
Huh? The story has a single big, evil computer in it. Nothing about a network at all, much less the internet.
The "mandatory" part doesn't make a lot of sense. You can't force authors to write books for free. And although a lot of free textbooks do exist already (see my sig), you can't guarantee that for a particular subject, the best book will always be a free book rather than a non-free.
But other than that, what you're suggesting seems similar to something California is doing now. Motivated by the California state budget crisis, Governor Schwarzenegger has announced a Free Digital Textbook Initiative, which has gathered a list of free, online high school math and science textbooks that are aligned with state content standards. The intention is to have the books used in classrooms in fall 2009. This article has some useful background, but it mistakenly suggests that the arduous state adoption process will be an obstacle to the FDTI; statewide adoption only applies to K-8, but FDTI is doing high-school books. There was a previous, unsuccessful effort called COSTP, which tried to produce a history textbook using Wikibooks. Here is a BBC article about the present effort, and here is a newspaper opinion piece by the Governor. This is a transcript of a speech by the Governor, with some interesting Q&A at the end. Twenty books were submitted (press release, links). The four books from traditional publisher Pearson are consumable workbooks, not actual textbooks.
Hmm...interesting, but I don't think it'll wash.
Friday was written in 1982, at which point ARPANET already existed, and the internet was already on its way. Most of the people using that kind of technology were in academia and the military, but it wasn't anything all that futuristic by then.
I've read For Us, The Living, which is the first fiction Heinlein ever wrote, ca. 1938. It wasn't published until 2003, so even if it had been extremely prescient, it wouldn't really be any more relevant than any other private thoughts of any individual that happened to be extremely prescient. I didn't remember anything internettish in it, and reading the link at nielsenhayden.com, it seems pretty clear to me that there wasn't. They've got what we would now refer to as videotelepony, fax machines, and pneumatic tubes. None of that has anything to do with the internet, which is a public network of general-purpose digital computers.
Actually Heinlein was pretty backward in stuff like this. For instance in Starman Jones (1953), the plot centers around the need for human spaceship pilots to navigate a spaceship by reading numbers out of tables printed in books, and then entering them into the ship's computer.
I think you're incorrect about Heinlein. If you look at his books, the closest I think he comes is The Moon Is a Harsh Mistress in 1966. A big central computer on the moon becomes self-aware, and he can project a synthesized voice and image over a video phone network. He's also networked to a lot of stuff, and can, e.g., make toilets run backwards. However, it's really not depicted as anything at all resembling the internet. All he really did was take existing time-sharing systems (the Dartmouth time-sharing system started in 1964) and extrapolate to the case where the central computer was self-aware, and the network spread across the whole moon. The way humans use the network in the story is always as nothing more than a video phone network. There is only one computer, and nobody ever transfers any digital data other than video telephony. It's true that the network is described as global (meaning global on the moon), but it's really only depicted as a telephone network, and a global telephone network already existed in 1966. A global network of computers would have been an innovation, but Heinlein doesn't depict the existence of any other computers on the network.
Probably "A Logic Named Joe," by Murray Leinster, is the most relevant example that predates the actual internet.
Okay, this is going to be wildly unpopular on slashdot, but I'm going to play devil's advocate and argue that the damages are not necessarily disproportionate to the actual harm.
It's definitely not correct to calculate the actual harm like this: 24 songs, multiplied by a market price of $1 per track on iTunes, giving 24x$1= $24. She wasn't being sued for obtaining single copies of the 24 songs illegally for her own use, she was being sued for giving them to other people -- possibly a very large number of other people.
Another possibility would be to take 24x$1, and then multiply by the number of people who actually downloaded the specific copies of the songs that she made available. There are a couple of problems with this. One is that nobody has any data on how many people downloaded those specific copies. Maybe zero people downloaded them. Maybe 10 people downloaded a particular song, each of them copied that copy 10 times for 10 friends, and so on, so that conceivably there are millions of mp3 files of that song out there that are all descended from Jammie Thomas's original. We just don't know, because the copies don't carry a history that allows us to know how they spread. You could argue that other people may have bought CDs of the same songs, ripped them, and put them online; again, there was no data about this at trial. And in any case, it's kind of morally obtuse to say that what she did isn't really so bad, because lots of other people were doing it as well. If you want to argue that what she did wasn't blameworthy, then you need to do it on other grounds, e.g., US copyright laws are unjust; well, I agree that US copyright laws are unjust, but we're talking about a court case here, so that's irrelevant. Suppose I set fire to the trees behind my neighbor's house, and around the same time 100 other people are also lighting matches and throwing them into the same bushes. My neighbor's house burns down, and he sues me. IANAL, but I doubt that the court is going to reduce the damages they make me pay by a factor of 100, on the theory that there were 99 other people attempting to do the same thing.
So it seems to me there's a pretty strong logical case to be made that you should calculate the damages like this: 24x$1xNxBxLxD. N is the total number of people who downloaded the song (and, as argued above, not just the number who got a copy directly from the ones Thomas put up). B is a factor that takes into account that only a fraction of the blame is hers, but I argued above that setting B=1 is reasonable (or at least not so unreasonable that it's obviously unconstitutional). L is a factor that measures how much the industry really loses when songs of a certain market value are downloaded. The industry usually assumes L=1, i.e., that every illegal download equals one lost sale. We actually don't know what L is. It's possible that 75% of downloaders would never have bought the song (e.g., they're 12-year-olds who have no money), in which case L=.25. It's even possible that L is negative (illegal downloads publicize and popularize the music, thereby driving legal sales). But these things are all impossible to determine, so I don't think it's completely crazy to say L=1, as a ballpark figure. (Again, the issue isn't whether it's exactly correct, the issue is whether it's so crazy that it gives a result that is obviously unconstitutionally disproportionate.) And finally D is the deterrent factor. When you sue someone for copyright violation, you're allowed to sue for more than the actual damages. As an example, if you write GPL'd software, in many cases the market value of the software is vanishingly small because you're intentionally giving it away for free; nevertheless if you satisfy certain technical requirements, you're allowed to sue for more than your actual damages, i.e., D can be much greater than one. (This is one reason that the GNU project encourages authors of GNU software to formally register a copyright, and sign over the copyrights to GNU. If you don't file a formal copyright re
The major/minor system is less than a thousand years old, and only occurs in Western music.
Thanks for the cool link :-) You're misusing the terminology a little -- the original NYT article is more correct.
Every sound can be broken down into a sum of sine waves. Usually, for basic physical reasons, those sine waves have frequencies that are all integer multiples (or nearly integer multiples) of the fundamental frequency. When they have this integer-multiple relationship, they're called "harmonics;" the more general term for the case where they're not integer multiples (anharmonic) is "partials." Any wind instrument that's made out of an air column is going to have integer-multiple harmonics, not anharmonic partials. So when you say that the harmonics are close to a modern flute, that's not really a useful statement; trivially, for physical reasons, any tone played on any wind instrument is going to have the same harmonics as the same note played on any other wind instrument. The only thing that will be different is the strengths of the harmonics.
What the expert quoted in the NYT article says is "The tones are quite harmonic." This is a different statement. It means that if you had two flutes like this one, and you played combinations of notes, they would sound good together. This has to do with how the scale is constructed. He also doesn't say the scale is the same as any particular modern one, just that it's a scale that sounds good in relation to itself.
The only cross-cultural universal we see today is that all cultures have what's called octave identification, meaning that, e.g., middle C and the C an octave above it are perceived as being similar, and able to play the same musical function. Most cultures don't have harmony at all -- that's mainly a function of Western music. Different cultures generally don't use the same scales. E.g., Beethoven, a Javanese gamelan orchestra, and a Delta blues musician use different scales in different ways. It wouldn't even make sense to interpret the expert's quote as saying that the scale is the same as today's scale, there's more than one scale used today.
Unfortunately I couldn't get the sound widget to play in my browser.
I'd tend to agree that this is useless voyeurism, except that there are some ethical issues that come up in transplants when the patient is very rich. The NY Times had an article about this today, and they specifically mentioned this hospital as one that had a very short average wait time of 3.8 months, compared to the national average of 12.3 months. "If you had access to a jet and had six hours to get anywhere in the country, you'd have a wide choice of programs," they quote one doctor as saying.
If you live in a remote rural area that's off the grid, you need a battery so you can have electricity at night, or when it's cloudy. The battery is big and expensive. That's a non-grid-tied system.
I use unison to sync my files on two desktop machines, a small-form-factor music server, and a web server. At various times I've had a mix of various OSes on the various machines, including Debian, Ubuntu, and FreeBSD. I use unison to sync them all. It's true that you don't want to sync all home directory files if one account is BSD and another is Debian. However, using USB sticks and email is, IMO, much more difficult, complicated, time-consuming, and error prone than using a file synchronizer. Every decent file synchronization program will have some way to control what gets synced and what doesn't. For instance, I have mine set up not to sync my .bashrc, which has to be different on BSD versus Debian. For syncing my home desktop with my music server, it's configured to sync the music files, whereas it won't sync music files between the desktop and the web server.
The OP says he's contemplating putting everything in a version control system. In my experience, version control systems are a hassle, and you don't want to use them for more things than you need to use them for. The most common reasons you might want a version control system are (1) aw, %#@!, I just deleted (or munged) that file, and (2) I want a three-year-old version of a file for some reason. For me, what works for #1 is to sync using unison very frequently, so that if I inadvertently delete a file, I can just get it back from another machine. For #2, that's what backups are for.
The problem with methods like thumb drives and email is that they're a pain, so you're not going to do them more than once a day or so. I probably run unison 10 or 20 times a day, because it's really quick and easy to do. That means that I can never lose more than about half an hour's worth of work, because that's how old my last sync was. It helps that unison is extremely efficient. For instance, if you have a 100 kB text file that you only changed one line of, it's smart enough to detect that, and just transmit a small amount of data over the network.
When I do have a good reason to use version control for a specific project, I use git. Now, git tends to make duplicate copies of large binary files, which you might think would be a problem. However, there's a nice trick, which is to use rzip rather than gzip for compressing your backup archives. Rzip can detect large, identical segments of a file that are far apart, so your backups that include git directories aren't any bigger than they would have been without the git directories.
I have a residential PV system.
This is incorrect. Unless you're in an area where grid power is not available, you buy a PV system that's grid-tied, and then if there's a blackout, you have no power. A non-grid-tied system is significantly more expensive and complicated, so you don't want one unless its your only option.
This is true, but the uncertainty argument isn't all in favor of getting PV. When people ask me how long my system will take to pay for itself, I always answer that I have absolutely no way of knowing, because I don't know what electricity will cost for the next 25 years, which is the design life of the system. If we get a big spike in electricity prices, I'll come out looking like a genius. But there are also other sources of uncertainty that work against PV. The price of the panels is expected to go down, and their efficiency is expected to go up. I don't know how much the price and efficiency will improve in the next 25 years, which is the design lifetime of my system. That's an uncertainty that argues against buying a PV system now.
I live in an area with expensive electricity and lots of sun. I have a south-facing roof with no shade. Given those factors, getting PV is a fairly reasonable investment, but there's no way to pin down exactly whether or not it's really going to look like a winner for me with hindsight in the year 2032 when the system's design lifetime is over.
What I do know is that (a) it's in the right ballpark to be a reasonable investment in purely financial terms, and (b) when my grandkids ask me what the **** I did to try to reduce global warming, I'll have at least one reasonable thing I can point to.
Yep. I bought a eeePC, and the wireless didn't work. Called Asus tech support, and they told me that it had the wrong card installed, and they didn't have a linux driver for that card. My only option was to return it.
To be fair, there are also some other factors that worked against them. It's a low-margin product, so they can't afford to put too much money into producing a beautifully integrated system. The language barrier contributes to the lack of documentation. If you look at the reviews on amazon, it seems that a lot of doofuses bought a eeePC with linux on it, then tried to install windows, and failed.
TFA says that the way sound is implemented in the kernel is basically okay, but there are problems with how the kernel's facilities are used at higher levels by applications, and with the way the whole thing is integrated by distros. I think he's basically correct.
As an example of what's not broke about the kernel, and doesn't need to be fixed, it's a good thing that we still have support for OSS. OSS allows you to do sound I/O in exactly the way you would expect to do sound I/O based on the fundamental design principles of unix. You just do open(), ioctl(), read() or write() on devices like /dev/dsp. If you couldn't do that, it would be a failure to do the obvious, straightforward stuff to handle sound in the Unix Way.
As an example of what is broken at higher levels: I run Ubuntu Jaunty. Sound works fine every time I boot the computer, and I get the bongo sound as the login screen comes up. Then when I log in, master playback is muted, and the volume is down at 1/31. Also, the way the Gnome icon shows me that sound is muted (a tiny red box with a white x in it) is the same as the way the network icon would show me that I'd disconnected my ethernet cable or something; in other words, it makes it look like it's not just muted, but actually broken. Here's my best attempt to characterize the bug: Here's a bug on launchpad that may or may not be the same thing:
Further instructions on the form:
16d. Please analyze your own handwriting for us, and supply a full report on whether the results show that you may be predisposed to workplace violence.
16e. Please build your own polygraph machine, administer the test to yourself, and let us know whether it turns up any proclivity for white collar crime.
I live in Orange County, California, which is famous as a bastion of Reagan-style conservatism. In the last general election, my congressman, Ed Royce, outdid his Democratic opponent in fundraising by more than 10 to 1, and won with 67% of the vote. Your prescription is not going to work here in my district. Vote the bum out? If you tell my neighbors that the NSA is reading people's email, they'll probably say that's great, because it's a good way to fight terrorism. My district isn't unusual, either. The reason incumbents in the US almost always get reelected is that we have a two-party system with geographically defined election districts, and party loyalty is highly correlated with geography.
It's a majoritarian fallacy to say that if the minority's rights are violated, the minority should just vote to have them not be violated anymore. The reason we have a constitution is to protect the rights of the minority, even when violating them is a very popular, majority position.
There are two problems. One is the one you refer to, that mail clients make encryption way harder than it needs to be. The other is that there's a network effect. I could figure out how to get encryption working with my own mail client, but that would do me absolutely no good, because the people I send mail to don't know how to read my encrypted mail, and don't want to be bothered. The ease of use problem is much easier to solve than the network effect problem.
Well, maybe.
First off, pretty much every time we get one of these "OMG!" stories on slashdot about a security flaw going unfixed, we find out that it's not nearly as bad as suggested by the slashdot summary. In this case, the description linked to from the slashdot article says: "The Java plug-in does not block applets from launching file:// URLs. Visiting a website containing a maliciously crafted Java applet may allow a remote attacker to launch local files, which may lead to arbitrary code execution." So that's quite a bit less scary than the slashdot summary makes it sound. If I'm understanding correctly, it apparently doesn't let the attacker launch any code the attacker choses. It only lets the attacker launch code that's already present on the user's filesystem. And doesn't the java sandbox model prevent java applets from writing to the filesystem? So the attacker really may have very little opportunity to execute arbitrary code of the attacker's choosing.
Second: the slashdot summary says, "Apple had previously advised users to turn off Java temporarily in their Web browsers." Wow, that sounds really awful. It makes it sound like a really serious problem. But wait, the apple page doesn't say this. According to the tidbits.com article, Rich Mogull is the one who says the fix is to disable applets. The link to Rich Mogull's advice is a link within tidbits.com.
No, there's nothing quantum-mechanical about the issue. The problem with causality violations is that they create paradoxes. What happens when you go back in time and kill yourself?
Thanks for the informative post. The thing is, if it's a proposal to build an Alcubierre drive, there are serious problems with that. And they're not just problems as in "that makes it hard to do," they're problems as in "it wouldn't even do what people have in mind when they imagine FTL."
I'll give a couple of examples, one using special relativity and one involving some general relativity, to amplify a little on what Geoffrey Landis said above.
Let's start with a couple of definitions. An "event" in relativity means a combination of time and place. Event B is defined as lying outside event A's light cone if the distance from A to B, in light-years, is greater than the time-difference between A and B, in years.
Example #1: Suppose that faster-than-light (FTL) were possible. Then it would be possible for event A to cause event B, where B lies outside A's light cone. You could simply travel in your FTL spaceship, starting at A and ending up at B, where you'd deliver a message. But according to special relativity, the time-ordering of events is not as absolute as in classical physics, because observers in different frames of reference disagree on the flow of time. Suppose the original setup was described according to one observer, O1, and now we have a second observer, O2, who is moving relative to O1 at an appreciable fraction of the speed of light. If the speed of the relative motion is high enough, then you can always get a situation where O2 says B happened before A, rather than after A. (This only happens if B is outside A's light cone.) So O1 says A caused B, but O2 says B caused A.
Example #2: In general relativity, wormhole is a possible way to travel between different places, but since time and space are treated on the same footing in general relativity, there's every reason to believe that if wormholes exist, they would also go between different times, i.e., they would be time machines. But let's suppose for the sake of argument that you come across a wormhole that only goes between different places, with both mouths being synchronized in time. This would seem like FTL without time travel. But such a wormhole can always be used for time-travel as well. One method is to use gravitational fields to accelerate one mouth of the wormhole in some direction, bring it to a stop, and then use a similar acceleration and deceleration to bring it back to where it started. When you do this, you get something exactly like the twin "paradox" of special relavitity; the wormholes' times are no longer synchronized. So now if your no-time-travel FTL has been turned into FTL with time travel.
There's nothing special about these two examples. The idea that FTL naturally makes time travel possible is tightly bound to the structure of relativity. Since time travel seems to lead to causality paradoxes (e.g., going back in time and killing yourself), the conclusion seems to be that FTL leads to paradoxes, and that makes physicists suspect that FTL isn't actually physically possible.
No, he didn't. All they did was find one very simple toy model in which there was no inconsistency. And even in their toy model, there are too many solutions, with no way to tell which is the one that actually would occur.
There are a lot of very difficult theoretical problems involved in trying to describe the structure of neutron stars. The classic picture of a star made of nothing but neutrons is probably not quite right, and is possibly qualitatively wrong in important ways. There's supposed to be an upper limit on the mass of a neutron star, and the theoretical uncertainties get greater as you get closer to this mass limit. E.g., it's possible that you get quark stars. We just don't know, because we don't know the behavior of the strong and weak nuclear forces with sufficient precision to be able to extrapolate to these extreme conditions.
Given all that uncertainty, which has existed for many decades, it's not at all surprising to me that there's a corresponding uncertainty about the conditions under which a neutron star is or isn't unstable with respect to collapse into a black hole. The paper, which is linked to from the end of the Technology Review article, is pretty heavy going. My field is nuclear physics, not relativistic astrophysics, and I had a hard time understanding it. The author's English is also pretty hard to understand, so it's hard to tell exactly what he's saying his conclusions are. But if you look at the end, he seems to be suggesting that black holes actually do not form.
I wonder to what extent existing observations constrain this idea. For instance, we know that the Sagittarius A* object at the center of our galaxy has a mass of at least 3.7 million solar masses and a radius of less than 6.25 light-hours. It would be interesting to know what he proposes this object is, if he says it's not a black hole.
This is literally true, but very misleading. Microsoft has ECMA bless .NET from time to time. Java has the Java Community Process. Yeah, sure, ECMA calls itself a standards organization, and the Java Community Process doesn't. If you look back at the history of Java, its big selling point from the beginning was that it was cross-platform, Sun fought intensely to make sure that it didn't get turned into a nonstandardized mess by MS, and Oracle's reference implementation is GPL'd. Microsoft, on the other hand, has demonstrated with OOXML that they see standards bodies as things that they can cynically manipulate.
The issue isn't that the software had a bug that had to be fixed. The issue is that the author of the software has shown himself to be untrustworthy by making his software interfere with other software, for the purpose of increasing his own financial gain from ads.