I believe that if you try to access google.com from China you'll get the same thing you'd get in the US. It's just that if you don't circumvent the firewall you'll see a lot of stuff censored. If you try to access google.cn from the US you get the same thing you would in China. A common tool used to circumvent the firewall is a proxy connection, with which Google won't know the access is coming from China anyway.
As far as I know the reason Google set up google.cn was because for most people in China access to google.com was really bad because of the firewall. By starting google.cn and playing by the Chinese government's rules they could operate a much more reliable (but self-censored) site and build their brand in the rapidly-growing Chinese market. google.com is still available in China to the same extent it was before, just google.cn is another option that will usually work "better" due to less government interference. I'm not saying it was necessarily the right thing to do from a humanitarian perspective, there's plenty of arguments both ways... but it certainly was the right thing to do as a long-term investment in Google's future.
Music, in your opinion, is a language or means to convey emotion. Why emotion? Do you say that emotion is the only thing that can be conveyed by music? Or is it that conveying emotion is a necessary condition for music? I think you'd find plenty of musicians throughout history that weren't very concerned with conveying emotion at all, while certainly there are others for whom it is paramount.
What I'm saying about music is that you can have a musical composition and a musical performance. When you record it and pump it into a store, there will very rarely be a connection with the listener that is musical. I agree that when the listener in the store stops and listens for a second, he chooses to be an audience and makes a musical connection. Maybe he has feelings communicated to him, maybe he has something other than feelings communicated to him. I also agree with your use of the word "choose": many people at concerts choose not to be audiences, and there really isn't music going on for those people at all. What I'm saying is that you can't misuse music because there is no music until the audience makes the connection. This might just be a matter of semantics; I once played for a conductor that told a story where he was directing an orchestra in a country he wasn't familiar with. He told the orchestra to take out their music and the group got very confused. After a while one of the trumpeters blew a note, then swatted at the air as if trying to pick it up. I guess the conductor was just using the wrong translation of "music"; in English we tend to use "music" to mean just about everything to do with music; I generally try to specify whether I'm talking about a composition, a performance, some sounds, or the Whole Enchilada.
Some people argue that music cannot happen through recordings. I'm not sure about that; I think that very often playback of recorded audio doesn't result in music but that it can. OTOH the performer-audience connection is completely and strictly one-way... and there are some other arguments that I don't remember. For the sake of not straw-manning them I won't go into it.
Yeah, indeterminacy and chance techniques were very important to Cage. More important than the bullshit "what is art" question, for sure. But GGP wasn't talking about indeterminacy, he was talking about the bullshit question, so indeterminacy wasn't really relavent. Whether Cage thought it was important or not, people asking the bullshit question have asked it in the context of 4'33" for years because 4'33" is the perfect pad from which to launch into bullshit arguments like mine. Furthermore, IIRC Cage was interested in the roles of performer and audience in a musical performance. Which ties into the bullshit question.
Some concept similar to indeterminacy/aleatoricness (I don't know a good noun for aleatoric... hm...) ties nicely into the discussion of the nature of games as an artistic medium. The actions of the player (simultaneously a performer and primary audience) are left open by the game's creators. The game of chess, one of the most successful games in history, is somewhat interesting in itself but the way people play it is what really makes it so fascinating.
Re:There's no such thing as art
on
Are Videogames Art?
·
· Score: 5, Insightful
Y'ever study John Cage? You've hit the nail precisely on the head! John Cage wrote a piece of music called 4'33", consisting of four minutes and thirty-three seconds of silence divided into 3 movements. Because it was performed as a work of music and accepted by its audience as a work of music, it was music. It has also been discussed ever since by musicians and by people that study music, adding weight to its status as a musical composition (it becomes music itself when it is performed and listened to). Meanwhile, consider the music that's pumped through speakers into stores. There is no performance, there is no attentive and active audience, and nobody cares about it. It's being played to present an atmosphere that will subtly convince consumers to buy more things. Even if what's being played is one of Beethoven's great symphonies, something with sound, with notes, with all kinds of recognizable musical elements, it's not being used as music (there is a composition, but only questionably a performance or audience); therefore its status as "music" is in question.
So your definition, as cynically as you offered it, is pretty much right on. Art requires artist and audience (these roles may overlap, or, as in much music, be separated further by tradition). That is all.
First, you don't need virtualization to "start fresh". You need backups.
So let's say you can easily back up to a known-good HD image (companies have been doing this for ages, often with the help of programs like Norton Ghost, which runs on a floppy disk under DOS, no fancy virtualization required). Put the antivirus folks out of business? Not so fast here. Don't viruses still infect your docs and executables? Won't users continue to create new documents and install new programs before realizing what's wrong? Won't users want to recover those? Won't users still be inconvenienced by reinstalling programs? As long as it's inconvenient to "start fresh" people won't do it. They'll prefer the pounds of dull and ineffective prevention over the ounce of bitter cure.
Let's say that the current profitable crop of malware doesn't infect documents like a virus does (I really don't know much about this). If people then find it really easy to just copy their $HOME and lay that over a known-good backup, thus cleanly avoiding the malware, guess what the malware folk will do: they'll make programs that act like viruses, infecting files and programs!
I haven't read Microsoft's statement that Word docs aren't for sharing, but I do know they put a lot of effort into creating and marketing their collaboration features. I think that Word docs are absolutely meant for sharing and collaboration within an organization that's standardized around MS-Office; in fact, if you have people within an organization collaborating on office-type documents there's probably no better way to do it than MS-Office's collaboration features. Are there other office suites that offer these features? I mean, I love using TeX+CVS for that kind of thing, but how many office workers are going to understand CVS (I think most could be trained to, they're not idiots, but MS-Office uses language that typical office workers are more familiar with while CVS uses terms that geeks like me are more familiar with... )?
What Word docs are bad for is trying to share a document with people outside of your organization that aren't going to collaborate on it. For one thing, traditionally you could easily wind up with metadata that you don't want getting out, such as revision history, stored in the documents. I don't know if recent versions of MS-Office have easy ways to clean all the invisible stuff out, but it's always a risk when you have revision control info stored in the documents themselves. For another, MS-Office docs aren't designed to render the same way every time the way Postscript and PDF files are. These aren't so much flaws in office as design decisions, and for the market that Microsoft is trying to sell to they seem like the right decisions to me. It's just a shame that Microsoft doesn't beat the "don't email people.docs" drum a bit harder. It's easy for me to see the proper place for Word docs, but I don't need Microsoft to tell me that. They've gotta get this message clearly across to their users.
Oddly enough, once when I was in college some dude sent me a promotional poster for some event in MS-Publisher format. He clearly hadn't thought about document formats for even a second before sending it out, so I didn't want to hit him with the full RMS-style rant... but I did let him know that not only could unwashed GNU/hippies like myself not open the file, but that even many MS Office-using knowledge-worker types would have a hard time with it as well. And that for pixel-perfect final posters he should be using Postscript or PDF. And a URL where he could find a Windows PDF print driver. I also made sure to include as many sweeping generalizations about the userbases of the various software packages as possible... or maybe that was just this/. post, I can't remember. If Microsoft's not going to educate the masses I guess it's up to all the people that get burned by 'em.
These people that did this study with real watches and lots of people sitting around... you don't think they actually didn't take into account the hand movements, do you? The study said that in the course of doing real editing, probably of a non-technical nature, that it actually is quicker to move your hand to the mouse, because the keyboard commands are really that much less efficient. I don't know how technical their users were, I doubt they were using Vim, and they probably didn't have to make huge multi-screen selections of the type that mice tend to make a mess of.
They probably also had their people sitting with proper posture and with all their equipment well-positioned. They probably had well-maintained mice with mousepads. They probably had enough room in their setup to allow for effective mouse use. When all these things are true it takes very little time to move your hand to the mouse, do some things quickly, then move back. If you're reaching for your keyboard or hunched over that makes it more difficult to move to the mouse. I have a trackball that allows me to operate without taking much space, but it's a bit slower than a mouse (all my best minesweeper times have come with a normal mouse). Laptop users have a whole slew of input devices that sacrifice something (usually speed and ergonomics) to deal with the limited space in which they can operate; on the other hand, it's often very efficient to switch from keyboard to mouse on a laptop.
I've read that some people did studies using actual stopwatches and found that using the mouse for things like selecting text and moving your input cursor is faster than using the keyboard on average. Using the keyboard is faster for some things, but mostly it's that using the keyboard feels faster than using the mouse. Sometimes when I'm bouncing around in Vim I realize that some of my operations would be much quicker easier if I used the mouse (Vim, of course, has a mouse language but I'm not all that familiar with it). It feels like there are more situations where the keyboard is quicker, but when the mouse is quicker it's a *lot* quicker. But I don't have stopwatch numbers to back that particular feeling up. It's hard to test things like that without a bunch of test subjects.
"Fortunately, law doesn't work like this. You cannot exercise judicial power for things outside a country's juristiction or if you're not a citizen of that country (you can be held accountable for age of consent laws without borders solely based on citizenship, for example if you're an Australian and go to Thailand and back, you can still be arrested for sexual crimes if you had sex with a 13 year old girl in Thailand)."
Just for the sake of argument, if an US company sets up a factory in the far east that operates in a manner that would be illegal in the US (with regards to such things as child labor, labor relations, wages, pollution controls, etc.) would that company be held accountable for those actions in the US? Even if the US laws don't make sense in whatever other country they set up in (the best example being a minimum wage law based on US costs of living)? Yeah, I'm sure the foreign company is technically a separate company, but this is a thought experiment so I don't care.
I think your point about exporting, in general, makes sense: people can take measures to prevent exports that their own country deems illegal. If we were really going to treat Internet exporting like physical exporting, though, the government would be cracking down on proxy servers that potentially broker traffic between embargoed nations and websites offering services. We generally don't do that, and it would probably be a horrible mess if we did, but I think that there should be a clearly codified reason why. We wouldn't want the government to randomly begin applying pressure on particular proxies it didn't like by surprise, or anything like that.
"It's not illegal (afaik) in the UK to export gambling services, just like its not illegal to export P2P clients, so why should they be bothered."
They should be bothered because a UK man was just arreseted for exporting gambling services to the US, and because some P2P company was somehow ordered by an Australian court to provide a modified program to Australians (I don't remember exactly how that was enforced, but it was on/. a while back). They should be bothered because people that distribute Falun Gong material aren't treated very nicely when they travel to China. Because, while they perhaps should be able to export anything their own countries allow from their domestic servers, the countries they're exporting to aren't seeing it this way.
This is why I believe we (the United States, but also every other country that wants to consider itself a civilized nation) need a clear policy for these situations: what should our response be to illegal material from other countries on the Internet? Right now we clearly have confusion, because if there was a clear policy that the people in his situation would be arrested our UK friend would probably never have traveled to the US.
In the case of selling most products internationally you have to actually make an effort to sell in another country. Even if you're shipping stuff sold "on the Internet" you're shipping it physically to a foreign address and should certainly make sure you're not shipping stuff to countries where it's illegal. When you open up a server on the Internet that content is available to everyone by default. In fact, there's not even really a reliable way to tell where the hits are coming from given proxies, VPNs and whatnot. The same problem exists with payment methods: yes, in most cases it's easy to determine the location of a payer given, say, a credit card billing address. But it's certainly possible to get money routed through foreign accounts or through online payment systems that don't list a shipping address. Perhaps when operating a site that accepts payments for controversial items one should mandate a payment system that includes a billing address and refuse to bill any countries but those where the service is known to be legal. That's not really very difficult, so you're right that this one guy has it easy.
But what about services that don't require payment? For example, the P2P software that's been deemed illegal in Australia. Or racist websites operating in the US that run afoul of Canadian and European laws. Falun Gong sites that are illegal in China. Various types of pornography that are illegal in various places in the world. IIRC it's still illegal to export cryptographic software from the US to certain embargoed nations under the same laws that prohibit the sale of weapons to these nations; I'm pretty sure if you sold any other weapon to such a country the Feds wouldn't take, "But he clicked a box on our website saying he wasn't from Iran," for an answer. We run into the problem that it's easy to cross national borders on the Internet and impossible to control which borders you're crossing. All you have to do to "export" something online is sign up with your neighborhood ISP. This is one of the reasons it's been so successful.
Expressing the laws of every country and state through a technological standard would be ridiculously challenging; even accurately representing the location of clients to every server would be pretty difficult without massive cooperation. If every country in the world tries to apply their import/export laws to Internet traffic we'll be creating massive amounts of work for server operators, who would likely overcompensate by trying to lock down their services to domestic viewers. Countries like Saudi Arabia and China that figure they probably can't force foreign webmasters to obey them set up firewalls, but those are expensive and hard to maintain. I don't have a total solution, but I think as a practical matter Internet traffic can't be treated like standard imports and exports.
Linux (the kernel) supports DRM exactly the same as the Windows NT kernel does: it doesn't do anything in particular to support or preclude it. Under Linux you could circumvent many types of DRM schemes with kernel modules that dump audio or video data to disk, in theory. Under Windows NT you could do the same. There are various other techniques you might use to either snoop out the unencrypted data or discover a program's keys with debuggers and disassemblers if you have kernel-level or even root-level access to your system. It's hard for software to hide, because computers have for so many years made programs examinable. Linux and current versions of Windows don't do much against these trends.
Vista is implementing a bunch of new techniques to make media-handling programs unexaminable, to disallow unsigned kernel modules, etc.
This is why it's surprising to hear that XP SP2 would have support for this stuff. Especially given the earlier announcement that 32-bit versions of Vista wouldn't support HD playback. I guess it's all up to the drives: if the drives allow XP software to authenticate and retrieve the content then XP will "support" HD playback.
I'm guessing they like that all Xboxen have the same exact hardware, whereas all old PCs don't. This way they can just create one Linux image and slap it on all the Xboxen without worrying about differences in hardware compatibility and performance wasting all their time.
I know I'll sound humor-challenged, but... Cygwin works on Windows 95, and there are various projects trying to get various versions of Gnome running under Cygwin. UTFG and all will become clear.
I'll jump on the bandwagon of people that ignore the fact that you specifically state that you're using WinXP Pro and suggest that you switch to a different OS. Scroll bars are on the left in every Plan 9 program I can think of. This is also true in those programs' ports to other OSes (such as wily, the acme clone for Unix). For that matter, most X11 terminal emulators put the scroll bars on the left, and so do some other old X programs...
Forget pop-quizzes just for the sake of attendance enforcement. Pop quiz to make sure that people are staying up with the material. That's the real problem with poor attendance: people don't keep pace with the class and prevent the class from building on material and moving at the proper pace. Random pop-quizzes already will get people in the door for fear of missing one, and I don't see how making them meta-quizzes helps with this. If people struggle with the pop-quizzes they'll be worried by a bad grade and come to office hours to try to catch up (even if the quizzes don't actually affect grades that much).
I have a laptop with a very loud startup beep that can't be disabled (one of my roomates said when he first heard it go off he thought it was the fire alarm). I run FreeBSD on the laptop, and FreeBSD has an even longer startup beep that I only recently learned how to disable (you have to edit some assembly code and rebuild the first-stage bootloader). I always carried around one of those 1/8"-to-1/4" headphone converters and stuck it in the headphone jack when booting my computer to disable the startup beeps.
People saw it sticking out the back of the computer and asked me what it was. I came up with lots of funny answers.
Since when did "professional" mean "difficult to make"? If the site conveys its content in a clear way who cares if you could have made it in high school? A web site that's simple to implement is a great thing, and extra technologies (that usually will increase development, maintenance and bandwidth costs) need to be justified in terms of how they actually make the site's experience better.
Many people have pointed out that Linux's firewall is in its kernel. That is true, and parent post gets off the mark trying to describe Windows as a technical monolith without clean interfaces between components when most other OSes are the same way.
But there still is a difference, it's just not technical: Windows as a product is a monolith, without divisions even where users would really appreciate them. You want one part, you get it all. One part is delayed, the whole thing is delayed. With, say, Ubuntu they'll still put out a new release if some of the major software included has major upcoming releases, and just allow users to upgrade later. This is possible because they're distributing free (the "beer" and "speech" aspects of this help equally) software, and largely distributing to people that are willing to go through these upgrades. And even if the distro won't package them most of the projects are independent enough in their development that you can upgrade them yourself. A 3d desktop for X may or may not be ready for the general public (ever), but you can find out whether it's ready for your system and use it today. With Windows you have to wait for all the other Vista features. ACLs in Linux may be a hassle for most people to set up, but if you want 'em you've got 'em. On Windows, to get Vista's account management abilities you have to wait for the rest of Vista to be done.
In the hypothetical case of someone that wanted to write a Free-Software driver for FreeBSD and Linux (that is, not the case of nVidia, which wants to write a closed-source driver for whatever operating systems are useful to it), putting it in the BSD tree doesn't work exactly: for it to gain the benefits of being in the Linux tree (which are many) it has to actually be in the Linux tree.
The author could offer it under both the BSD license to the BSD tree and GPLv2 to the Linux tree. But contributions added by other Linux hackers would be stuck under GPLv2 and not available to the BSD version. At any rate, the benefits of being in the BSD mainline kernel tree aren't as great as those for Linux because at least FreeBSD keeps binary module compatibility. So I think most people would probably choose the Linux tree given the choice. I guess there's always the public domain: you could release the code to the public domain, and put a note in your code asking particularly the Linux devs to just release their modifications to the public domain independent of the GPLv2-license that they extend to the Linux kernel version. If they don't, that's their choice.
That said, I think the differences between the two are large enough that most drivers couldn't share a whole lot of code between the OSes.
The Windows key is very useful, and the fact that no applications use it is what makes it useful: it allows for a wide variety of windowing system-specific keybindings that won't conflict with application keybindings. That's exactly the way I set up my FVWM bindings, all on mod4, and no matter what crazy-ass bindings my apps use (and it wouldn't be a Unix app if it didn't have crazy-ass bindings) they don't conflict with my crazy-ass bindings because just about no apps use mod4 for shortcuts (a pox on any that do).
Windows differs from my setup (which is clearly perfect) in a few ways. First, it uses alt-tab, whereas I use mod4+tab for consistency (and my keyboard's two meta keys are each bigger than its one alt key anyhow). Second, it brings up the Start menu when you tap the Windows key, which is the part that's pure evil. IMHO modifier keys should strictly be modifier keys. This also goes for alt (compare, say, GTK+'s alt behavior with that of most Win32 programs).
But the overriding point is that Microsoft hardware doesn't really have much to do with the Windows key. Unix vendors and also Apple have had similar keys long before Microsoft introduced theirs (though Apple's key is also the primary key used for app bindings). And IMO its failings have entirely to do with software. If you want to disable it in Windows there's a registry hack to do that, which can be found by searching the wb if you're lucky.
1. Well if you're lucky you just recompile. If you're not lucky you fix stuff first. I could try to argue that a stable binary kernel module interface is important, but I'm honestly not sure that it's best for Linux. I've read the arguments for both sides, and you can have a successful kernel either way. One problem with the Linux way of wanting everyone to put their drivers in the main tree is that it makes things difficult for driver projects that target multiple OSes and it forces drivers to be GPLv2 (which is something that will just never happen for a ton of drivers out there; it also means that once a driver gets into the mainline tree it can't be the basis for, say, a mainline BSD driver). The advantages that it brings for users and kernel developers are real (on my box at home all my modules are mainline, and I enjoy this greatly; furthermore, the freedom to do things that break binary module compatibility is just one less obstacle in the way of a better kernel), but they only apply to the drivers that are in the mainline kernel tree. It's like the minimum wage, which helps people with jobs but is sometimes argued to hurt those without. As Linux is today, as a Linux user, I love having a tree full of good drivers, and that wouldn't be there without the unstable binary module interface IMO.
It sucks for users that when you have a closed-source Linux driver the changes to reflect the changing module interfaces get tied in with the changes in driver features. No matter how many configurations you compile for you'll never get all of them. But I think I've butchered this question enough. Maybe I should have trolled against Linux instead. Moving on.
2. W = J/s. The two-square centimeter chip dissipates 100 Watts (period). Or maybe it doesn't, I don't keep track of those things too precisely myself (it's not really my department). I'll take your word for it; it doesn't get any easier when there are four of them sitting right next to eachother in one box, but somehow those contraptions manage to run for quite a long time crunching through 3d programs without overheating or crashing. It gets pretty loud (what with the helicopter engines and all; fortunately most of the realy hardcore systems are heavy enough that they don't lift off), but them's the breaks. YMMV. Anyhoo, the abacus I've been using to calculate the packets to send down the line is getting very loud and hot (damn friction!), so it's probably about time I stop.
But wouldn't aren't there already plenty of organizations that do the same thing in lots of different industries? What about the (RI|MP)AA? The BSA? All these groups that/. loves to hate behave this way, and they aren't being investigated for antitrust. I'm sure if one went about it right one could probably create an organization of contractors to help contractors protect their rights. A quick look at Wikipedia indicates that such a group might be called a Trade Organization. Of course, if they couldn't get a critical mass of contractors to join the group they'd just lose jobs to non-members that were willing to take the contract risk and if they did get a critical mass they might wind up getting corrupt and trying to force non-members to join or perish.
I work at nVidia (driver development... please don't kill me, herd of vicious/.ers), and I was talking at lunch the other day to one of the guys that does a lot of interviews. He mentioned this one question he loved to ask, and said the reason he loved it is because there was no right answer and not even really any good answer. He asks the interviewees to basically think completely out loud, display their whole problem-solving process while working, and has a list of several approaches that are good ones to take into the problem but that all ultimately fail. The more of these a candidate goes through before giving up completely the better he's done.
He asked me the question at my interview. I remember asking him a lot of questions while trying to solve it. In any other context but an interview I'd probably have recognized the kind of excercise it was, but in the interview context I was frantically trying everything I could to come up with something that would come close to working. Nifty technique.
Slashdot Mirror
I believe that if you try to access google.com from China you'll get the same thing you'd get in the US. It's just that if you don't circumvent the firewall you'll see a lot of stuff censored. If you try to access google.cn from the US you get the same thing you would in China. A common tool used to circumvent the firewall is a proxy connection, with which Google won't know the access is coming from China anyway.
As far as I know the reason Google set up google.cn was because for most people in China access to google.com was really bad because of the firewall. By starting google.cn and playing by the Chinese government's rules they could operate a much more reliable (but self-censored) site and build their brand in the rapidly-growing Chinese market. google.com is still available in China to the same extent it was before, just google.cn is another option that will usually work "better" due to less government interference. I'm not saying it was necessarily the right thing to do from a humanitarian perspective, there's plenty of arguments both ways... but it certainly was the right thing to do as a long-term investment in Google's future.
Music, in your opinion, is a language or means to convey emotion. Why emotion? Do you say that emotion is the only thing that can be conveyed by music? Or is it that conveying emotion is a necessary condition for music? I think you'd find plenty of musicians throughout history that weren't very concerned with conveying emotion at all, while certainly there are others for whom it is paramount.
What I'm saying about music is that you can have a musical composition and a musical performance. When you record it and pump it into a store, there will very rarely be a connection with the listener that is musical. I agree that when the listener in the store stops and listens for a second, he chooses to be an audience and makes a musical connection. Maybe he has feelings communicated to him, maybe he has something other than feelings communicated to him. I also agree with your use of the word "choose": many people at concerts choose not to be audiences, and there really isn't music going on for those people at all. What I'm saying is that you can't misuse music because there is no music until the audience makes the connection. This might just be a matter of semantics; I once played for a conductor that told a story where he was directing an orchestra in a country he wasn't familiar with. He told the orchestra to take out their music and the group got very confused. After a while one of the trumpeters blew a note, then swatted at the air as if trying to pick it up. I guess the conductor was just using the wrong translation of "music"; in English we tend to use "music" to mean just about everything to do with music; I generally try to specify whether I'm talking about a composition, a performance, some sounds, or the Whole Enchilada.
Some people argue that music cannot happen through recordings. I'm not sure about that; I think that very often playback of recorded audio doesn't result in music but that it can. OTOH the performer-audience connection is completely and strictly one-way... and there are some other arguments that I don't remember. For the sake of not straw-manning them I won't go into it.
Yeah, indeterminacy and chance techniques were very important to Cage. More important than the bullshit "what is art" question, for sure. But GGP wasn't talking about indeterminacy, he was talking about the bullshit question, so indeterminacy wasn't really relavent. Whether Cage thought it was important or not, people asking the bullshit question have asked it in the context of 4'33" for years because 4'33" is the perfect pad from which to launch into bullshit arguments like mine. Furthermore, IIRC Cage was interested in the roles of performer and audience in a musical performance. Which ties into the bullshit question.
Some concept similar to indeterminacy/aleatoricness (I don't know a good noun for aleatoric... hm...) ties nicely into the discussion of the nature of games as an artistic medium. The actions of the player (simultaneously a performer and primary audience) are left open by the game's creators. The game of chess, one of the most successful games in history, is somewhat interesting in itself but the way people play it is what really makes it so fascinating.
Y'ever study John Cage? You've hit the nail precisely on the head! John Cage wrote a piece of music called 4'33", consisting of four minutes and thirty-three seconds of silence divided into 3 movements. Because it was performed as a work of music and accepted by its audience as a work of music, it was music. It has also been discussed ever since by musicians and by people that study music, adding weight to its status as a musical composition (it becomes music itself when it is performed and listened to). Meanwhile, consider the music that's pumped through speakers into stores. There is no performance, there is no attentive and active audience, and nobody cares about it. It's being played to present an atmosphere that will subtly convince consumers to buy more things. Even if what's being played is one of Beethoven's great symphonies, something with sound, with notes, with all kinds of recognizable musical elements, it's not being used as music (there is a composition, but only questionably a performance or audience); therefore its status as "music" is in question.
So your definition, as cynically as you offered it, is pretty much right on. Art requires artist and audience (these roles may overlap, or, as in much music, be separated further by tradition). That is all.
First, you don't need virtualization to "start fresh". You need backups.
So let's say you can easily back up to a known-good HD image (companies have been doing this for ages, often with the help of programs like Norton Ghost, which runs on a floppy disk under DOS, no fancy virtualization required). Put the antivirus folks out of business? Not so fast here. Don't viruses still infect your docs and executables? Won't users continue to create new documents and install new programs before realizing what's wrong? Won't users want to recover those? Won't users still be inconvenienced by reinstalling programs? As long as it's inconvenient to "start fresh" people won't do it. They'll prefer the pounds of dull and ineffective prevention over the ounce of bitter cure.
Let's say that the current profitable crop of malware doesn't infect documents like a virus does (I really don't know much about this). If people then find it really easy to just copy their $HOME and lay that over a known-good backup, thus cleanly avoiding the malware, guess what the malware folk will do: they'll make programs that act like viruses, infecting files and programs!
I haven't read Microsoft's statement that Word docs aren't for sharing, but I do know they put a lot of effort into creating and marketing their collaboration features. I think that Word docs are absolutely meant for sharing and collaboration within an organization that's standardized around MS-Office; in fact, if you have people within an organization collaborating on office-type documents there's probably no better way to do it than MS-Office's collaboration features. Are there other office suites that offer these features? I mean, I love using TeX+CVS for that kind of thing, but how many office workers are going to understand CVS (I think most could be trained to, they're not idiots, but MS-Office uses language that typical office workers are more familiar with while CVS uses terms that geeks like me are more familiar with... )?
.docs" drum a bit harder. It's easy for me to see the proper place for Word docs, but I don't need Microsoft to tell me that. They've gotta get this message clearly across to their users.
/. post, I can't remember. If Microsoft's not going to educate the masses I guess it's up to all the people that get burned by 'em.
What Word docs are bad for is trying to share a document with people outside of your organization that aren't going to collaborate on it. For one thing, traditionally you could easily wind up with metadata that you don't want getting out, such as revision history, stored in the documents. I don't know if recent versions of MS-Office have easy ways to clean all the invisible stuff out, but it's always a risk when you have revision control info stored in the documents themselves. For another, MS-Office docs aren't designed to render the same way every time the way Postscript and PDF files are. These aren't so much flaws in office as design decisions, and for the market that Microsoft is trying to sell to they seem like the right decisions to me. It's just a shame that Microsoft doesn't beat the "don't email people
Oddly enough, once when I was in college some dude sent me a promotional poster for some event in MS-Publisher format. He clearly hadn't thought about document formats for even a second before sending it out, so I didn't want to hit him with the full RMS-style rant... but I did let him know that not only could unwashed GNU/hippies like myself not open the file, but that even many MS Office-using knowledge-worker types would have a hard time with it as well. And that for pixel-perfect final posters he should be using Postscript or PDF. And a URL where he could find a Windows PDF print driver. I also made sure to include as many sweeping generalizations about the userbases of the various software packages as possible... or maybe that was just this
These people that did this study with real watches and lots of people sitting around... you don't think they actually didn't take into account the hand movements, do you? The study said that in the course of doing real editing, probably of a non-technical nature, that it actually is quicker to move your hand to the mouse, because the keyboard commands are really that much less efficient. I don't know how technical their users were, I doubt they were using Vim, and they probably didn't have to make huge multi-screen selections of the type that mice tend to make a mess of.
They probably also had their people sitting with proper posture and with all their equipment well-positioned. They probably had well-maintained mice with mousepads. They probably had enough room in their setup to allow for effective mouse use. When all these things are true it takes very little time to move your hand to the mouse, do some things quickly, then move back. If you're reaching for your keyboard or hunched over that makes it more difficult to move to the mouse. I have a trackball that allows me to operate without taking much space, but it's a bit slower than a mouse (all my best minesweeper times have come with a normal mouse). Laptop users have a whole slew of input devices that sacrifice something (usually speed and ergonomics) to deal with the limited space in which they can operate; on the other hand, it's often very efficient to switch from keyboard to mouse on a laptop.
I've read that some people did studies using actual stopwatches and found that using the mouse for things like selecting text and moving your input cursor is faster than using the keyboard on average. Using the keyboard is faster for some things, but mostly it's that using the keyboard feels faster than using the mouse. Sometimes when I'm bouncing around in Vim I realize that some of my operations would be much quicker easier if I used the mouse (Vim, of course, has a mouse language but I'm not all that familiar with it). It feels like there are more situations where the keyboard is quicker, but when the mouse is quicker it's a *lot* quicker. But I don't have stopwatch numbers to back that particular feeling up. It's hard to test things like that without a bunch of test subjects.
"Fortunately, law doesn't work like this. You cannot exercise judicial power for things outside a country's juristiction or if you're not a citizen of that country (you can be held accountable for age of consent laws without borders solely based on citizenship, for example if you're an Australian and go to Thailand and back, you can still be arrested for sexual crimes if you had sex with a 13 year old girl in Thailand)."
Just for the sake of argument, if an US company sets up a factory in the far east that operates in a manner that would be illegal in the US (with regards to such things as child labor, labor relations, wages, pollution controls, etc.) would that company be held accountable for those actions in the US? Even if the US laws don't make sense in whatever other country they set up in (the best example being a minimum wage law based on US costs of living)? Yeah, I'm sure the foreign company is technically a separate company, but this is a thought experiment so I don't care.
I think your point about exporting, in general, makes sense: people can take measures to prevent exports that their own country deems illegal. If we were really going to treat Internet exporting like physical exporting, though, the government would be cracking down on proxy servers that potentially broker traffic between embargoed nations and websites offering services. We generally don't do that, and it would probably be a horrible mess if we did, but I think that there should be a clearly codified reason why. We wouldn't want the government to randomly begin applying pressure on particular proxies it didn't like by surprise, or anything like that.
/. a while back). They should be bothered because people that distribute Falun Gong material aren't treated very nicely when they travel to China. Because, while they perhaps should be able to export anything their own countries allow from their domestic servers, the countries they're exporting to aren't seeing it this way.
"It's not illegal (afaik) in the UK to export gambling services, just like its not illegal to export P2P clients, so why should they be bothered."
They should be bothered because a UK man was just arreseted for exporting gambling services to the US, and because some P2P company was somehow ordered by an Australian court to provide a modified program to Australians (I don't remember exactly how that was enforced, but it was on
This is why I believe we (the United States, but also every other country that wants to consider itself a civilized nation) need a clear policy for these situations: what should our response be to illegal material from other countries on the Internet? Right now we clearly have confusion, because if there was a clear policy that the people in his situation would be arrested our UK friend would probably never have traveled to the US.
In the case of selling most products internationally you have to actually make an effort to sell in another country. Even if you're shipping stuff sold "on the Internet" you're shipping it physically to a foreign address and should certainly make sure you're not shipping stuff to countries where it's illegal. When you open up a server on the Internet that content is available to everyone by default. In fact, there's not even really a reliable way to tell where the hits are coming from given proxies, VPNs and whatnot. The same problem exists with payment methods: yes, in most cases it's easy to determine the location of a payer given, say, a credit card billing address. But it's certainly possible to get money routed through foreign accounts or through online payment systems that don't list a shipping address. Perhaps when operating a site that accepts payments for controversial items one should mandate a payment system that includes a billing address and refuse to bill any countries but those where the service is known to be legal. That's not really very difficult, so you're right that this one guy has it easy.
But what about services that don't require payment? For example, the P2P software that's been deemed illegal in Australia. Or racist websites operating in the US that run afoul of Canadian and European laws. Falun Gong sites that are illegal in China. Various types of pornography that are illegal in various places in the world. IIRC it's still illegal to export cryptographic software from the US to certain embargoed nations under the same laws that prohibit the sale of weapons to these nations; I'm pretty sure if you sold any other weapon to such a country the Feds wouldn't take, "But he clicked a box on our website saying he wasn't from Iran," for an answer. We run into the problem that it's easy to cross national borders on the Internet and impossible to control which borders you're crossing. All you have to do to "export" something online is sign up with your neighborhood ISP. This is one of the reasons it's been so successful.
Expressing the laws of every country and state through a technological standard would be ridiculously challenging; even accurately representing the location of clients to every server would be pretty difficult without massive cooperation. If every country in the world tries to apply their import/export laws to Internet traffic we'll be creating massive amounts of work for server operators, who would likely overcompensate by trying to lock down their services to domestic viewers. Countries like Saudi Arabia and China that figure they probably can't force foreign webmasters to obey them set up firewalls, but those are expensive and hard to maintain. I don't have a total solution, but I think as a practical matter Internet traffic can't be treated like standard imports and exports.
Linux (the kernel) supports DRM exactly the same as the Windows NT kernel does: it doesn't do anything in particular to support or preclude it. Under Linux you could circumvent many types of DRM schemes with kernel modules that dump audio or video data to disk, in theory. Under Windows NT you could do the same. There are various other techniques you might use to either snoop out the unencrypted data or discover a program's keys with debuggers and disassemblers if you have kernel-level or even root-level access to your system. It's hard for software to hide, because computers have for so many years made programs examinable. Linux and current versions of Windows don't do much against these trends.
Vista is implementing a bunch of new techniques to make media-handling programs unexaminable, to disallow unsigned kernel modules, etc.
This is why it's surprising to hear that XP SP2 would have support for this stuff. Especially given the earlier announcement that 32-bit versions of Vista wouldn't support HD playback. I guess it's all up to the drives: if the drives allow XP software to authenticate and retrieve the content then XP will "support" HD playback.
I'm guessing they like that all Xboxen have the same exact hardware, whereas all old PCs don't. This way they can just create one Linux image and slap it on all the Xboxen without worrying about differences in hardware compatibility and performance wasting all their time.
I know I'll sound humor-challenged, but... Cygwin works on Windows 95, and there are various projects trying to get various versions of Gnome running under Cygwin. UTFG and all will become clear.
I'll jump on the bandwagon of people that ignore the fact that you specifically state that you're using WinXP Pro and suggest that you switch to a different OS. Scroll bars are on the left in every Plan 9 program I can think of. This is also true in those programs' ports to other OSes (such as wily, the acme clone for Unix). For that matter, most X11 terminal emulators put the scroll bars on the left, and so do some other old X programs...
Forget pop-quizzes just for the sake of attendance enforcement. Pop quiz to make sure that people are staying up with the material. That's the real problem with poor attendance: people don't keep pace with the class and prevent the class from building on material and moving at the proper pace. Random pop-quizzes already will get people in the door for fear of missing one, and I don't see how making them meta-quizzes helps with this. If people struggle with the pop-quizzes they'll be worried by a bad grade and come to office hours to try to catch up (even if the quizzes don't actually affect grades that much).
You probably should escape the slash in your substitution string there.
I have a laptop with a very loud startup beep that can't be disabled (one of my roomates said when he first heard it go off he thought it was the fire alarm). I run FreeBSD on the laptop, and FreeBSD has an even longer startup beep that I only recently learned how to disable (you have to edit some assembly code and rebuild the first-stage bootloader). I always carried around one of those 1/8"-to-1/4" headphone converters and stuck it in the headphone jack when booting my computer to disable the startup beeps.
People saw it sticking out the back of the computer and asked me what it was. I came up with lots of funny answers.
Since when did "professional" mean "difficult to make"? If the site conveys its content in a clear way who cares if you could have made it in high school? A web site that's simple to implement is a great thing, and extra technologies (that usually will increase development, maintenance and bandwidth costs) need to be justified in terms of how they actually make the site's experience better.
Many people have pointed out that Linux's firewall is in its kernel. That is true, and parent post gets off the mark trying to describe Windows as a technical monolith without clean interfaces between components when most other OSes are the same way.
But there still is a difference, it's just not technical: Windows as a product is a monolith, without divisions even where users would really appreciate them. You want one part, you get it all. One part is delayed, the whole thing is delayed. With, say, Ubuntu they'll still put out a new release if some of the major software included has major upcoming releases, and just allow users to upgrade later. This is possible because they're distributing free (the "beer" and "speech" aspects of this help equally) software, and largely distributing to people that are willing to go through these upgrades. And even if the distro won't package them most of the projects are independent enough in their development that you can upgrade them yourself. A 3d desktop for X may or may not be ready for the general public (ever), but you can find out whether it's ready for your system and use it today. With Windows you have to wait for all the other Vista features. ACLs in Linux may be a hassle for most people to set up, but if you want 'em you've got 'em. On Windows, to get Vista's account management abilities you have to wait for the rest of Vista to be done.
In the hypothetical case of someone that wanted to write a Free-Software driver for FreeBSD and Linux (that is, not the case of nVidia, which wants to write a closed-source driver for whatever operating systems are useful to it), putting it in the BSD tree doesn't work exactly: for it to gain the benefits of being in the Linux tree (which are many) it has to actually be in the Linux tree.
The author could offer it under both the BSD license to the BSD tree and GPLv2 to the Linux tree. But contributions added by other Linux hackers would be stuck under GPLv2 and not available to the BSD version. At any rate, the benefits of being in the BSD mainline kernel tree aren't as great as those for Linux because at least FreeBSD keeps binary module compatibility. So I think most people would probably choose the Linux tree given the choice. I guess there's always the public domain: you could release the code to the public domain, and put a note in your code asking particularly the Linux devs to just release their modifications to the public domain independent of the GPLv2-license that they extend to the Linux kernel version. If they don't, that's their choice.
That said, I think the differences between the two are large enough that most drivers couldn't share a whole lot of code between the OSes.
The Windows key is very useful, and the fact that no applications use it is what makes it useful: it allows for a wide variety of windowing system-specific keybindings that won't conflict with application keybindings. That's exactly the way I set up my FVWM bindings, all on mod4, and no matter what crazy-ass bindings my apps use (and it wouldn't be a Unix app if it didn't have crazy-ass bindings) they don't conflict with my crazy-ass bindings because just about no apps use mod4 for shortcuts (a pox on any that do).
Windows differs from my setup (which is clearly perfect) in a few ways. First, it uses alt-tab, whereas I use mod4+tab for consistency (and my keyboard's two meta keys are each bigger than its one alt key anyhow). Second, it brings up the Start menu when you tap the Windows key, which is the part that's pure evil. IMHO modifier keys should strictly be modifier keys. This also goes for alt (compare, say, GTK+'s alt behavior with that of most Win32 programs).
But the overriding point is that Microsoft hardware doesn't really have much to do with the Windows key. Unix vendors and also Apple have had similar keys long before Microsoft introduced theirs (though Apple's key is also the primary key used for app bindings). And IMO its failings have entirely to do with software. If you want to disable it in Windows there's a registry hack to do that, which can be found by searching the wb if you're lucky.
1. Well if you're lucky you just recompile. If you're not lucky you fix stuff first. I could try to argue that a stable binary kernel module interface is important, but I'm honestly not sure that it's best for Linux. I've read the arguments for both sides, and you can have a successful kernel either way. One problem with the Linux way of wanting everyone to put their drivers in the main tree is that it makes things difficult for driver projects that target multiple OSes and it forces drivers to be GPLv2 (which is something that will just never happen for a ton of drivers out there; it also means that once a driver gets into the mainline tree it can't be the basis for, say, a mainline BSD driver). The advantages that it brings for users and kernel developers are real (on my box at home all my modules are mainline, and I enjoy this greatly; furthermore, the freedom to do things that break binary module compatibility is just one less obstacle in the way of a better kernel), but they only apply to the drivers that are in the mainline kernel tree. It's like the minimum wage, which helps people with jobs but is sometimes argued to hurt those without. As Linux is today, as a Linux user, I love having a tree full of good drivers, and that wouldn't be there without the unstable binary module interface IMO.
It sucks for users that when you have a closed-source Linux driver the changes to reflect the changing module interfaces get tied in with the changes in driver features. No matter how many configurations you compile for you'll never get all of them. But I think I've butchered this question enough. Maybe I should have trolled against Linux instead. Moving on.
2. W = J/s. The two-square centimeter chip dissipates 100 Watts (period). Or maybe it doesn't, I don't keep track of those things too precisely myself (it's not really my department). I'll take your word for it; it doesn't get any easier when there are four of them sitting right next to eachother in one box, but somehow those contraptions manage to run for quite a long time crunching through 3d programs without overheating or crashing. It gets pretty loud (what with the helicopter engines and all; fortunately most of the realy hardcore systems are heavy enough that they don't lift off), but them's the breaks. YMMV. Anyhoo, the abacus I've been using to calculate the packets to send down the line is getting very loud and hot (damn friction!), so it's probably about time I stop.
But wouldn't aren't there already plenty of organizations that do the same thing in lots of different industries? What about the (RI|MP)AA? The BSA? All these groups that /. loves to hate behave this way, and they aren't being investigated for antitrust. I'm sure if one went about it right one could probably create an organization of contractors to help contractors protect their rights. A quick look at Wikipedia indicates that such a group might be called a Trade Organization. Of course, if they couldn't get a critical mass of contractors to join the group they'd just lose jobs to non-members that were willing to take the contract risk and if they did get a critical mass they might wind up getting corrupt and trying to force non-members to join or perish.
I work at nVidia (driver development... please don't kill me, herd of vicious /.ers), and I was talking at lunch the other day to one of the guys that does a lot of interviews. He mentioned this one question he loved to ask, and said the reason he loved it is because there was no right answer and not even really any good answer. He asks the interviewees to basically think completely out loud, display their whole problem-solving process while working, and has a list of several approaches that are good ones to take into the problem but that all ultimately fail. The more of these a candidate goes through before giving up completely the better he's done.
He asked me the question at my interview. I remember asking him a lot of questions while trying to solve it. In any other context but an interview I'd probably have recognized the kind of excercise it was, but in the interview context I was frantically trying everything I could to come up with something that would come close to working. Nifty technique.