Chapter 3 is far and away the most interesting in the book. It sets forth the so-called 10,000 hour rule, and in its course, shows why Bill Gates and the Beatles succeeded for essentially the same reason. Gladwell begins by noting that musical geniuses such as Mozart, and chess grandmasters, both achieved their status after about 10 years. 10 years is roughly how long it takes to put in 10,000 hours of hard practice. 10,000 hours is the magic number of greatness.
I used to be really good at DDR. In Konami's official Internet Ranking, I placed 10th place in North America, and 98th place World Wide. I hit my peak after about 2 years of practice.
I'm a bit past the 10th year of practice now (I started playing around 1997-1998), and I'm nowhere near my previous level. I can't even claim to be within the top 10 of my province anymore, let alone all of North America.
The fact that Asian languages in many cases use of shorter and more logical words for numbers confers a strong early advantage which, like the age advantage in the hockey player example, snowball significantly over time.
I really don't know what the author is referring to here. I think the system used by most western cultures is the best numbering system I know. You can judge the magnitude of a number based on the number of digits it contains. You can't do that with roman numerals, and you can't do that with Asian numbering systems. Hopefully I won't need to convince you that the system used by the Romans was terrible.
In Asian languages 234 is written as äOEç(TM)¾äåå, or literally translated: "[2][100][3][10][4]". Doesn't this seem reminiscent of the Roman system? "CCXXXIV"? The real innovation with the western system was that the position of a digit gave its magnitude. 234 = 2 * 10^2 + 3 * 10^1 + 4 * 10^0. This is what made it so much easier to work with over the Roman system, where you basically had some rules about adding and subtracting digits together (C means 100, so CC means 100 + 100; IV means 5 - 1 = 4). The Asian system is similar to the roman system, but uses multiplicate and addition (äOE means 2 and ç(TM)¾ means 100, so äOEç(TM)¾ means 200; ç(TM)¾äOE means 100 + 2).
(In physics we preferred the term "not going to happen" as opposed to "impossible".)
Well, we do use "impossible" for stuff which actually is impossible. I don't often hear "not going to happen"; instead, around here, it's "VLP" for "Very Low Probability" or "negligible probability" (and it's debatable about which of the two is lower probability than the other). We don't use "NP" for "Negligible Probability" because a lot of the people in my circle of geeks are also comp-sci people, and "NP" already has a meaning.
Most big printers can do Post Script, and PS is a turing complete language, so in theory there is nothing preventing you from making a PS file-sharing program to run on the printers.
I don't know much about Post-Script, but if there is no API for networking in the language, that might sort of kill any attempt to write a file-sharing program in that language.
People tend to forget that Turing-Complete only means that it can run any program that could be run on a Turing Machine. While TMs are powerful, recall that TMs don't have keyboard, mcie, screens, speakers, network cards, CD drives, etc. so there are many programs that you are running on your computer right now which, in fact, cannot be implemented on a Turing Machine.
That's why the "Turing-Complete" argument typically used when talking about algorithms, rather than complete programs.
Lessig's book lists four reasons for piracy, and how each could affect sales. It's an excellent read.
I googled for it and it's at http://www.sslug.dk/~chlor/lessig/freeculture/ but it's kinda long. Can you list what the 4 reasons are, or at least on which of the 200 pages we can find those 4 reasons?
I don't think I've ever seen a music torrent complete in less time than it would take me to rip a CD and grab the track listing from CDDB.
You can queue up multiple CDs to download before going to bed. You can't queue up multiple CDs to rip (unless you got multiple CD drives, or a robot arm to load the CDs for you).
Regardless, there are probably about ten people alive who have "pirated" music simply to avoid ripping it from a CD they already own.
I torrented music I already owned a couple years ago after a harddrive crash. Why didn't I have a backup of my mp3s? 'Cause I knew I could easily get it again via torrents, and in the worst case, I still got the original CDs anyway.
(Actually some of the mp3s I lost are songs I composed myself, and I didn't have a backup of those, so they're "gone forever" and I'm really kicking myself for that.)
Ahem, isn't any digital format, including CD "lossy" to some extent?
No. "Lossy" here has a very specific meaning. It refers to, when compressing data, throwing away or modifying some of the data to make it more easily compressible.
I'll give several examples of "data loss" which are not lossy. Some of them may sound facetious, but my intent is really to highlight how many different points "data loss" can occur, even beyond simply using a digital format.
If someone plays a guitar into a microphone, and the microphone is crappy and "loses some data", we usually don't use the term "lossy" here, because we were not compressing data (we were recording data). If your speakers are crappy and can't reproduce the sounds of the guitar accurately, again we usually don't use the term "lossy". If your hearing is kind of crappy, and you're sitting in the same room as the guitarist, who's playing live, and you don't hear every single subtle nuance in his playing, we're losing data, but we don't use the term "lossy". If the guitarist is unskilled, and he has this concept of how the song should sound in his mind, but his fingers don't play it quite the way he imagined it, we're losing data, but we don't call it lossy.
Also note that if the original artist is composing everything digitally via software, and not using any recording equipment at all, then it's quite possible to use digital formats and have zero loss of information.
And if you have a really hardcore artist who directly creates a.mp3 file using a hex editor, then it's plausible that the resulting file contains EXACTLY the sound the artist envisioned, and so you could say that you have a "lossless mp3 recording", though most people wouldn't know what you meant by that and you'd have to explain how the file was created.
We have to show them that we can produce art of the same, or better, quality without the protections of traditional copyright. We need to show them that copyright is unnecessary or harmful. We already have a decent start on this: artists are increasingly releasing their work to be shared freely, a significant portion of the software market is dominated by GPL or BSD licensed software, and people everywhere are making and distributing content on sites like YouTube. We've not quite reached parity yet with the traditional media; but it's close.
Note that the only reason GPL works is because of copyright, though. By default, you're not allowed to use other people's code (due to copyright violation). GPL grants you the right to use certain code as long as you follow certain restrictions (e.g. that you release your own code under GPL as well).
If copyright is eliminated entirely, then GPL will "fail" in the sense that people could then take the source code posted on the internet, and use it without complying to the GPL, and there'd be no way to sue them for doing so.
(And just to show that I'm trying to clarify something here, and not just doing a strawman argument -- nor arguing with you at all, for that matter, -- I'll explicitly highlight the fact that there's another possibility that I'm not bothering to address, which is that we don't eliminate copyright altogether, but simply "fix it somehow" so that GPL still works, but everything that makes copyright sucks disappears.)
My opinion is that if a place won't hire me over petty personal stuff, I don't need to work for them that badly.
My opinion is that if I was born into a poor family and the job market sucks, then you DO need to work for them badly, even if they do petty personal stuff, or hell, even if they do outright illegal stuff like pay beneath minimum wage or not declare their income for tax purposes.
Interesting stories and characters are important, but they must be balanced by varied and entertaining gameplay. The lack of either will be a limiting factor in how many people return to play once the primary plot is completed.
Neither of the factor is a must for a game to be a success. World of Warcraft for example. It has no story, weak characters and gameplay which been obsolete for about 10 years. Halo is another example.
The "Warcraft" and "Halo" worlds are actually pretty well fleshed out with lots of detail, so I don't think these are the best examples for "successful games with no story".
Better examples, IMHO, include: Tetris, Dance Dance Revolution, Rock Band, most flight sims and racing games.
Fallout III actually had a more correct approach to giving the player a true open world choice in that the entire landscape was available to be explored the moment you exile yourself from the vault. And every sidequest and other task is available to be completed as the player's own judgment and they can go in any direction and order. You can even choose your character's name, gender, race, and some facial features.
Choosing your name, gender, race and some facial features isn't really that impressive in terms of "truly open world", IMHO.
What bugged me about Fallout 3 is how many of what are supposed to be "roleplaying perks" did not actually do much for roleplaying at all. I'm talking about the "Kid at heart" and "Seducer/Seductress" perks (not sure if that's the exact name, I'm reciting this from memory). The opportunity to use these perks were extremely rare, and even then only for "one-off" situations. Never actually molded the full story.
Plus the "Karma" system in Fallout 3 removed any and all moral ambiguity. They specifically showed which actions were "good" and which were "bad". There were no dilemma to go through, no self-examination or personal growth.
Space Quest, Maniac Mansion, King's Quest, etc were quite "open" games in the sense that players were encouraged to try anything they could think of to solve puzzles.
Just my personal opinion here, but any game in which, when you try the "wrong" thing, kills you off and presents "Gee, I sure hope you have an earlier save" message box, doesn't really encourage you to try anything you can think of.
"Companions of Xanth" was a bit better in this regard in that they made custom responses to even the most absurd actions (e.g. "Talk to table" yields "Sorry, the table has taken a vow of silence.") and while a lot of things did kill you in Xanth, instead of the normal dialog with only "Quit", "New Game", "Load", there was also an "undo last action" button, so dying by experimentation became a lot less painful and a lot more fun.
The "Monkey Islands" series was even better (in this particular metric of "encourage players to try everything"), because there was no way to die from trying things. I believe the series had 4 games, and there was only 1 game you could die in, and even then it was only via inaction (don't do anything for 10 real-life minutes), rather than from experimentation.
If the user leaves before a task completes, you don't have anything to reduce.
Google's implementation of MapReduce already takes this into account. Haven't you heard of how they just have a bunch of vanilla x86 networked together, and when one of them fails, they just throw it away plug in a new one.
Something to think about:
If I'm sending names to your pc, what can I derive from that list without having the entire list?
Frequency of each name? Frequency of characters in names? Bayesian probability of one character following another in names? Number of names of a particular length?
Each worker would compute the stats for their chunk of work (the "Map" part of MapReduce), and then send the results back to the server to be aggregated (the "Reduce" part of MapReduce).
Some of these may seem interesting, but then again, what interesting data can you derive at all from a list of names, even if you had the whole list?
Javascript really isn't suited for this kind of thing, even with worker threads, for two reasons I can think of. First, web clients are transient... they'd have to report back often in case the user clicks away.
I don't see why web clients being transient is a problem. The whole point of the MapReduce algorithm is that each worker (the web clients in this case) don't need to know anything about what the other worker is doing, what the system as a whole is doing, nor what it had done with any past job.
You don't need to have every band member play to a metronome to sync them up, even if they're recorded individually and with isolation. If they aren't all playing at the same time, as long as the rhythm track is done first, everyone plays to that.
In a way, what you call "the rhythm track" is essentially what this "click track" concept refers to. In some (many?) cases, the track is just two different percussive sounds played at regular intervals (essentially mimicking a metronome). In other cases, it might be a synth or MIDI rendering of the song. In yet others, it might be a "single-track" version of the song, and the musicians are replaying along with it just for the purpose of having a multitrack recording."
I have no idea what you were talking about when you said 'you can't just tell everyone to play 120bpm and then have them sync up.
I'm referring to the scenario of you telling someone to play at "120bpm", without providing them with any tempo-aid at all. One band member might end up playing at 119bpm instead, and another at 121bpm, and thus they won't be synchronized.
Sorry, I really don't understand your post. When you say "You don't", what is it that "I don't do" that you are referring to? And how are the existence of headphones relevant?
Even if a drum track sounds the same for say one section of a song there will be small variations in rhythm and volume over time that to human ears just sounds more natural.
It would also be not much fun for the drummer as he would just be laying down loops instead of an entire track.
As a drummer, the presence or absence of a click track doesn't really affect the "fun". If the song has a repetitive drum pattern played over and over throughout, the song is not fun to play; and this is true regardless of whether I have to get the timing perfectly right with a click track, or if I just play the same old pattern loosely. What makes a song fun (for me) is having lots of interesting variations throughout the song.
Actually one of the more fun things for me to do as a drummer is to use a click track for some odd time signature I haven't really practiced or internalized yet (say 11/8), and try to improvise over the click. The click itself (on most metronomes the "1" will sound different from all the other beats) help me know what the "feel" of the time signature is supposed to be, and my challenge is to see how quickly I can adapt and play around with that feel.
The only time I use a metronome is when I'm drumming on my own, to improve my own performance, but I could never play a whole song with the invariable click in my ears (actually that's not true, there is one song where I used a metronome, but only because I had a continuous snare drum pattern and was (am?) just not good enough to keep the rythm).
I've toyed with the idea of creating a "whoosh" track, a less fucking annoying and more forgiving version of a click-track, to help my own occasionally shaky rhythm-keeping. Dig, with something like a white-noise swell whooshing in time instead of the click. I truly don't mind playing with headphones on, even if usually just to give me a mix of everyone else's monitors that can compete with my own volume, but, man, that unforgiving click is worse than having Roger Waters glaring at you balefully.
One of my most uncomfortable drumming moments was during a live show where I was using a click track which was just a tad too quiet. Everything was going perfectly fine until I hit a mini-drum-solo/fill a bit too quickly, and when I came out of it, I so I was off the rest of the song. I had to really focus to play everything with a 3/16th delay after when I actually heard the click in my headphones.
If the metronome is "click", the snare is "tack", and the two simultaneous is "clack", then the song, in my mind, went from "clack, clack, clack, clack" to "cltack, cltack, cltack, cltack". Kind of gave it a nice jazz-swing feel, but I wouldn't want to go through that again.
Much music has always been explicitly shit, regardless of when it was made. Go back and take a look at the charts for any year you care to name, and probably 95% of the artists will be people you've never heard of...because they were shit, had their fifteen minutes, and are now long-forgotten.
Not only that, but I bet 95% of the artists never made it to the top 5%.
I think in some ways, the modern recording tools, have helped kill good music in many ways, it can really mask the lack of talent in todays musicians.
I guess it depends on whether you care more about the music or the musician. If all you care about is "good music", then who cares whether it was even a human who played the music, or if it was all computer generated with synthesizers?
Maybe there's an artist who can't play any instrument at all, but is a great composer, and so all his songs are "played" by software. No human has ever played the song, but the music sounds great.
Maybe there's a mathematician who can't play instruments and can't compose, but figured out an algorithm which generates beautiful melodies. No human has ever composed the song or played the song, but the music songs great.
Maybe there's a programmer who can't play instruments, can't compose, and can't figure out complex equations, but wrote a genetic algorithm that figured out the equations to compose the songs. No human ever conceived of the math, composed the song, or played the song, but the music sounds great.
However to leverage much of the flexibility and power of a digital recording you need a click.
Really? Why?
It has more to do with multitrack recording than digital recording. When you do a multitrack recording, each track is recorded individually. Usually this means that each band member will be recorded in isolation, and it's very rare for every band member to have such a developped internal metronome such that you can just tell them "play at 120 beats per minute", and when you combine all the tracks, they'll all be synced up.
Instead, what is done is that you set a metronome to play at 120bpm, and each bandmember will play along to the metronome.
Actually, the pattern is that it alternates. So the next good console will be easy to program (i.e. XBox360), and the one after that will be hard to program (i.e. PS4).
Slashdot Mirror
I present to you: Burger Space.
(äOE means 2 and ç(TM)¾ means 100, so äOEç(TM)¾ means 200; ç(TM)¾äOE means 100 + 2).
Oh gosh; it's 2009 and Slashdot still doesn't support Unicode?
I am Asian.
Chapter 3 is far and away the most interesting in the book. It sets forth the so-called 10,000 hour rule, and in its course, shows why Bill Gates and the Beatles succeeded for essentially the same reason. Gladwell begins by noting that musical geniuses such as Mozart, and chess grandmasters, both achieved their status after about 10 years. 10 years is roughly how long it takes to put in 10,000 hours of hard practice. 10,000 hours is the magic number of greatness.
I used to be really good at DDR. In Konami's official Internet Ranking, I placed 10th place in North America, and 98th place World Wide. I hit my peak after about 2 years of practice.
I'm a bit past the 10th year of practice now (I started playing around 1997-1998), and I'm nowhere near my previous level. I can't even claim to be within the top 10 of my province anymore, let alone all of North America.
The fact that Asian languages in many cases use of shorter and more logical words for numbers confers a strong early advantage which, like the age advantage in the hockey player example, snowball significantly over time.
I really don't know what the author is referring to here. I think the system used by most western cultures is the best numbering system I know. You can judge the magnitude of a number based on the number of digits it contains. You can't do that with roman numerals, and you can't do that with Asian numbering systems. Hopefully I won't need to convince you that the system used by the Romans was terrible.
In Asian languages 234 is written as äOEç(TM)¾äåå, or literally translated: "[2][100][3][10][4]". Doesn't this seem reminiscent of the Roman system? "CCXXXIV"? The real innovation with the western system was that the position of a digit gave its magnitude. 234 = 2 * 10^2 + 3 * 10^1 + 4 * 10^0. This is what made it so much easier to work with over the Roman system, where you basically had some rules about adding and subtracting digits together (C means 100, so CC means 100 + 100; IV means 5 - 1 = 4). The Asian system is similar to the roman system, but uses multiplicate and addition (äOE means 2 and ç(TM)¾ means 100, so äOEç(TM)¾ means 200; ç(TM)¾äOE means 100 + 2).
green != blue
It is if you add yellow pigment to your blue. green=blue+yellow (in pigment)
5 != 4
It is if you add 1 to your 4. 5=4+1 (in arithmatic)
(In physics we preferred the term "not going to happen" as opposed to "impossible".)
Well, we do use "impossible" for stuff which actually is impossible. I don't often hear "not going to happen"; instead, around here, it's "VLP" for "Very Low Probability" or "negligible probability" (and it's debatable about which of the two is lower probability than the other). We don't use "NP" for "Negligible Probability" because a lot of the people in my circle of geeks are also comp-sci people, and "NP" already has a meaning.
Most big printers can do Post Script, and PS is a turing complete language, so in theory there is nothing preventing you from making a PS file-sharing program to run on the printers.
I don't know much about Post-Script, but if there is no API for networking in the language, that might sort of kill any attempt to write a file-sharing program in that language.
People tend to forget that Turing-Complete only means that it can run any program that could be run on a Turing Machine. While TMs are powerful, recall that TMs don't have keyboard, mcie, screens, speakers, network cards, CD drives, etc. so there are many programs that you are running on your computer right now which, in fact, cannot be implemented on a Turing Machine.
That's why the "Turing-Complete" argument typically used when talking about algorithms, rather than complete programs.
Lessig's book lists four reasons for piracy, and how each could affect sales. It's an excellent read.
I googled for it and it's at http://www.sslug.dk/~chlor/lessig/freeculture/ but it's kinda long. Can you list what the 4 reasons are, or at least on which of the 200 pages we can find those 4 reasons?
I don't think I've ever seen a music torrent complete in less time than it would take me to rip a CD and grab the track listing from CDDB.
You can queue up multiple CDs to download before going to bed. You can't queue up multiple CDs to rip (unless you got multiple CD drives, or a robot arm to load the CDs for you).
Regardless, there are probably about ten people alive who have "pirated" music simply to avoid ripping it from a CD they already own.
I torrented music I already owned a couple years ago after a harddrive crash. Why didn't I have a backup of my mp3s? 'Cause I knew I could easily get it again via torrents, and in the worst case, I still got the original CDs anyway.
(Actually some of the mp3s I lost are songs I composed myself, and I didn't have a backup of those, so they're "gone forever" and I'm really kicking myself for that.)
Ahem, isn't any digital format, including CD "lossy" to some extent?
No. "Lossy" here has a very specific meaning. It refers to, when compressing data, throwing away or modifying some of the data to make it more easily compressible.
I'll give several examples of "data loss" which are not lossy. Some of them may sound facetious, but my intent is really to highlight how many different points "data loss" can occur, even beyond simply using a digital format.
If someone plays a guitar into a microphone, and the microphone is crappy and "loses some data", we usually don't use the term "lossy" here, because we were not compressing data (we were recording data). If your speakers are crappy and can't reproduce the sounds of the guitar accurately, again we usually don't use the term "lossy". If your hearing is kind of crappy, and you're sitting in the same room as the guitarist, who's playing live, and you don't hear every single subtle nuance in his playing, we're losing data, but we don't use the term "lossy". If the guitarist is unskilled, and he has this concept of how the song should sound in his mind, but his fingers don't play it quite the way he imagined it, we're losing data, but we don't call it lossy.
Also note that if the original artist is composing everything digitally via software, and not using any recording equipment at all, then it's quite possible to use digital formats and have zero loss of information.
And if you have a really hardcore artist who directly creates a .mp3 file using a hex editor, then it's plausible that the resulting file contains EXACTLY the sound the artist envisioned, and so you could say that you have a "lossless mp3 recording", though most people wouldn't know what you meant by that and you'd have to explain how the file was created.
We have to show them that we can produce art of the same, or better, quality without the protections of traditional copyright. We need to show them that copyright is unnecessary or harmful. We already have a decent start on this: artists are increasingly releasing their work to be shared freely, a significant portion of the software market is dominated by GPL or BSD licensed software, and people everywhere are making and distributing content on sites like YouTube. We've not quite reached parity yet with the traditional media; but it's close.
Note that the only reason GPL works is because of copyright, though. By default, you're not allowed to use other people's code (due to copyright violation). GPL grants you the right to use certain code as long as you follow certain restrictions (e.g. that you release your own code under GPL as well).
If copyright is eliminated entirely, then GPL will "fail" in the sense that people could then take the source code posted on the internet, and use it without complying to the GPL, and there'd be no way to sue them for doing so.
(And just to show that I'm trying to clarify something here, and not just doing a strawman argument -- nor arguing with you at all, for that matter, -- I'll explicitly highlight the fact that there's another possibility that I'm not bothering to address, which is that we don't eliminate copyright altogether, but simply "fix it somehow" so that GPL still works, but everything that makes copyright sucks disappears.)
My opinion is that if a place won't hire me over petty personal stuff, I don't need to work for them that badly.
My opinion is that if I was born into a poor family and the job market sucks, then you DO need to work for them badly, even if they do petty personal stuff, or hell, even if they do outright illegal stuff like pay beneath minimum wage or not declare their income for tax purposes.
Interesting stories and characters are important, but they must be balanced by varied and entertaining gameplay. The lack of either will be a limiting factor in how many people return to play once the primary plot is completed.
Neither of the factor is a must for a game to be a success. World of Warcraft for example. It has no story, weak characters and gameplay which been obsolete for about 10 years. Halo is another example.
The "Warcraft" and "Halo" worlds are actually pretty well fleshed out with lots of detail, so I don't think these are the best examples for "successful games with no story".
Better examples, IMHO, include: Tetris, Dance Dance Revolution, Rock Band, most flight sims and racing games.
Fallout III actually had a more correct approach to giving the player a true open world choice in that the entire landscape was available to be explored the moment you exile yourself from the vault. And every sidequest and other task is available to be completed as the player's own judgment and they can go in any direction and order. You can even choose your character's name, gender, race, and some facial features.
Choosing your name, gender, race and some facial features isn't really that impressive in terms of "truly open world", IMHO.
What bugged me about Fallout 3 is how many of what are supposed to be "roleplaying perks" did not actually do much for roleplaying at all. I'm talking about the "Kid at heart" and "Seducer/Seductress" perks (not sure if that's the exact name, I'm reciting this from memory). The opportunity to use these perks were extremely rare, and even then only for "one-off" situations. Never actually molded the full story.
Plus the "Karma" system in Fallout 3 removed any and all moral ambiguity. They specifically showed which actions were "good" and which were "bad". There were no dilemma to go through, no self-examination or personal growth.
Space Quest, Maniac Mansion, King's Quest, etc were quite "open" games in the sense that players were encouraged to try anything they could think of to solve puzzles.
Just my personal opinion here, but any game in which, when you try the "wrong" thing, kills you off and presents "Gee, I sure hope you have an earlier save" message box, doesn't really encourage you to try anything you can think of.
"Companions of Xanth" was a bit better in this regard in that they made custom responses to even the most absurd actions (e.g. "Talk to table" yields "Sorry, the table has taken a vow of silence.") and while a lot of things did kill you in Xanth, instead of the normal dialog with only "Quit", "New Game", "Load", there was also an "undo last action" button, so dying by experimentation became a lot less painful and a lot more fun.
The "Monkey Islands" series was even better (in this particular metric of "encourage players to try everything"), because there was no way to die from trying things. I believe the series had 4 games, and there was only 1 game you could die in, and even then it was only via inaction (don't do anything for 10 real-life minutes), rather than from experimentation.
If the user leaves before a task completes, you don't have anything to reduce.
Google's implementation of MapReduce already takes this into account. Haven't you heard of how they just have a bunch of vanilla x86 networked together, and when one of them fails, they just throw it away plug in a new one.
Something to think about: If I'm sending names to your pc, what can I derive from that list without having the entire list?
Frequency of each name? Frequency of characters in names? Bayesian probability of one character following another in names? Number of names of a particular length?
Each worker would compute the stats for their chunk of work (the "Map" part of MapReduce), and then send the results back to the server to be aggregated (the "Reduce" part of MapReduce).
Some of these may seem interesting, but then again, what interesting data can you derive at all from a list of names, even if you had the whole list?
Javascript really isn't suited for this kind of thing, even with worker threads, for two reasons I can think of. First, web clients are transient... they'd have to report back often in case the user clicks away.
I don't see why web clients being transient is a problem. The whole point of the MapReduce algorithm is that each worker (the web clients in this case) don't need to know anything about what the other worker is doing, what the system as a whole is doing, nor what it had done with any past job.
You don't need to have every band member play to a metronome to sync them up, even if they're recorded individually and with isolation. If they aren't all playing at the same time, as long as the rhythm track is done first, everyone plays to that.
In a way, what you call "the rhythm track" is essentially what this "click track" concept refers to. In some (many?) cases, the track is just two different percussive sounds played at regular intervals (essentially mimicking a metronome). In other cases, it might be a synth or MIDI rendering of the song. In yet others, it might be a "single-track" version of the song, and the musicians are replaying along with it just for the purpose of having a multitrack recording."
I have no idea what you were talking about when you said 'you can't just tell everyone to play 120bpm and then have them sync up.
I'm referring to the scenario of you telling someone to play at "120bpm", without providing them with any tempo-aid at all. One band member might end up playing at 119bpm instead, and another at 121bpm, and thus they won't be synchronized.
Sorry, I really don't understand your post. When you say "You don't", what is it that "I don't do" that you are referring to? And how are the existence of headphones relevant?
Even if a drum track sounds the same for say one section of a song there will be small variations in rhythm and volume over time that to human ears just sounds more natural.
It would also be not much fun for the drummer as he would just be laying down loops instead of an entire track.
As a drummer, the presence or absence of a click track doesn't really affect the "fun". If the song has a repetitive drum pattern played over and over throughout, the song is not fun to play; and this is true regardless of whether I have to get the timing perfectly right with a click track, or if I just play the same old pattern loosely. What makes a song fun (for me) is having lots of interesting variations throughout the song.
Actually one of the more fun things for me to do as a drummer is to use a click track for some odd time signature I haven't really practiced or internalized yet (say 11/8), and try to improvise over the click. The click itself (on most metronomes the "1" will sound different from all the other beats) help me know what the "feel" of the time signature is supposed to be, and my challenge is to see how quickly I can adapt and play around with that feel.
The only time I use a metronome is when I'm drumming on my own, to improve my own performance, but I could never play a whole song with the invariable click in my ears (actually that's not true, there is one song where I used a metronome, but only because I had a continuous snare drum pattern and was (am?) just not good enough to keep the rythm).
I've toyed with the idea of creating a "whoosh" track, a less fucking annoying and more forgiving version of a click-track, to help my own occasionally shaky rhythm-keeping. Dig, with something like a white-noise swell whooshing in time instead of the click. I truly don't mind playing with headphones on, even if usually just to give me a mix of everyone else's monitors that can compete with my own volume, but, man, that unforgiving click is worse than having Roger Waters glaring at you balefully.
One of my most uncomfortable drumming moments was during a live show where I was using a click track which was just a tad too quiet. Everything was going perfectly fine until I hit a mini-drum-solo/fill a bit too quickly, and when I came out of it, I so I was off the rest of the song. I had to really focus to play everything with a 3/16th delay after when I actually heard the click in my headphones.
If the metronome is "click", the snare is "tack", and the two simultaneous is "clack", then the song, in my mind, went from "clack, clack, clack, clack" to "cltack, cltack, cltack, cltack". Kind of gave it a nice jazz-swing feel, but I wouldn't want to go through that again.
Much music has always been explicitly shit, regardless of when it was made. Go back and take a look at the charts for any year you care to name, and probably 95% of the artists will be people you've never heard of...because they were shit, had their fifteen minutes, and are now long-forgotten.
Not only that, but I bet 95% of the artists never made it to the top 5%.
I think in some ways, the modern recording tools, have helped kill good music in many ways, it can really mask the lack of talent in todays musicians.
I guess it depends on whether you care more about the music or the musician. If all you care about is "good music", then who cares whether it was even a human who played the music, or if it was all computer generated with synthesizers?
Maybe there's an artist who can't play any instrument at all, but is a great composer, and so all his songs are "played" by software. No human has ever played the song, but the music sounds great.
Maybe there's a mathematician who can't play instruments and can't compose, but figured out an algorithm which generates beautiful melodies. No human has ever composed the song or played the song, but the music songs great.
Maybe there's a programmer who can't play instruments, can't compose, and can't figure out complex equations, but wrote a genetic algorithm that figured out the equations to compose the songs. No human ever conceived of the math, composed the song, or played the song, but the music sounds great.
etc.
However to leverage much of the flexibility and power of a digital recording you need a click.
Really? Why?
It has more to do with multitrack recording than digital recording. When you do a multitrack recording, each track is recorded individually. Usually this means that each band member will be recorded in isolation, and it's very rare for every band member to have such a developped internal metronome such that you can just tell them "play at 120 beats per minute", and when you combine all the tracks, they'll all be synced up.
Instead, what is done is that you set a metronome to play at 120bpm, and each bandmember will play along to the metronome.
Actually, the pattern is that it alternates. So the next good console will be easy to program (i.e. XBox360), and the one after that will be hard to program (i.e. PS4).