Terrific insight! However, you missed the part in the article about the difficulty of "resting on your laurels" in such a system. In the current system, Michael Jackson can behave like a total freak and produce lousy music, and folks can't vote with their feet, because there's nowhere to go. In the post-big-music-industry world (as was true in the pre-big-music-industry world), there will still be superstars, but there's more hope that the superstars will be folks with actual talent and respect for the public. No guarantee, but more hope.
Re:What is an example that can't run in parallel?
on
Forget Moore's Law?
·
· Score: 1
Am I missing anything important here?
Yes. Remember that alpha-beta prunes exponentially. With optimal move ordering, it will reduce the number of nodes for a search to depth d from b**d to b**(d/2). Consider a game like chess, where the average branching factor is around 10 or so. Thus, to search to a depth of 12 without alpha-beta requires something like 1e12 nodes. To search to a depth of 12 with alpha-beta requires examining 1e6 nodes, as in your example. Note that if you miss even one good early prune, you could easily search an extra 1e10 or more nodes: that means that all of your 1e4 processors would be doing wasted work the whole time, and you would take just as long as your sequential opponent! The situation gets worse as the search gets deeper you need an number of processors exponential in the search depth to keep up.
In short, while it should be better to examine more boards than one during a given timestep, this is only true if you can find more than one board worth examining during this timestep. Standard minimax alpha-beta is inherently sequential: study the algorithm carefully, and you will be convinced of this.
For more information on parallel game tree search, hit the U. Alberta Games Group web site.
Re:What is an example that can't run in parallel?
on
Forget Moore's Law?
·
· Score: 2, Informative
Can someone please give an example of a computing task that CANNOT be subdivided into smaller tasks and run in parallel on many processing elements?
The technical issue here is known as "linear speedup". Take chess, for example: the standard search algorithm for chess play is something called minimax search with alpha-beta pruning. It turns out that the alpha-beta pruning step effectively involves information from the entire search up to this point. With only a subset of this information, exponentially more work will be needed: a bad thing.
How do parallel chess computers such as Deep Blue work, then? Very fancy algorithms that still get sublinear but interesting speedups at the expense of a ton of clever programming. This is a rough explanation of why today's PC chess programs are probably comparable with the now-disassembled Deep Blue: the PC chess programmers can use much simpler search algorithms, concentrating on other important programming tasks, also a 10x speedup in uniprocessor performance has a 10x search speed increase, whereas using 10x slow processors isn't nearly so effective. Note that Deep Blue was decommissioned largely because of maintenance costs: a lot of rework would have to be done to make Deep Blue take advantage of Moore's Law.
That said, many tasks are "trivially" parallelizable. Aside from the pragmatic problem of coding for parallel machines (harder than writing serial code even for simple algorithms), there is also the silicon issue: given a transistor budget, are manufacturers better spending it on a bunch of little processors or one big one? This is the real question, and so far the answer is generally "one big one". YMMV. HTH.
(BTW, why can't I use HTML entities for Greek alpha and beta in my Slashdot article? What are they protecting me from?)
Most glyph / font file management code is hopelessly unoptimized for Far East fonts: designers tend to make simplifying assumptions about character set size, font overlap, and glyph rendering that are expensive to correct later.
The latest fontconfig/Xft code has been carefully designed to do a really nice job with Far East glyphs and fonts, as well as most other languages on the face of the planet. Keith Packard is to be congratulated, as usual, for his leading-edge work on this. The problem with Qpe being closed-source is that there is no reasonable alternative for integrating this work except to be patient and hope that Trolltech does it: then hope there is an upgrade path.
In the L.A. area, check out the free TrafficDodger, which has been around for some time now. I used to go to school/work where it was developed: it's got some nice AI search algorithms.
The way I understand it, as with copyright, a trademark need not be registered to be protected. Registering a trademark for the name Locutus in connection with action figures is likely to establish Paramount's right to the name in other contexts. With an obviously invented name like "Locutus", the onus may be on the defendant in an infringement suit. Keep in mind that the Lanham Act is quite broad: the "dilution" argument may provide a basis for a successful suit in this situation.
Besides, remember the golden rule of lawsuits: if you are Ian, and you are sued by Paramount, you lose, because Paramount can afford to fight the suit forever out of pocket change and run you bankrupt before justice is done. Much better to pick a name that is either clearly conventional or clearly unique, to avoid trouble from the beginning.
I hope Ian has gotten permission to use the name Locutus, which is, no surprise, a trademark of Paramount. Info below.
(BTW, why does Slashdot not allow <pre> tags but allow text-only postings and the obvious <tt>...<br> thing? What a pain.)
Word Mark LOCUTUS
Goods and Services IC 028. US 022.
G & S: toys; namely, action figures and accessories therefor, poseable figures, dolls. FIRST USE: 19930600.
FIRST USE IN COMMERCE: 19930600
Mark Drawing Code (1) TYPED DRAWING
Serial Number 74462053
Filing Date November 12, 1993
Published for Opposition August 23, 1994
Registration Number 1862622
Registration Date November 15, 1994
Owner (REGISTRANT) PARAMOUNT PICTURES CORPORATION
CORPORATION DELAWARE
5555 Melrose Avenue Los Angeles CALIFORNIA 900383197
Type of Mark TRADEMARK
Register PRINCIPAL
Affidavit Text SECT 15. SECT 8 (6-YR).
Live/Dead Indicator LIVE
Re:That�s what tripwire-integrit is for!
on
Kevin Mitnick Answers
·
· Score: 4, Insightful
Please don't take these comments personally: they are directed at everyone here with similar attitudes.
First, if you weren't managing a reasonable-sized installation in 1989, you are not qualified to comment on best practice: you have no idea what it was like to do so. (I was, and do.)
Tripwire?? Tripwire barely existed yet, and no one used it. Secret-key encryption was more-or-less illegal because of governmental administrative rules. Public-key encryption was the legal property of RSA Inc., and as a result no reasonable tools were built and distributed that used it. There was no DSA. So all the 'leet security tools everyone seems astounded jg wasn't using were not even close to options. No SSH, obviously. Not even encrypted telnet.
Backups? Backups were on the fancy new 1GB Exabyte tape drives, if you were lucky. The less-lucky were still using 9-track tape. Restore times were literally overnight: disks were slow, and much of the restore process was (get this) CPU-limited. The number of operators skilled enough to perform restoration from backups was small: that job paid better at the time than web-monkeys got at the height of the.com era.
And remember: more or less the whole organization (DEC, for Pete's sake, one of the largest computer companies in the world at the time) was down while the restore completed, due to the joys of centralized computing.
Do I think that Mitnick "got what he deserved"? No: he was persecuted and prosecuted beyond the extent his legal and ethical crimes deserved. Do I think he caused a lot of very clueful people lots of legitimate grief for no good reason? You bet. Even he seems to think that. Hopefully you can understand the compatibility of these two viewpoints, and not just blame the victims for Markoff's and the government's sins.
I'm probably just being dense, but I'm not seeing how the Worst Abuse of the Rules winner this year actually abuses the rules in any way. Just because it's a compiler to x86 native code? If that's all, it doesn't seem very abusive. The compiler itself appears to be perfectly good (well okay, maybe that's an exaggeration) portable C code, and to not depend on x86 code for its correct operation.
It's definitely a way cool program, but I don't get the fit to the category. Suggestions?
For those who aren't aware, Carl Worth and Keith Packard's Xr library (the link is pretty stale, sorry) works with the X Render extension to provide a really nice 2D rendering model. They've done some cutting-edge work here, including high-quality anti-aliased compositing, high-speed rendering, high-quality splines, and a Postscript-like C API. Postscript/PDF and SVG support is an explicit goal of Xr: xpdf has already been modified to render with Xr (and Xft), and it looks great!
Watch for an upcoming announcement soon, when the project is sufficiently complete.
Doesn't anyone in the news media and the general public understand that the label "Czar" is not a compliment? The original Czars were ruthless tyrants whose treatment of the average Russian was so bad that it made Communism look attractive. While this may be an accurate assessment of the role played by our current crop of "Czars", the concept of having more of them seems like it ought to be self-criticizing to me.
Then again, if the U.S. had an Education Czar, maybe more Americans would know some history...
...the USB key stuck in the back looks like a likely point of failure.
Conceivably. Anyone who is running one of these drives without backups somewhere is even more insane than the folks running un-encrypted drives without backups. The backups themselves can easily be encrypted, so there's no need for major security risk. If your key dongles stop working or your drive fries, you'd better have some way of getting the bits back from outside, 'cause they're not coming from the platter.
OTOH, what is "64-bit/ 40-bit DES" supposed to be? Presumably this means the drive supports "40-bit watered-down DES keys" and "64-bit normal DES blocks". So I guess I'm wrong: this drive is designed to be break-inable in an emergency. Great. I'll wait until they offer 3DES or AES-128 options, thanks.
In the meantime, check out the BSD Cryptographic disk driver cgd: SW on-disk encryption at the block level.
I left NWLink DSL a couple of weeks ago over their mediocre uptime and high costs, especially bandwidth costs.
They were down quite a few hours per month lately (that I noticed) and their tech support was not so good. They seemed to have a lot of router configuration problems: there were frequent router loops. This may have been partly due to their prime (only?) feed to most of the world being alter.net, which at least in this neck of the woods is quite slow and tangled. NWlink claimed at one point that they were just finishing up some big network reconfiguration, and things should get better: when things didn't seem to, that was the end for me.
I should have waited to switch: it would be nice to dump them now in protest over their DDOSing of SDF:-).
I chose NWLink several years ago because at the time their prices were good, and my previous provider, NW Nexus, had been bought out twice and become expensive and quite unreliable in the process. Now I'm with DSL Only, and so far they seem great. ISP service is $18/month for 640/256 DSL, with no bandwidth charges, no restrictions on use, a static IP, and two mailboxes. It's been reliable so far, and they are direct to a local exchange that in turn is direct to my work and to Internet2, so ping times and bandwidth are excellent.
It would be nice to be done with ISP switches forever: maybe this is it. The other good news is that this was the easiest ISP switch so far: Qwest seems to have it figured out now, and it was completely routine.
The Liquid War folks claim their game is "truly original". Isn't it pretty much directly taken from the much-older XBattle, but with the units scaled down? Or am I missing something?
All it takes is for one false positive on a Really Important Email and be accidentally deleted to totally destroy the value of any filtering system.
Huh? You drive around in cars all the time, in spite of the fact that if that system fails (which it not infrequently does) in the wrong way and at the wrong time...you die.
Technology occasionally fails. The only way to avoid technological failure is to avoid technology. (You'll still have failures: they just won't be technological ones.)
If you anticipate receiving a communication so Really Important that the consequences of accidentally spam-filtering it are catastrophic, you shouldn't be using e-mail anyhow. I would guess that my personal spam filtering has about the same average false-positive rate as the rate of drops of mail by my software, hardware, and upstream mail providers. At least with the spam-filtered messages, I can save them around and do post-mortem on them.
"Dear openoffice.org, because bits are infinitely replicable at zero cost, please send my copy of Open Office, free of charge, to this IP address. Sincerely, po8."
Worked for me! Might want to try it yourself some time...it's pretty cool. Very "real world", in a manner of speaking.
I wish you happiness, my friend, but I'm afraid I have real-world responsibilities and can't play on this thread any longer. Take care!
Data has one very important property that it shares with physical systems and all other things: scarcity.
Surely scarcity is not an inherent property of bits? After all, bits are infinitely replicable at zero cost!
Legally you can't get bits from just anywhere. This is what economists call an "artificial scarcity". There is no physical or information-theoretic law that keeps bits scarce. Instead, bits are scarce (in the Western world) because people with guns enforce that scarcity. The people with guns "should" go away. This would make me and (to judge by the number of people that are willing to risk the wrath of people with guns to work around the artificial scarcity) many of those I share the planet with happier.
Do we buy and sell books based not on the value of the ink and paper but the contents? Yes, we do. So are books "stuff?" Yes, they are.
The fact that you can have a financial transaction around bits doesn't make them stuff. After all, you can have financial transactions about such abstract concepts as real estate, honor, or liability. IMHO, to call liability "stuff" pushes the boundaries of the word beyond any useful meaning.
It's not clear on what basis a traditional ink-and-paper book is priced. Can't be entirely the bits; some publishers make good money selling books they also make freely available online. I would suggest that books, like music CDs, mainly represent a payment for services: I am willing to give the store owner $7 for a paperback in return for such considerations as providing me with a convenient venue for access, a convenient carrier for the bits, a reasonable assurance of some minimum quality control and evaluation (e.g. none of the pages are missing, the language is reasonably-edited, etc.), a convenient way to donate some money to the author, and other such services. Change any of these factors substantially, and the book price I will tolerate changes too. Change the actual content, the bits, and my price tolerance is unlikely to change: when I buy the book, I typically haven't even accessed most of the bits yet.
Data has a very different set of properties than the physical systems used to store and transfer it. This difference represents an opportunity and a challenge. The legal and political world would like to simplify their lives by ignoring the difference: I am not happy about this. I want different rules and different laws for bits and stuff.
Are you free to do whatever you want with other people's stuff?
Stuff? What stuff? There's no stuff here. Hand me a megabyte of music. No, no, not the floppy disk, or USB device. If there's a megabyte of music in there, I sure can't see it. The megabyte itself. OK, maybe that megabyte is like atoms or something: I can't see it because it's too small. I'll take it out of the disk and put it on my disk. Now I've got the megabyte of music. But wait, you still have it too! Maybe there were two of them in there, so I'll take out the other one. Nope, now I have two megabytes of music and you still have yours. I know, let's weigh the disk before and after, using the most precise conceivable balance scale. Nope, no matter how many times I take a megabyte of music out of the disk, it still weighs exactly the same. This sure isn't like any stuff I've ever played with before.
Bits are something or other, but they aren't stuff. Get used to it.
Slashdot Mirror
Apparently, this is Elwood P. Dowd from after the injection.
Terrific insight! However, you missed the part in the article about the difficulty of "resting on your laurels" in such a system. In the current system, Michael Jackson can behave like a total freak and produce lousy music, and folks can't vote with their feet, because there's nowhere to go. In the post-big-music-industry world (as was true in the pre-big-music-industry world), there will still be superstars, but there's more hope that the superstars will be folks with actual talent and respect for the public. No guarantee, but more hope.
Am I missing anything important here?
Yes. Remember that alpha-beta prunes exponentially. With optimal move ordering, it will reduce the number of nodes for a search to depth d from b**d to b**(d/2). Consider a game like chess, where the average branching factor is around 10 or so. Thus, to search to a depth of 12 without alpha-beta requires something like 1e12 nodes. To search to a depth of 12 with alpha-beta requires examining 1e6 nodes, as in your example. Note that if you miss even one good early prune, you could easily search an extra 1e10 or more nodes: that means that all of your 1e4 processors would be doing wasted work the whole time, and you would take just as long as your sequential opponent! The situation gets worse as the search gets deeper you need an number of processors exponential in the search depth to keep up.
In short, while it should be better to examine more boards than one during a given timestep, this is only true if you can find more than one board worth examining during this timestep. Standard minimax alpha-beta is inherently sequential: study the algorithm carefully, and you will be convinced of this.
For more information on parallel game tree search, hit the U. Alberta Games Group web site.
Can someone please give an example of a computing task that CANNOT be subdivided into smaller tasks and run in parallel on many processing elements?
The technical issue here is known as "linear speedup". Take chess, for example: the standard search algorithm for chess play is something called minimax search with alpha-beta pruning. It turns out that the alpha-beta pruning step effectively involves information from the entire search up to this point. With only a subset of this information, exponentially more work will be needed: a bad thing.
How do parallel chess computers such as Deep Blue work, then? Very fancy algorithms that still get sublinear but interesting speedups at the expense of a ton of clever programming. This is a rough explanation of why today's PC chess programs are probably comparable with the now-disassembled Deep Blue: the PC chess programmers can use much simpler search algorithms, concentrating on other important programming tasks, also a 10x speedup in uniprocessor performance has a 10x search speed increase, whereas using 10x slow processors isn't nearly so effective. Note that Deep Blue was decommissioned largely because of maintenance costs: a lot of rework would have to be done to make Deep Blue take advantage of Moore's Law.
That said, many tasks are "trivially" parallelizable. Aside from the pragmatic problem of coding for parallel machines (harder than writing serial code even for simple algorithms), there is also the silicon issue: given a transistor budget, are manufacturers better spending it on a bunch of little processors or one big one? This is the real question, and so far the answer is generally "one big one". YMMV. HTH.
(BTW, why can't I use HTML entities for Greek alpha and beta in my Slashdot article? What are they protecting me from?)
Most glyph / font file management code is hopelessly unoptimized for Far East fonts: designers tend to make simplifying assumptions about character set size, font overlap, and glyph rendering that are expensive to correct later.
The latest fontconfig/Xft code has been carefully designed to do a really nice job with Far East glyphs and fonts, as well as most other languages on the face of the planet. Keith Packard is to be congratulated, as usual, for his leading-edge work on this. The problem with Qpe being closed-source is that there is no reasonable alternative for integrating this work except to be patient and hope that Trolltech does it: then hope there is an upgrade path.
Thanks!
Oops! Thanks.
In the L.A. area, check out the free TrafficDodger, which has been around for some time now. I used to go to school/work where it was developed: it's got some nice AI search algorithms.
The way I understand it, as with copyright, a trademark need not be registered to be protected. Registering a trademark for the name Locutus in connection with action figures is likely to establish Paramount's right to the name in other contexts. With an obviously invented name like "Locutus", the onus may be on the defendant in an infringement suit. Keep in mind that the Lanham Act is quite broad: the "dilution" argument may provide a basis for a successful suit in this situation.
Besides, remember the golden rule of lawsuits: if you are Ian, and you are sued by Paramount, you lose, because Paramount can afford to fight the suit forever out of pocket change and run you bankrupt before justice is done. Much better to pick a name that is either clearly conventional or clearly unique, to avoid trouble from the beginning.
I hope Ian has gotten permission to use the name Locutus, which is, no surprise, a trademark of Paramount. Info below.
(BTW, why does Slashdot not allow <pre> tags but allow text-only postings and the obvious <tt>...<br> thing? What a pain.)
Word Mark LOCUTUS
Goods and Services IC 028. US 022.
G & S: toys; namely, action figures and accessories therefor, poseable figures, dolls.
FIRST USE: 19930600.
FIRST USE IN COMMERCE: 19930600
Mark Drawing Code (1) TYPED DRAWING
Serial Number 74462053
Filing Date November 12, 1993
Published for Opposition August 23, 1994
Registration Number 1862622
Registration Date November 15, 1994
Owner (REGISTRANT) PARAMOUNT PICTURES CORPORATION
CORPORATION DELAWARE
5555 Melrose Avenue Los Angeles CALIFORNIA 900383197
Type of Mark TRADEMARK
Register PRINCIPAL
Affidavit Text SECT 15. SECT 8 (6-YR).
Live/Dead Indicator LIVE
Please don't take these comments personally: they are directed at everyone here with similar attitudes.
First, if you weren't managing a reasonable-sized installation in 1989, you are not qualified to comment on best practice: you have no idea what it was like to do so. (I was, and do.)
Tripwire?? Tripwire barely existed yet, and no one used it. Secret-key encryption was more-or-less illegal because of governmental administrative rules. Public-key encryption was the legal property of RSA Inc., and as a result no reasonable tools were built and distributed that used it. There was no DSA. So all the 'leet security tools everyone seems astounded jg wasn't using were not even close to options. No SSH, obviously. Not even encrypted telnet.
Backups? Backups were on the fancy new 1GB Exabyte tape drives, if you were lucky. The less-lucky were still using 9-track tape. Restore times were literally overnight: disks were slow, and much of the restore process was (get this) CPU-limited. The number of operators skilled enough to perform restoration from backups was small: that job paid better at the time than web-monkeys got at the height of the .com era.
And remember: more or less the whole organization (DEC, for Pete's sake, one of the largest computer companies in the world at the time) was down while the restore completed, due to the joys of centralized computing.
Do I think that Mitnick "got what he deserved"? No: he was persecuted and prosecuted beyond the extent his legal and ethical crimes deserved. Do I think he caused a lot of very clueful people lots of legitimate grief for no good reason? You bet. Even he seems to think that. Hopefully you can understand the compatibility of these two viewpoints, and not just blame the victims for Markoff's and the government's sins.
I'm probably just being dense, but I'm not seeing how the Worst Abuse of the Rules winner this year actually abuses the rules in any way. Just because it's a compiler to x86 native code? If that's all, it doesn't seem very abusive. The compiler itself appears to be perfectly good (well okay, maybe that's an exaggeration) portable C code, and to not depend on x86 code for its correct operation.
It's definitely a way cool program, but I don't get the fit to the category. Suggestions?
For those who aren't aware, Carl Worth and Keith Packard's Xr library (the link is pretty stale, sorry) works with the X Render extension to provide a really nice 2D rendering model. They've done some cutting-edge work here, including high-quality anti-aliased compositing, high-speed rendering, high-quality splines, and a Postscript-like C API. Postscript/PDF and SVG support is an explicit goal of Xr: xpdf has already been modified to render with Xr (and Xft), and it looks great!
Watch for an upcoming announcement soon, when the project is sufficiently complete.
Doesn't anyone in the news media and the general public understand that the label "Czar" is not a compliment? The original Czars were ruthless tyrants whose treatment of the average Russian was so bad that it made Communism look attractive. While this may be an accurate assessment of the role played by our current crop of "Czars", the concept of having more of them seems like it ought to be self-criticizing to me.
Then again, if the U.S. had an Education Czar, maybe more Americans would know some history...
In the words of immigrants in a Firesign Theatre piece, "We were running away from poverty, injustice, the law, and the army."
I love those guys.
Conceivably. Anyone who is running one of these drives without backups somewhere is even more insane than the folks running un-encrypted drives without backups. The backups themselves can easily be encrypted, so there's no need for major security risk. If your key dongles stop working or your drive fries, you'd better have some way of getting the bits back from outside, 'cause they're not coming from the platter.
OTOH, what is "64-bit/ 40-bit DES" supposed to be? Presumably this means the drive supports "40-bit watered-down DES keys" and "64-bit normal DES blocks". So I guess I'm wrong: this drive is designed to be break-inable in an emergency. Great. I'll wait until they offer 3DES or AES-128 options, thanks.
In the meantime, check out the BSD Cryptographic disk driver cgd: SW on-disk encryption at the block level.
I left NWLink DSL a couple of weeks ago over their mediocre uptime and high costs, especially bandwidth costs. They were down quite a few hours per month lately (that I noticed) and their tech support was not so good. They seemed to have a lot of router configuration problems: there were frequent router loops. This may have been partly due to their prime (only?) feed to most of the world being alter.net, which at least in this neck of the woods is quite slow and tangled. NWlink claimed at one point that they were just finishing up some big network reconfiguration, and things should get better: when things didn't seem to, that was the end for me. I should have waited to switch: it would be nice to dump them now in protest over their DDOSing of SDF :-).
I chose NWLink several years ago because at the time their prices were good, and my previous provider, NW Nexus, had been bought out twice and become expensive and quite unreliable in the process. Now I'm with DSL Only, and so far they seem great. ISP service is $18/month for 640/256 DSL, with no bandwidth charges, no restrictions on use, a static IP, and two mailboxes. It's been reliable so far, and they are direct to a local exchange that in turn is direct to my work and to Internet2, so ping times and bandwidth are excellent.
It would be nice to be done with ISP switches forever: maybe this is it. The other good news is that this was the easiest ISP switch so far: Qwest seems to have it figured out now, and it was completely routine.
FYI. As always, YMMV.
The Liquid War folks claim their game is "truly original". Isn't it pretty much directly taken from the much-older XBattle, but with the units scaled down? Or am I missing something?
All it takes is for one false positive on a Really Important Email and be accidentally deleted to totally destroy the value of any filtering system.
Huh? You drive around in cars all the time, in spite of the fact that if that system fails (which it not infrequently does) in the wrong way and at the wrong time...you die.
Technology occasionally fails. The only way to avoid technological failure is to avoid technology. (You'll still have failures: they just won't be technological ones.)
If you anticipate receiving a communication so Really Important that the consequences of accidentally spam-filtering it are catastrophic, you shouldn't be using e-mail anyhow. I would guess that my personal spam filtering has about the same average false-positive rate as the rate of drops of mail by my software, hardware, and upstream mail providers. At least with the spam-filtered messages, I can save them around and do post-mortem on them.
Deep Blue was dismantled in 2001, and part of it donated to the Smithsonian in 2002.
Deep Blue was dismantled several years ago.
"Dear openoffice.org, because bits are infinitely replicable at zero cost, please send my copy of Open Office, free of charge, to this IP address. Sincerely, po8." Worked for me! Might want to try it yourself some time...it's pretty cool. Very "real world", in a manner of speaking. I wish you happiness, my friend, but I'm afraid I have real-world responsibilities and can't play on this thread any longer. Take care!
Data has one very important property that it shares with physical systems and all other things: scarcity.
Surely scarcity is not an inherent property of bits? After all, bits are infinitely replicable at zero cost!
Legally you can't get bits from just anywhere. This is what economists call an "artificial scarcity". There is no physical or information-theoretic law that keeps bits scarce. Instead, bits are scarce (in the Western world) because people with guns enforce that scarcity. The people with guns "should" go away. This would make me and (to judge by the number of people that are willing to risk the wrath of people with guns to work around the artificial scarcity) many of those I share the planet with happier.
Do we buy and sell books based not on the value of the ink and paper but the contents? Yes, we do. So are books "stuff?" Yes, they are.
The fact that you can have a financial transaction around bits doesn't make them stuff. After all, you can have financial transactions about such abstract concepts as real estate, honor, or liability. IMHO, to call liability "stuff" pushes the boundaries of the word beyond any useful meaning.
It's not clear on what basis a traditional ink-and-paper book is priced. Can't be entirely the bits; some publishers make good money selling books they also make freely available online. I would suggest that books, like music CDs, mainly represent a payment for services: I am willing to give the store owner $7 for a paperback in return for such considerations as providing me with a convenient venue for access, a convenient carrier for the bits, a reasonable assurance of some minimum quality control and evaluation (e.g. none of the pages are missing, the language is reasonably-edited, etc.), a convenient way to donate some money to the author, and other such services. Change any of these factors substantially, and the book price I will tolerate changes too. Change the actual content, the bits, and my price tolerance is unlikely to change: when I buy the book, I typically haven't even accessed most of the bits yet.
Data has a very different set of properties than the physical systems used to store and transfer it. This difference represents an opportunity and a challenge. The legal and political world would like to simplify their lives by ignoring the difference: I am not happy about this. I want different rules and different laws for bits and stuff.
Are you free to do whatever you want with other people's stuff?
Stuff? What stuff? There's no stuff here. Hand me a megabyte of music. No, no, not the floppy disk, or USB device. If there's a megabyte of music in there, I sure can't see it. The megabyte itself. OK, maybe that megabyte is like atoms or something: I can't see it because it's too small. I'll take it out of the disk and put it on my disk. Now I've got the megabyte of music. But wait, you still have it too! Maybe there were two of them in there, so I'll take out the other one. Nope, now I have two megabytes of music and you still have yours. I know, let's weigh the disk before and after, using the most precise conceivable balance scale. Nope, no matter how many times I take a megabyte of music out of the disk, it still weighs exactly the same. This sure isn't like any stuff I've ever played with before.
Bits are something or other, but they aren't stuff. Get used to it.