Slashdot Mirror

Mips sued Lexra. Lexra's chips didn't implement the patented instructions in the Mips instruction set but Mips argued that since the unimplemented patented instructions faulted, like other invalid instructions, it was theoretically possible for a user of the chip to emulate them in software and thus violate the patent. Lexra went bust without that argument being decided. The Intel Transmeta case was settled. So it's not clear what would happen if Intel sued either Microsoft or a hardware vendor over an emulator which allowed patented instructions to run on ARM

http://probell.com/lexra/

And before you talk about that being the same as banning compilers and assemblers it's not the same thing. An assembler or compiler that generates NEON instructions from assembler or C is not the same legally as a JIT compiler which generates them from x86 binaries in order to run them on ARM.

As someone once put it "Bits have color". The "color" in this case is intent - even though assemblers, compilers and JITs generate NEON instructions only the JIT does it to violate a patent. Intent matters legally, even if it doesn't in computer science and engineering.

http://ansuz.sooke.bc.ca/entry...

XCode could care LESS

It's not about what the Xcode software technically implements. It's about what a BSA audit could uncover. True, the bits of a licensed copy of Xcode downloaded from Apple are exactly the same as the bits of an infringing copy obtained through sneakernet. But an audit would uncover that the bits are a different color, and Apple has the right to sue over the use of incorrectly colored bits. Bit color is a legal construct, not a technical construct. This might be discovered if someone in your office is discovered to be using Xcode on a computer for which no successful Xcode installation is recorded in Mac App Store.

Just because something isn't Open Sores, doesn't mean that it's DRMed.

Nor does just because something isn't subject to technical DRM mean it's legally free to redistribute to all comers.

I still control whether I give you the original or a copy of it

Well, I suppose that it's possible that you might unplug your hard drive, put it in a cardboard box, and mail it to me, in response to a download request, but that's surely too unusual to care about.

Because the law defines making copies as a form of infringement, defines copies as material objects, and because we lack the ability to send a material object through the net, you cannot transmit an original copy of a work to me online. All you can do is give me the information I need to create a new copy on my end.

Very few times will you ever have the ability to determine if the file on my server or computer is copied or deleted

It's irrelevant whether you delete the file once I've downloaded it. The Copyright Act doesn't treat a copy followed by a deletion as not being copying. It doesn't matter in the least how many copies actually exist in the end, only what the provenance of the copies is. There is an essay called 'What colour are your bits?' which you may find helpful.

it is transferred to your system

It is not, in any legally meaningful way, transferred anywhere.

Please take a look at this page, which discusses the outcome of the ReDigi case, and includes a copy of the opinion. ReDigi tried to sell used music files, going through the sort of copy and delete rigamarole as you suggest. They got shut down hard because it's utter nonsense as far as the legal system is concerned.

This is oddly close to what I think DRM ought to be: advisory, not enforcing. Remove the accountability aspect, not least because it's a farce that leaves the most recent honest party holding the bag, and you have my concept of an ideal DRM engine: provenance meta-tags that let you know what color your bits are, which you can use if it affects you or ignore if it doesn't, leaving no rights-holder the wiser no matter what course you take.

Accountability-oriented DRM, which prevents no action but forces your use of certain combinations of certain colors of bits into public record, would be prone to false positives. Pulled in some GPL code to a local build of 7-Zip? Chances are the other code doesn't have a GPL exception to allow linking against the non-Open unrar, so the resulting software likely may not be conveyed (in the GPLv3 sense) at all, but creating or using the resulting combined work won't infringe anyone's rights and shouldn't require you posting public notice that you have created such a combined work if you have no intent to convey it.

What a strange thing! I guess I am allowed to time-shift broadcast TV, and I am allowed to space-shift broadcast TV. I can rent an antenna, and I can rent a VCR (PVR).

I cannot retransmit (time or space shifted or not) a broadcast to other parties (which is the difference here CATV rebroadcast to all CATV clients).

Now I have to read the arguments! About the only thing left is having an agent do the time or space shifting for me! And, of course, I can't really figure out is why the AGENT is in court for this. If my neighbour asks me to rent her some roofspace and rent her an antenna AND a VCR and then asks me to record a TV show... for which I may charge a bit for the service. And the TV network comes after someone, why would that be me? I would be inclined to laugh.

I think my lawyer would have a good laugh too. We refer you to the reply given in the case of Arkell v. Pressdram.

I guess I am not allowed to sell my labour freely in the USA. Now I REALLY have to follow this. I am personally guilty of renting antennas, and PVR (equivalent) to provide people with recordings. I never pressed a "record" button -- my customer went on-line to a web page and selected the recording themselves (using MythTV 10 years ago). I would deliver the recorded program(s) via disk drives or flash drives.

After all, if I have multiple tuners and I am not using them all, why CAN'T I RENT THEM OUT.

The only problem would have been an event like the "Superbowl" where I would have needed to have ALL my tuners capturing the same content. Instead of being efficient, you know, and sharing... Because WHERE the bits come from is important in Copyright law. See http://ansuz.sooke.bc.ca/entry...

As long as Aereo uses an antenna and receiver PER USER, the bits should be the right colour. And subject to the users rights. Including time and space shifting. Aereo wouldn't be rebroadcasting. IF Aereo IS IN THE WRONG then the question is why. As far as I can tell, they are not even being an agent for the user. They are simply renting an antenna and receiver. The actual Copyright material is NOT being shared, from Aereo's perspective. And yes, cloud storage would be at risk. For example, I quite enjoy using Kobo. I may purchase a book from Kobo WHICH IS Copyrighted. Of course. I then download to my reading device. The bits have the right colour at Kobo's end, and they have the right colour at my end. I should be able to do with those bits ANYTHING that Copyright law permits me to. And I do. There is no DRM in OTA broadcast, and typically there is DRM in Kobo electronic books. If *I* turn around and share the book, Kobo wouldn't be legally liable. The author would come after me for that. So why is Aereo being attacked here?

If the bits are simply coloured "copyrighted" and it IS authorized to the user, what else should Aereo do? Simply, Kobo is selling access to authorized bits as well, and would be AT THE SAME RISK. And, it goes deeper. Since Copyright is automatically assigned on creation, you would have NO IDEA what is ok to look at, here or touch.

Colour me completely confused.

Wow, that idea is about as old as dirt, and has been thrashed out long ago....
A great post from 2004 talks about this and other ideas of Copyright in a digital age.

What Colour are your bits?
http://ansuz.sooke.bc.ca/entry/23

Go read "What Colour are your bits?".

You are not ripped off as no information you create belongs to you. You are simply its creator, but not its owner. You typically are also the holder of a copy. But that copy is in no way different than any other copy.

I am not stepping into any minefield here, I am just pointing out facts. They are pretty obvious unless you are brainwashed. If you really have trouble to understand the nature of things, I recommend reading "What Color Are Your Bits": http://ansuz.sooke.bc.ca/entry/23

Sure, the content-industry would like things to be different. They try very hard to change reality, but ultimately they will fail. Of course, if you have this fantasy that a specific piece of information somehow can "belong" to you, you will conclude that you are being ripped off. But if you redefine your perception of reality arbitrarily, that is entirely your fault. While a copy of the information can belong to you, the information cannot. And it did exist long before you "created" the work by crating the first copy of it, as all information is eternal. Just representing it is a different matter.

However, does the data 1101 hold up to a copyright claim in court?

Of course it does. If it is a part of a copyrighted file, it absolutely holds up. See What colour are your bits?.

You can try to get away with calling it fair use. Good luck.

(Yes, it's absolutely ridiculous that bits have colour. I am not saying that the law is moral or even sane.)

The algorithm tries to compute a function from the bits to some condition (legal, not legal).

However, the legality of the usage of those bits is not (and can never be) just a function of the bits themselves... the way the bits were acquired, and the context in which they are being used or distributed, are important. (For example: maybe the user is the copyright owner, or has a license from the copyright owner to use those bits, but there is no way for the algorithm to reliably know this).

Read this: What colour are your bits?

The problem is that the copyright status of information is a "color", a meta-property that cannot be deduced solely from the bits that make up the information. So any technological systems we build that try to derive copyright status of the bits as a function of the bits themselves, is doomed to encounter problems like this.

Years ago, this problem was explained quite well.

That brings to mind the "colour" of bits: http://ansuz.sooke.bc.ca/entry/23. The concept of "legality" ultimately cannot be properly applied to bits and so attempts to censor based on legality are doomed to fail although I'm sure they'll manage to piss off a lot of people trying.

It's all in the color of the bits.

As the GP pointed, he's time-shifting. Nothing illegal here.

That isn't how copyright law works. You don't get to download from someone else even if you own the exact same bits from a legal source. See What Colour are your bits?.

The end result is as if I'd ripped them myself but without the hassle of actually doing it.

What color are your bits?

What they appear to be accusing MU of is having deduplication and therefore knowing that there were multiple links to the same file, only taking down the links actually identified in a DMCA 512(c) takedown, and having actual knowledge that other links to the same file were also copyright-infringing and doing nothing about it.

Uh, no, they didn't know (at least, not legally) that the other links were infringing.

If I take a picture and upload it to e.g. Flickr, and then someone else downloads it from my profile and uploads to his, that copy is infringing and mine isn't, even though it's the same file.

Whether a file is infringing depends on its colour, not just on its bits.

One might argue (...) that a million monkeys typing randomly could have typed that same occurrence.

Yes, but that wouldn't have the right colour.

This is covered a bit in TFA. They reference a case that does remote DVR storage and make the distinction that having one copy that you send to multiple people becomes a broadcast, whereas keeping an individual copy for each person that uploads it is actually streaming the individual their own copy. So essentially, for your work case, you would have to have an individual disc, file, whatever medium stored the content you were streaming, and make sure that you were demonstrably sending a different copy of the same bits to different members. (This article What Colour are your bits seems relevant.) The implication appears to be that Google and Amazon will not be able to use any sort of duplication-based compression to save space when multiple people upload the same track, if they don't want to run afoul of broadcast licensing restrictions.

The legality of the file is not a property of the file itself, and cannot be determined from the file's content. If I buy an MP3 on Amazon, I can legally use it. If I put it on bittorrent and you download it, you have the same file as I do, but the RIAA says you're not allowed to use it.

This idea is explored in more details in the following blog post What Colour are your bits?

That's a great article. It explains the tech to the non-technical and vice versa.

The legality of the file is not a property of the file itself, and cannot be determined from the file's content. If I buy an MP3 on Amazon, I can legally use it. If I put it on bittorrent and you download it, you have the same file as I do, but the RIAA says you're not allowed to use it.

This idea is explored in more details in the following blog post What Colour are your bits?

That doesn't mean it makes any sense from a technical or scientific point of view. The only reason that is the law is because special interests have decided to go with delusional impossible ideas to protect their profit engine.

The legality of the file is not a property of the file itself, and cannot be determined from the file's content. If I buy an MP3 on Amazon, I can legally use it. If I put it on bittorrent and you download it, you have the same file as I do, but the RIAA says you're not allowed to use it.

This idea is explored in more details in the following blog post What Colour are your bits?

There is no difference between a byte to my phone versus a byte to other device.

Yes there is. The bytes to your phone are a different color than the ones being sent to your other device.

I can encode any information as a rather large decimal (base 10) number, numbers can't be patented or copyrighted. In fact, I've even written a program that encodes and decodes in such a way (arbitrary bit-length & radix integer math) -- It's terribly inefficient in decimal mode; in Hexadecimal (base 16) it's blazingly fast, but it doubles the output size... You can avoid the size bloat by encoding & decoding your NUMBERS in base 2 --- Oh, wait binary numbers are what's claimed as infringing copyright. (How is this not a 1st amendment issue?)

If I understand copyright law correctly (yeah, fat chance), numbers can be copyrighted if sufficient creativity* has been necessary to produce it. If you need, say, a book in order to produce the number, all of the creativity used in producing that book is needed to produce your number, so yes, it can be covered by copyright. I think http://ansuz.sooke.bc.ca/entry/23 covers the points of your post pretty decently.

*Probably not the right word, but I don't know the English equivalent of the Danish "værkshøjde".

Maybe we should make a website that generates sequential numbers from 1->arbitrarily large, discounting those that are known to be things like the CSS key. If any other number is used as a cryptographic key in a commercial product, they would clearly be violating our copyright on that particular arbitrarily large number, right?

No, because those numbers wouldn't have the right colour.

http://ansuz.sooke.bc.ca/entry/23

Or, "Provenance matters".

Your CD arises because you feel a false sense of ownership over your work. Maybe this thing written by the smartest guy I know will help: http://ansuz.sooke.bc.ca/entry/23 YMMV.

If media files could be successfully watermarked and traced back to their primary sources,

That's essentially DRM, and like DRM, it won't ever work; because bits don't have an inherent source (don't have Colour), and anything that can be added can also be removed, and then that process can be automated and the script/program shared.

That depends on its colour.

Is there an algorithm that can distinguish between an original copyrighted work and a fair-use derivative for audio or video?

Of course not, because bits have no color.

They won't stop until all your bits are the right color.

Matt Skala's modest proposal was apparently written before this story broke. Yes, his satire is outpacing reality, but only just barely.

I tried the SNES emulator on the thing, with ROMs of those games (pre-dumped -- not pretentious enough to build or buy my own dumper)

Then technically, you've used your R4 for piracy. If you dump your own Super NES Game Paks (which will become easier once Retrode is out), you're protected under 17 USC 117 and foreign counterparts. But if you obtained bit-identical files over the Internet, then you've copied the copyrighted code libraries written and licensed by Nintendo for use in Super NES games without authorization. Unlike patent and trademark law, copyright law cares about the provenance of a copy.

If you download an already-ripped mp3 when you already own the CD and could have ripped and encoded it yourself, could your action be found fair use, yet the actions of the site who provided you the mp3 be found as infringing? I love the bizarro world of copyright infringement, where for example a bit is more than just a 1 or 0.

What Colour are your bits?.

He thought mathematically turning bits into something else could alter their color, a classic mistake when programmers try to apply computer science to copyright law.

The problem you have is not that you have a wrong idea of programming, but that you have a wrong idea of mathematics. Most people only get educated into a tiny, tiny, tiny fraction of mathematics and subconsciously think that something like the quadratic equation is the height of mathematical expression complexity.

However, nothing stops a mathematician from examining the consequences of much larger systems, such as might look like a program, and in fact mathematicians do. There's research into large numbers, research into large proof systems that make most programs look like small beans (go read the Principia Mathematica... or rather, go skim the Principia Mathematica and run screaming), research into all sorts of things that are definitely math but look a lot more like a program than you think if you've only been education with conventional primary school mathematics, or even if you've gotten a bachelor's degree in computer science, which does not generally cover the Curry-Howard isomorphism. (I didn't even get it in my masters program, I had to learn it myself.)

There is no feasible way of drawing a distinction between mathematics and programs. You might be able to draw a legal one, but as always happens when you try to introduce colors to bits, the coloring just won't stand up to the sort of sandblaster scrutiny that will be applied by the plaintiffs and defendants.

You might observe that few programmers appear to be thinking mathematically when they program, to which I'd follow up with an observation that no, no they don't and boy does it show! But note carefully that's a characteristic of the programmer, not the program. The programmer may not understand math, and may crank out a mathematical system of breathtaking worthlessness and with few or none of the properties the programmer would have found desirable (like "actually doing what it's supposed to do"), but it is not written into the definition of mathematics that something is only mathematics if it is "useful" or "good". It's still all just a term-rewriting system when you get down to it, and any attempt to draw a barrier around "term rewriting system" and "real-world program" is simply doomed to failure.

It's all just assembler, and all assembler can do is basic arithmetic, moving numbers around in memory in a manner very easy to characterize mathematically, and some very mathematical conditionals. You might be able to fool yourself into thinking you've somehow transcended these primitives into a "non-math" domain... but you haven't.

Hosting a bunch of images doesn't do any good unless you have a text (or at least searchable) description of what you're downloading. Without context, warehoused information is useless. And these PNG files are just different representations of the same quasi-legal information (that is, they're still colored bits.

That was my initial reaction as well.

Instead of a .torrent file you've got a PNG, but I'm not sure how much that helps anyone.

I don't think the complaint was ever that the information contained in the .torrent file was somehow infringing copyright or breaking laws... I believe the argument has been made that there's nothing actually copyrighted/illegal in the .torrent file itself, and judging from the results of recent court cases that argument doesn't seem to be working terribly well. The PNG still contains the same information as the torrent... Still enables you to download the same files... Would, therefor, still be vulnerable to the same lawsuits and takedown notices - wouldn't it?

It isn't a text file, so it might be harder to locate the incriminating evidence with a simple text search... But you'll need to indicate what's contained in the file somewhere, or nobody is going to know what they're downloading. You'll still need a searchable index, or a header on the forum post, or at least a line or two of text saying what the picture gets you.

So... How is this actually better than a .torrent?

Hosting a bunch of images doesn't do any good unless you have a text (or at least searchable) description of what you're downloading. Without context, warehoused information is useless. And these PNG files are just different representations of the same quasi-legal information (that is, they're still colored bits.

Anyone who hasn't seen it already should read What Colour Are Your Bits? which discusses this in detail.

I wouldnt quite agree, as the .torrent files are checksums derived from either DVDrips or screencaps, which are both derivative works.

Go read up on what lawyers talk about the The Colour of the Bits. It's rather interesting, but also states indirectly that one cannot know this type of colour easily.

I wonder what NewYorkCoultryLawyer would say about this..

Several similarly silly schemes actually exist (see the OFF System, and lawyers tend to be unimpressed - they all amount to encrypting the communication and/or obfuscating what you're transferring. So if what you're transferring is copyrighted, you're still infringing on that copyright.

See What Colour are your bits?.

If they ever brought me to court, I'm going to ask what the difference is between getting the movie in the mail and ripping it myself.

Ripping it yourself from where? You're talking about two completely different things here--having possession of a disc and making a digital copy, and getting possession of the disc in the first place.

But since I think I understand what you're getting at, here's an interesting read for you:
http://ansuz.sooke.bc.ca/lawpoli/colour/2004061001.php

It's not very sexy, but it's fascinating and readable. I remember coming across it in Dillons bookshop, not knowing the name, and flicking through. Half an hour later, when I realised the time, I knew I had to buy it! Other books go into more exhausting detail (Knuth in particular), or cover a wider range (Knuth again!), or more modern ideas or languages. But Sedgewick is a great read, and I've been through it several times.

It covers all the basics (maths, searching, sorting, strings, graphs, and touches on FFTs and hardware and optimisation), and gives enough detail that you could go off and write some programs yourself. But more importantly, it explains them: how each algorithm works, what it's trying to achieve, how it behaves, and why. And it's because it explains the ideas so well that I'd recommend it. After every section I felt I'd learned something -- not because I had to, but for the sheer pleasure of understanding something new and interesting.

Other recommendations: Effective Java (a staggering amount of insight into the language), Thinking in Java (by someone who understands that language is more than just syntax), Deep C Secrets (again a pile of insight, interspersed with anecdotes and some rather off-the-wall diversions), Programming Pearls and More Programming Pearls (problem-solving in bite-sized chunks -- a little dated but still interesting). Plus I've already mentioned Knuth. K&R is well done, though narrow in scope. I find Design Patterns useful, but more for clarifying things I've already seen than for learning new things. I've never actually read The Mythical Man-Month, but people I respect mention it, so I'm sure it's well worth reading too!

Of course, times being what they are, especially in this field, a lot of interesting stuff is on-line. Some hat should go without saying hereabouts include the latest Jargon File, some of Eric Raymond's books, and more online documentation and archives than anyone but Google can cope with.

Other interesting articles include The Programmer's Stone, a guide to writing Unmaintainable Code, The Ten Commandments for C Programmers (annotated edition), Ken Thompson's Reflections on Trusting Trust, What Colour are your Bits?, and Guy Steele's Growing a Language.

Aside the difficulty of factoring out primes in big numbers, would trading the primes and frequency violate copyright?

I imagine that would depend on the color of the primes. If those primes were derived from from the music, then yes, there is infringement; otherwise no. If you do not understand the reference, read the linked article. It is excellent reading.

Simple, they'll just check the Colour of the bits being downloaded!

Anyone interested in copyright who hasn't already read What Colour are your Bits really should do so.

Now on to the file system in this article. I thought of almost exactly this system a couple of years ago, so I already considered exactly this legal analysis. The creator is mostly mistaken about escaping copyright with the randomized files. Exactly one out of every three data chunks will be infected by copyright-colour. Under the right circumstances it is impossible for anyone to ever determine which of the three data blocks carries the copyright-colour. This actually has some interesting legal implications.

In civil court the standard of proof is that you have to show a greater than 50-50 chance that the person is "guilty" of infringing the copyright. If you have three data chunks for some copyrighted file spread across three different servers redistributing those chunks, and you don't know which chunk is carrying the copyright colour, then you can't successfully sue any of the three in civil court for copyright infringement. You would have to find one server serving up at least two chunks for the same file to demonstrate more than a 50% chance that they are infringing.

However this file system does not operate in an ideal manner to reliably apply the above legal situation. As I said, under the right circumstances it is impossible for anyone to determine which of the three data blocks carries the copyright-colour. However instead of using two completely random colourless data chunks when creating the new colour-carrying third chunk, this system grabs two old chunks from the system to fill the role of the two colourless blocks. This has the advantage that you effectively triple your disk storage ability. It also has the disadvantage that it is then specifically the third block added to the system that is carrying the copyright-colour for that file. In general it will probably be pretty difficult for someone trying pin down which of the three chunks is the colour-carrying chunk from some specific file, it does open up several possible avenues of attack to figure it out. And then they could sue the person distributing that particular chuck.

-

Except that is probably bullshit to copyright lawyers.

There's a great explanation of why in this essay, What Colour are
your Bits. It's actually about another system based on the same sort of
ideas.

http://ansuz.sooke.bc.ca/lawpoli/colour/2004061001.php

Under the lawyer's rules, Colour is not a mathematical function of
the bits that you can determine by examining the bits. It matters where
the bits came from. The scrambled file still has the copyright Colour
because it came from the copyrighted input file. It doesn't matter that
it looks like, or maybe even is bit-for-bit identical with, some other
file that you could get from a random number generator. It happens that
you didn't get it from a random number generator. You got it from
copyrighted material; it is copyrighted.

But the OFF is a bit different from standard encryption. In the OFF,
every data block is produced by a random-number generator.

The only thing that would have Colour, from the lawyer's point of view,
would be the instructions at the URL, as indicated on the OFF wiki. So
the data blocks in the OFF would be Colourless even to the lawyer
in the example above.

For an excellent explanation of why this is legally stupid, see What Colour Are Your Bits?

Replying to my own post, but this IS just a sort of encryption - their main claim being because the data is encrypted, it's not copyright.

As has been pointed out below, the data transferred is not the thing copyrighted - it's what it represents. So it's an arduous and painful encryption, with high overhead, easy to crack and no plausible benefit. With some hand-wavy 'it annuls all badness from bad things' explanation.

Except that is probably bullshit to copyright lawyers

There's a great explanation of why in this essay, What Colour are your Bits. It's actually about another system based on the same sort of ideas.

http://ansuz.sooke.bc.ca/lawpoli/colour/2004061001.php

The fallacy of Monolith is that it's playing fast and loose with Colour, attempting to use legal rules one moment and math rules another moment as convenient. When you have a copyrighted file at the start, that file clearly has the "covered by copyright" Colour, and you're not cleared for it, Citizen. When it's scrambled by Monolith, the claim is that the resulting file has no Colour - how could it have the copyright Colour? It's just random bits! Then when it's descrambled, it still can't have the copyright Colour because it came from public inputs. The problem is that there are two conflicting sets of rules there. Under the lawyer's rules, Colour is not a mathematical function of the bits that you can determine by examining the bits. It matters where the bits came from. The scrambled file still has the copyright Colour because it came from the copyrighted input file. It doesn't matter that it looks like, or maybe even is bit-for-bit identical with, some other file that you could get from a random number generator. It happens that you didn't get it from a random number generator. You got it from copyrighted material; it is copyrighted. The randomly-generated file, even if bit-for-bit identical, would have a different Colour. The Colour inherits through all scrambling and descrambling operations and you're distributing a copyrighted work, you Commie Mutant Traitor.

To a computer scientist, on the other hand, bits are bits are bits and it is absolutely fundamental that two identical chunks of bits cannot be distinguished. Colour does not exist. I've seen computer people claim (indeed, one did this to me just today in the very discussion that inspired this posting) that copyright law inescapably leads to nonsense conclusions like "If I own copyright on one thing, and copyright inherits through XOR, then I own copyright on everything because everything can be obtained from my one thing by XORing it with the right file." That sounds profound only if you're a Colour-blind computer scientist; it would be boring nonsense to a lawyer because lawyers are trained to believe in and use Colour, and it's obvious to a lawyer that the Colour doesn't magically bleed to the entire universe through the hypothetical random files that might be created some day. You could create the file randomly, but you didn't. Maybe you could create a file identical to the complete works of Shakespeare by XORing together two files of apparently random garbage. "Why, so can I, or so can any man;" but that doesn't mean that I am William Shakespeare.

I strongly recommend that you go read "What Colour are your bits?", which discusses this exact scenario.

However since MP3s are actually being sold now through online music stores, you would have to argue that these are degraded compared to the actual product being sold.

On the other hand, when some user converts the CD tracks into MP3 and puts them on P2P, and the MP3 found on webstores aren't the same product. At all.
That would be claiming copyright infringement on some picture you took with your camera of some public monument - on the ground that an artist is selling a poster of the same monument and your photograph infringes the artist's copyright because they both depict the same monument. It doesn't work this way.
What is important is not what the piece of work represents. What is important is what you duplicated. Which was the "source" of your data. The colour of the bits which compose the data file.
Your photograph isn't related to the artist's poster. You took picture of some monument, your photograph is related to it and as the monument is public, you can take pictures of it.

To go back to MP3 : you took a CD, you ripped it and compressed it to MP3. Those MP3s are descendant of the CD - the CD was the source of data of which you made a copy. The MP3s aren't bit for bit identical to the original material that you duplicated. They are degraded. You're safe.

The fact that some MP3 exists on some on-line store have nothing to do with the copyright law. The copyright law is only about duplicating content and at no point in time in this exemple the online store did come into play. The duplicated thing was a CD, copyright law concerns it.

You say you couldn't upload an 8gb ISO of a DVD, but isn't that 'degraded' from the original masters or even the HD-DVD version?

Again, doesn't matter. You are *NOT* making a copy of the master or the HD-DVD. the bit that you convert aren't colored from those.
What you make a copy of is your DVD.
When proprely done, a 8GB ISO of ripped DVD is supposed to be bit-for-bit equivalent (at least regarding the multimedia content that is protect by copyright law - the file header and other metadata might change, but the data that will ultimately be sent to the screen is kept un-changed).
As opposed, for example, to shrinking (using K9Copy or something similar) the movie to fit a 4GB disc (in that case, the quality was decreased to fit in a tighter place).
Or for example, recompressing it using X264 - the conversion is lossy and it mathematically provable that degradation has resulted.

The fact that the 8GB iso is degraded when compared to some master copy at the studio that you don't have access to doesn't count. What count is the source that was used in the process of copying.
Exact copy is "no no!", degraded using lossy process is tolerated.

Certainly if you bought a 256kbit MP3 from Amazon and shared it you wouldn't consider that 'degraded' since it is exactly what you purchased, right?

Exactly, in that case it won't be considered 'degraded' because what you are making and distributing copies of it the exact data as you received it (no matter what exists on some other CD that wasn't involved in the current copying).

That's what I meant in the end of my previous posting :
- If companies continue to distribute music in CDs and movies as DVDs (or not-much compressed MPEG2 streams on digital boardcast networks), people have an incentive to make degraded copies using lossy compressors (to make data use less space). And those copies could be tolerated on P2P network under Italian law (as long as done not-for-profit).

- If companies switches their distribution channels to MP3s and more compressed digital broadcast (H264), the situation change.
There's no incentive for people to re-compress the media : what they got is already compressed.
But they don't have the right to distribute these files under any copyright law (except so