Suppose I created a website that cataloged every possible sequence of letters in order.
You would start with the letter "a", could swipe right to read "b", and further to find "aa", or "ab" (appearing in dictionary order, as you would observe on the rows of an excel sheet).

My website has a feature where you can instantly read any text given a numerical ID. The IDs could, for example, be ordered as follows:

catalog of text with their respective IDs

Where individual texts in the catalog are separated by red bars, and the blue numbers are their respective IDs.

This website happens to be a harbor for internet piracy. One user, for instance, discovered the numerical ID referring to the entire text of Harry Potter and the Sorcerer's Stone, and has distributed the ID to their friends so that they may read the novel at their leisure.

It's unclear to me where the copyright infringement occurs. Is it the website for "storing" every possible non-public domain text? Is it the user for finding and sharing a number that has apparent meaning to this website?

  • "Is it (copyright infringement) for "storing" every possible non-public domain text?" Ah, yes. Why the confusion? Even if there's a password or secret code, the website owner is still infringing. – BlueDogRanch Aug 28 '20 at 18:31
  • @BlueDogRanch That's strange to me. The website doesn't have to store anything, it can just generate the text on the fly provided the ID. A person with 1 computer programming class under their belt could make that and paste the program in this very comment. – bitconfused Aug 28 '20 at 18:40
  • 2
    Also, the concept of illegal numbers may also be relevant here. – Michael Seifert Aug 28 '20 at 18:42
  • Since you are generating these sequences in order, you start with any number of as. Not just a lot, but all numbers. Turns out that "a", "aa", "aaa"... "aaaaaa..aaa" are not protected by copyright. You never get out of that rut, so you never have the opportunity to infringe copyright. – user6726 Aug 29 '20 at 14:50
  • 2
    @BlueDogRanch - if text is generated randomly and happens to include a copyrighted poem, the poem has not been copied. To copy one must have access to the original. Otherwise it is a coincidental production of the same text. – George White Aug 29 '20 at 17:06

It's not clear to me that your website is, per se, illegal. (Though good luck finding a webhost with a hard drive large enough to store it. Harry Potter and the Philosopher's Stone is roughly 38,000 characters. If I've done my calculations correctly, storing all of the strings of this length or shorter would require something like 1053760 GB of storage. Give or take.)

The main problem is in distributing the "index" of the string containing the text of Harry Potter and the Sorcerer's Stone. This index is just the integer obtained by taking the entire text as a string and treating it as a number in base 26 (A = 1, B = 2, etc.) In other words, all of the information contained in the book is also contained in the index number, and in a relatively simple way.

Distributing this number is therefore entirely equivalent to distributing the text; you don't need the website at all. It's entirely analogous to e-mailing the text to someone else; in that process, the text is transformed into a number (in binary, under a different encoding scheme) along the way. But you couldn't seriously argue in a court of law that you were not distributing the text, you were just telling a friend about a very large number.

Just to back this up with some actual law citation: US copyright law gives the copyright owner the exclusive right to make and distribute copies of the work. And a "copy" is defined as follows:

“Copies” are material objects, other than phonorecords, in which a work is fixed by any method now known or later developed, and from which the work can be perceived, reproduced, or otherwise communicated, either directly or with the aid of a machine or device.

In other words, J.K. Rowling (and/or her publishers) have the exclusive rights to distribute that integer, since the work is "fixed" in it by its very nature, and since the work can be reproduced from it.

  • 1
    Interesting! The citation really helps. If I'm generalizing correctly, do you think a generic link to a real website committing piracy would be considered a "copy", where the website is the "device" that communicates the original work? I'm still having trouble distinguishing a large base-26 number from a random link like "www(dot)illegalwebsite(dot)com/Harry_Potter" since they both reproduce the work with the help of the website. Is distributing the link the same as distributing a copy? – bitconfused Aug 28 '20 at 19:22

It was noted in a comment that one possible implementation of such a website would be to dynamically generate the text from the ID:

The website doesn't have to store anything, it can just generate the text on the fly provided the ID. A person with 1 computer programming class under their belt could make that and paste the program in this very comment.

Your website, in this case, is effectively interpreting an encoding format. Another possible (more efficient, probably) encoding format would be, say, a ZIP file. It would not be a copyright violation, per se, to create a website that decompresses ZIP files that could potentially contain copyrighted content, but it would be a violation to distribute said ZIP files, as they are just another way of encoding the content. Similarly, it would be a copyright violation to distribute the "ID" in your system, as the ID would contain the copyrighted content.

There's a tiiiiiny catch here: if it can be proven that the primary purpose for you to create your website was to facilitate the distribution of copyrighted material, you may have some liability under some jurisdictions, or at least something that looks enough like liability to force you to defend a very expensive copyright lawsuit.


To prove infringement the rightsholder needs to show you actually copied their work, and did not create it independently. Normally the way this is done is by showing that you had access to the original, or that the end result is substantially similar. So it is theoretically possible this would not be infringement, because your creation did not involve copying their work, you created it independently, and you can prove it. But in practice it is doubtful, it is not mathematically feasible to happen to create a duplicate of an existing creative work like a book by formula like this. If each integer could be stored in one atom, recording all combinations to cover 100 characters would require more atoms than there are in the observable universe. And a typical book has on the order of 100000 characters. So the only way to really find this ID is to start with the finished book, and create a copy by encoding it.

Both the website, and any user who shares this ID would be liable for infringement if they do any of the following, reproduce the work, produce derivative works, distribute copies, or display the copyrighted work.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.