Update on Slow rays of hope

Story: Oracle v. Google - Judge Alsup Rules APIs Not Protected By CopyrightTotal Replies: 13
Author Content

Jun 01, 2012
7:46 AM EDT
My apologies for the unusual information notes on the story. When I first posted this on LXer, the story had only just broken on Groklaw and there was only the two line statement as to Alsup's judgement....Groklaw has the lot now.

Given the enormous thread dealing with this case under the heading "Slow rays of hope":


this judgement by Alsup could not have come at a better time. It's another move by the courts that little by little is detaching software that should be unencumbered from the monopolies of corporations. PJ's comments indicate that while Oracle may appeal, there is little chance that such an appeal will be successful. Apart from Google's enormous win, only Oracle's lawyers have done well out of this case I think. What utter waste, and for what ? Control ? Money ? Power ?

And one last thing.....PJ mentions it with what I can only imply is "amazed delight" and I'd echo her sentiments absolutely. Judge Alsup is also a person who has reasonable knowledge of software programming and if ever there was a need in any case that deals with software patents/copyrights, that is a crucial one. Alsup's ability to navigate through this case was definitely helped by his good understanding of software construction and I firmly believe that accurate judgements on these highly complex issues can be best dealt with by persons who have a foot in both the legal and software programming circles.

Jun 01, 2012
8:12 AM EDT
Glad to see that decision. It is completely correct and reflects the way I learned copyright law back in the early 90s.

The classic example is recipes, which cannot be copyrighted. Ditto for phone books, though that might make many younger folks do a quick, "Wazzat?"

Do a recipe in a nice cheery and creative way? Your nice cheery creativity can be copyrighted. The recipe itself, however, can be copied by somebody else and published in a different cheery way or a bland and ugly way. It ain't the proper subject for copyright.

Ditto for APIs.

That, by the way, has always been the underlying basis of my contention that the EFF overstates the case when it claims that dynamic linkage is sufficient to bring programs under the GPL.


Jun 01, 2012
10:35 AM EDT
> The classic example is recipes, which cannot be copyrighted. Ditto for phone books,

My understanding was that while the individual data could not be copyrighted, the compilation of the data could. Is that no longer correct? In fact, I think it was referred to as a "compilation copyright".

Jun 01, 2012
11:41 AM EDT
A compilation can, but does not necessarily, represent a creative act on the part of the compiler, so...you can (almost) certainly copyright a recipe book of very uncreative recipes.

An interesting exception to that involved a company trying to assert a copyright on white page telephone books, but...a telephone book really needs all the names if it is to be useful, and those names likewise need to be alphabetized, so....where is the creative input?

Jun 01, 2012
12:04 PM EDT
> A compilation can, but does not necessarily, represent a creative act on the part of the compiler,

Ah. So yes, a compilation is copyrightable but only if some aspect of the compilation is creative.

That makes sense.

Jun 01, 2012
12:05 PM EDT
It seems to me that a lot of the confusion in the issues of copyright and patents revolves around the basic meanings of the terms used to define them.

The relevant laws may be well defined superficially, but what do the words "creative" and "innovative" really mean?

How different does something have to be before it is "innovative"?

What is the critical essence of the difference between something that is "creative" and something that is not?

You could go nutty trying to think those things through.

Jun 01, 2012
12:56 PM EDT
@skelband --

They do have different standards, and for good reason.

In copyright, you need only a tiny bit of creativity, but your protection is very thin. It does give you protection against somebody substantially copying your work (say, translating it to chinese, adjusting the colors of a picture, etc), but it also allows somebody to develop exactly the same work independently (though, good luck proving that). One more thing: copyrights are automatic. There is no copyright examiner and the question arises only in court post-hoc.

The exceptions in copyright tend to be things that are purely functional: lists, apis, etc, or where function more or less demanded the result, ie, how many ways can your really do this thing?

Another, very similar, exception applies to things like a zebra head mounted on the wall. To portray a zebra head, you have very little choice but to represent zebra features. Copyrightable elements are limited to those things not required to represent a zebra.

Patents are more concerned with inventions and the standard is higher. Creativity is not enough. The invention must also be non-obvious to the ordinary skilled practitioner of the art. Patents are subject to court challenges just like copyrights, but must also clear the hurdle of going through the patent examination process before they are granted. This is appropriate when you consider that patents grant an absolute monopoly.

Jun 01, 2012
1:00 PM EDT
IANAL, but my understanding is that "innovation" is (in legal terms) something different than "invention" -- and that one (ie. "invention") is protected by patents, while the other isn't.

(If that even makes a difference in this discussion)

Jun 01, 2012
1:18 PM EDT
I guess I wasn't trying to draw a parallel between patents and copyrights. They are clearly very different.

My point was that the words that we use to define them are difficult to pin down in a way that means the same to all people.

I guess that is a constant issue in legal circles where the challenge is always to be particularly OCD about what the words used mean in the very strictest sense.

However, a lot of the debate about patents and copyrights are outside of the legal circle.

So my point stands. What does "creative" mean? If as an artist, I make a chair that is a functional chair, how different from a standard chair does it have to be before it is creative?

If I come up with a new idea that I think might be patentable, how "unobvious" does it have to be before it is fit for patent? If I ask one practitioner, they may deem it revolutionary. If I ask another, they may say that it is obvious. The definitions of these terms rely on judgment and I'm always wary of law that relies too heavily on judgment to determine the nature of the law itself.

If you murder someone, the court expends its effort on determining if you did it, or not. They don't have to debate the fundamental nature of murder to achieve its aim. Killing someone is illegal. If you do it, you're guilty. Sure, the court may have to mull over your motivation, if you were of sound mind and all that, but at the end of the day, the definition of the crime is as precise as it can be. The recent Oracle/Google case was all about trying to pin down the nature of the law itself. I think must cause us to pause and wonder if the law itself is actually the problem.

It seems to me that the crux of a lot of what's wrong with copyright and patent is its imprecise nature. We debate endlessly about whether or not something is truly creative, or non-obvious, or innovative. I think this is at the heart of the problems with these laws.

Jun 01, 2012
3:12 PM EDT
"We debate endlessly about whether or not something is truly creative, or non-obvious, or innovative. I think this is at the heart of the problems with these laws."

I believe that the reason for all this imprecision is that trying to restrict the flow of information is unnatural. Laws created to restrict people's peaceful actions end up being eternally bickered over, since they don't actually satisfy anyone.

Jun 01, 2012
10:04 PM EDT
Dinotrac, I am not sure I understand why you say the EFF overstates its case when it says dynamic linkage is sufficient to bring programs under the GPL.

You are saying that code is copyrightable as a creative authorship (i.e. as a particular expression or illustration of an inventive idea or method of operation, as opposed to the idea itself which would fall under the realm of patents not copyrights).The API would be part of an idea,system,or method of operation, so therefore COULD be patentable if it was considered inventive enough (and I have seen the patents for dotNet which do include API's).

So far that is consistent with the Oracle vs Google decision that copyright protection is not appropriate for the API's.

However, EFF is not claiming copyright protection for API's of linked sw,but rather for the creative authorship of code of the linked sw,which seems to me consistent with the above argument.

You may say they only link to symbols,but in practical terms,that is only a function of when the linking happens - compile time vs runtime. It would otherwise be a trivial operation to bypass any gpl protection by adding a trivial interface layer ,would it not?

Jun 01, 2012
10:38 PM EDT
Which is what nVidia does, n'est-ce pas?

Jun 02, 2012
11:13 AM EDT
@lg --

APIs generally fail the creativity test for copyright and should not themselves be patentable. Frankly, I can't even imagine a way for an API to qualify as an invention in patent law terms. That is different, by the way, from saying that patent claims don't include reference to APIs. You would not believe how many patent claims make reference to using a database, but they do not claim the database as an invention (duh!).

Here's why I think the EFF overstates its case:

The combination of a program with linked libraries does, arguably, create a derived work. This is most clear in the case of static linkage where the linked libraries are actually incorporated into the executable.

In the case of dynamic linkage, the program generally includes header files that define structures and calls that related to the linked libraries. If, however, those are not copyrightable, the program without the libraries is not a derived worked and violates no copyright, hence would not be subject to the GPL simply because the libraries were GPL'd. That might even be true if the headers, etc were copyrightable, if such use were construed to be fair use under the law.

If you have read the GPL, it doesn't actually place any contraints on users' ability to do with GPL'd work as they please. It merely places a requirement on those who distribute GPL'd software: provide the source.


1. If the program is distributed without the libraries, no source needs to be provided at all. Users are free to obtain the libraries and run the program because the GPL expressly permits users to do that.

2. If the program is distributed with the libraries, source code for the libraries must be provided. EFF claims that source code for the program must also be provided because the program must be GPL'd, but I believe that to be over-reaching unless the header files are copyrightable and their use is not fair use.


Jun 04, 2012
10:18 AM EDT
Quoting: Glad to see that decision. It is completely correct and reflects the way I learned copyright law back in the early 90s.

Ditto for me... I think I had two classes on the subject, one on Copyright, Patent and Trademark, and another on technology, law and ethics that also dealt somewhat with this subject (and also with right to die cases, etc.).

Posting in this forum is limited to members of the group: [ForumMods, SITEADMINS, MEMBERS.]

Becoming a member of LXer is easy and free. Join Us!