Oracle is, of course, appealing it’s loss on the recent Java copyright cases. Oracle, MS, NetApp, and EMC are all running scared that they won’t be able to copyright their APIs anymore. And they should be. An API is the visible portion of the application to a developer. API is like the letters and words in a language though. They’re completely meaningless symbols and strings by themselves and embody little, or nothing in terms of function and meaning.
In the dead tree world it would be sort of like copyrighting the words. Individual words, not even really phrases. It’s like trying to argue that their copyright isn’t on say Homer’s Iliad, but also on every individual word and heading, by themselves. That the mere synopsis, the cliff notes, of the Iliad is their copyright. And if someone looks at the table of contents or outline and creates an ORIGINAL STORY based on that outline, they’re doing something illegal. An API is like that…An outline, the cliff notes. Independent API implementations (stories) will hit those same cliff notes and outline points, but the story, how they got there, especially having never read the Iliad (not having their source code), will be different. In some cases it will be identical, there are certain data structures and manipulations that are common, but the whole story will be quite different, but you could still look at the outline of this new story, and agree, that largely it’s the same story.
From a computing standpoint WINE and Windows are exactly this. WINE implements the ABI/API of Windows and various Windows components. How they do it is *wildly* different. But from the standpoint of someone who only looks at the outline (the program running under WINE or Windows) – it is substantially identical. This is one of the powers of modern computing.
That (among other things) is why I’m on the Google side of the fence here. Oracle and their corporate cronies are trying to strengthen making it illegal to reverse engineer. Or indeed to even be able to compete against them in any shape or form. They want to be able to sue anyone who finds a bug in their software. Bugs cost them money. And if they remain undiscovered they don’t have to spend man hours (or months or years) fixing them. That’s how the corporate mindset thinks. I doubt any of the individual developers share this viewpoint, but as a company, thats got to be part of the thought process here.
They’d never say that publicly or in court mind you, but you can bet that this argument has been made. They’ve tried it time and time again, making reverse engineering illegal. It scares corporate america that hackers (no matter the color of the hat) could reverse engineer what they’re doing. Heck I’m not even talking necessarily about disassembling the code. I’m talking about looking at a public API, and from that, discerning the behavior and creating a compatible (and in their mind) competing implementation. They want to put the kibash on that.
If they can start to argue successfully that APIs are copyrightable and that ANYTHING implementing that API is potentially infringing on their copyrights, they get power over any potential competitor. If the Android/Google v. Oracle case stands, APIs are not copyrightable.
It’s the difference between saying it is the method, the way in which something achieves the end result over just how it looks to the outside. Right now they can, do, and I argue should, have strong protection against the method they chose to use. But they want to be able to say that the API itself is theirs and theirs alone, and anyone else who attempts to use it needs to agree to their copyright, and be subject to them entirely.