A GPL-enforcement suit against VMware

Posted Mar 13, 2015 15:03 UTC (Fri) by enyst (guest, #92308)
In reply to: A GPL-enforcement suit against VMware by dlang
Parent article: A GPL-enforcement suit against VMware

I don't understand how is that relevant here. It appears that VMWare did much more, it redesigned a large part of the software to talk to another API they made in-house, then kept the software implementing it proprietary. If my understanding is correct, by doing so, it's highly likely the whole work (emcompassing the proprietary parts) is a derivative of the original.

Do I misunderstand the case? I don't see the relevance of Oracle v. Google here.

A GPL-enforcement suit against VMware

Posted Mar 13, 2015 17:19 UTC (Fri) by dlang (guest, #313) [Link] (3 responses)

> I don't see the relevance of Oracle v. Google here.

I was replying to this part of the comment

> but mere use of an API can never make something a derivative work of something that presents that API.

A GPL-enforcement suit against VMware

Posted Mar 14, 2015 4:27 UTC (Sat) by butlerm (subscriber, #13312) [Link] (2 responses)

In Oracle v. Google, Google has re-implemented a entire, rather extensive API. You could make a reasonable case that any re- implementation of such an extensive and detailed interface amounts to a derived work, even if the similarity is strictly speaking only superficial.

The rather more questionable (if not absolutely ridiculous) claim, however, is that any consumer of an API is a derived work of the original description or implementation of the API. Under that standard, which appears to have no basis in law anywhere, anything compatible with X would be considered a legally restricted derivative of X.

No one would be able to write, license, or sell a Java program to anyone anywhere without some kind of license from Oracle to do so, for example. One might imagine that Oracle is just the kind of company that might require a commercial license to distribute Java programs, or even to compose one in a text editor.

So besides the sheer audacity of the claim that mere consumption of an API makes any program a derived work of the API or of the original implementation of the API, it is the sort of thing that if adopted by the courts would create a greater obstacle to progress in the field than practically anything else imaginable.

A GPL-enforcement suit against VMware

Posted Mar 14, 2015 5:26 UTC (Sat) by Cyberax (✭ supporter ✭, #52523) [Link]

> No one would be able to write, license, or sell a Java program to anyone anywhere without some kind of license from Oracle to do so, for example. One might imagine that Oracle is just the kind of company that might require a commercial license to distribute Java programs, or even to compose one in a text editor.
Stop. Giving. Ideas. To. Oracle.

A GPL-enforcement suit against VMware

Posted Mar 19, 2015 13:05 UTC (Thu) by enyst (guest, #92308) [Link]

> In Oracle v. Google, Google has re-implemented a entire, rather extensive API. You could make a reasonable case that any re- implementation of such an extensive and detailed interface amounts to a derived work, even if the similarity is strictly speaking only superficial.

This is what Oracle lawyers have argued in the trial court.

It's a shocking theory, to say the least. I wish more people would realize that this is what Oracle is arguing, that any reimplementation of a "creative enough" API would be a derivative work merely by being a reimplementation.

If this was becoming the copyright law, then almost any clean room reimplementation would be presumably infringing.

I note that the "creative enough" -only standard of copyrightability would mean only slightly more "choices" than just random numbers or alphabetical lists.

Saying 'extensive' or 'detailed' doesn't give a criterium. The criterium Oracle is arguing for is Feist (slightly more than alphabetical lists).

> The rather more questionable (if not absolutely ridiculous) claim, however, is that any consumer of an API is a derived work of the original description or implementation of the API. Under that standard, which appears to have no basis in law anywhere, anything compatible with X would be considered a legally restricted derivative of X.

The two are related, the consumer replicates names too. Can't 'use' an interface without writing those names in a form or another; as many names as one needs.

According to Pamela Samuelson, consumers are potentially infringing:
http://goo.gl/DCoSqv