Derived work? Maybe...
|
| Author | Content |
|---|---|
| dinotrac Feb 05, 2005 12:47 PM EDT |
This whole notion of a driver for Linux being a derived work seems silly to me, but...I haven't seen how much of Linux is required to write a driver. Here's the deal: A derived work is always a combination of two works: 1. The original work from which a new work will be derived and 2. The new work of authorship which is not part of the original, and, from which a new work also will be derived. Putting the two seperate works together creates a new work, the derived work. That seems like a trivial recital of the obvious, but let's see where that "trivial" recital leaves us: 1. The combination of the kernel with a new driver is a derived work. 2. The driver by itself may or may not be a derived work. Let's explore number 2: If it is impossible to write a Linux driver without incorporating any copyrightable elements of the Linux kernel, then the driver itself is a derived work of the kernel. That is different, by the way, from saying that the Linux driver can be used apart from the kernel. If, however, you write a driver without incorporating copyrightable kernel elements, then the driver is not a derived work of the kernel, only the combination of kernel and driver is a derived work. What throws the whole mess up in the air is that it's pretty hard to write a driver without including assorted header files. Whether or not those header files are sufficient to make the driver a derived work depends on the level of authorship (creative input) in the headers that is seperate and apart from the purely functional content. And ... sigh ... that is what court cases get built on. |
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!