Messages in this thread | | | Date | Tue, 20 Oct 1998 11:53:37 -0700 | From | (Jim Gettys) | Subject | Re: [Only Tenuously On Topic] Profibus for Linux - license advice wanted. |
| |
Fundamentally, software has to have some sort of copyright statement to be useful to anyone other than the author.
The short reason why is that without knowing who claims authorship, someone recieving it has no way of knowing even who to ask for permission, or under what terms it may be used. Even here, experience shows that redistribution generally should be done with explicit permission, rather than just relying on what may be an inaccurate copyright statement at the top.
So "public domain" software is essentially useless. I learned this lesson early in the X Window system development, when it this was pointed out to me by an IBM'er of the day the reasoning behind the IBM policy against even touching any software without copyright.
Ok, given that you have have a copyright, what then? You need to understand the terms of the copyright (and implications of those terms). And you need to understand what your aims are by your copyright, and whether they would interact badly in the environment you are trying to achieve.
When I wrote the original X Window System copyright (don't blame me for recent brain-damage), the aim was to enable third party system and ISV software developers to use X as widely as possible; GPL's add restrictions that make it difficult for many commercial companies to take code and "add value".
On the plus side, this encouraged a wide variety of applications to be developed and marketed that might have been inhibited by a GPL style copyright (note that the GPL was still in the future, I believe; I wrote the draft of the X copyright sometime around 1986 if my memory serves me); on the down side, there was no guarantee that improvements would be available to the X community. Note that there is still often great incentive for companies to donate improvements to X (or BSD style) copyrighted code, just so they don't have to maintain the code afterwards. Experience showed that a large amount of code was given back to MIT (and subsequently the X consortium) by a large set of companies, in large part, so that they would not have to "improve" the code forever in a proprietary release. I would say that the X copyright succeeded in its aims quite well.
Experience also showed that the copyright/availability/license terms of Motif (and CDE) greatly inhibited their adoption, to the point of helping seal the fate of X Workstations vs. PC's and Windows (though the GUI wars before this are even more to blame than the bad business terms). (Sigh...) So there is an object lesson here to be careful about.
If you believe your value is primarily in your libraries and tools on top, then it would seem like the easy solution is to "give away" the kernel code under some unrestricted copyright; if you want to guarantee any improvements are available, a GPL may be most appropriate (particularly with the existing GPL on the Linux kernel). Your library and tools on top might have a very different copyright, if sources are available.
If your primarly value is in the kernel code, you have a harder problem.
Ultimately, it is up to you to set the terms of a copyright and/or licensing terms. Please do so carefully, with some real thought. It is not something to do without some real thought as to the long term consequences.
- Jim
-- Jim Gettys Digital Industry Standards and Consortia Compaq Computer Corporation Visting Scientist, World Wide Web Consortium, M.I.T. http://www.w3.org/People/Gettys/ jg@w3.org, jg@pa.dec.com
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |