Tobias DiPasquale on Mon, 21 Jul 2003 07:21:16 -0400 |
On Sun, 2003-07-20 at 18:24, Barry Roomberg wrote: > Code that is released GPL can't have the GPL revoked. But an IDENTICAL > copy of the code sitting on the original author's hard disk is still > controlled by the author. That particular piece of code is not GPL. Yes, this is the distinction between a "release of an item" and the "item" itself. More on this below. > Both copies have have "Copyright" associated with them. But the GPLed > copy has a license that is associated with it as it leaves the author's > possession and is given to someone else. As long as the recipient abides > by the license, the GPL is intact and "controlls" what is allowed to happen > to the code. If the recipient breaks the GPL, then the Copyright is still > intact and then dramatically limits what the person can do with the code. > They really are not allowed to do anything with it without the author's > permission. The license only restricts the parties receiving the licensed item, not the copyright holder. The copyright holder is still free to do what they wish with their item. However, if they have accepted copyrighted code (patches, modules, new features, etc.) from the outside and incorporated it into said item, they are no longer the sole copyright holder for the item and therefore must obtain a consensus before making any license changes. Everyone knows about SourceForge, right? Well, SourceForge got around this by requiring that all submissions be copyrighted to them and contacted all previous submitters to obtain the copyrights to their works, as well. Then, in version 3.0, they modified the license to proprietary once they were sole copyright owners. Versions prior to 3.0 are still open source and in fact have been used as the basis for both the Savannah and GForge software indeces. > The original (not GPLed copy) can then be used for any purpose that the > copyright holder wishes, with any license at all. > > The issue here beomes a possibility of forking. If another developer takes > the GPL code, makes a modification, and makes it public, and sends the > changes back to the original author, there is problem for the original author. Technically, if I were to release some code and then you submitted a patch to me, which I accepted, I could no longer revoke the GPL license from that __COMPLETE WORK__ without your permission, as well. However, in the cases I mentioned in my previous email, the companies in question did indeed hold copyright over the COMPLETE WORK and thus had to consult with no one before changing the license. Code that was previously released as GPL could have been forked at a later date by anyone as the license still applied to it; the revocation can only apply to future releases of an item. > Outside coders picked up the last released GPL version. This now > competes with the commercial version. This is what I was talking about above. Sistina could not revoke the GPL on code its already released, only on the code itself (meaning, future releases). It would be technically impossible to "recall" already released GPL code, anyway (esp. in this day and age of Google-caching). > Last time I looked, the LVM in the kernel is written by Sistina corp. > I won't use it. I wasn't aware of this story. I know LVM is one of the big show-stoppers for enterprise adoption of Linux right now; it would be a shame to have a socially-unconscious company in charge of such a critical feature of the kernel. -- Tobias DiPasquale 88FA 30C9 1E63 CFE2 CBD8 37C4 DA1C E2BF 1D26 F036 http://cbcg.net/ Attachment:
signature.asc
|
|