Date: Tue, 18 Feb 1997 13:02:45 -0500 (EST) From: "John S. Dyson" <toor@dyson.iquest.net> To: obrien@NUXI.com (David O'Brien) Cc: chat@freebsd.org Subject: Re: GPL Message-ID: <199702181802.NAA04246@dyson.iquest.net> In-Reply-To: <19970218020731.GM57190@dragon.nuxi.com> from "David O'Brien" at Feb 18, 97 02:07:31 am
next in thread | previous in thread | raw e-mail | index | archive | help
> J Wunsch writes: > > As David O'Brien wrote: > > > Say someone has written fooquix and from version 0.01 to 0.49 it was > > > GPL'ed. Then they decided they wanted to make some $$$ from it. So the > > > next release (say 0.50) was binary only. Now obiviously 0.50 is derived > > > work based on the GPL'ed code of 0.49. > > > > > > Is this allowable, or once software is under GLP it stays there? > > > > It gets problematic for everything that other contributed under the > > terms of GPL from version 0.01 through 0.49. Either they all agree in > > the new copyright, or you've got a problem. :) > > So in theory, the author of every little patch needs to be consulted? > Hum... :-) > (This is not an opinion of -core, and none of my opinions unless specifically stated are that of -core or any other body, except mine :-)). That is where the original author looses -- he started a waterfall of GPL, that he ends up being in the same boat as the other contributors (and looses the freedom to make other arrangements for the derived works.) Everyone is pretty much in the same boat with GPL when it becomes net-owned. It is mostly the big changes and patches that need to be cleared the the other authors for a change in licensing terms. Imagine if you write a kernel, for example, and declare it to be under GPL. Imagine also that lots of other people add to that kernel, under GPL. You (the original owner) would have difficulty making the parallel licensing terms for other use, without either chasing down each individual author and get their okay, or stripping out the other GPLed code. Note, however, that alot of people use GPL for philosophical reasons. You (the author of that kernel) are at their mercy, and likely out of luck. The only way that the author of the kernel can keep control is to get written assignment of the code. Just because you might add the code yourself doesn't free the contributed code from GPL. (In that case, you are just acting as a repository maintainer.) There are cases where GPL can be useful, and seems to me to be a reasonable usage of it. Imagine that you have written a cool set of drivers for MSDOS machines. You are also agressively maintaining the source code, and accept modifications with written assignment of ownership or copyright. You, the author, maintains control so that you can also distribute under other license terms. In this case, GPL is much better than Shareware, because you can safely give away the source AND can accept payment to relieve the customer of redistribution encumberances. The original author of the package might even give such relief to those who contribute significantly to the package. There are not-so-subtile ways to subvert some of the intent of GPL, and frankly my opinion is that one should try to follow the INTENT of the original author. If an author really meant for the source code and all of it's derivatives to be redistributed, I think that it should be done. GPL is not such a bad thing that one should ignore it (and perhaps suffer legal consequences.) GPL should be applied, understood, and followed carefully like any other license. Note that it is the INTENT and desire for the code that I add to BSD be used in any ethical and legal way that the user desires. I don't feel the need for someone to have to disclose their derived works, but it is in the spirit of good will that those who use the code, when it wouldn't hurt their business, will contribute back to the original distribution. In fact, it might even make their code maintenence easier. John
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199702181802.NAA04246>