1 |
On Sun, Feb 22, 2004 at 03:50:08PM -0600, Jay Maynard wrote: |
2 |
> <grumble> I promised myself I'd stay the heck out of this one...but I just |
3 |
> can't let Jon's comment slide unrebutted, and so I'll somment on Drake's |
4 |
> message while I'm here. |
5 |
> |
6 |
> On Sun, Feb 22, 2004 at 04:33:03PM -0500, Jon Portnoy wrote: |
7 |
> > On Sun, Feb 22, 2004 at 12:50:34PM -0800, Drake Wyrm wrote: |
8 |
> > > The problem is that the XFree license and the GPL are now |
9 |
> > > incompatible. For smooth progression of overall Linux development, one |
10 |
> > > of these licenses must now change. This does not necessarily imply that |
11 |
> > > the one which changed most recently needs to change back. So much focus |
12 |
> > > has been applied to the "offending" portion of the new XFree license; |
13 |
> > > perhaps we should lend the same critical eye to our beloved GPL. One |
14 |
> > > possible analogy for the situation is that the XFree license demonstrated |
15 |
> > > a bug in the GPL. |
16 |
> |
17 |
> I've been arguing for over a decade that the GPL is buggy. Its deliberate |
18 |
> incompatibility with other open source licenses is one symptom of that. |
19 |
> |
20 |
|
21 |
You mean "deliberate incompatibility with licenses imposing additional |
22 |
restrictions." |
23 |
|
24 |
According to the FSF it's perfectly compatible with the following |
25 |
licenses: |
26 |
http://www.gnu.org/licenses/license-list.html#TOCGPLCompatibleLicenses |
27 |
|
28 |
I license my code under the GPL because I absolutely agree with |
29 |
everything the GPL states. If I didn't, I would use a different |
30 |
license. It's that simple: if you don't want your code to be strictly |
31 |
free software, don't use a license that makes your code strictly free |
32 |
software. You can spin it any way you want, but the copyright owner |
33 |
picks the license, and they pick the license they _want to use_. If |
34 |
people are picking licenses they haven't read or don't understand, that |
35 |
is not a bug in the license but rather in the copyright owner. |
36 |
|
37 |
> > I think it's perfectly reasonable for a license to insist on people not |
38 |
> > placing further restrictions on the code. This is what prevents people |
39 |
> > from making free software nonfree. |
40 |
> |
41 |
> NO!!! |
42 |
> |
43 |
> You CANNOT make existing freely available software non-freely-available. |
44 |
> Period. It cannot be done. Copyright laws will not allow it. Regardless of |
45 |
> what anyone does with a piece of code, no matter what license itt's under, |
46 |
> that code cannot be made other than freely available. Non-viral licenses |
47 |
> allow people to make THEIR OWN WORK other than freely available, but this is |
48 |
> exactly as it should be. |
49 |
|
50 |
I have an app 'foo' under the GPL. I link it to 'bar' under a |
51 |
proprietary license. Because they're now linked, foo is a derived work |
52 |
of bar. If I then redistribute those in a binary set, I have just |
53 |
violated the license on foo. |
54 |
|
55 |
Yes, that piece of GPL-licensed code cannot be made nonfree. When |
56 |
they're linked, they can. That's why additional restrictions are a |
57 |
problem. If we distribute a GRP set with GPL apps linked to libs with |
58 |
additional restrictions, we are violating the license terms set down by |
59 |
any of the GPL apps that're linked. Something under the BSD license |
60 |
*can* have additional restrictions, including proprietary restrictions. |
61 |
|
62 |
If you don't like the GPL, don't write your code under it; don't tell |
63 |
those of us who license their code under the GPL *because we like it* |
64 |
that we're somehow wrong. If you don't like a core component being under |
65 |
a license you don't like, rewrite the component from scratch under a |
66 |
different license. That's what the freedesktop people are doing with X. |
67 |
|
68 |
You are absolutely right; copyright law will not allow you to violate |
69 |
copyright licenses. The GPL is a license that forbids additional |
70 |
restrictions, so it's illegal to try to impose additional restrictions. |
71 |
The BSD license is a license that _doesn't_ forbid additional |
72 |
restrictions, so it's _not_ illegal to impose additional restrictions. |
73 |
Did you misunderstand what I was saying? I was saying that what keeps |
74 |
free software from being redistributed under nonfree terms is the "no |
75 |
additional restrictions" clause (read section 6 of the GPL; the only |
76 |
additional restrictions allowed are specified by section 8, which is |
77 |
very limited in scope). The BSD license, on the other hand, does not set |
78 |
down very strict terms and permits source/binary redistribution with |
79 |
additional restrictions. |
80 |
|
81 |
As for "THEIR OWN WORK," as you put it, the copyright holder can |
82 |
relicense GPL-licensed code under another license any time they please. |
83 |
I can release an application I write under the GPL, again under the BSD |
84 |
license, and again under a proprietary license. Lots of people use dual |
85 |
license approaches. Nobody can *retroactively* change a license, of |
86 |
course, but that's an entirely different subject. I think you're a |
87 |
little confused about this. A license does not place restrictions on the |
88 |
actual copyright holder; the actual copyright holder can do whatever |
89 |
they please by virtue of being the copyright holder. They can grant |
90 |
other people permission to do whatever they please by distributing it to |
91 |
that person under a totally different license. |
92 |
|
93 |
Does that clarify things for you? Let me know if it doesn't; I'm willing |
94 |
to elaborate further. |
95 |
|
96 |
> |
97 |
> > The BSD license approach is not at all intended to keep free software |
98 |
> > free; |
99 |
> |
100 |
> Wrong. If this were not the case, we wouldn't have the BSDs today. Sun would |
101 |
> have made them proprietary a long time ago. |
102 |
|
103 |
You misunderstand what I mean. See above; you cannot _retroactively_ |
104 |
change a license on anything, but the BSD license permits other people |
105 |
to take your code and use it in a proprietary project. In other words, |
106 |
that code can be made nonfree. This is not the purpose of the GPL; your |
107 |
entire issue seems to be that the GPL is a license intended to |
108 |
perpetuate freedom, whereas the BSD license doesn't have the same |
109 |
ideals involved. |
110 |
|
111 |
> |
112 |
> > it's intended to let people create proprietary software using free |
113 |
> > software code. |
114 |
> |
115 |
> Thus guaranteeing the maximum freedom for everyone. This is the true spirit |
116 |
> of free software, and why I think the FSF has it all wrong - and is being |
117 |
> intellectually dishonest in calling their position "free". |
118 |
|
119 |
You mean "this guarantees the freedom to deny freedom." I do not want my |
120 |
code used in such a way, which is why I do not use the BSD license for |
121 |
anything nontrivial. If someone *does* want their code used in such a |
122 |
way, wouldn't they use a BSD-style license? Isn't it fair to let the |
123 |
copyright holder make that decision? Isn't that an improvement on Jay |
124 |
Maynard making that decision for them? |
125 |
|
126 |
If you want your code used in proprietary products, use a license that |
127 |
permits it to be used in proprietary products. If you (like me) do not |
128 |
want that, do not use a license that permits it to be used in |
129 |
proprietary products. If you do not want your code used in free software |
130 |
products, use a license that does not permit it to be used in free |
131 |
software products. And so on. |
132 |
|
133 |
What you seem to actually be suggesting is that everyone should use |
134 |
licenses that do things they don't want them to do because that's what |
135 |
*you* believe. |
136 |
|
137 |
> |
138 |
> > What you're actually suggesting is reforming the way the entire free |
139 |
> > software community feels about freedom and licensing. |
140 |
> |
141 |
> Speak for yourself, pilgrim. |
142 |
> |
143 |
> > Have you really thought this through? |
144 |
> |
145 |
> More than the frothing Stallmanites have. |
146 |
> |
147 |
|
148 |
Please tell me I didn't just waste all that time explaining copyright |
149 |
and licensing basics for a troll. |
150 |
|
151 |
> > Basically you're stating that it should be acceptable to place |
152 |
> > proprietary restrictions on what was at one point free software. |
153 |
> |
154 |
> No, he's not, because you cannot change the restrictions under which a piece |
155 |
> of code is available. You can place additional restrictions on your code, |
156 |
> but the original code is now and forever available under the same terms as |
157 |
> it was when the proprietary fork was taken. |
158 |
> |
159 |
|
160 |
Please study up on derived works and linking so you can understand what |
161 |
I'm saying. |
162 |
|
163 |
> > If people wanted their code to be treated that way, they wouldn't license |
164 |
> > it under the GPL. |
165 |
> |
166 |
> Some folks don't have any choice. |
167 |
|
168 |
Really? Somebody's holding a gun to their head and forcing them to write |
169 |
code under the GPL? |
170 |
|
171 |
> Further, there are LOTS of folks out there that hold the same wrong belief |
172 |
> you do, and choose the GPL in light of that error. |
173 |
|
174 |
Sorry, none of my beliefs are wrong. Unfortunately, your understanding |
175 |
of copyright law and licensing seems to be very much lacking, which |
176 |
makes it very difficult to hold an intelligent discussing about real |
177 |
world licensing issues. |
178 |
|
179 |
> |
180 |
> > Nobody's forcing anybody to put their code under one |
181 |
> > license or another. |
182 |
> |
183 |
> It is the explicit aim of the GPL to have all software, everywhere, under |
184 |
> any conditions, licensed under their anticommercial terms. |
185 |
|
186 |
That's just silly. |
187 |
|
188 |
Perhaps you didn't mean the GPL; if you meant "the explicit aim of the |
189 |
FSF" you may be closer to the mark. However, they are not forcing anyone |
190 |
to do anything. They are not forcing you to use the GPL on your code, |
191 |
they are not forcing you to link to GPL-licensed code, they are not |
192 |
forcing you to use GPL-licensed applications. What they *are* doing is |
193 |
attempting to explain to people why they feel that free software is |
194 |
better for society. They can't _force_ you to do anything. |
195 |
|
196 |
> |
197 |
> > I would not use a license that allows additional restrictions for most of |
198 |
> > what I do; I only license relatively trivial things under BSD-style |
199 |
> > licenses. |
200 |
> |
201 |
> That is your prerogative. However, if you base that on the erroneous belief |
202 |
> that someone could take that software and make it no longer freely |
203 |
> available, I suggest you reexamine your position. |
204 |
|
205 |
I suggest you reexamine the issue of derived works in copyright law and |
206 |
then reexamine what I said, particularly the parts about licenses that |
207 |
allow additional restrictions versus those that don't and how that |
208 |
affects the perpetuation of freedom in software. |
209 |
|
210 |
By not allowing additional restrictions on things I write, I am |
211 |
preventing other people from taking my work, adding their work, and |
212 |
never letting other people benefit from the opportunity to take both my |
213 |
work and their work together, add their own work, and release that back. |
214 |
Under a BSD-style license, they can take my work, add their work, and |
215 |
release it under a proprietary license. If the copyright holder feels |
216 |
that's how they want their code to be used, that's the licensing scheme |
217 |
they'll use. |
218 |
Under a GPL-style license, they can take my work, add their work, and |
219 |
have to release their work too so that somebody else can take my work |
220 |
and their work, add their own work, and so on. That's how I want my code |
221 |
to be treated, so that's the licensing scheme I use. |
222 |
|
223 |
-- |
224 |
Jon Portnoy |
225 |
avenj/irc.freenode.net |
226 |
|
227 |
-- |
228 |
gentoo-dev@g.o mailing list |