Re: [Fwd: [PATCH 8/12] LED: Add LED device support for ixp4xx devices]

Re: [Fwd: [PATCH 8/12] LED: Add LED device support for ixp4xx devices]

[Andrew Morton]

Richard Purdie <rpurdie@rpsys.net> wrote:
>
> +MODULE_AUTHOR("John Bowler <jbowler@acm.org>");
> +MODULE_DESCRIPTION("IXP4XX GPIO LED driver");
> +MODULE_LICENSE("MIT");

MIT license is unusual.  There's one other file in the kernel which uses it
and that's down in MTD where nobody dares look.

I don't know whether MIT is GPL-compatible-for-kernel-purposes or not.  Help.

Re: [Fwd: [PATCH 8/12] LED: Add LED device support for ixp4xx devices]

[Felipe W Damasio]

    Hi,

Andrew Morton wrote:

> MIT license is unusual. There's one other file in the kernel which uses it
>
>and that's down in MTD where nobody dares look.
>
>I don't know whether MIT is GPL-compatible-for-kernel-purposes or not.  Help.
>

    Well, checking FSF's licenses page, there's an entry about a license
which *might* be the same as this:

"X11 License

This is a simple, permissive non-copyleft free software license,
compatible with the GNU GPL.
...
This license is sometimes called the "MIT" license, but that term is
misleading, since MIT has used many licenses for software."

<http://www.fsf.org/licensing/licenses/index_html/view?searchterm=software%20licenses>

    So my unreliable opinion is yes: This is
GPL-compatible-for-kernel-purposes.

    Cheers,

Felipe Damasio

Re: [Fwd: [PATCH 8/12] LED: Add LED device support for ixp4xx devices]

[Linus Torvalds]

On Sun, 5 Feb 2006, Andrew Morton wrote:
>
> Richard Purdie <rpurdie@rpsys.net> wrote:
> >
> > +MODULE_AUTHOR("John Bowler <jbowler@acm.org>");
> > +MODULE_DESCRIPTION("IXP4XX GPIO LED driver");
> > +MODULE_LICENSE("MIT");
> 
> MIT license is unusual.  There's one other file in the kernel which uses it
> and that's down in MTD where nobody dares look.
> 
> I don't know whether MIT is GPL-compatible-for-kernel-purposes or not.  Help.

The FSF considers the normal MIT license (the original X license) to be 
GPL-compatible, but suggests not using the name (because there have been 
multiple licenses used at MIT).

The real problem is that it will taint the kernel, because the kernel 
won't _recognize_ it as being GPL-compatible. See module.c: function
"license_is_gpl_compatible()". 

For that reason, if no other, I would suggest changing it to

	"Dual MIT/GPL"

and adding that to the list of recognized licenses (the "Dual xxx/GPL" 
because by the time it is linked into the kernel it _will_ be GPL for any 
license that is compatible with the GPL).

Otherwise most kernel developers will discard any bug reports due to the 
oops showing the kernel as "tainted".

		Linus

Re: [Fwd: [PATCH 8/12] LED: Add LED device support for ixp4xx devices]

[Arjan van de Ven]

On Sun, 2006-02-05 at 19:20 -0800, Andrew Morton wrote:
> Richard Purdie <rpurdie@rpsys.net> wrote:
> >
> > +MODULE_AUTHOR("John Bowler <jbowler@acm.org>");
> > +MODULE_DESCRIPTION("IXP4XX GPIO LED driver");
> > +MODULE_LICENSE("MIT");
> 
> MIT license is unusual.  There's one other file in the kernel which uses it
> and that's down in MTD where nobody dares look.
> 
> I don't know whether MIT is GPL-compatible-for-kernel-purposes or not.  Help.

would be really nice if the author would at least dual license it under
the GPL as well (and thus also granting the patent rights if any); in
which case it's a clear "GPL when used with linux, but optionally MIT if
used outside linux")

Re: [Fwd: [PATCH 8/12] LED: Add LED device support for ixp4xx devices]

[Dan McDonald]

On Mon, 6 Feb 2006, Arjan van de Ven wrote:

> On Sun, 2006-02-05 at 19:20 -0800, Andrew Morton wrote:
>> Richard Purdie <rpurdie@rpsys.net> wrote:
>>>
>>> +MODULE_AUTHOR("John Bowler <jbowler@acm.org>");
>>> +MODULE_DESCRIPTION("IXP4XX GPIO LED driver");
>>> +MODULE_LICENSE("MIT");
>>
>> MIT license is unusual.  There's one other file in the kernel which uses it
>> and that's down in MTD where nobody dares look.
>>
>> I don't know whether MIT is GPL-compatible-for-kernel-purposes or not.  Help.
>
> would be really nice if the author would at least dual license it under
> the GPL as well (and thus also granting the patent rights if any)

It would be really nice, but that doesn't make the rights to any patents 
granted unless 'GPLv2 or any later version' is explicitly specified. The 
default GPL license is v2 and only v2. There was a bug thread about this 
on LKML earlier.


Dan

Re: [Fwd: [PATCH 8/12] LED: Add LED device support for ixp4xx devices]

[Arjan van de Ven]

On Sun, 2006-02-05 at 23:00 -0800, Dan McDonald wrote:
> 
> On Mon, 6 Feb 2006, Arjan van de Ven wrote:
> 
> > On Sun, 2006-02-05 at 19:20 -0800, Andrew Morton wrote:
> >> Richard Purdie <rpurdie@rpsys.net> wrote:
> >>>
> >>> +MODULE_AUTHOR("John Bowler <jbowler@acm.org>");
> >>> +MODULE_DESCRIPTION("IXP4XX GPIO LED driver");
> >>> +MODULE_LICENSE("MIT");
> >>
> >> MIT license is unusual.  There's one other file in the kernel which uses it
> >> and that's down in MTD where nobody dares look.
> >>
> >> I don't know whether MIT is GPL-compatible-for-kernel-purposes or not.  Help.
> >
> > would be really nice if the author would at least dual license it under
> > the GPL as well (and thus also granting the patent rights if any)
> 
> It would be really nice, but that doesn't make the rights to any patents 
> granted unless 'GPLv2 or any later version' is explicitly specified. The 
> default GPL license is v2 and only v2. There was a bug thread about this 
> on LKML earlier.

wrong. v2 already is a statement that the author doesn't have patents on
the code, or alternatively that the author has granted you the patent
rights and will not sue you for using the code. it's also a statement
that the author doesn't have patent licenses from 3rd parties that
he/she cannot sublicense.

v3 clarifies and expands on this most likely, but v2 already offers
basic patent trap protection.

RE: [Fwd: [PATCH 8/12] LED: Add LED device support for ixp4xx devices]

[John Bowler]

From: Andrew Morton [mailto:akpm@osdl.org]
>MIT license is unusual.  There's one other file in the kernel which uses it
>and that's down in MTD where nobody dares look.

Well, I look at the MTD code a lot... That's why I used it (i.e. that was the
existence proof I found that it's fine ;-)

>I don't know whether MIT is GPL-compatible-for-kernel-purposes or not.  Help.

Nothing it that license precludes the code being redistributed under the
GPL.  (It is a verbatim copy of
http://www.opensource.org/licenses/mit-license.php).

>MODULE_LICENSE("Dual MIT/GPL");

Is fine with me, indeed, given the license text, I don't believe I retain
any rights to *prevent* such a change (and it was certainly not my intent
to prevent redistribution under a more restrictive license).  Using 'MIT'
or substituting some more specific tag must be fine for the same reason
(it's a name of a license, not a license itself).

John Bowler <jbowler@acm.org>