License compatibility of MS-PL and MS-CL (Was: (RE: Groklaw's OSI item (was: When will CPAL actually be _used_?))

[Chris Travers]

Matthew Flaschen wrote:
> Chris Travers wrote:
>   
>> However, suppose we extend the BSDL slightly to
>> conform to the language of the GPL3 but place an equivalent restriction as
>> the MS-PL:
>>
>> * Any portion of this code which has not been substantively altered
>>     
>
> What does "substantially" mean?
>   
Suppose *substantially" to mean "sufficient for copyright protection?"
>   
>> must bear a notice stating that it is still original, is authored by <author>,
>> and has been publically licensed by the author under the terms of this
>> modified BSD license.  One may use additional comments to indicate ends of
>> such sections.
>>
>> Would such a BSDL now be incompatible with the GPL3?
>>     
>
> I don't see why.  It says the code is still "under the terms of this
> modified BSD license", but it doesn't say it can't be under another
> compatible license as well.
>
>   
>> Would this be allowed under the 7b legal/attribution notices?
>>
>> Now, what affect does this have in practice?  In my view, it identifies
>> which sections of the code can safely be extracted without infringing on the
>> GPL work's copyrights (since they don't extend to the excerpt).
>>     
>
> Maybe.  It depends what substantially means.  If "insubstantial"
> modifications still can have copyright protection, then they may be
> under GPLv3 only.  This effectively means the BSD notice is lying in
> saying all the code in the section is "still original" and under BSD.
>   

Again, suppose we rephrase to be "altered to a form sufficient for new 
copyright protections?"  (the intent is to exclude typo corrections from 
reverting the license since those only create nonliteral copies)
>   
>> Now, how does this *functionally* differ from the MS-PL?
>>     
>
> Only the MS-PL says "only under this license".  Even this modified BSD
> (that seems to run counter to the purposes of BSD) has nothing similar.
>   
Ok, so if we take my second hypothetical and state that you must 
identify code which you are only using by permission and don't have 
copyright claim to (itself), then how is this different?  In effect, you 
are just preserving the license on those chunks of code since you are 
announcing that they are copyrighted elsewhere and used with permission 
which has been granted to the public.

In short you are saying, "I don't own this code and cannot control what 
you do to it.  The author has given a public license to the code.  See 
the top of the file for details."

This is, of course, assuming that the MS-PL does not lay claim to any 
derivative or collected works and that "the software" means "the 
accompanying software."
>   
>> Therefore is the MS-PL really any less compatible with the GPL3 than
>>     
> the BSDL?
>
> Yes, because of the use of the word "only".
>   

I am not sure.  In this example, perhaps with some slight rewording, I 
have effectively prevented sublicensing of the code itself because I ask 
you to attribute copyrights to my code to me and advertise to the 
developer the public license I have granted.  I can do this without 
including "only" in the license, and I can do this using *only* 
attribution requirements and legal notices.

The reason is that my specific code is *only* under licenses I grant it 
anyway.  There is a difference between the author's license to the 
excerpt granted to the public and the project's license to the file as a 
whole granted to the public.  If I take an excerpt from the file as a 
whole, then I am bound by the authors' licenses for any code in that 
excerpt (and if there are more than one, probably the file as a while 
license as well).

The reason is that my code copyrights can't be enforced by those who 
alter it anyway.

The questions are:
1)  If I come up with something equivalent simply asserting my copyright 
and advertising my license on blocks of code still exclusively copyright 
by me, is it compatible with the GPL3?  (seems to be)
2)  Is this functionally different than the MS-PL (assuming that "the 
software" only applies to the source code of "the accompanying software" 
and not to copyrighted elements of that code present after modification).

Best Wishes,
Chris Travers

-------------- next part --------------
A non-text attachment was scrubbed...
Name: chris.vcf
Type: text/x-vcard
Size: 171 bytes
Desc: not available
URL: <http://lists.opensource.org/pipermail/license-discuss_lists.opensource.org/attachments/20070829/f79da36a/attachment.vcf>