[tei-council] Fwd: TEI licensing issues

[Martin Holmes]

HI Piotr,

I don't think TEI is too big to be abused; I'm just having trouble 
imagining what kind of abuse we could be subject to. What could someone 
do with our schemas, ODDs and stylesheets that could constitute abuse? 
All of our code is XML data and non-compilable scripting, as far as I 
can see; it's not as though someone could steal it, compile a new 
application, and release it as a commercial app without mentioning the 
TEI. There would be no point; the only value in using our stuff is that 
you're buying into the long-term viability and community support that 
comes along with using TEI.

On the other hand, if people CAN build viable commercial businesses 
around our schemas and stylesheets, then great -- that's good for us. 
And whatever licensing we choose, there's nothing we can do to prevent 
people from writing (say) a commercial C++ application which processes 
TEI files, even using TEI stylesheets, without opening up their code; 
our licensing would only apply to our libraries, and those would be 
external to any compiled app, so they wouldn't infect the commercial C++ 
code with GPL-itis anyway.

The absolute worst-case scenario I can imagine is that someone takes the 
text of the Guidelines, tweaks it a bit, makes a book out of it, and 
sells it. There's absolutely no value in the book unless it mentions who 
and what the TEI is/are, and there's no marketability in the resulting 
product unless they're supplying some additional value somehow -- in 
which case, they're helping to propagate and publicize our work anyway.

Can anyone think of any realistic abuse scenario, if we put everything 
in the public domain?

Cheers,
Martin

On 11-09-12 11:02 AM, Piotr Bański wrote:
> Hi Martin,
>
> I'm a bit afraid that giving up copyright (that's what public domain
> means) would invite abuse, and I'm not convinced that the TEI is widely
> known everywhere. For example, looking at mine and related work on
> applying TEI to linguistic markup (and yes, I am aware of the roots and
> the history of the TEI), I recall the results of the survey that Damir
> Ćavar did before his panel presentation at the last TEI-MM, where he
> quoted language-technology people going "TEI? Does that still exist?!".
> Andreas Witt has recently experienced this too. I wish that this is
> going to change, but so far, the picture of the TEI being a giant beyond
> abuse is well, just a picture :-)
>
> And one more issue, pretty grim, is that relicensing appears to require
> agreement of the contributors, who contributed under a different
> license. XEmacs practically died from that.
>
> Thirdly, GNU GPL is a good and widely supported license, it would be a
> pity to give it up.
>
> Best,
>
>    P.
>
> On 12/09/11 19:02, Martin Holmes wrote:
>> I've added links to everything specific, so not including e.g. "LGPL"
>> without a version number.
>>
>> I'm honestly wondering why we should be looking at such a broad range of
>> licences. Could we consider the possibility of putting absolutely
>> everything in the public domain (CC0)? Do we have anything to lose,
>> other than attribution? And it seems to me that anyone using TEI stuff
>> for any project would most likely be doing so _because_ it's a
>> widely-known and broadly-supported standard, so they'd gain nothing from
>> avoiding attribution.
>>
>> Cheers,
>> Martin
>>
>> On 11-09-12 09:18 AM, James Cummings wrote:
>>> On 11/09/11 09:21, Laurent Romary wrote:
>>>> Before we move to other issues here. Could someone with a good
>>>> knowledge in the domain of licenses summarizes the options,
>>>> considering the various type of objects we have to deal with
>>>> (ODD, schemas, doc, xslt...). We could have these put together
>>>> in the wiki, before we make a comprehensive proposal to the
>>>> board. Laurent
>>>
>>>
>>> I don't think I qualify as such, but since no one else has done
>>> so, I have done a first-draft at a table of objects the TEI-C
>>> produces, Current licences, and proposed Licences.
>>>
>>> http://wiki.tei-c.org/index.php/Council-licensing
>>>
>>> Warning: I may have got stuff entirely wrong, please feel free to
>>> correct, bonus points for adding in links to the licenses themselves.
>>>
>>> Feel free to continue debate here and/or on discussion tab on the
>>> wiki.
>>>
>>> Currently the stylesheets take the content of<availability>   for
>>> the output schemas when generated by Roma. I don't think that we
>>> should license the roma-output schemas differently than the ODDs
>>> themselves (and not sure how we would store different licences
>>> for different outputs in availability).
>>>
>>> My desire would be to consistently license across the board as
>>> much as possible, with the most open licences possible.
>>> Multi-licensing is fine if it makes it easier for people (like
>>> George) to use and doesn't confuse users.
>>>
>>> -James
>>>
>>
>
>

-- 
Martin Holmes
University of Victoria Humanities Computing and Media Centre
(mholmes at uvic.ca)