I like that solution, since it not only avoids the license conflict but
would keep the descriptions up-to-date automatically.

Now, to dig even deeper into the MediaWiki API docs...

On Tue, Feb 19, 2013 at 12:39 AM, Ben Companjen <[email protected]>wrote:

> +1 to that. I had been thinking along the same lines, but didn't write it
> down.
> Of course it isn't a solution for all the works that already contain
> Wikipedia content.
>
> But even more important I think is that the licence that applies is made
> very clear. If it is CC0, add a note that copyrighted content cannot be
> added.
> Can you shed a light on that, Anand?
>
> Ben
> On Feb 19, 2013 9:29 AM, "Anand Chitipothu" <[email protected]> wrote:
>
>> Another solution is to add wikipedia ID to work page and fetch the
>> wikipedia description while rendering the page.
>>
>> That way the content will not be part of OL data. No licensing issues.
>>
>> Anand
>>
>> On 13-Feb-2013, at 2:31 AM, Ben Companjen wrote:
>>
>> I totally agree with Tom here.
>>
>> I have always believed OL data was released under CC0 (and I hope it will
>> (continue to) be), mostly because of the message in the edit form.
>> Theoretically me agreeing to share info with OL under CC0 doesn't mean that
>> OL must/will share it as CC0 again, but it is the most prominent licencing
>> agreement. Let's be clear about it, soon.
>>
>> I confess that I too have taken a bit of Wikipedia and put in an author
>> profile (with attribution) - I didn't copy a whole page, but I don't know
>> if I can call it "fair use" either.
>>
>> Ben
>>
>>
>> On 12 February 2013 21:07, Tom Morris <[email protected]> wrote:
>>
>>> I was repeating CC0 without checking (partly because I thought I'd heard
>>> that before).  Actually, the edit page *DOES* say CC0 "By saving a change
>>> to this wiki, you agree that your contribution is given freely to the world
>>> under CC0. Yippee!"  What it should probably also say is "and you have the
>>> rights to make this grant."  However, that's in conflict with the license
>>> statement below...
>>>
>>> On Tue, Feb 12, 2013 at 12:36 PM, Karen Coyle <[email protected]> wrote:
>>>
>>>>
>>>> The OL "license terms" are the IA "license terms" -- thus:
>>>>    http://archive.org/about/terms.php
>>>
>>>
>>> The "license" which is linked to from the site itself is at
>>> http://openlibrary.org/developers/licensing and it says, in part:
>>>
>>> "The Internet Archive does not assert any new copyright or other
>>> proprietary rights over any of the material in the Open Library database.
>>> There may be existing rights issues on some contributions and in some
>>> jurisdictions. "
>>>
>>> which is, quite frankly, a huge cop out.  That effectively says that no
>>> one can use the information because you have no idea what  rights and
>>> restrictions apply.  The only thing I can guess is that they either didn't
>>> have the CC0 requirement in the early days or they imported data of dubious
>>> provenance early on.
>>>
>>> The only reasonable way to run a shared database like this is the way
>>> Wikipedia, Freebase, etc do it.  That is, decide what your license is going
>>> to be, then only accept contributions which are acceptable under that
>>> license.  People will still break the rules, but at least you've made an
>>> effort and are covered.
>>>
>>>
>>>> It is not CC0, because most of the info in OL is not owned by OL/IA.
>>>> Only a rights owner can assign a CC license.
>>>>
>>>> OL already pulls in descriptions from Wikipedia and sources them:
>>>>    http://openlibrary.org/authors/OL29497A/Herman_Melville
>>>
>>>
>>> That was added by hand by user Winnie
>>> http://openlibrary.org/authors/OL29497A/Herman_Melville?m=history
>>>
>>>
>>>> I believe that this fulfills the "attribution - share alike" of
>>>> Wikipedia.
>>>>
>>>
>>> I disagree because there is no requirement on downstream consumers that
>>> they also license that text under CC-BY-SA.  If that were allowed you could
>>> do "license washing" by taking licensed text from Wikipedia, pouring it
>>> into Open Library and then taking the OL dump and claiming that there was
>>> no license attached.
>>>
>>> Either the entire database needs a single homogeneous license that
>>> humans can deal with or there needs to be machine readable licensing
>>> information attached to subsets of the data.
>>>
>>> The "we don't know what the license is and you'll need to figure it out
>>> on your own" is useless from the point of view of someone who wants to
>>> reuse the information.
>>>
>>> Tom
>>>
>>>
>>>>
>>>> kc
>>>>
>>>> On 2/12/13 11:16 AM, Tom Morris wrote:
>>>> > On Tue, Feb 12, 2013 at 1:58 AM, John Shutt <[email protected]
>>>> > <mailto:[email protected]>> wrote:
>>>> >
>>>> >
>>>> >     I noticed that a lot of books on Open Library don't have
>>>> >     descriptions, so I've started working on NondescriptBot
>>>> >     <https://github.com/pemulis/nondescript-bot>, which would make it
>>>> >     easy to pull book summaries from Wikipedia, reformat them, and add
>>>> >     them to Open Library. I haven't written any code yet (except for
>>>> the
>>>> >     login, which was adapted from IdentifierBot
>>>> >     <
>>>> https://github.com/dmontalvo/IdentifierBot/blob/master/fastadder.py>),
>>>> >     but you can see the basic outline in the comments
>>>> >     <
>>>> https://github.com/pemulis/nondescript-bot/blob/master/nondescriptbot.py
>>>> >.
>>>> >
>>>> >     Before I go any further, I want to see if anyone knows if this bot
>>>> >     would be okay from a licensing standpoint. Wikipedia entries are
>>>> >     licensed under CC-BY-SA
>>>> >     <
>>>> https://en.wikipedia.org/wiki/Wikipedia:Text_of_Creative_Commons_Attribution-ShareAlike_3.0_Unported_License
>>>> >,
>>>> >     which requires attribution, while Open Library content is supposed
>>>> >     to be licensed under CC0
>>>> >     <https://creativecommons.org/publicdomain/zero/1.0/>, which
>>>> waives
>>>> >     all rights. It's trivial to put a CC-BY-SA disclaimer at the
>>>> bottom
>>>> >     of a description, but I don't know if it's permitted to add
>>>> content
>>>> >     to OL that falls under that license.
>>>> >
>>>> >
>>>> > No, you can't use a copyrightable amount of text which is CC-BY-SA
>>>> > licensed on a CC0 site.  Part of the license is that you need to
>>>> enforce
>>>> > it for sub-licensees & reusers, which there's no way to do with a CC0
>>>> work.
>>>> >
>>>> > You could paraphrase or reword the description, but that's clearly
>>>> not a
>>>> > job for a bot.  You could also extract a small enough amount of text
>>>> > that it would fall under "fair use" guidelines and then link back to
>>>> > Wikipedia for the full text.  If nothing else, links to Wikipedia
>>>> would
>>>> > be useful (provided that their reliable).
>>>> >
>>>> >     Assuming this bot is allowed, it would be awesome to get advice
>>>> and
>>>> >     pull requests from other developers! I'm coming into this project
>>>> >     with very limited knowledge of Python, so I'm sure there will be
>>>> >     plenty of places where my code could be improved.
>>>> >
>>>> >
>>>> > I'm happy to help with Python as well as OpenLibrary or Wikipedia
>>>> APIs.
>>>> >
>>>> > Tom
>>>> >
>>>>
>>>
>>> _______________________________________________
>>> Ol-tech mailing list
>>> [email protected]
>>> http://mail.archive.org/cgi-bin/mailman/listinfo/ol-tech
>>> To unsubscribe from this mailing list, send email to
>>> [email protected]
>>>
>>>
>> _______________________________________________
>> Ol-tech mailing list
>> [email protected]
>> http://mail.archive.org/cgi-bin/mailman/listinfo/ol-tech
>> To unsubscribe from this mailing list, send email to
>> [email protected]
>>
>>
>>
>> _______________________________________________
>> Ol-tech mailing list
>> [email protected]
>> http://mail.archive.org/cgi-bin/mailman/listinfo/ol-tech
>> To unsubscribe from this mailing list, send email to
>> [email protected]
>>
>>
> _______________________________________________
> Ol-tech mailing list
> [email protected]
> http://mail.archive.org/cgi-bin/mailman/listinfo/ol-tech
> To unsubscribe from this mailing list, send email to
> [email protected]
>
>
_______________________________________________
Ol-tech mailing list
[email protected]
http://mail.archive.org/cgi-bin/mailman/listinfo/ol-tech
To unsubscribe from this mailing list, send email to 
[email protected]

Reply via email to