[okfn-discuss] Annotating Open Images with licence and authorship to prevent copyfraud

Thu Aug 8 13:59:27 UTC 2013

On 8 August 2013 14:59, Peter Murray-Rust <pm286 at cam.ac.uk> wrote:

>
>
>
> On Thu, Aug 8, 2013 at 11:31 AM, Emanuil Tolev <emanuil.tolev at gmail.com>wrote:
>
>> Right, so the image metadata sections of image formats are insufficient
>> for this purpose?
>>
>
> Yes. If you print the paper (and many people do) the metadata gets lost
> immediately
>
I can see how people (well, ill-intentioned publishers I guess) might just
>> mangle that section as it's much easier to edit than actual pixels... I
>> also guess you considered this option, but am interested in why it was
>> rejected.
>>
>
> If you look at an electronic image can you tell just by looking at it
> whether it has any metadata? I can't. And most people don't have tools
> where they know how to look for the metadata of a whole number of formats?
> I don't.
>
> And when the PNG is incorporated into the PDF is the image metadata
> preserved? I have no idea but I doubt it.
>
> And when someone snips an image the MD disappears immediately.
>

I should've given the issue some more thought before suggesting that idea.
I agree, all excellent points.

> And what is wrong with adding the metadata visually to the image? We do it
> to the text. As a simple example it would immediately mean that when we
> re-use Wikimedia we automatically can see the most important provenance.
>

Nothing at all! I suppose I just had the initial informatics reaction
"nonono don't touch the data", but in this case that is what the users -
authors - want, so such tools would be welcome. It's a part of producing
the data. We take great care to produce a ton of metadata in the form of
references, this is just more metadata.

And yes, we definitely do it to text. I hadn't thought of it like that.

The result itself is actually easily achievable on almost any platform /
software stack ever made. This initiative might benefit from a brief, crisp
recipe-style project page where developers of authoring tools and even
people who host images (publishers..) are encouraged to offer this
functionality. E.g. "this is why this is good and how you implement this
feature with the 3 most popular image manipulation libs on X platform".

The end goal for the tool developers being something like tell TeXMaker "do
this to all images in this TeX project", or PLoS marking all images in a
CC-BY paper as such automatically (possibly at the time of submitting the
manuscript if at all possible, so the authors can see it there and then -
don't have to approve the automatic changes before publication happens). As
long as it's as seamless as possible and doesn't introduce clunky workflow
steps, people might take to doing this on a greater scale.

Of course a web service is a good idea, I'm just thinking of where
authoring happens.

Greetings,
Emanuil

P.S. Actually a LaTeX package is probably much better for the TeX case,
this way it isn't tied with an editor.

> In general non-visible metadata gets lost very rapidly in any
> transformation chain.
>
>>
>>
>> On 7 August 2013 23:30, Peter Murray-Rust <pm286 at cam.ac.uk> wrote:
>>
>>> Brilliant - this was exactly what I was indicating - you have done it
>>> much better.
>>>
>>>
>>>
>>
> --
> Peter Murray-Rust
> Reader in Molecular Informatics
> Unilever Centre, Dep. Of Chemistry
> University of Cambridge
> CB2 1EW, UK
> +44-1223-763069
>
> _______________________________________________
> okfn-discuss mailing list
> okfn-discuss at lists.okfn.org
> http://lists.okfn.org/mailman/listinfo/okfn-discuss
> Unsubscribe: http://lists.okfn.org/mailman/options/okfn-discuss
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.okfn.org/pipermail/okfn-discuss/attachments/20130808/fc2e328b/attachment.html>