Hi Wikimedia Commons folks!
Metadata Games (MG) is a FOSS project that we've been working on at
Tiltfactor Lab at Dartmouth College. It's a bit of an experimental
project using online games to help with the collection of metadata for
images in various collections (libraries, archives, etc...).
There's a bit of basic information up on the Tiltfactor website about
the background and main goals of the project:
And (because talk is cheap :-), source code is up on Gitorious here:
If you promise not to explode our server too much, you can point your
browser over to http://metadatagames.com and see a version of the
software in action. Right now that's pointing to a test install with
some pictures from Dartmouth's archives. We have some great
black-and-white images from old Winter Carnivals, as well as some more
modern color photographs from around the campus. We're mostly done
collecting data with that particular test, so feel free to play around
with the images and games we have up there.
The MG system currently has a couple of different games for single and
multi-player tagging of images. In the next couple of months we're
hoping to add support for more media types (including audio and video)
as well as making some big improvements to the backend of the system
so that we can scale-up for big installs.
We're currently collaborating with a couple of different groups
including the Rauner Library here at Dartmouth, and we're eager to see
more groups benefit from tagging media with MG. Sam Klein pinged me
about working with Wikimedia, and we'd definitely be excited to
collaborate on improvements or expansions to the current system. The
simplest way to use MG would just be to funnel images from Commons
through the system, and then export and use the highest-ranked tags. I
haven't been very active in the Wikimedia community for the last few
years, so I'm not quite up to date on all of the projects percolating
out there, but there might also be some more creative ways in which
the MG system could be employed :-)
(User:Womble on various wikimedia sites)
Over the next year, the Missouri Botanical Gardens plans to identify
and extract illustrations from the BHL's 39.3 million scanned pages as
part of the Art of Life project , and then to publish those
illustrations to the Wikimedia Commons  (as well as to Flickr 
and ArtStor). My colleagues and I have spent the last few months
developed a metadata schema to provide structured information
describing an image -- subjects, "agents" (i.e. publishers, painters,
engravers and writers) and inscriptions. Within the Commons, we've
created a template to handle this structured data, which we call
"Information Art of Life" (based on the ubiquitous Information
Since the BHL doesn't have the resources to comprehensively describe
all the images itself, our plan is for BHL staff members to minimally
describe the illustrations and then to rely on the Commons community
to improve metadata, descriptions and categorization. So when images
are uploaded to the Commons from the BHL, they will have basic
metadata in their "Information Art of Life" templates and basic
categorization, and nothing else. We hope to encourage users of BHL
illustrations (artists, biologists, humanities scholars, library staff
and educators, among others) to take it from there, improving the
metadata, descriptions and categorization on the uploaded images.
However, as many of them would not have much experience with
Wikipedia, we fear that the learning curve in understanding the
Commons' template-based metadata system might turn away potential
To make it easier for non-Wikimedians to contribute, we have been
considering developing tools to simplify updating these templates,
such as by creating user scripts  to provide a form based interface
to our template; maybe something visually similar to the Index page
form that the ProofreadPage extension creates on Wikisource . Do
such tools already exist for the Commons somewhere? What do you think
would be the easiest way to simplify the ways in which non-Wikimedians
can use the Commons' cataloging system?
Thanks so much for your attention!
 Based on an external links search, see:
 An example of an index page form created by the ProofreadPage
extension on Wikisource:
Kaldari has been polishing the work done during Google Summer of Code
by Ankur Anand to support importing correctly licensed Flickr
photo(-sets) using Upload Wizard. You specify a photoset URL, and
Upload Wizard should treat it like a batch upload. You can test the
http://mwreview.wmflabs.org/wiki/index.php/Main_Page (you'll need to
make a new test user account)
Just click the "Import from Flickr" button on the first page to get started.
Since UW performs a license check, UW will also add a
"VerifiedByUploadWizard" template which should help with the long term
validation of licenses for Flickr imported content.
* Getting a better 'source' value for each image - ideally we want the
regular Flickr URL, not the farm server URL
* Getting the description for each image, this may require separate
calls to the Flickr API.
* Making the 'author' value a link to the Flickr account
* Supporting the feature to copy metadata across a whole batch, which
is shown for regular batch uploads
Now's a good time to start playing with it. You can leave feedback
here: http://mwreview.wmflabs.org/wiki/index.php/Talk:Flickr_testing -
or file in Bugzilla against the UploadWizard extension.
VP of Engineering and Product Development, Wikimedia Foundation
Support Free Knowledge: https://wikimediafoundation.org/wiki/Donate
I'm assuming you're referring to the editable version, though I'm not
sure why the English version (say) has had its text converted to paths
(font issues?). Sure, I can try that and see.
As a side note, you've got some serious licensing issues there: you
can't just "add" licenses - CC-BY has to go as a start - and assuming
we're calling this a derivative of all of the source files you're in
fact into illegal territory because the GPL isn't compatible with a
CC-BY-SA license AFAIK.
[Apologies for the cross-post]
After a short delay while I sorted out a Wikimedia Labs account, I am
pleased to announce that version 2.0 of the TranslateSvg extension is
officially available for testing .
TranslateSvg enables the easy translation of virtually any (currently
93.1%, but increasing all the time) SVG image containing text, with
the result embedded into the SVG file so that graphical updates
instantly propagate to all language versions .
Available for testing are three images to give you a feel for the
interface. There's likely to be one future change - the introduction
of an extra dialog box - but it's 99% feature complete. Well, until
*you* tell me what's wrong with it :)
So what are you waiting for? Find ten minutes and get yourself to
Google Summer of Code student