Over the last year, we have seen some discussion about if and how Wikidata can
be useful for Wikimedia Commons. One aspect of this is maintaining meta data as
I think that the technology of Wikidata would be very useful for this, but that
the meta data should be maintained on Commons itself. On behalf of the Wikidata
development team, I just posted a proposal for this:
We hope you regard this as an invitation to discuss the proposal and to identify
use cases that we do not cover with it.
Please use the proposal's talk page as a central place for discussion about
Wikidata and media meta data.
Please invite others involved with the Wikidata or Commons projects or otherwise
interested in media meta data to take part in the invitation.
I was just writing about file formats, thinking about the discussion
on Commons of its scope (and the many free formats that we still do
And what should pop up in my mailbox but a discussion about the next
free video codec the Xiph Foundation is working on. This looks pretty
amazing, and is worth a read:
Samuel Klein @metasj w:user:sj +1 617 529 4266
Is there some kind of CSS tag I can use to mark a caption as hidden on visual browsers but visible on screen readers?
My experience with A/B testing is that visible captions badly hurt the viability of any commercial image site. Layouts that break up the physical flow of the images seem to make many people feel mentally, physically or emotionally uncomfortable. (Most just leave) This is why even Flickr has changed to a hidden caption format.
I, for one, liked the old Flickr better because it was faster, for one thing.
Wikipedia Commons is an 800 pound gorilla, not all the start-ups and conglomerates that are equally desperate to seduce the marginal user away from their competitors. We can do something different.
Sent: Monday, June 10, 2013 10:31 PM
To: bawolff+wn(a)gmail.com ; Wikimedia Commons Discussion List
Subject: Re: [Commons-l] Redesigning the <gallery> tag.
Requiring a hoover to see a caption becomes an impediment to non sight readers who rely on various readers.
This format is used on sites like Flickr, Pininterest, Google+ etc.. while many user spend there time just scraping these sites do we really need to look them
How does an editor decide what image sizes are,
IMHO rather than doing this to <gallery> tag something thats frowned upon in GA, A class and FA processes on en.WP anyway this would be better utilised within the category structure where the more important elements of image size, and name could be displayed with the hover giving you linking options.
On 11 June 2013 10:20, bawolff <bawolff+wn(a)gmail.com> wrote:
That's how its currently implemented (with an attribute named mode). How it all works (Or even if it gets adopted), will depend on the sort of feedback I get.
You're 100% right, that this would not work well with the COM:QIC workflow, and there are many places where the captions are important. But I would argue there are many places where people just have captions as a place holder text since there is a spot for a caption (Many [not all] galleries on commons fit this description). In particular, the captions on the auto-generated galleries in category listings are pretty useless, and could easily be replaced with something triggered on hover.
Even in encyclopedic contexts, a gallery is used as a collection of pretty pictures, and the caption is a sort of "see also" text to provide further context for if one of the images catches the reader's eye.
On Mon, Jun 10, 2013 at 10:53 PM, Daniel Schwen <lists(a)schwen.de> wrote:
If this should find its way into core, it should be made optional
(using an attribute of <gallery>)! Please be aware that this would
instantly destroy gallery applications like COM:QIC. Also from a
usability standpoint hiding the captions is probably not always a good
thing. Especially in the context of an encyclopedia rather than a
"pretty picture" site (like flickr).
On Mon, Jun 10, 2013 at 6:36 PM, bawolff <mailto:bawolff%2Bwn@gmail.com> wrote:
> [Cross posting to commons-l and design list]
> Hi all,
> Over the last couple days, I've been looking at making our <gallery> tag a
> lot more slick. Compared to other websites, our galleries look very "boxy"
> imo. This isn't that bad when it comes to icon/clip art type media, but for
> photos I feel we could really do better. So I've tried to make the galleries
> be more compact, with rows of images all the same height but different
> width, and the caption visible on hover. The only downside to this is the
> borders of the gallery become a little ragged, which mostly looks fine to me
> (Possibly could be dealt with with js. Would involve a bit of double loading
> the images though).
> A demo speaks louder than words, hence:
> (contrast with
> (contrast with
> http://commons.wikimedia.org/wiki/Commons:Featured_pictures/Animals/Fish )
> * http://tools.wmflabs.org/bawolff/gallery/index.php?title=Main_Page
> Additionally the wiki is open to editor (You need to register an account
> first), so please don't hesitate to experiment.
> Anyhow, this is all still a work in progress, and I would really appreciate
> your feedback/criticism/hate mail/love letters/etc.
> Commons-l mailing list
Commons-l mailing list
Photo Gallery: http://gnangarra.redbubble.com
Gn. Blogg: http://gnangarra.wordpress.com
Commons-l mailing list