You can clone with
HTTPS or Subversion.
The stacking of min-width: 400px, min-width: 800px, etc., should behave like it would in CSS. I was expecting all the images to be downloaded at higher viewport dimensions – as all stylesheets would be if using the examples media values – but only the larger one was.
ah, issue #1. :)
I commented on the thread ( http://www.w3.org/community/respimg/2012/02/21/a-sample-picture-implementation/ ). I think there are maybe a couple of ways to interpret this. I'm not sure which model is most intuitive.
@necolas do you still feel the same? I'm kinda thinking the way it behaves currently is most responsible. The idea here after all is to get the right - one - request, rather than many...
The important thing is for it to match what the media attribute values are expected to do for video and audio. It seems that it will have to be identical to how media queries function in CSS - http://dev.w3.org/html5/markup/source.html#source - which means the current behaviour of the polyfill is wrong. But talking to some browser / spec people would clear up any uncertainity...because you're right that it does seem to leave the door open for people to accidentally initiate multiple requests.
Thanks for your ideas here, @necolas . Old thread :) We're closing this out as Picturefill moves to support and match the new standard markup landing in browsers. Thread here: #125