I have a weird issue in Chrome (tested on version 26), haven't had a chance to test it on Safari. Each time I load a media element (<video> or <audio>), Chrome will start two HTTP request. The first one will stay pending forever (I guess this is the "meta-data", "partial content" request. But the point is that it stay pending) The second one to the same file is ok and goes on and close after the loading is over. The problem here is that the first request stay pending until I close the browser page. So at some point, if I load multiple media, Chrome will break and stop downloading anything because every available request is occupied by a pending request. I've check to reproduce the issue, and it is happening on both Video.js and MediaElements.js frontpages. To confirm, open your network tab when loading the page, you'll see the first pending request. Then press play on the video, and you'll see the second request working, but the first one will stay pending forever. I've also tried it on Youtube.com with the HTML5 video mode, and the same problem was occuring.
Created attachment 199139 [details] Added a reduced test case with a single bare <video> element and a src attribute.
Safari and Chrome use entirely different media back ends, so this bug is almost Chrome specific.
I can confirm the same happens with the Qt port, at least according to the web inspector.
I can confirm I'm getting the same problem. It seems to happen if there is a slow connection involved somewhere. For us it happens when our CDN network is still pulling the file back from the origin server, so it's only able to deliver it slowly. Once the full file is available, it goes faster and then plays. Can also reproduce this by forcing a slow network connection. If you restart chrome this sometimes fixes the problem, though that might be a false positive http://stackoverflow.com/questions/16137381/html5-video-element-request-stay-pending-forever-on-chrome I can confirm that I do not get the same problem when using a flash based player such as FlowPlayer
(In reply to comment #4) > I can confirm I'm getting the same problem. > What browser do you seeing this in?