W3C home > Mailing lists > Public > public-media-fragment@w3.org > September 2009

Re: great news from YouTube on frag addressing

From: Silvia Pfeiffer <silviapfeiffer1@gmail.com>
Date: Thu, 24 Sep 2009 17:34:28 +1000
Message-ID: <2c0e02830909240034kaf8cadai1e6feba60f3820ae@mail.gmail.com>
To: Jack Jansen <Jack.Jansen@cwi.nl>
Cc: Media Fragment <public-media-fragment@w3.org>
On Thu, Sep 24, 2009 at 6:12 AM, Jack Jansen <Jack.Jansen@cwi.nl> wrote:
>
> On  21-Sep-2009, at 05:30 , Silvia Pfeiffer wrote:
>
>> Hi Jack,
>>
>> So, I tried this and you need a really long video to trigger it. I did
>> it on http://www.youtube.com/watch?v=v_UyVmITiYQ the Google Wave demo.
>> Browsed to a different video and came back. Then checked the cookies
>> (which I suspected how they would do it in the current given browser
>> capabilities). Here it is:
>>
>> Cookie: FcZDK.resume
>> Value: v_UyVmITiYQ:2553
>>
>> I believe the 2553 is the time offset that I stopped at in seconds: 42:33
>> .
>>
>> However, I don't think we need to have to worry about this, since it's
>> not done with URLs.
>
> Hmm. This behaviour is exactly why I think we *do* need to worry about it.
>
> We (well, I, at least) have always assumed that the unadorned URL would be
> interpreted as "the whole resource". Now it seems there's at least one
> important implementation that does not always interpret it this way: its
> interpretation depends on some cookie. I'm not sure about how to call this
> interpretation, but I think I'd settle for something like "what the server
> thinks is convenient for the user".
>
> Assuming the URL-police thinks this is acceptable behaviour (or,
> alternatively, if youtube is powerful enough to override the URL-police) it
> would be good if the MFWG catered for this possibility, by defining a media
> fragment syntax that tells the server to return "the whole resource, and
> please refrain from any processing you might consider convenient, thank
> you". Maybe an explicitly empty time range ("t=") or something like that.

The YouTube URL is to a Web page and not to a video. Thus, we don't
really have to worry about it, since we are only doing media fragment
URIs.

We can continue to assume the anadorned URL refers to the full video.
YouTube does so, too, but uses javascript and the flash player to do
the offset. That does not contradict our approach.

Cheers,
Silvia.
Received on Thursday, 24 September 2009 07:35:33 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 21 September 2011 12:13:34 GMT