On Wed, Mar 17, 2010 at 4:02 PM, Marco Schuster marco@harddisk.is-a-geek.org wrote:
I hope no one is ever insane enough to use this. Imagine those people with cellphones and no data transfer flat (~70% of mobile internet users) - their bills will skyrocket when even a single video is set to auto-preload.
On Wed, Mar 17, 2010 at 4:40 PM, Marco Schuster marco@harddisk.is-a-geek.org wrote:
What about people who use tethering? A browser can't determine if he connects via a flatrate connection (DSL, company network) or via tethering and a mobile connection.
The attribute is a hint, which user agents can respect or ignore as they choose. Mobile user agents will probably buffer much more conservatively than desktop UAs -- that's up to them. Other users who don't like auto-buffering might be able to turn it off in their browser options, or switch to a browser that allows them to do that (as with any other feature they want). Or use a browser that doesn't autobuffer at all by default.
On top of that, browsers can feel free to autobuffer only the first few seconds, up to the point where they think they can play through the whole video -- not the whole thing. Ten seconds of video now and again is not going to exhaust anyone's bandwidth budget.
In real life, the vast majority of people who aren't using cell phones aren't going to use up any bandwidth limit they might have by occasional video downloads, even downloading the whole thing. They would prefer faster video startup via autobuffering (if they're likely to play the video). This is evidenced by the fact that video sites actually do this -- if users were harmed more than helped by autobuffering (or thought they were), popular video sites would avoid it. We should use it for the same reason, wherever we think the user is likely to watch the video -- e.g., on the video page itself.
The fact that one can imagine a small minority of users who might hypothetically be harmed by autobuffering does not mean that it's insane.