On Wed, Nov 26, 2008 at 7:22 AM, Maik Merten <[EMAIL PROTECTED]> wrote: > Silvia Pfeiffer schrieb: >> In any case - if you (and also Chris Double) are satisfied with the >> estimates you're getting for file duration/length - I'll stop arguing >> for it. It would be nice to hear some experimental evidence about how >> well it's doing, e.g. for typical movie trailers, so we can lay that >> argument to bed knowing we've done our homework. > > I now also account for the bytes in the buffer to compute the current > playback byte-position and added a status bar showing playback progress. > > http://people.xiph.org/~maikmerten/demos/bigbuckbunny-applet-javascript.html > > Given that I estimate the duration with outrageous crudeness (no > averaging) and also compute the slider position without any means to > ensure it doesn't jump around wildly I think it's pretty smooth - so > perhaps estimating the duration is in fact a viable fallback if one > doesn't want to determine the duration "properly".
Nice work! The duration is indeed jumping quite a bit between 8min and 12 min and even at the end still has a gap of actual end time of 9m54s while the estimate is still at 10m44s. Players like YouTube's player display the duration of the file which is very useful for a consumer to estimate if they actually have the time to spend on watching the video. So, even if we don't use the duration/length attribute for calculating the timeline, it may well be useful metadata for display purposes. BTW: are you planning to implement seeking on the timeline, too? It would probably not too bad given the smoothness of the slider position. Cheers, Silvia.
