- From: Shadow2531 <shadow2531@gmail.com>
- Date: Sat, 27 May 2006 07:06:16 -0400
On 5/26/06, Anne van Kesteren <fora at annevankesteren.nl> wrote: > In Opera we now throw INDEX_SIZE_ERR when someone calls loop(x) on the > Audio object with x being either 0 or negative. This seems to be > consistent with how similar "errors" are treated in the <canvas> > interface and makes sense to us. Our suggestion would be to put this > into the specification, unless someone has a better suggestion. > > (Note that loop() without arguments can be used for infinite repetition.) I might seem picky, but I don't really like how loop() takes a *playcount* param. loop() = infinite loop(1) = play once loop(2) = play twice It's the loop(n) that bugs me. I see loop(1) and think that the file is going to play once and then loop once. (The first time it plays is not a looped play.) Luckily, I saw the spec and read that the parameter is NOT how many times to loop, but how many times to play, hence *playcount*. Not seeing the spec, if I wanted to *specifically* set the file to only play once, I'd do loop(0), which to me would mean, play once and not loop at all. Loop is usualy a boolean value anyway. The following makes more sense to me: playcount() = infinite playcount(1) play once playcount(2) play twice playcount(-x) throw error playcount(0) throw error ( Or don't throw an error and just don't play. Can't think of a use case for that, but maybe someone can.) On a side note, the Windows Media Plugin param PlayCount works like this: playcount = 0 - infinite playcount = n - number of times to play I can see people doing loop(0) to try to turn off looping if they previously set loop(99) for example. Getting rid of loop() and using playcount() would clear things up IMO. Of course, I could just be crazy. :) I agree though, an error should be thrown for loop() for anything less than 1. -- burnout426
Received on Saturday, 27 May 2006 04:06:16 UTC