Re: Proposal to update the spec

On 2015-07-09 12:03, Ingar Mæhlum Arntzen wrote:
> Hi Francois and CG members.
>
>
> Thank you Francois for your excellent work and recent proposal. We very
> much share your opinion that CG specifications must detail how the
> timing object
>
>  1.
>
>     connects to online timing objects
>
>  2.
>
>     connects to media elements

I submitted two pull requests for these first two points.

TimingProvider interface:
https://github.com/webtiming/timingobject/pull/7

First stab at associating a media element with a timing object:
https://github.com/webtiming/timingobject/pull/8

I had to shuffle things around and to revisit terms and interfaces 
defined in the spec to be able to introduce a proper TimingProvider 
interface. I kept that interface minimal, on purpose. Since the user 
agent will not be in a position to control the quality of the code 
produced by a timing resource provider, it cannot "trust" it so reducing 
the number of things that a timing resource provider could do wrong is 
important, especially considering that regular Web app developers may 
act as timing resource providers in practice.

Typically, I dropped the "query" method on that interface in favor of a 
"now" method that simply exposes the clock used by the provider. The 
goal was not to end up in a situation where the provider fails to report 
a change on the current state vector and still uses that information to 
compute new state vectors, leading to chaotic moves.

Similarly, I stuck to a simple "one error closes the timing object" 
mechanism for errors. Essentially, if something goes wrong, the timing 
object switches to a "closed" state where it cannot be used anymore. 
That is consistent with the way WebSockets are defined for instance.

I listed a few open issues inline but I'm confident that most of the 
things I wrote will need to be fixed or improved, and that interfaces 
and procedures will have to be modified in particular. I still hope this 
goes in the right direction :)

Thanks,
Francois.

Received on Friday, 10 July 2015 16:38:40 UTC