- From: François Daoust via GitHub <sysbot+gh@w3.org>
- Date: Tue, 19 Apr 2016 14:28:12 +0000
- To: public-secondscreen@w3.org
> One problem with having the receiver generate the identifier: we won't be able to resolve a new PresentationConnection on the controller in a "connecting" state, since we don't yet have the identifier from the receiver. Yes, or the `id` attribute would need to return `null` as long as the identifier is not known, but that looks bad. Anyway, my initial worry misses one counter-example: if the identifier is generated by the receiving user agent, two receiving user agents could also still create the same identifier for a given presentation connected to two different tabs of a controlling user agent. In such a case, a page running on the controlling user agent would not be able to select the right presentation connection when calling `reconnect`. I suggest to close this issue: no matter where the identifier gets generated, the main issue is indeed that we probably want to reduce the likelihood that the identifier may clash with another one created in its vicinity. And it does make things more complicated from a spec perspective since the spec expects the identifier to be available right away on the controlling side. I created issue #279 (Specify the algorithm for generating a presentation identifier) to discuss the generation of the presentation identifier. -- GitHub Notification of comment by tidoust Please view or discuss this issue at https://github.com/w3c/presentation-api/issues/253#issuecomment-211945119 using your GitHub account
Received on Tuesday, 19 April 2016 14:28:16 UTC