W3C home > Mailing lists > Public > public-secondscreen@w3.org > January 2015

Inviting Mozilla to share implementation feedback

From: Rottsches, Dominik <dominik.rottsches@intel.com>
Date: Thu, 22 Jan 2015 13:35:01 +0000
To: "public-secondscreen@w3.org" <public-secondscreen@w3.org>
CC: "schien@mozilla.com" <schien@mozilla.com>, "gyeh@mozilla.com" <gyeh@mozilla.com>, "jonas@sickin.cc" <jonas@sickin.cc>, "kikuo@mozilla.com" <kikuo@mozilla.com>
Message-ID: <1421933700.12857.1.camel@intel.com>
Hi Kilik, Shih-Chiang, Gina, Jonas, 

I'd like to congratulate Mozilla's team on the great progress on
implementing Presentation API. It's been very interesting to read the
technical discussion in Mozilla's bug 1069230.

We're currently quite active in edging out certain aspects of the API.
We have just recently defined the most important algorithms, but we're
still far from done. 

Here's a list of our open issues:
https://github.com/w3c/presentation-api/issues


I'd like to invite Mozilla's engineers to share their implementation
feedback on Presentation API: What has been working well, where are
clarifications needed, where was the spec difficult to implement and
could be easier, what is missing, etc.

I am especially curious to hear about more requirements, recommendations
for the messaging part.

Trying to summarize from the bug report:

* The implementation focused on the communication channel and what the
method signatures centered around this topic should look like:
https://bugzilla.mozilla.org/show_bug.cgi?id=1080474#c22


Mozilla took an interesting approach in extending the communication
options:

PresentationSession has a 

readonly attribute PresentationSocketChannel channel;

with an extended set of possible types that can be sent:

void send (DOMString data);
void send (Blob data);
void send (ArrayBuffer data);
void send (ArrayBufferView data);

and its own readyState:
https://bugzilla.mozilla.org/attachment.cgi?id=8527264&action=diff#a/dom/webidl/PresentationSocketChannel.webidl_sec2


Is this what you would prefer the final specification to look like? Does
this approach work well for you?

* The discussion on whether there should be a flag for "available" came
up again, and Anton referenced our previous conversation on the subject.

* ...More?

The working group would be very curious to hear some feedback and
learnings, thanks very much in advance,

Dominik

Received on Thursday, 22 January 2015 13:35:35 UTC

This archive was generated by hypermail 2.3.1 : Thursday, 22 January 2015 13:35:36 UTC