See also: IRC log
<scribe> scribe: anssik
<scribe> scribeNick: anssik
[Dominik shows some slides on the screen, material to be shared after the meeting.]
<jcverdie> are the slides available on the CG page or somewhere?
<jcverdie> ooops thx anssik:)
Q: what is the difference in the definition of the "second screen" in this context re Web & TV?
A: in the TV context, TV is the primary, whereas in this proposal the UA is the primary driver
<stakagi> Is chromecast contained in the use case of second screen?
<jcverdie> (rumors in the room from web&tv fellows who tend to disagree with the previous statement)
<olivier> confirmation that a UA could connect to many screens. whether the application would be able to discover/select many targets is uncertain
<jcverdie> scribe: jcverdie
demo of a firefoxOS game projected on a laptop
the mobile phone turns into a cntrol center while the game is displayed on the big screen
(public can see the demonstrator is not a very good driver :p)
@@@: how many lines of codes changed to enable remote display
anssik: not that many, circa 200 lines
Marcus Altman: what is displayed on the second screen?
anssik: it's a navigator window.
it's an URL
... the second demo is a picture gallery
(image with small description on mobile, fullscreen image on secondscreen)
anssik: demo of a hacked youtube
page. opens a window on second screen and launches youtube
... the demo needs a network connection faster than 33.6 Kbps. Not working here
<anssik> is the only visible browser window shared, or is the whole OS screen to be shared?
dominik: we're using the OS facilities @@@
hongbo: for instance we're using Android SDK to control second display
anssik if you want to take over feel free ;)
dominik: we are trying to use a
common ground and rely on OS facilities
... for instance most desktop OS show something as soon as you connect, it's fine
doug: if I want to share a part
of my screen e.g. a keynote and a comment window
... I'd do this using CSS w/ different styles
<anssik> DougS: I want to share only a part of the screen, not the whole screen, I'd like to use CSS for that
<anssik> Dominik: we've considered that option
dominik: our initial solution uses API so you can author your content
giuseppe: it's two windows, not 2 divs
ddavis: url for the demo ?
dominik: these demo depend on specific versions of chromium and we haven't released the code
visible online webscreens.github.io/presentation-api
dominik: abstracts physical
... a little bit like fullscreen API for nearby tv
... here's the IDL of the API preview
(dominik describes the IDL)
dominik: there's a boolean
displayAvailable to tell you have something available
... and an event to be notified when there are some changes
(now a sample code of the two pages code, one for 1st screen, one for second screen)
dominik: you can use post messages to communicate between screens
giuseppe: when the CG was create it seemed that you were not considering the case where there are two user agents running, just one. Now you're saying this case is considered
dominik: i think it's in
... we have tried a variation of this API in such a case, new questions come up
giuseppe: there is already this discussion in other groups
ddavis: the device discovery is done separately
dominik: it uses the OS to do this, it's not part of the spec
ddavis: what about availability?
dominik: in a simple way. it basically replaces windows.open on another screen
anssik: it's automatically the "open new window move it to second screen"
<Zakim> jcverdie, you wanted to ask about capability check of second screen in the IDL
jcverdie: what about capabilities of the second screen?
dominik: we struggle about the
requirements (nearby, ...)
... but that could be added
mounir_: you can't really now where is the second screen. @@@ (missed the end of the question)
dominik: our goal is a simple API
anssik: we should discuss this in the context of SysApps WG
mounir_: this is not something very sensitive except fingerprinting [...]
anssik: our priority is to make it available
richt: are you planning to specify how it works (discovery, connectivity), sounds like black magic so far
dominik: in our idea, that would
be an extension to the OS and from the browser it would just be
a secondary display
... community group
... mozilla expressed interest and joined
<ddavis> richt: UPnP is encumbered - there could be patent issues if you're thinking of the user agent incorporating this.
dominik: we're done part of
implementation but we're eager to get more information and use
... slides will be made available from the session wiki page
<olivier> session wiki page -> http://www.w3.org/wiki/TPAC2013/session-second-screen-web
dominik: thanks attendance
philipp h: have you got a timeframe to move to a WG?
anssik: when there's enough momentum. when we have a better idea of what work we have to do, and we get enough support
richt: which WG?
anssik: any group with good participation
dino: what is a "good level of support"?
anssik: we need interoperable implementations for instance
@@@: are you assuming a single second screen?
anssik: right now it's a
... but it's possible to perform the same API twice and eventually get a second second screen (pun intended)
jcverdie: i am concerned by the
fact you say now you are considering the case of two UA, while
you said in the mailing list initially you didn't
... you need to clarify which goal you are pursuing, I am afraid of potential overlap
dominik: i agree we need to clarify
giuseppe: may be it is worth exploring, but you need to clarify
jcverdie: to be clear, i'm just asking for clarification, i'm not saying i'm against it
dino: sometimes your second
screen has more power and it makes sense it opens a UA
... also you need to poll for new displays, detect if some new display goes online or offline, if there's delay in the display on 2nd screen, ...
dominik: we'll need additional events yes
dino: likely a big use case will be to display a video
dino: may be just ask for a specific object to go to the second screen
giuseppe: it would be like a "requestSecondScreen" similar to request Fullscreen
israelh: what about the UI of the
user agent (such as geolocation popup, ...)
... where would the question go?
dominik: at this point we don't expect any interaction on the second screen, everything would need to be covered on 1st screen but no use case so far
This is scribe.perl Revision: 1.138 of Date: 2013-04-25 13:59:11 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Succeeded: s/@@@/anssik/ Succeeded: s/someone in the back of the room/Marcus Altman/ FAILED: s/annsik/dominik/g Succeeded: s/what about permission?/this is not something very sensitive except fingerprinting [...]/ Succeeded: s/anssik/dominik/ Found Scribe: anssik Inferring ScribeNick: anssik Found ScribeNick: anssik Found Scribe: jcverdie Inferring ScribeNick: jcverdie Scribes: anssik, jcverdie ScribeNicks: anssik, jcverdie WARNING: No "Present: ... " found! Possibly Present: ChristianFuhrhop DougS Kiyoshi Takahiro Tomoyuki Yudong aizu anssik ddavis dino dominik doug drott giuseppe giuseppep gmandyam gwm hisa hongbo https israelh jcverdie joined left mounir_ nigel olivier razybon_ richt sangwhan scribeNick shan stakagi trackbot webscreens wonsuk youenn_ ywu You can indicate people for the Present list like this: <dbooth> Present: dbooth jonathan mary <dbooth> Present+ amy Agenda: http://www.w3.org/wiki/TPAC2013/session-second-screen-web Got date from IRC log name: 13 Nov 2013 Guessing minutes URL: http://www.w3.org/2013/11/13-webscreens-minutes.html People with action items:[End of scribe.perl diagnostic output]