W3C

- DRAFT -

Cloud Browser

27 Apr 2016

See also: IRC log

Attendees

Present
JF, Kaz, Alex, Harrison, John, So, Steve
Regrets
Chair
Alexandra
Scribe
kaz

Contents


<scribe> scribe: kaz

<scribe> scribenick: kaz

alex: would look at the plan
... bullet points on my notes

-> https://lists.w3.org/Archives/Public/public-web-and-tv/2016Apr/0003.html Alex's note

Architecture

alex: DOM tree within the cloud browser
... can mention that
... but should we add a chapter on that?

jf: you have to expose the DOM tree to some of the users
... the problem is that if I can't see or can't hear
... I can't see the dancing picture, etc., on the screen
... in that case should use the accessibility APIs for DOM
... HTML5 guys also have similar problems with Canvas
... what to do with shadow DOM
... would provide this kind of feedback for the discussion

alex: can you provide resources on Shadow DOM?

<JF> https://www.w3.org/TR/2015/WD-shadow-dom-20151215/

jf: just pasted the link to Shadow DOM spec
... the other thing we have to think is we can't assume every user can use TV remote

alex: we need to identify communication use cases
... apart of that, maybe questions from Harrison and Kaz
... we should add a section for DOM tree including shadow dom

<JF> +1 to thinking about modalities - that is a high-level (related) accessibility issue as well

kaz: this discussion made me think that we may want to think about user preferred modality for cloud browsers in addition to directly handle DOM tree on the local side

alex: ok
... that was DOM tree discussion
... and what the functionality should be?
... what would be the best way for the group to handle this?

-> https://www.w3.org/2011/webtv/wiki/Main_Page/Cloud_Browser_TF/UseCases cloud browser use case wiki

alex: we're discussing cloud browser architecture and functionalities
... some of the rendering result sent from the cloud browser to the local side
... this would be rather confusing because we're thinking about APIs
... what we have done is hard-coded things
... cloud browser for gaming and remote desk top
... this is the first approach
... and then would think about cable infrastructure, etc.
... some of them are kind of artificial, though
... what are the basic functions
... video rendering, scrambling, etc.
... basic set of functionalities

alex: if we talk about cloud browser on every single cloud
... how to put the architecture chapter?
... would make things easier
... should we extend the document?

harrison: your question sounds reasonable
... we can split out all our requirements
... one question is from our diagram we only two components
... we should have additional components
... to make things clearer

-> https://www.w3.org/2011/webtv/wiki/Main_Page/Cloud_Browser_TF#architecture architecture wiki

alex: shows updated diagrams

harrison: cloud browser is the key player
... from this diagram, cloud browser does everything

-> https://www.w3.org/2011/webtv/wiki/File:Ss-cp.png SS CP

-> https://www.w3.org/2011/webtv/wiki/File:Ds-lp.png DS LP

jf: wondering what the cloud browser client is
... less clear about the interaction
... how to click the link?

alex: you can receive any kind of results

jf: still not sure about local runtime environment

kaz: maybe we should clarify the lifecycle of cloud browsing session

jf: would be useful

alex: please give me an action
... basic motivation of cloud browsing is that the client is not an actual browser

kaz: maybe we could call "Cloud Browsing server" for the "cloud browser"
... and "Cloud Browsing client" for the "cloud browser client"
... what do you think?

jf: your cloud browser server processes the rendering task

alex: would distribute the diagrams to make things clearer
... got another comment from Paul Higgs
... got a comment offline
... wanted to ask the group for opinions

jf: we can start with an elevator speech, i.e., brief description within a few minutes

kaz: +1

<JF> FWIW, I am a good target audience for the "elevator speech" audience

kaz: will you update the diagrams on the wiki with these updated picture?

<Zakim> JF, you wanted to say that dancing pixels is almost the least complicated issue

<JF> +1 to that Kaz, as it helps make the elvator speech clearer

kaz: we can keep the original name "cloud browser" but would be better to add an alias like:
... cloud browser (cloud browsing server)
... local player (cloud browsing client)

<alexandra_mikityuk> what if "Cloud Browser Server"?

<JF> the key word is "server"

alex: another two months for use case discussion
... and would add clearer definition too

harrison: one question
... capturing video encoding functionality
... visual output downloaded to the set top box
... capturing the video data
... video encoding procedure
... my question is if the cloud browser is in charge of the interaction?

kaz: would suggest we clarify the interaction during the session lifecycle
... maybe we can generate a ladder diagram

harrison: will do
... another question is graphics library for composition
... how/who handles that capability?

kaz: video streaming is generated on the cloud browsing server side
... and 2-D graphics generation (could be hardware accelerated) is done by the local player side

alex: will put proposals on the wiki
... lifecycle, etc., as well

kaz: great

harrison: would be great to send by email as well

alex: also would assign an action item
... to our action tracker
... tx for your cooperation
... great discussion

[ adjourned ]

Summary of Action Items

Summary of Resolutions

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.144 (CVS log)
$Date: 2016/04/28 04:10:36 $