- From: Mark Watson <watsonm@netflix.com>
- Date: Thu, 5 Jan 2012 16:13:09 +0000
- To: Bent G Christensen <benchris@cisco.com>
- CC: Giuseppe Pascale <giuseppep@opera.com>, "public-web-and-tv@w3.org" <public-web-and-tv@w3.org>
On Jan 5, 2012, at 6:43 AM, Bent G Christensen wrote: > Great to see action on the TV profile. > > On 1/5/12 4:56 AM, "Giuseppe Pascale" <giuseppep@opera.com> wrote: > >> On Thu, 05 Jan 2012 10:40:48 +0100, Giuseppe Pascale <giuseppep@opera.com> >> wrote: >> >>> >>>> 2) We have found it key to making our HTML5 applications work on >>>> consumer electronics devices that we consider performance as well as >>>> conformance. It may be that a given device supports a given feature but >>>> it is just too slow to be useful. Performance requirements would need >>>> to be defined in terms of specific performance tests/benchmarks and >>>> would provide guidance to implementors as to where to focus >>>> optimization efforts. >>>> >>> >>> Completely agree, but I envision this as a second (set of) document. >>> First, I would like to write the functional architecture. Second step >>> could be to deal with performances. W3C is probably not the right place >>> to write a normative performances document, but what we could discuss is >>> probably how to measure such performances (and which one are relevant >>> for the industry) leaving to other groups to deal with a minimal >>> performance requirements. Given the limited bandwidth I would start >>> with this once the profile document is at least half way through. Unless >>> someone is willing to drive this second effort, in this case, we can >>> discuss them in parallel. >>> >> >> Giving it a second thought, we may want at least for now add a section in >> the profile document about this (and split it later only if it grows) >> As said this would be more to point out which performance metrics should >> be considered, and in some cases suggesting (not mandating) minimal >> performance requirements. >> >> I also think that "performance" is to be intended in a wider sense that >> just "speed". e.g. features that rely on disk space (app-cache) may be >> enabled but unusable if the memory available on the device is limited. > > I agree on defining the metrics to measure "performance" / goodness factor > of the implementation. We could even designate an appropriate score for each > such quality giving the implementers an easy way to measure and compare > implementations. As with conformance requirements, another purpose of such performance work is to provide guidance to implementors on which features they should prioritize for optimization based on opinions from the potential users of the platform. Of course, as with conformance, it only works if we can agree ;-) I think it's likely the first wave of web apps for TVs will be related to content discovery and so there may well be a common set of functionality that is needed for those applications. ...Mark > > /Bent > >> /g >> >> >>>> ...Mark >>>> >>>> >>>> On Jan 4, 2012, at 8:07 AM, Giuseppe Pascale wrote: >>>> >>>>> *** NOTE: from now on, all mails related to the TV profile SHALL be >>>>> tagged with [profile] *** >>>>> >>>>> Hi all, >>>>> I'm happy to see interest in working together on a TV profile document. >>>>> >>>>> As mentioned by several people (me included), we need to agree on the >>>>> scope of such document and the goals we want to achieve. >>>>> Based on the discussion on this list, I put together a first draft >>>>> >>>>> http://dvcs.w3.org/hg/webtv/raw-file/3587643883c4/tvprofile/tv.html >>>>> >>>>> Please take a look at the introduction section of this document and >>>>> let me know what do you think. Feel free to point out typos and to >>>>> suggest additions or changes to the text. >>>>> >>>>> Once there is consensus on the goals, we are ready to move on to the >>>>> next steps. Meanwhile I'll keep refining the document adding more text >>>>> in the other sections so that we have starting point for further >>>>> discussion >>>>> (if you have already suggestions or comments on the rest of the >>>>> documents, feel free to send them to this list, but PLEASE start a NEW >>>>> discussion thread) >>>>> >>>>> One last point: many of you have mentioned that a dialogue with other >>>>> organizations is needed. I agree. >>>>> I would suggest we wait to have a clear agreement on the goals and >>>>> maybe a more detailed structure of the documents before starting to >>>>> reach out to relevant groups. Let me know if you think we should reach >>>>> out to this groups earlier (or later) instead. >>>>> >>>>> Thanks, >>>>> /g >>>>> >>>>> >>>>> >>>>> -- >>>>> Giuseppe Pascale >>>>> TV & Connected Devices >>>>> Opera Software >>>>> >>>>> >>> >>> >> > > -- > Bent G Christensen, Ph.D. > SP, Video Technology Group, > Cisco Systems, Inc. > > E: benchris@cisco.com > T: +1.770.236.4900 > W: http://neuromancer.sciatl.com > >
Received on Thursday, 5 January 2012 16:13:42 UTC