See also: IRC log
<stakagi> hi i am satoru
<scribe> ScribeNick: dom
Schuki: light agenda today
Schuki: update from Marcos' task force
<schuki> https://github.com/w3c-webmob/installable-webapps
Marcos: I've been traveling, so
we've slowed down a bit
... lots of feedback on the document we've built on
public-webapps
... I'll be filing more bugs based on this
... we would like to get a couple more people involved to
help
... still plenty of work to do, but it's coming along
Schuki: can you an update on the manifest file more specifically?
marcos: we've been trying to
define what the manifest file should be
... we've looked at the what current Web apps do
... and based on this, we've defined what would need to be in a
manifest
... or even if a manifest is needed at all given all that's
already defined
... we have now filed a transition request to publish that
document as FPWD
... the first version is probably richer than what we want in
the end for IPR reasons
... based on my discussions with the Patent and Standards
Interest Group
<marcosc> http://w3c.github.io/manifest/releases/FPWD.html
marcos: the FPWD should be
published shortly
... work is continuing on the document, based on use
cases
... we've also made a call out to developers to get feedback on
some of the aspects of the manifest
<schuki> https://gist.github.com/marcoscaceres/7783977
marcos: with plenty of retweets
<marcosc> https://gist.github.com/marcoscaceres/7783977
<bryan> natasha, one topic we could add is mobile-effective UI features/design for privacy preferences management - we can add some ideas to the manifest discussion on that
marcos: we got a very good range
of feedback
... incl. from non-W3C people
<dka> Anything from the feedback that potentially changes your thinking on the manifest?
schuki: bryan mentions on IRC privacy-features management
marcos: send feedback and ideas on the github repo
<marcosc> bryan: http://w3c-webmob.github.io/installable-webapps/#security-and-privacy
dka: any thing from the feedback that is triggering substantive changes?
marcos: some people quite opposed
to the notion of a manifest
... suggestions to avoid inventing new names but instead to
re-use the ones defined in HTML
... I had proposed to use the <script> element and got
feedback this would break a lot of stuff
... feedback on no need for an inline manifest
declaration
... great feedback in general
dka: there was a proposal that Alex Russel had made on how to incorporate permissions in manifest
marcos: I'm hoping he'll raise this to our repo once he has a chance
dka: this could come out of a discussion we're planning at an upcoming TAG meeting with Dom
<Zakim> bryan, you wanted to ask about the manifest privacy info
bryan: to wrap that up, related
to what DKA was talking about,
... I haven't seen yet the details of how the capabilities and
intents of applications will be disclosed in the manifest
... similar to what was done in widgets or in WAC
... would be interesting to look more into that
marcos: it would be useful for this group to report on what has been done in previous projects, what academic research has shown in this field
<marcosc> bryan: https://github.com/w3c-webmob/installable-webapps/issues/12
marcos: this would inform the work on permissioning
bryan: would be good to document what people have been using and their experience with that
<marcosc> bryan: https://github.com/w3c/manifest/issues/75
<marcosc> https://github.com/w3c/manifest/issues/75#issuecomment-29461650
marcos: ben francis has made a bunch of research on existing runtimes
<bryan> thanks, obviously a lot to catch up on - we'll review and comment
marcos: another mozilla's person
indicated possible interest in looking into this
... which matches one of the task forces of this group
... bryan, you'd be most welcome to help document what's been
done in this space
<schuki> https://github.com/w3c-webmob/netinfo
Schuki: Marcos started this new
github repo looking at use cases and requirements for the
network info API
... that API has been discussed in a number of groups
... Marcos and others are documenting the use cases that native
apps use for that kind of API
... e.g. dropbox making it optional to sync on cellular
data
... or big downloads only on wifi
... We understand the limitation of network API assuming that
network type being a good characterization of network
quality
... We have reasonable info on iOS
... would be useful to get similar data on Android and Windows
Phone
bryan: there was a discussion a
while back in DAP about taking a different approach
... which I intend to bring back in this thread
... an app could indicate what it's OK with: I'm OK with
delayed transaction to benefit from newly available
bearers
... another approach: if you have a shared connection among
several apps with aggregate transport
... this might address a number of the things that app
developers might need
... somewhat different approach: not obtaining info, but
expressing intent
... I'll try to summarize it for the current discussion
schuki: hopefully this can be integrated in the current use cases and requirements of the doc
marcos: ideally, you would find
existing examples of apps taking this approach
... at the bare minimum, Web apps should be able to catch up
with native
<bryan> as a second note, the approach Marcos is taking re focusing what techniques are actually using today, is very good
marcos: examples that support use cases on some platform are important
<bryan> the use cases / approaches I am speaking of are not necessarily common today, just new ideas that could help optimize network use
<Zakim> bryan, you wanted to menion I've been watching the discussion, and have not chimed in yet with the different approach discussed on the DAP list a while back, i.e. letting the app
<schuki> https://github.com/w3c-webmob/offline/tree/gh-pages
Schuki: I have uploaded a light
structured document on the offline topic
... In its current state, there is a little of scope creep
going on here
... lots of things could be covered
... this is first and foremost a way to get people to start
contributing
<schuki> http://w3c-webmob.github.io/offline/
schuki: help from everywhere
would be welcome
... esp. pull requests
<schuki> http://www.youtube.com/watch?v=Z7sRMg0f5Hk
schuki: you can also raise issues
to help orient the document
... I want to add stuff from Jake Archibald's recent talk on
offline
... the document covers current situation, existing gaps, and
upcoming solutions
... some bits about privacy and security
... any question on this?
marcos: I'm worried that we would
duplicate what has been done in the ServiceWorker
explainer
... I'm wondering if we should take the reference
implementation from Mozilla
<bryan> marcos, are you talking about https://github.com/slightlyoff/ServiceWorker/blob/master/explainer.md ?
marcos: and focus this document
around experience gathered in implementing various use cases as
already identified
... as a way to gather feedback on using the API
schuki: great idea, in light of
this upcoming release from Mozilla
... so we would refer back to the explainer doc
<marcosc> https://tbpl.mozilla.org/?tree=Try&rev=db542e5dc66b
schuki: and evolve this toward building experience feedback
<bryan> Re the explainer "handle all resource requests for an application" does not actually mean that all server functions are emulatable, offline, using service workers.
<bryan> Service workers appears to me to be an interesting but as-yet-unverified solution to all offline use cases.
marcos: on the right hand side
https://tbpl.mozilla.org/?tree=Try&rev=db542e5dc66b
you can find various build labels
... these are service worker builds in Firefox nightly
... I haven't managed to get them to work yet, but I'll try
again
... download the build by clicking on the "B" matching your
platform
... and then click on move to build dir
bryan: android 2.2 means 2.2+?
marcos: not sure, given there is also a 4.2 build
<marcosc> Nikhil Marathe
marcos: the guy working on this is happy to work with us and to get feedback
<bryan> 4.2 for x86 not arm
marcos: he's giving a talk today on this
<bryan> but we'll check it out - when is the FFOS patch coming out?
<schuki> http://www.w3.org/wiki/Mobile/Work#TASK_FORCE:_Permissions
Schuki: there are a number of
discussions on this topic across task forces
... it came up in installable, in offline
<schuki> dom: my current plan is to document the issues that have been discussed so far
<schuki> ... then highlight the inconsistencies so far
<schuki> ... and then to enlist the TAG's help in driving the topic in the task force
<bryan> permissions is probably related to what I mentioned earlier, the accepted permissions is something the UI will enable, and needs to be conveyed to the user in some effective way, with either granular or total opt-in/out
<schuki> ... dka has invited me to talk in the TAG meeting
<schuki> dom: we'll work out together the next best steps
schuki: at TPAC, one of the best
messages out of the TAG was: "we're here to help you, come and
talk to us"
... and we want to do this with permissions topic
... I expect Dom will report back after the TAG meeting
... which I guess means the task force won't really start
before that meeting
schuki: one of our other task forces is on scrolling
<schuki> http://www.w3.org/wiki/Mobile/Work#TASK_FORCE:_Scrolling
schuki: there has been lots of
talk on twitter regarding scrolling
... I'll keep linking stuff as I find them
<schuki> http://www.w3.org/wiki/Mobile/Meetings
schuki: Last piece of AOB is
about meetings
... we've decided to move to a monthly meeting rhythm instead
of fortnightly
... given that work seems to be running smoothly on
github
... we would like to use teleconference calls for getting
updates on the task forces
... for which a monthly meeting seems to be sufficient
... aside of that, marcos and I are always hanging out on
IRC
... Teleconferences: some people love them, others hate
them
... if you feel very strongly that this change isn't quite
right, let us know
... we're alternating between 8am and 3pm UTC to accommodate
Asian vs US participants
... again, we welcome feedback on this
... these teleconferences will be on task force updates
... but task forces should feel free to set up their own
teleconferences if they feel it's useful
... We now have an icalendar feed for our teleconferences to
import them in your agenda
<bryan> have a great holiday all!
schuki: This is our last teleconference this year: happy new year to all