Mobile Web Best Practices Working Group F2F Meeting Day 2

DKA: [welcomes participants]

<jeffs> welcomes in progress at TPAC face-to-face

DKA: Spent yesterday on Content Transformation
... Made a lot of progress, which is good, but we now only have one day to run 1.5 days of schedule.
... I'd like to clean our morning session so that we can address Mobile Web Applications Best Practices

Jo: We could rather do a few easy things such as mobileOK Scheme, mobileOK Basic Tests, then we have Stéphane
... in short try to leave the afternoon for Mobile Web Applications Best Practices.

DKA: [playing with the agenda]

<DKA> http://docs.google.com/Doc?docid=dd3jk8v_128gb3xp5hq&hl=en

mobileOK Basic Tests 1.0

<jo> http://lists.w3.org/Archives/Public/public-bpwg/2008Oct/0050.html

<dom> mobileOK draft 45

<jo> mobileOK Basic 1.0 Tests

Jo: I'd like to start with a first resolution on mobileOK Basic Tests. I sent an email shortly before the meeting.
... Changes in the latest draft are pretty limited.

Jo: the only difference is the insertion of the HTTPS section, removed from the HTTP Response section because it did not quite fit there. This was triggered by a comment from the Web Security Context Working Group.

<Zakim> francois, you wanted to note that Thomas said "close enough"

Jo: I'd like us to resolve that we request transition of this document to Proposed Recommendation.

francois: I sent the link to Thomas yesterday who said "close enough".

<dom> PROPOSED RESOLUTION: Request Advancement of mobileOK Basic Tests to PR, skipping CR on the basis that implementation experience informed the return to Last Call, and those issues have been dealt with.

<nacho> +1

<DKA> +1

<jeffs> +1

<rob> +1

<abel> +1

<DKA> +1 for jo (in absentia)

<SeanP> +1

<dom> +1

<dom> PROPOSED RESOLUTION: Request Advancement of mobileOK Basic Tests to PR, not waiting until the checker catches up with latest small change, skipping CR on the basis that implementation experience informed the return to Last Call, and those issues have been dealt with.

dom: independently of the update of the Checker?

Jo: I'd prefer so. What do you advise?

dom: yup.

RESOLUTION: Request Advancement of mobileOK Basic Tests to PR, not waiting until the checker catches up with latest small change, skipping CR on the basis that implementation experience informed the return to Last Call, and those issues have been dealt with.

<jeffs> +1

DKA: wonderful outcome, and the resolution is taken before 9:30am :-)

mobileOK Scheme

Jo: OK, back to mobileOK Scheme.

<jo> mobileOK Scheme

Jo: A bit of history, two years ago, one mobileOK, splitted into mobileOK Basic and Pro, yada yada. I ended up being the editor of the Scheme document.
... The document explains what are the relations between the best practices and mobileOK, how to claim for mobileOK confromance.
... It is a relatively short document

[scrolling through the document]

Jo: The "claiming conformance" part will have to be revised on the light of the legal license we talked about yesterday.
... The section details how put a machine-readable claim in a page.
... The RDFa solution is a bit misleading since an RDFa doc is by definition not mobileOK.
... but it could be put in a representation that is not the mobileOK representation. Anyway.
... In short, I think the document is the one we should produce but the details need to be sorted out.
... If you have other views, please say so.

DKA: do you need a machine-readable claim to claim conformance? Does the presence of the image constitute a claim in itself?

Jo: That's one thing we need to resolve.
... We may consider that putting the logo is enough.

Kai: If we have a machine-readable claim, it makes propagation mobileOK easier.

DKA: Taking it from the perspective of a site manager of company X.
... They know that their resources are mobileOK, and now they want to set the claim.
... What's the workflow?

<Zakim> Kai, you wanted to point out that maschine readable claim may aid in propagating mobileok

Jo: I think it's important not to trivialize the notion of a claim.
... It's an assertion made by someone that something is true at a given point in time. A logo doesn't say "Jo said that resource X is mobileOK on 20 Oct 2008".
... The notion of trust is important
... POWDER contains more information, so that could be the reason to push this forward.

jeffs: could we mandate some string in the ALT attribute that goes with the logo?

Jo: I don't really want us to go down that path.

DKA: The barrier to adoption is higher with POWDER, because people will then have both to understand mobileOK AND POWDER. That may not be as easy as it seems.

Kai: POWDER would help for groups of resources as well. You can group resources in one POWDER file.

DKA: I don't think that's POWDER against logo. It's more how to do it for Mom's and Pop's web sites. And then POWDER could be used for Pro sites.
... What can we resolve today about this?

<hendry> how about the validator maintains a registry of MobileOK conforming sites?

Jo: I appreciate the point on simplicity, but I think we should emphasize the fact that POWDER contains more information.

<Zakim> Kai, you wanted to point out that there is a section in the licence which says that groups of resources can be claimed to be mobileok. this may need maschine readability as well.

DKA: So the workflow would be: you check a page, and you get the POWDER file that you could put at the root of your server. Maybe that should be part of the Checker's results then.

Kai: if we want people to adopt mobileOK, it has to be true.

jeffs: I push my students to run their pages through the validator, and what they get as a result is a short piece of text that they can put in their page. Easy. And the thing is it puts the brand out there.

Andrew: adding the date the logo could help improve the information carried out by the logo

Hendry: maybe we could maintain a registry of mobileOK web sites.
... I don't see it really working otherwise. The assertion won't be always true.

Jo: I acknowledge that Kai has a point.
... In response to Jeffs, the point in mobileOK is that there are little details we want to push forward. It's a statement that at some future point when you resolve a URI, it will be mobile-friendly.
... Running a checker is not necessary (although most probable).
... In particular, it's not so much a claim saying "I passed the tests" even though it is by definition.

<DKA> PROPOSED RESOLUTION: With regard to MobileOK Scheme, it must be rationalized with the new terms & conditions of the logo usage. Some machine-reable version of the claim must be present - the logo is not enough on its own.

dom: I think POWDER is a nice thing to encourage, but having the logo is the most important thing to have for people in the short term.
... Personally, I would certain go to say that the logo is enough.

<DKA> PROPOSED RESOLUTION: With regard to MobileOK Scheme, it must be rationalized with the new terms & conditions of the logo usage. The logo is enough to claim conformance. However, we will encourage checkers to provide end users with not only the logo but also instructions for claiming with POWDER to encourage the spread of POWDER-based claims.

<dom> (is mobileOK scheme targeted to be a WG Note?)

<dom> [the latest draft of mobileOK scheme says "This document was developed by the Mobile Web Best Practices Working Group. The Working Group expects to advance this Working Draft to Recommendation"]

<jeffs> (reluctantly) agree with Kai about including date

DKA: having the checkers return both the logo and the POWDER file would help go in the direction we want people to go in the end.

<Zakim> Kai, you wanted to propose to add the date

Kai: My main concern is really about the value of the logo. I would be happy to use the logo. I would just suggest that we take the idea to add the date.

<DKA> PROPOSED RESOLUTION: With regard to MobileOK Scheme, it must be rationalized with the new terms & conditions of the logo usage (including putting the date in the meta-data of the logo). The logo is enough to claim conformance. However, we will encourage checkers to provide end users with not only the logo but also instructions for claiming with POWDER to encourage the spread of POWDER-based claims.

<Zakim> jo, you wanted to respond to Dom

Jo: what we have produced here is an image without substance.

Kai: I agree. But that's not without substance.

<Kai> There is a licence connected to it, which gives it more weight

dom: I agree. The question is: are we going to go around and chase people that use POWDER or the logo incorrectly?

<Zakim> hendry, you wanted to say how about at least watering down "claim conformance" to "checked with" (less substance)

dom: This would really depends on whether people really use mobileOK, but it's independent of the fact that the claim is made using a machine-readable claim or not.

Hendry: I was suggesting that we tone down "claim" to "checked with".

Jo: I think it makes it as valuable as the sticker my kids get at the dentist: "I brushed my teeth this morning".

<rob> +1

<jeffs> +1 on resolution

<dom> +1

<nacho> +1

DKA: I think the proposed resolution here allows for the most basic needs and the evolution towards a more complex scenario.

<jo> 0

<abel> +1

<hendry> -1

<SeanP> +1

<Kai> 0

<DKA> +1

Kai: Just to say that since you can provide the claim in a POWDER form along with the text for the logo, we may want to consider that to make POWDER mandatory.

RESOLUTION: With regard to MobileOK Scheme, it must be rationalized with the new terms & conditions of the logo usage (including putting the date in the meta-data of the logo). The logo is enough to claim conformance. However, we will encourage checkers to provide end users with not only the logo but also instructions for claiming with POWDER to encourage the spread of POWDER-based claims.

<DKA> �1

<Zakim> Jo, you wanted to respond to Dom's question on future status of document

dom: will this be going to be a note or a Rec?

Jo: A note in my view.

[break]

Korean Task Force report

<seungyun> see http://www.w3.org/2005/MWI/BPWG/Group/TaskForces/Korean/

<francois> Seungyun: I will explain the status of the Korean TF created last March.

<JonathanJ> http://www.w3.org/2005/MWI/BPWG/Group/TaskForces/Korean/#documents

seungyun: want to avoid fragmentation of standards
... wanted to create a trial service report
... gap analyis, requirements for new standards
... showing roadmap
... want to issue issue a report in January 2009
... two documents - status of trial and 1st draft of gap analysis
... goal of the trial service is to defin the feasibility of the web standard
... have many member to invoke the trial service
... will do a three year project
... 2nd phase is an activation stage for apps and contents
... 3rd phase is to complete the mobileok eco system
... there are 14 organisations participating
... this year we are developing some key components such a DDR, Test and Certifiaction server...
... web browser, web portal

seungyun: using device inonformantion
... showing service scenario
... issues regarding device context key
... real world deployment issues
... no key context of device
... in Korea URI based devilery
... Korean operators do it their way, instead of using standards
... interested in providing mobileOK services
... want to provide clearer path for operators to conform to standards
... no trustmark for mobileOK conformance
... Koreans considering using their own logo/trustmark for mobileOK

<JonathanJ> http://docs.google.com/View?docid=ddkw3489_43cfkqjsdm

seungyun: content adaption different meaning to CT
... content adaption includes media transformation
... types of mobile device ISSUE. legacy devices not being supported
... video media is very important in Korea, mobile IPTV is popular. ISSUE re supporting this in mobile
... trial in development. In december there will be further results published.

<dom> mobileOK logo

jo: restate issue re logo/trustmark

<Kai> DKA: will rigo's information yesterday address this issue?

seungyun: cannot use logo because our rules are different
... we need to resolve fragmentation issues between Korea & W3C

jo: asks what is the requirement for different devices (asking for clarification)

seungyun: contents providers use UA strings to provide services
... trying to employ standard UA prof approach

jeffs: wondering about different media types like flash & mobileOK

seungyun: look&feel is very important for mobileOK adoption using visual media

Kai: suggests using DDC and asks about minimum requirements

seungyun: addressing DDC problem. considering profile based DDC.
... size issues different in Korea

<Kai> seungyun: we are setting the page size limit at 50 K rather than 20K

hendry: issues surround trustmark, validity, suggests upto date realtime checking of mobileOK conformance

seungyun: discussing certificate with Figure 3: Basic service scenario for mobileOK trial service

jo: comments the koreans have taken mobileOK to a greater level of sophistication than we have

<Kai> JonathanJ: showing the gap analysis document

JonathanJ: going through figure 1
... analyis of gaps, three of them
... three types of gaps explained in figure
... and relations to activities in W3C

<Kai> jonathanJ: gap 1 - perception differences, gap 2 .- activity differences, gap 3 - diffs in standards

JonathanJ: three main inputs; mobileOK forum, forum's member requirements & industry at large
... gap1 market requirements, gap2: standardization scope; gap3: standards
... focused on gap2 & gap3
... figure 3 activies in the W3C
... table 1: comparison chart of differences between elements of mobileOK in Korea and W3C
... mobileOK basic tests 1.0 <---> K-mobileOK basic tests 1.0

jo: asks for clarification re character sets

JonathanJ: supporting both EUC-KR & UTF-8

seungyun: UTF-8 has the emphasis, though we test *both*

JonathanJ: DDC <---> K-DDC 1.5
... trustmark issue again raised
... increate value to 50 for external resources
... page size increase to 50K
... we need to implement DDC
... new document K-MWBP 1.5 addressing 20 items

<dom> doesn't the web compatibility test do this? http://www.w3.org/2008/06/mobile-test/

JonathanJ: type2 relasonship with full browsing. in Korea many way for "full browsing" to be achieved
... organising deployment task force
... interested in addressing mobile widgets

DKA: asks to follow widget standards

jonathanJ: yes, will follow standards
... K-G3-002 evolution of DDC --> K-DDC 1.5
... html4 included along with xhtml
... higher minumum screen sizes

jo: how can screen size be tested?

jonathanJ: tests not considered
... PNG is added
... supporting ECMAscript 3
... & AJAX
... & SSL
... & DOM L1/2/3 & events
... Mobile Web Best Practises translated ( no modifications )
... CT Guidelines not started
... looking at 2.0 applications
... [APPENDIX2: Comparison Table: MWBP 1.0 and K-MWBP 1.5]
... Black icon is defined
... finalised at end of year (gap analysis)
... & proposal coming too

seungyun: overal picture is we support more

DKA: can we take some elements and import it?

jo: we want to use your work
... there are practical issues, what can we help you with & how can we work together?

jonathanJ: possibility of back translating additions

seungyun: wanting to avoid fragmentation by working together

jo: suggests liasing telcons
... in a month's time?

seungyun: every two weeks from there?

jonathanJ: not happy, but ok with this

<jo> ACTION: Dan to liaise with Seungyun and Jonathan re setting up a call in 4 weeks with a view to having 2 weekly calls in the morning European time [recorded in http://www.w3.org/2008/10/21-bpwg-minutes.html#action01]

<trackbot> Created ACTION-870 - Liaise with Seungyun and Jonathan re setting up a call in 4 weeks with a view to having 2 weekly calls in the morning European time [on Daniel Appelquist - due 2008-10-28].

Mobile Web for Social Development (MW4D) Presentation

steph: Some background information...
... 1.5bn web users in the world; 4.5bn to go
... biggest hole in Africa

jo: (and Greenland!)

steph: particulary the poorest on <US$2/day
... The point isn't to connect people to the web; the point is access to essential services
... There are two trends:
... 1 is to provide low-cost laptops and internet connections

<dom> Stéphane's slides on MW4D

steph: 2 is to use mobile phones; 3.5bn people have access to a mobile phone, 80% of the population is covered with GSM signal
... Some impressive success stories, eg farmers finding markets for their produce instead of vice-versa

<jo> [stephane pumps m-pesa ...]

steph: So how can we help and encourage development of access to such services?
... Four strands: 1) understanding the technology channels; 2) people from these communities who know the needs, education issues etc;
... 3) the tools to support these people and 4) raising awareness of the whole issue and who can help
... The MW4D group runs to June 2009
... and expects to produce a handbook
... and a roadmap for actions for organisations that will lower barriers to MW4D
... and to identify a set of resources for all the players
... There is already a MW4D Factsheet available to everyone

http://www.w3.org/2008/MW4D/

scribe: The heart of the group at the moment is the wiki (link from the homepage above)
... There have been workshops in India and Brazil so far, the next will be in Africa

jo: Is the BBC involved?
... Lots of their mobile traffic is from Africa

steph: not yet!

dka: Is there any way BPWG can help?

steph: we're not gathering training material yet - but clearly mobile web is one channel and ths will be coming soon

jeffs: This is an issue I want to push - a broad base of people who know how to make these services will drive this

hendry: how did it become *mobile* web?

dom: the web foundation programme "Web for Society" is broader

steph: have to recognise the idea of an Internet cafe in some places doesn't exist
... but GSM penetration is good in these places
... Even the needs of electricity for mobiles vs Internet PCs are fundamental

dka: Remember Mobile Tech for Social Change in San Francisco on US Election Day

<DKA> http://barcamp.org/MobileTechForSocialChangeSanFrancisco

<DKA> (sign up now!)

steph: The huge plus point for mobile services is that you can build them on an existing platform

Addendum to the Mobile Web Best Practices (BP1.5)

<Kai> http://lists.w3.org/Archives/Public/public-bpwg/2008Oct/att-0001/ED-mobileOK-pro10-tests-20080731.htm

Kai: We've reduced the test-like language in the document and moved it to an addendum

<dom> "mobileOK best practices" don't exit; we have "mobileOK Basic Tests" and the "Mobile Web Best Practices"

Kai: So the purpose (of the addendum) is now more precise
... Test format pass/fail/warn criteria are removed
... and tests are now evaluation criteria
... Now does the group have any feedback on this?
... since te group previously asked for this change

<dom> [I certainly prefer the new format and orientation]

jo: I think this doc needs to be an addendum to MWBP but I don't think it has much (or anything) to do with mobileOK
... these evaluations apply not just to mobileOK but are more widely applicable
... for example if people see this as a supplement to mobileOK (which is only about the DDC) then some of the contents will be confusing
... for example some of this document is useful for an iPhone experience which is not likely to be mobileOK

Kai: where (aside from the title) are the strong references to the mobileOK Basic Tests?

jo: the BP doc itself loosely contains tests
... the introduction links to this doc

Kai: Jo, can you please provide clearer text for the introduction?

<Zakim> dom, you wanted to ask what will be done when mobileOK basic defines a DDC-independent test (e.g. images_specify_size)

dom: the switch from test to evaluation is an improvement

<DKA> +1 to Dom - moving to "evaluation" from "test" a good step.

<DKA> �1 to Jo

dom: avoiding reference to checker and mobileOK is nore difficult
... eg best practices on image sizes are impossible to talk about without having a device in mind

dka: I don't see the need to expunge references to mobileOK basic
... nobody's going to mistake the references to mean it's normative

Kai: I agree with Dan, I don't see the risk
... and I see advantages in cross-refererences in this family of documents

francois: The Purpose section can clarify this - the rest of the document can still include references

jo: Now think this document should be titled "Supplementary evaluations for mobileOK Basic" (!!!?!)

francois: mobileOK Basic is based on DDC

jo: What's the "call to action" with this document?
... ie what are we asking people to do here?

Kai: the Purpose states this
... it is to provide guidelines on how to get better than mobileOK Basic

jeffs: and it is stuff that requires human intervention, it isn't machine-testable

dom: These aren't "guidelines" though, they are "evaluation procedures"

dka: We're not setting this up as mobileOK Pro though - that isn't clear

francois: I prefer Jo v1
... we're adding to the lack of clarity about what relates to what

<DKA> PROPOSED RESOLUTION: The group agrees with the approach to BP 1.5 started by Kai.

francois: so clarification of the Purpose and the relationship between the 3 documents is most important

<DKA> PROPOSED RESOLUTION: The group agrees with the approach to BP 1.5 proposed by Kai (removing "test" language and positioning it as an addendum to BP 1.0).

<jeffs> +1

Kai: I'm not sure these evaluations are completely separate from the DDC

<nacho> +1

<abel> +1

+1

<DKA> PROPOSED RESOLUTION: The group agrees with the basic approach to BP 1.5 proposed by Kai (removing "test" language and positioning it as an addendum to BP 1.0). Jo to work with Kai on clarifying 1.1 "Purpose" and present back to group.

<jeffs> +2

<francois> +1

+1

<Kai> +1

<dom> ACTION: Jo to work with Kai on clarifying purpose text of bp 1.5 [recorded in http://www.w3.org/2008/10/21-bpwg-minutes.html#action02]

<trackbot> Created ACTION-871 - Work with Kai on clarifying purpose text of bp 1.5 [on Jo Rabin - due 2008-10-28].

<abel> +1

RESOLUTION: The group agrees with the basic approach to BP 1.5 proposed by Kai (removing "test" language and positioning it as an addendum to BP 1.0). Jo to work with Kai on clarifying 1.1 "Purpose" and present back to group.

<nacho> +1 (CTIC is very happy with the great work from Kai pending editorial issues)

Kai: Scope needs to change as well then

jo: Yes; we'll cover that in this review
... There is naff-all point in doing further work on this doc unless the group reads and reviews it

dom: I can reformat the main body of the document if it's not going to change

<dom> ACTION: Dom to work on reformulating 4 "tests" of BP 1.5 [recorded in http://www.w3.org/2008/10/21-bpwg-minutes.html#action03]

<trackbot> Created ACTION-872 - Work on reformulating 4 \"tests\" of BP 1.5 [on Dominique Hazaël-Massieux - due 2008-10-28].

dom: can I start on the two tests and pass them back for review?

Kai: Yes, thanks

Mobile Web Application/s Best Practices

adam about to present his documents with discussion points about the BP2 document

<jo> PROPOSED RESOLUTION: The name of the document is: Mobile Web Application Best Practices

<DKA> +1

<jeffs> +1

<rob> +1

<SeanP> +1

+1

RESOLUTION: The name of the document is: Mobile Web Application Best Practices

<jeffs> +1

<aconnors> Latest Editor's Draft of MWABP

aconnors: some questions after comments in the document I am using to present
... better to start by top level headings and refactor the doc and see if there is something important missing
... some text on SVG from Abel
... does that sound a reasonable plan?
... personalization, retaining info between sessions
... dimishing the need for user input
... sensibility to network connection quality
... section on conservative use of resources

[reading through the google doc...]

<aconnors> Adam's List of Issues

scribe: yellow section treated in the last F2F; need for refactoring
... some formal resolutions needed
... one web section, wondering whether this is a top leve section
... UI section, maybe needs refactoring... some stuff needs more level of detail, maybe some additions

Jo: Input from the Korean TF needed

[aconnors adds it to the google doc]

BP2 - One web section

jo: intent of One Web section?

aconnors: thematic consistency, basically
... my concern is that the top level headers are not appropriate and consistency of the documents in itself and against the BP 1.0

[someone please correct the scribe on his mistakes]

jo: more than editorial work needed.. there are stuff here that we did not have in BP1.0... a lot more practical stuff

[nacho please asks people to speak a bit louder (not adam and jo :-) )]

aconnors: adding a section in the gDoc on User Experience, mentioning One Web issues within it

<DKA> PROPOSED RESOLUTION: In BP2.0, "user interface" section becomes "user experience" and "Offer A Consistent View Across Multiple Devices" section goes into "User Experience"

<DKA> PROPOSED RESOLUTION: In BP2.0, "user interface" section becomes "user experience" and "Offer A Consistent View Across Multiple Devices" section goes into "User Experience" (thereby removing "one web" ase a top section).

hendry: I see UI as HTML+CSS thing

[anybody improving my scribing on KHendry?]

+1

<DKA> +1

<SeanP> +1

<jo> +1

hendry: i do not see the sense of the One Web issue

DKA: it is an important preamble

francois: how about a first preamble on One Web and then insisting and clarifying within the User Experience subsection?

DKA: right
... but let's get into resolutions

RESOLUTION: In BP2.0, "user interface" section becomes "user experience" and "Offer A Consistent View Across Multiple Devices" section goes into "User Experience" (thereby removing "one web" ase a top section).

<jeffs> +1

BP2 - Conservative use of resources

dka: conservative use of resources... from an operator perspective, as more users reach plain data rates it might not bring benefit to operators but it does to the user.. although it fosters the usage of mobile web so in the end it is a win-win

jo: responsiveness is a key point here... smaller apps being rendered faster

aconnors: i agree but hard to word it

dka: i volunteer to do that

<scribe> ACTION: DKA to provide some words on conservative use of resources [recorded in http://www.w3.org/2008/10/21-bpwg-minutes.html#action04]

<trackbot> Created ACTION-873 - Provide some words on conservative use of resources [on Daniel Appelquist - due 2008-10-28].

[phillip hoschka introducing himself]

[Rotan hanrahan doing the same]

aconnors: we should recap on where we are not, rather than start refactoring the doc

jeffs: about user awareness and control, it should talk about the user retaining control on his PIM

BP2 - 3.8.3 Use Device Classification to Simplify Content Selection/Adaptation

aconnors: renaming "personalization" section into "retaining information on personalization"
... "handling device capabilit variation" section, need for explaining the need for device clasification to facilitate adaptation

dka: referring contributions in the mailing list about device capability and also section 3.8.3 of the current draft of the BP2.0 doc

jo: original intention in my mind of this section is device clasification in terms of not having classes so use your own classification (but do it)

Rotan: we had a little bit progress about classification (Structures oldie document) and the idea is to write device classification so others can read it... usage of ontologies for modelling that
... the definition itself is up to the web developers and other stakeholders, but it has to be made in a way that can be understood by others
... there is a bit of art in this process so it is impossible to "standardize" on what a tablet, or a smartphone is

jeffs: conversation with blind people made me think about this... maybe the most important are things like input mechanisms

rotan: if your focus is accessibility, you'll classify based on multimodality and other issues of interest for accessibility
... but others interested in other issues than accessibility will create other categories

dka: let's focus on the doc
... maybe examples of categorization needed in the terms in the current draft (good, better, best bullets in the draft BP 2.0 doc)
... maybe reference DDR work as the source for classification information

<DKA> PROPOSED RESOLUTION: In BP2: Group agrees with the approach taken in 3.8.3 - provide examples of classifications but do not define classifications.

nacho: yes to classification in the bp2.0 just in terms of examplification... and clarifying this

+1

<rob> +1

<abel> +1

<jeffs> +1

<SeanP> +1

<Kai> +1

<DKA> +1

RESOLUTION: In BP2: Group agrees with the approach taken in 3.8.3 - provide examples of classifications but do not define classifications.

BP2 - 3.8.5 Provide for Both Graceful Degradation & Progressive Enhancement of CSS

aconnors: graceful degradation... more detail than what we might need in this section sometimes

<scribe> ACTION: aconnors to refine the wording on graceful degradation trimming and making more explicit the text [recorded in http://www.w3.org/2008/10/21-bpwg-minutes.html#action05]

<trackbot> Created ACTION-874 - Refine the wording on graceful degradation trimming and making more explicit the text [on Adam Connors - due 2008-10-28].

hendry: what about fallbacks? (when some feature is missing)

[nacho can barely hear jeffs]

jeffs: should encourage graceful degradation... when some technology is not supported, provide an alternative content/app to the browser.. exploit device capabilities that the server side knows somehow

hendry: i'd like some words about AJAX here

aconnors: the concept is like the one in google web toolkit... some JS to detect stuff and then the "real" JS for what the browser actually has to do

jeffs: not just degrade, also talk about enhancing in the doc

aconnors: it is useful to recommend to take decisions on the device too (not only server-side)

jeffs: for example, i am offline so proceed accordingly

jo: there are deeper underlined questions, about what is the guidance promoting here

[nacho begs on his knees to jo to scribe his words later as it was hard for him to follow his words]

Jo: any assessment you make about the norm is prejudiced by the market you're in, who you're trying to serve, etc
... your judgment is likely to be fragile in any case
... so the best advice to give is to tell people to take this into account

adam: should the section on handling devices variation have guidance on what kind of detection to trust?
... e.g. favor server-side detection, and fall back to on-device decision
... including things that can not be determined a priori
... would be good to identify which of these priorities are

Jo: e.g. battery level, signal coverage

JeffS: I think pushing for graceful degradation is important
... we should advice developers to be prepared to vary based on their mid-point device

Adam: you can do adaptation by having a script that adapts given values to the devices that host it
... or have different bundle of javascript/stylesheets served on a device per device basis
... do we want to promote one over the over?
... it seems like something that is extremely dependent of the applications

Jo: we can point out the trade off
... although developers would more likely to be looking for more concrete advices

DKA: still think it's worth highlighting
... if we discuss it more

Jo: we need to point out the variables: time to download, latency, ...

Adam: the cost of maintenance is also an important one
... so we would refactor 3.8.5 to say to favor server-side capabilities detection
... take into account that some capabilities cannot be determined server side
... balance adding device conditional statements to a single applications vs having different bundles per devices or devices classes

<DKA> PROPOSED RESOLUTION: BP2 - refactor 3.8.5 to prefer server-side capability detection ; in some cases, client side cap detection is necessary, in which case: balance adding device specific statements into a single applications or having different modules / bundles ...

DKA: "some of my best friends are media queries"

<aconnors> PROPOSED RESOLUTION: BP2 - refactor 3.8.5 into the form of 3 recommendations: 1) prefer server-side capability detection ; 2) in some cases, client side cap detection is necessary (example); 3) balance adding device specific statements into a single applications against serving separate device-specific modules / bundles.

<nacho> +1

<DKA> +1

<aconnors> +1

<jo> +1

<jeffs> +1

<francois> +1

<abel> +1

RESOLUTION: BP2 - refactor 3.8.5 into the form of 3 recommendations: 1) prefer server-side capability detection ; 2) in some cases, client side cap detection is necessary (example); 3) balance adding device specific statements into a single applications against serving separate device-specific modules / bundles.

Device Description Repository (DDR) Simple API interlude

Rotan: after many years of work, the DDWG is coming to a close
... the charter concludes at the end of this month
... hopefully coinciding with the publication of the DDR API as W3C Rec
... the API comes with a core vocabularies that contain essential properties for basic content adaptation, particularly in the mobile space
... but it is only illustrative
... it defines properties such a height, width, accepted mime types
... you can use these properties to query the repository
... which can run in front of any kind of system (e.g. a database)
... The API is a server-side API, described for the Java language but usable in a variety of other languages, incl. IDL, WSDL, C#
... the API is expected to be used by anybody providing adaption technologies
... the existing solutions operate in silos
... we're hoping this group will be able to refer to this API when talking about adaptation
... This API is qualified as "simple", and is quite easy to implement
... it can serve as an illustration of what more complex systems could do
... There are other ways to do adaptation: DCCI (developed by the DIWG/UWAWG)
... that maps devices capabilities into a DOM structure available to javascript
... there is OMA's DPE (Device Profile Evolution), where a server and a device work in concert and enables the server to retrieve information from the server on the fly
... e.g. whether the user has turned the sound off
... so useful for information not available a-priori (thus not relevant for the DDR use case)
... The vocabulary is defined in a very simple way, with references to simple type (string, booleans, etc)
... it doesn't have a formal definition for its concepts, though
... we're hoping that the ongoing work in UWA to provide a common ontology could be used to provide formal semantics for our vocabulary
... We expect new vocabularies will be created - e.g. the Korean market place could use a new vocabulary for its use
... The danger might be that without a common ontology, digression between these vocabularies could appear
... so having this common ontology with units, dimensions of scripts, etc is critical

<abel> Delivery Context Ontology latest public draft-->http://www.w3.org/TR/dcontology/

Rotan: We will be keeping our Wiki alive even after the end of the group's charter
... with commitment from editors to maintain it

Seungyun: where are the existing implementations of the API?

Rotan: there is a Java module derived from the spec
... available to developers
... there is also an IDL description, and OWL description, coming a c# description

<francois> Java representation of the DDR Simple API

DKA: so UWA has a work item to continue some of your work?

Rotan: yes, the ontology is a continuation/complement of our work
... it should be usable to formally define our core vocabulary
... and should reduce the risks of incompatibilities across vocabularies

Jo: what's the ETA for the first version of the ontology?

Rotan: UWA is meeting at the end of the week - would know more afterward
... we're hoping to have a sufficiently clear ontology in the next 6 to 12 months to describe the core vocabulary
... but we need it to work across the various ubiquitous web applications
... I intend to continue pushing for this work in UWA

Liaison with ETSI - Statement from Bruno

Bruno wants us to review the work of ETSI

Jo: [reads statement]

<jo> text from bruno:

<jo> A team of mobile usability experts in ETSI Human Factors I am leading are

<jo> developing a set of guidelines for generic user interface elements for

<jo> 3G/UMTS mobile devices, services and applications.

<jo> Detailed information and an earlier draft is available at

<jo> http://portal.etsi.org/stfs/STF_HomePages/STF322/STF322.asp

<jo> The work provides:

<jo> 1) Infrastructure and device-related guidelines:

<jo> 2) Guidelines for services, media and applications: and

<jo> 3) Guidelines for other (related) areas.

<jo> A topic of 2) is "Mobile Internet access and development guidelines", where

<jo> deliverables are a strong reference and starting point.

<jo> This and possibly, several other topics addressed may be of interest to MWBP

<jo> experts. We'd like to warmly invite those interested to review the stable

<jo> draft (to be released by the end of the week) and provide their comments and

<jo> feedback.

<jo> An availability announcement will be made to the list(s).

<jo> For further information and any questions, contact bruno@vonniman.com"

Jo: It is likely that there are some things in here that apply to Mobile Web Best Practices

Jeff: Seems like a short amount of time to review
... Is is Ok to share this with some of the ACI folks?

<jo> ACTION: Jeff to scope current draft and see what aspects may be of interest to us. [recorded in http://www.w3.org/2008/10/21-bpwg-minutes.html#action06]

<trackbot> Sorry, couldn't find user - Jeff

Jo: I think it is in the public domain.

<jo> ACTION: Sonstein to scope current draft and see what aspects may be of interest to us. [recorded in http://www.w3.org/2008/10/21-bpwg-minutes.html#action07]

<trackbot> Created ACTION-875 - Scope current draft and see what aspects may be of interest to us. [on Jeffrey Sonstein - due 2008-10-28].

BP2 - 3.8.4 Provide Alternatives to Client-Side Scripting

Adam: Continuing the theme of handling device definition variability
... The question is do we want to say "Provide Alternatives to Client Side Scripting"
... My concern is that no all devices can do this. It may just be ignored if people can't follow it.
... Not all apps have non-scripting counterparts.

DKA: Could we say something that if your app must have Javascript, we give them a nice message.
... The "upgrade your browser" message is not great, but better than junk on the screen.

Adam: What is the recommended response?

Jo: use 406 response
... message goes in the body.

Rotan: Sometimes the error message itself is a source of errors.

DKA: Error message should be mobileOK.

Francois: Not all mobile browsers handle 406. This could confuse the user.
... Some browsers just say there is a connection problem.
... just a generic error.

Jo: Isn't it better practice to send a response "javascript not supported"?

Adam: No, we want to prefer server side detection.
... For an app that requires JS, it is best just to say "Your device cannot support this app."

Jo: 406 is the right option.

Francois: Unfortunately, it is problematic for some browsers.

Jo: How about 200 response with an auto-refresh to a 406.

Adam: Should I put in 406 for now?

Jo: How about a placeholder for now.

Jeff: A point of tension here.

Kai: We are dealing with possible bad app design. May be a bad assumption that a device has JS.

Jo: What about a game that requires JS?

Adam: Not all web apps can work without JS.

Kai: shouldn't have a link to the app in the first place.

Jo: What if my friend sends a link.

Jeff: I think a message to the user would be best.

Kai: What about the <noscript> element.

Adam: You might have to download something large to find the <noscript>
... May be better to detect at the server side.

Jeff: Should we list out the cases?

Adam: Could be a separate BP: Always include <noscript>

Jo: Need to separate cases where JS is not supported and where it is turned off.

Adam: I was think of just a simple BP where <noscript> should always be included.

<jo> PROPOSED RESOLUTION: Include noscript wherever you have script

<aconnors> +1

<rob> +1

RESOLUTION: Include noscript wherever you have script

<jeffs> +1

<aconnors> PROPOSED RESOLUTION: Change 3.8.4 Provide alternatives to Java-Script to be "Return 406 if you detect on server that javascript not supported" -- with a note that if a non-javascript counterpart to your application can possibly be supported this is of course desirable.

Jeff: What happened to the <noscript>?

Adam: separate resolution.

Jo: <noscript> is for if JS is turned off

<aconnors> PROPOSED RESOLUTION: Change 3.8.4 Provide alternatives to Java-Script to be "Return 200 and provide explanatory note + navigation if you detect on server that javascript not supported" -- with a note that if a non-javascript counterpart to your application can possibly be supported this is of course desirable.

Jo: We should say return 200 instead of 406 since we are a BP group, not a theory group. Include an informative note.

Rotan: Wouldn't a good CT be able to adapt a 406 to a 200?

Jo: Good point.

Jeff: If there is CT, here is the route you take, if not here is the route you take.

Jo: Unless you know the 406 will be handled, return a 200.

Kai: Search engine will be confused by a 200 and think it is a real page. Best to use real error code.

Francois: But we are talking about user experience.

<aconnors> PROPOSED RESOLUTION: Change 3.8.4 to: Ideally support a non-javascript version of the application. If there is no sensible non-javascript counterpart, inform the user with: 406 if you are confident end-user will receive the message, 200 and explanatory note if not.

Kai: We had this exact problem where 200 error messages were being indexed.

<jo> +1

<rob> +1

<nacho> +1

<abel> +1

RESOLUTION: Change 3.8.4 to: Ideally support a non-javascript version of the application. If there is no sensible non-javascript counterpart, inform the user with: 406 if you are confident end-user will receive the message, 200 and explanatory note if not.

<jeffs> +1

<DKA> +1

Jo: we now have 406 error support.

BP2 - 3.8.2 Use Reliable Methods for Determining Script Support

Adam: 3.8.2 User reliable methods to determine script support. Do we want to reword this?
... seems like a loaded statement.
... maybe this gets absorbed in the previous resolution.
... I suggest that 3.8.2 will go away to become a part of 3.8.6.

<jeffs> PROPOSED RESOLUTION: merge 3.8.2 and 3.8.6

<jo> PROPOSED RESOLUTION: Fold section 3.8.2 into subsequent sections where is is more naturally covered

<rob> +1

<jo> +1

<jeffs> +1

RESOLUTION: Fold section 3.8.2 into subsequent sections where is is more naturally covered

PhilArcher: Still interested in BPWG, but as able to contribute.

Adam: Would like to receive feedback on the new sections: user interface and latency.

Jo: Let's do user interface.

BP2 - 3.7 User Interface

Adam: 3.7 user interface.
... Is there anything we want to add to this?
... covers both UI things and using things like URI schemes, etc.
... Any comments on these BPs?

<aconnors> http://www.w3.org/2005/MWI/BPWG/Group/Drafts/BestPractices-2.0/ED-mobile-bp2-20081008

Jeff: We're encouraging them to use scripting.

Adam: BP1 is targeted at the DDC; this is targeted at something higher.
... In an ideal mode you'd have a JS and non-JS version so you'd pass BP1.

Jo: Wording needs to be better, but no objections to 3.7.2
... Should say "If scripting is available then...". If not available, then ....

Adam: There are a lot of BPs that assume scripting is available.

Jo: Anything else on 3.7.2?

Adam: would like the title to be better.
... current title seems broad.

Jo: Seems true when we are trying to avoid latency.
... Can you wordsmith it?

Adam: Don't have any particular opinions on this section, just want to discuss it because it is new.

Kai: Has anyone talked to the Multi-Modal Interaction Group?
... The are focusing on the correct mode of input whatever the situation is.

Adam: The focus of this document is say things like don't put links a long way apart if there is no pointing device.

Jo: Could we go through this. Looks good. Avoid saying things like "more recent" to avoid dating it.
... We need to be clear on the words we use for markup--don't use "tags", use "elements". Use ECMAScript instead of JavaScript.

Kai: We have some example code that uses ActiveX.

Jo: This should be in an appendix.
... For 3.7.3, can we change "Don't Move the Focus" to "Maintain Focus"

Jeff: How about "Preserve the focus"

Jo: I thought this was about reflowing and keeping the focus from disappearing for a while.

Adam: What is it we want to tell people to avoid?

Jo: Use of the focus method is bad practice. Restore the focus to where it was if it goes away.

Adam: We can reword it so that we preserve focus.

<DKA> PROPOSED RESOLUTION: we agree.

<jeffs> +1

BP2 - 3.7.5 Use hash URLs to Preserve Browser History and Enable Deep Links

Adam: On to 3.7.4
... 3.7.5

Jo: Would "Provide direct links to embedded views" be a better title?

Jeff: How about "internal links"?

Jo: Title needs to change.

DKA: I like "deep links"
... Web apps have a tendency to not allow deep links.

Rotan: Need to be careful with deep links since they assume an app that is hierarchical, which is not always the case.

Jeff: What about hash URLs?

<DKA> http://en.wikipedia.org/wiki/Deep_link

Jo: Fragment URIs would be better terminology.

Phil: I like "fragment URIs" because hash URIs is ambiguous.

Jo: This would be a good one to have an example for in the appendix.

<DKA> http://en.wikipedia.org/wiki/Deep_link#Court_rulings

<jeffs> http://en.wikipedia.org/wiki/Internal_link

Adam: Should we have examples in an appendix?

Jo: Yes.

<aconnors> PROPOSED RESOLUTION: Change title of 3.7.5 to: Use Fragment URIs for Embedded Views

<jeffs> +1

<francois> +1

<rob> +1

Jo: I find 3.7.6 as overly UA-Prof oriented.

RESOLUTION: Change title of 3.7.5 to: Use Fragment URIs for Embedded Views

<aconnors> PROPOSED RESOLUTION: Add section to Appendix to for examples, and link to example for 3.7.5 here.

<jeffs> +1

<rob> +1

RESOLUTION: Add section to Appendix to for examples, and link to example for 3.7.5 here.

BP2 - 3.7.6 Use URI Schemes for Device Functions

Jo: This makes reference to WTAI and UAProf.
... Isn't this overly general?
... Actually what we are saying is to decorate telephone numbers with a tel: URI to facilitate click-to-call.
... WTAI as far as I am aware has limited implementation.

Rob: More phones use wtai than use tel:, especially low-end handsets.

Jo: There are lots of features of WTAI, like "put this in my address book" etc... How widely supported are those features?

Jeff: Note that WTAI link doesn't point to the right place...

Francois: I read a document (not W3C) mobile web best practices that most modern devices support both.

DKA: (to Seungyun) do Korean browsers support tel:?

Seyngyun: yes. We don't use WAP-related functions [such as WTAI].

Jo: I think we should encourage the use of tel: because it's an RFC (RFC 3966).

<jeffs> http://www.ietf.org/rfc/rfc3966.txt

Jo: I don't think that the reference to UAProf is relevant.
... There are already references to UAProf to finding device capabilities.

PROPOSED RESOLUTION: Change 3.7.6 - encourage use of tel: discourage use of wtai:. Eliminate reference to UAProf.

<jeffs> +1

+1

<seungyun> +1

<SeanP> +1

<rob> +1

PROPOSED RESOLUTION: Change 3.7.6 - encourgae use of tel: discourage use of wtai:. Eliminate reference to UAProf. Title should be "Enable 'Click-To-Call'"

<aconnors> +1

<SeanP> +1

RESOLUTION: Change 3.7.6 - encourgae use of tel: discourage use of wtai:. Eliminate reference to UAProf. Title should be "Enable 'Click-To-Call'"

<jeffs> +1

Adam: Any other discussion around user experience?

Dan: Support Scott's input on different input modes.

Jo: Likewise.

BP2 - Latency

Adam: Regarding javascript optimization -- it would be useful if anyone is an expert on eg. webkit -

Jo: Maybe this could be a useful appendix?

Adam: it will be a list of tips.
... Maybe this is too much detail?

Jo: Where does application best practice stop and good programming style start?

Dan: I think this is very important in the mobile context - developers need to know it because they might not consider these issues if they are used to desktop PCs...

Jo: Impact on application-perceived latency and battery life of these kinds of optmiizations.

Jeff: there is no imperical work on this afaik

<scribe> ACTION: Adam to send to Jeff list of test cases that Jeff can inflict on a grad student. [recorded in http://www.w3.org/2008/10/21-bpwg-minutes.html#action08]

<trackbot> Created ACTION-876 - Send to Jeff list of test cases that Jeff can inflict on a grad student. [on Adam Connors - due 2008-10-28].

<JonathanJ> late comment for 3.7.6 - I think we need review this part of apple's document - https://developer.apple.com/webapps/docs/documentation/AppleApplications/Reference/SafariWebContent/AppleApplicationsURLSchemes/chapter_950_section_1.html#//apple_ref/doc/uid/TP40006513-SW1

BP2 - JSON vs. XML - the final conflict

Jo: Are we advocating use of JSON over XML?

Adam: JSON is far better than XML.

Dom: I don't think anyone would object to saying "send a more compact format to a mobile device."

Adam: Also it's quicker to parse JSON.

PROPOSED RESOLUTION: We recommend to use a compact format (e.g. JSON) as opposed to a less compact one (e.g. XML) for XHR requests.

PROPOSED RESOLUTION: We recommend to use a compact / natively processable format as opposed to a less compact one for XHR requests.

Dom: Some javascript engines are starting to have a native JSON parser in them - but other than that the EVAL function is used which can be a security issue.

<jeffs> +1

Warning Warning Aooooga.

Jo: 3 minute update coming up on POWDER. Then we can review proceedings and ajourn.

Jo: In the next 10 minutes: POWDER, accessibility, next f2f, brief review of this meeting, wrap

Pre-discussion on Accessibility documents

Jo: Yeliz's document is done.

Jo: Re. Alan's document, situation is that it can't be brought to a conclusion yet because pending some resolutions in the Education and Outreach Working Group (EOWG)?

Francois: The problem was the "addressing both" page.

Jo: Either way, I'd like this work item to be concluded. If there's consensus on this, [I'd like to resolve] that we have made all contributions we are able to within this group.

Dan: yes I'd like to take a resolution.

Jo: We must verify with Shawn [from EO group].

Jeff: Just so everyone knows - we [RIT] are interviewing people from various disabled groups....
... we are trying to come up with a set of recommendations to developers...
... could be made available to W3C as an input if they wish to take it up...

POWDER for mobileOK

Phil: POWDER is based on the idea of labeling websites - reference TBL quote from 15 Sept 2008.
... We should be done by the end of this year.
... [presents description resource code] Any powder document must have an attribution element, and an issued by element. That links to somewhere. This is all XML. You can process powder in a pure XML environment without having to parse RDF.
... IRIset grouping [is the key part of the POWDER spec] - can be as complicated or simple as you like.
... [something complicated about math and logic]
... [presents MobileOK claim POWDER example]

<dom> mobileOK logo

Phil: MobileOK claim can be transformed into an OWL document.
... You don't ever have to handle that - just showing OWL document to prove that it does have tha flexibility if needed.
... We MAY issue a 2nd last call this week.
... We will meet our CR exit criteria already [so we may move directly to PR].

Rotan: One of the pieces of data in there is supported by: can you have multiple such?

Phil: yes

Jeff: Is there a schema?

Phil: yes.
... Kevin Smith from Vodafone wrote it and it's really clever.

<dom> Powder WG home page

Accessibility documents

[follows above discussion, in presence of Shawn and Andrew from the Education and Outreach Working Group (EOWG).

Jo: 2 things - confirmed status update on documents from Alan and Yeliz and then where we go next

Shawn: On "shared web experiences" document that Yeliz is doing - it's a WAI resource, not a note and not a doc so can be modified more easily [but is basically done] will be published soon.
... On the "relationship between..." document that Alan is working on - once WCAG 2 is finalized, we can go back to this document and be ready to call this done.

Jo: What is the solution of the problem to the "together" section?

Shawn: Alan said mapping is too complex to do a mapping table - so what can we tell people who want to do both? We agreed to tell people to first look at "shared web experiences" for an overview - next WCAG 2, then look at the mapping in this document between wcag 2 and BPs, then look at the MWBP.
... Those steps are in the current version of the document. We will reword slightly.
... that's where EO is. Our question is: where are you [BP]?
... What level of review do you want to have at what point?

PROPOSED RESOLUTION: We are happy with both documents. Any changes to Alan's document that will happen next will be as a result of changes to WCAG 2. In essence we are done.

Shawn: With "shared experiences" we thought we were done and then we got a lot of comments... So a little considered that Alan's document hasn't got that level of review. I would like to go ahead and publish the version after this next set of changes since it's better than what's out there.
... we'll need Alan to make one more round of changes.

Dan: I think we need to build some kind of review in for Alan's document.

<francois> ACTION: daoust to create a poll on Alan's document once it's published as an updated working draft with a review period of 4 weeks [recorded in http://www.w3.org/2008/10/21-bpwg-minutes.html#action09]

<trackbot> Created ACTION-877 - Create a poll on Alan's document once it's published as an updated working draft with a review period of 4 weeks [on François Daoust - due 2008-10-28].

Next F2F meeting

Dom: Some budget issues coming up with regard to team participation in f2f meetings...

Jo: We assume we will want to meet in June.
... Before then we will need to move to last call on BP2 sometime this year. We'll want to review lc comments sometime in the new year. CT will be through a 2nd last call by then. early new year would be a good option. February is a bad option ebcause of MWC in barcelona.
... March is too late.

Dan: I don't think March is too late.

Jo: yes it is.

Dan: nu-uh

[Discussing week of January 26th for next f2f meeting]

[26th is national holiday in Korea - sunday, monday tuesday]

Dan: Ok - week beginning Feb 2nd?

[we will take resolution next meeting]

<dom> ACTION: Francois to create a poll on having f2f in seoul on week of February 2nd [recorded in http://www.w3.org/2008/10/21-bpwg-minutes.html#action10]

<trackbot> Created ACTION-878 - Create a poll on having f2f in seoul on week of February 2nd [on François Daoust - due 2008-10-28].