W3C

Mobile Web Best Practices Working Group Teleconference

11 Nov 2008

Agenda

See also: IRC log

Attendees

Present
francois, tomhume, rob, SeanP, Eduardo, jo
Regrets
Chair
francois
Scribe
rob

Contents


Welcome to Eduardo Casais

francois: Welcome Eduardo Casais
... already known to us from a host of useful comments on the mailing-list

[round of introductions]

Eduardo: I work for a small mobile content developer in Switzerland, previously worked for Nokia where I was involved in many things including UA-Prof

<francois> New draft of CT

New Draft

jo: I could talk at length on this!

<francois> Jo's changelog and discussion

jo: I hope it includes all the resolutions so far
... Hope to gather together anything missing (eg rewriting HTTPS links holes) in the next 2-3 days
... considerable polish is required but would be a waste of time until the substance is stable, which may take another draft
... it will be useful if everyone gives this draft a deep review
... santiy-check and make sure it says what we agreed
... check for consistency
... tighten the nuts and bolts (eg any shoulds that could be musts?)
... and check if clarity can be improved

francois: everyone should send comments to the mailing-list
... just like Eduardo has been doing

Content-Location

francois: introduced by Rob on the mailing-list.

rob: Yes. Don't think there is anything to do, but comes from a long discussion, so I wanted to check whether there was a reason not to include a Content-Location header in the response passed downstream to the phone

jo: do we need to propose text around this?

francois: Rob's said he doesn't think there is anything to propose here

Eduardo: RFC says the value of Content-Location also defines the base URI of the entity

francois: that's also one of my fears
... and <link> is more correct

rob: I just wanted to see if anyone has a need for this at all
... and so far I've heard no reason to want it

francois: could be a way to pass the canonical URI to the client for bookmarking

rob: I tried that on a real phone and it doesn't work

francois: correct, no-one uses that right now

<jo> PROPOSED RESOLUTION: No need to mention Content-Location header

francois: does anyone want us to go to TAG to check?

<francois> +1

<tomhume> +1

+1

RESOLUTION: No need to mention Content-Location header

OMA Standard Transcoding Interface

francois: no overlap at the moment between OMA doc and our doc
... OMA doc is an interface for a server, not for a proxy
... so there is no incompatibility and no overlap
... but there is some media-transcoding vocab defined that could be useful in future work

jo: can we resolve LC-2051 now then?

<jo> PROPOSED RESOLUTION: ref LC-2051 This is out of scope for our document but may have interest for a general transformation vocab

+1

<tomhume> +1

<francois> +1

<SeanP> +1

<jo> +1

RESOLUTION: ref LC-2051 This is out of scope for our document but may have interest for a general transformation vocab

<francois> Close ACTION-868

<trackbot> ACTION-868 Review OMA STI to see if there's something relevant for CT for LC-2051 closed

Draft responses to "resolved no" comments

francois: reminder to everyone to propose responses back to contributors ASAP

LC-2053 - classes of devices

<francois> LC-2053

francois: postponed LC-2053 response recently
... about classes of device
... and section 4.1.5 of the old draft

<francois> LC-2053

<francois> http://www.w3.org/2006/02/lc-comments-tracker/37584/WD-ct-guidelines-20080801/2053

jo: Eduardo can you please clarify what you want from LC-2053

<jo> ACTION: casais to review LC-2053 and clarify to group [recorded in http://www.w3.org/2008/11/11-bpwg-minutes.html#action01]

<trackbot> Created ACTION-880 - Review LC-2053 and clarify to group [on Eduardo Casais - due 2008-11-18].

Unclear form encoding must be preserved for the server

francois: again triggered by one of Eduardo's comments
... current wording is unclear about if a CT-proxy must roll-back encoding changes made in responses when a form is submitted

jo: which exceptions are we talking about?

francois: section 4.1.5

<francois> Alteration of HTTP Header Values

francois: the previous draft talked about Content-Encoding changes in the body, this has been removed

jo: we previously talked about transforming request bodies but decided we didn't add much (if anything) to the requirements that already exist
... do we have anything to say on something happening today that we want to stop happening?

Eduardo: the mowser transcoder doesn't handle Character-Encoding properly
... eg UTF-8 characters end up as Latin-1

jo: mowser hasn't been updated for a while and has lots of bugs, this is a known bug

Eduardo: Vodafone ES and PT transcoders don't handle numerical entities well

jo: but these are clear bugs, not debatable ambiguities
... so avoiding carelessness or error is not part of our specification

francois: we talked about adding Encoding to the appendix E list
... as merely a list of heuristics

<jo> PROPOSED RESOLUTION: On the subject of character encoding, we have revisited it and we still can't think of anything useful other than "avoid bugs" when transforming between character encoding so we have decided to leave it

francois: was written as "recoded or restructured"

jo: I remember something about this but could not find a resolution to follow when writing the latest draft
... so here is one:
... I don't think we need to talk about transforming the encoding of a request body

<jo> PROPOSED RESOLUTION: On the subject of character encoding, we have revisited it and we still can't think of anything useful other than "avoid bugs" when transforming between character encoding and to note that this is an example of a heuristic that the proxy may take into account when transforming content if it thinks that the encoding provided may mis-operate when presented on the client

francois: the only case is when the encoding of the response has been changed and the client is then submitting a form from that response

Eduardo: this isn't a heuristic, it's a rule

rob: yes, the rule is if you change Character-Encoding in one direction (server-to-phone) you have to change it in the other direction (phone-to-server) as well

Jo: we're not in the business of specifying how to transform images, HTML etc, so we don't need to specify this either
... this is not our job to write a "building transcoders for dummies" book

francois: I second that point, we don't need to expand on that in the Guidelines

jo: do we want to add this to 4.2.8.1? Because it's not a heuristic

<jo> PROPOSED RESOLUTION: On the subject of character encoding, we have revisited it and we still can't think of anything useful other than "avoid bugs" when transforming between character encoding (which we don't intend to say) and add it to the list in 4.2.8.1 so that character encoding is specifically referred to

+1

<francois> +1

<SeanP> +1

<tomhume> +1

<jo> PROPOSED RESOLUTION: On the subject of character encoding, we have revisited it and we still can't think of anything useful other than "avoid bugs" when transforming between character encoding (which we don't intend to say) but add it to the list in 4.2.8.1 so that character encoding is specifically referred to

RESOLUTION: On the subject of character encoding, we have revisited it and we still can't think of anything useful other than "avoid bugs" when transforming between character encoding (which we don't intend to say) but add it to the list in 4.2.8.1 so that character encoding is specifically referred to

<jo> ACTION: jo to enact resolution on 4.2.8.1 ref adding character-encoding to the list of format, layout, dimensions etc. [recorded in http://www.w3.org/2008/11/11-bpwg-minutes.html#action02]

<trackbot> Created ACTION-881 - Enact resolution on 4.2.8.1 ref adding character-encoding to the list of format, layout, dimensions etc. [on Jo Rabin - due 2008-11-18].

Eduardo: do we need to point out the requirements of what the server expects?

<jo> PROPOSED RESOLUTION: do not discuss alteration of request body in respect of character encoding

francois: do we need to mention that altering the request-body isn't envisaged except "rolling back" changes like Character-Encoding?

<francois> +1

<SeanP> +1

<tomhume> +1

rob: there are a few other changes that need rolling back too - for example pasting back together inputs that got split amongst sub-pages

+1

RESOLUTION: do not discuss alteration of request body in respect of character encoding

Summary of Action Items

[NEW] ACTION: casais to review LC-2053 and clarify to group [recorded in http://www.w3.org/2008/11/11-bpwg-minutes.html#action01]
[NEW] ACTION: jo to enact resolution on 4.2.8.1 ref adding character-encoding to the list of format, layout, dimensions etc. [recorded in http://www.w3.org/2008/11/11-bpwg-minutes.html#action02]
 
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.133 (CVS log)
$Date: 2008/11/11 17:05:59 $