RE: [widgets] Minutes from 30 October 2008 Voice Conference

Hi all,

I think there is considerable interest in a broad community in the topic of ZIP based packages, specifically MIME types for them and intra-package URI references within them, and possibly for standardizing metadata as well.

Procedurally, I don't think it is appropriate to attempt to resolve these issues in the WebAPP working group, if only because a number of the affected groups have little additional overlap with WebAPPS. I know the W3C TAG has discussed the URI issues at some point.  I'm not sure if the overhead of starting a new W3C working group focused specifically on this topic is too high, but if so, an IETF activity with W3C participation might be a way of getting broader participation, as well as getting additional IETF involvement in the MIME/URI issues.

Larry




From: Arthur Barstow [mailto:art.barstow@nokia.com]
Sent: Friday, October 31, 2008 5:40 AM
To: Jon Ferraiolo; Richard Cohn
Cc: Marcos Caceres; Bill McCoy; Larry Masinter
Subject: Re: [widgets] Minutes from 30 October 2008 Voice Conference

Jon - thanks for your e-mail and contact information!

Richard - thanks for this information. Ideally, it would be best if Adobe would directly participate in our effort i.e. join the WebApps WG. If that isn't going to happen then we would appreciate any feedback Adobe has on our suite of Widgets specs [Widgets], with the top priority ATM being our Packaging and Configuration spec, latest Editor's Draft is:

 <http://dev.w3.org/2006/waf/widgets/>

-Regards, Art

[Widgets] <http://www.w3.org/2008/webapps/wiki/Main_Page#Widgets>


On Oct 30, 2008, at 12:50 PM, ext Richard Cohn wrote:


For what it's worth, here's some info on Microsoft's URI scheme:

http://msdn.microsoft.com/en-us/magazine/cc163372.aspx#S9

I think we considered doing something like this but decided not to because from a practical standpoint, it's difficult to introduce new schemes (unless, say, you own a browser :-) ). On the other hand, I seem to recall that we used some sort of scheme during the processing of a Mars document but didn't persist it in the file. I can't speak for the people currently responsible for Mars, but I suspect we'd be interested. I'll forward this message within Adobe.

Richard

At 12:14 PM 10/30/2008, Jon Ferraiolo wrote:


Hi Art and Marcos,
Regarding the widget URI scheme, you might also want to reach out to Adobe to find out whether there is anyone from the Adobe Digitial Editions world (ebooks/epubs) or PDFXML world (Adobe Mars) who might have an opinion on the pros and cons of a widget URI scheme. When I left Adobe 2.5 years ago, both ADE and PDFXML were using pretty much the name ZIP packaging approach that is being used by W3C Widgets. There was definitely talk about inventing some sort of zip: protocol for relative referencing within the ZIP package, but I don't believe it ever became part of any specs.

I used to have the Microsoft ZIP spec memorized (the one that they use for XML Paper Specification) but now I can't even remember the name of the spec. They might have invented a new protocol for addressing objects inside the ZIP package.

I have copied two of my old buddies at Adobe, Richard Cohn and Bill McCoy. Probably neither of them will respond directly, but they might be able to get other people from Adobe (or other companies) who are actively involved in ADE or PDFXML to comment on the widget URI scheme.

Jon

PS Hi Richard and Bill!


<b0166e.jpg> Arthur Barstow <art.barstow@nokia.com<mailto:art.barstow@nokia.com>>
Arthur Barstow <art.barstow@nokia.com<mailto:art.barstow@nokia.com>> Sent by: public-webapps-request@w3.org<mailto:public-webapps-request@w3.org>

10/30/2008 07:10 AM
<b0167d.jpg>
To

public-webapps <public-webapps@w3.org<mailto:public-webapps@w3.org>>
<b0167d.jpg>
cc

<b0167d.jpg>
Subject

[widgets] Minutes from 30 October 2008 Voice Conference


The minutes from the October 30 Widgets voice conference are
available at the following and copied below:

 < http://www.w3.org/2008/10/30-wam-minutes.html>

WG Members - if you have any comments, corrections, etc., please send
them to the public-webapps mail list before November 6 (the next
Widgets voice conference); otherwise these minutes will be considered
approved.

-Regards, Art Barstow


   [1]W3C

      [1] http://www.w3.org/

                               - DRAFT -

                       Widgets Voice Conference

30 Oct 2008

   [2]Agenda

      [2] http://lists.w3.org/Archives/Public/public-webapps/
2008OctDec/0201.html

   See also: [3]IRC log

      [3] http://www.w3.org/2008/10/30-wam-irc

Attendees

   Present
          Art, Arve, Claudio, Mark, Marcos, Josh, Bryan

   Regrets
          Thomas, David, Jere

   Chair
          Art

   Scribe
          Art

Contents

     * [4]Topics
         1. [5]Agenda review
         2. [6]Annoucements
         3. [7]URI scheme
         4. [8]Version String
         5. [9]ID attribute
         6. [10]DigSig
         7. [11]AOB
     * [12]Summary of Action Items
     _________________________________________________________



   <timeless> zakim +??P18 is Marcos

   Date: 30 October 2008

   <scribe> Scribe: Art

   <scribe> ScribeNick: ArtB

   <arve> +??P19

   <timeless> zakim +39.011.228.aaaa is Claudio

Agenda review

   AB: any changes?

   [None]

Annoucements

   AB: any annoucements?

   <timeless> # who is on the phone?

   <timeless> who is on the phone?

   AB: Workshop deadline is now Nov 5

   <timeless> Zakim: who is on the phone?

   AB: who plans to submit a Position Paper?

   <Bryan> very noisy

   <arve> ArtB: Arve just spoke, and I said we were planning on
   submitting a position paper

   AB: am I the only one that cannot understand anything that is being
   said?

   <Bryan> I can't understand either

   <arve> muting me didn't help

   <arve> I can't understand a word being said

   <Bryan> I hear a 2nd conversation

   AB: everyone hang up and re-dial, please !

   <Bryan> OK

   <marcos> Arve, you are very noisy

   AB: is anyone going to submit a PP for the workshop?

   MC: I will

   AB: how about Vodafone?

   MP: no

   Arve: I believe Opera will submit a paper

   JS: no

   Bryan: we may submit something but we won't be present

   CV: we won't submit a paper but are very interested in the outcome

URI scheme

   AB: we had a good conversation with TAG last week
   ... would like to know what people think are the next steps for this
   issue

   MC: I think we have enough technical arguments to push forward
   ... we do need to fleshout the reqs
   ... We may also want to coordinate with other groups
   ... e.g. the ODF group
   ... they need a similar URI scheme
   ... for packaging

   <timeless> ack

   BS: we recognize (in OMA) that some type of URI scheme for widget
   interaction is needed

   CV: I agree with Marcos and Bryan
   ... I think a widget-specific URI scheme would be useful

   JS: I haven't changed my mind
   ... agree we need to flesh out the reqs

   <marcos> Arve, you might need to type out your answer

   <arve> ArtB: Your assesment of my opinion is essentially correct

   AB: I'd like to understand the ODF coordination point

   MC: I've had some conversations
   ... I don't want to block on them or create a dependency
   ... I will continue to talk with them

   AB: are there some actions we can assign?

   MC: think we need to look at the implications vis-a-vis the API spec
   ... we use the widget URI scheme to resolve the DOM at run time
   ... this affects the APIs we will define

Version String

   AB: Marcos
   [13] http://lists.w3.org/Archives/Public/public-webapps/2008OctDec/01
   83.html
   ... where are we on this?

     [13] http://lists.w3.org/Archives/Public/public-webapps/
2008OctDec/0183.html

   MC: there were no strong objections

   AB: OpenAjax recommend we consider their model?

   MC: they proposed another way to write the scheme
   ... they have a proc model for version strings
   ... Arve showed their model has problems
   ... I want to follow the KISS principle

   AB: I propose we agree with Marcos' version string proposal
   ... any objections?

   JS: what about leading zeros?

   MC: they are just opaque strings

   AB: Josh, please enter an example

   <timeless> MIDlet Suite Versioning suggests:

   <timeless> Major.Minor[.Micro] (X.X[.X])

   <arve> Does this mean any string difference is "a new version"

   MC: we aren't adding that complexity
   ... If they are differen, then they are different

   <timeless> do we need to suggest that we're aware that leading zeros
   are ignored by MIDlet

   <timeless> and that people should avoid using leading zeros (or at
   least inconsistently)

   <arve> What I actually meant is that "new" is that the UA, or the
   server, can decide whether it's "newer" or no

   <arve> +t

   AB: Marcos, what do you think about JS' recommendation?

   MC: we can recommend a format
   ... and that there is no special processing

   <arve> yes

   <Bryan> +1

   AB: can we live with the model Marcos has proposed?

   JS: yes

   <timeless> yes

   <claudio> yes

   RESOLUTION: Marcos' proposal for version string is acceptable

ID attribute

   AB:
   [14] http://lists.w3.org/Archives/Public/public-webapps/2008OctDec/01
   84.html
   ... there was no follow-up discussion
   ... see also [15] http://krijnhoetmer.nl/irc-logs/webapps/20081027

     [14] http://lists.w3.org/Archives/Public/public-webapps/ <http://lists.w3.org/Archives/Public/public-webapps/%A0>
2008OctDec/0184.html
     [15] http://krijnhoetmer.nl/irc-logs/webapps/20081027

   MC: the question is about whether xml:id should be used or we define
   our own ID attribute

   JS: could "name" be used?

   MC: we already have a name element
   ... we would have to rename that to "title" element

   <timeless> ack

   BS: are we planning to use the title attribute in a semantic way?

   MC: no, just a name

   <claudio> +q

   AB: Marcos, should you followup and make the proposal you and Josh
   just discussed?

   MC: yes, I can do that
   ... we are currently following what other people are doing
   ... but I'd like to hear from others

   CV: will there some semantics about the widget in the config doc?
   ... Req #12 is related to widget semantics

   MC: no, not at this point

   CV: so the manifest is extensible?

   MC: yes, can add other elements

   <timeless> ack

   MC: the name element could be use in that use case

   BS: there is a core set of metadata attributes already defined
   ... and it is extensible

   MC: right, via using another namespace

   AB: are you still looking for more input, Marcos on your ID
   attribute proposal?

   MC: I can make the change if people are OK with it

   <arve> did the channel just go dead?

   MC: my fear is confusing widget authors

   <arve> I'll have to give up on this, all audio just disappeared

   <marcos> MC: widgetid

   AB: what is your proposal?

   <marcos> MC: uid

   <marcos> MC: name

   <marcos> CV: wid

   AB: I am mostly indifferent

   MC: it is a URI to identify the widget

   JS: could use href

   MC: but that implies something that http can get

   <timeless> ack

   BS: so you want something that is unique, right

   MC: yes

   BS: what about uniqueid then?

   MC: yes, we could
   ... that's what I meant by "uid"

   <marcos> arve, do you have an opinion?

   MC: Are we providing at leas a non-normative suggestion about how to
   add semantics to the widget?

   <arve> leaning towards making it "just a string"

   <arve> I do not like the notion of saying it's an ID

   <marcos> arve, what would you call it?

   <marcos> ok, no probs

   AB: I propose you make a proposal on the mail list with a default
   resolution

   MC: OK

DigSig

   AB: I will take 4.a and 4.b agenda items to the mail list

   MC: Mark and I have been making some edits
   ... need comments from XMLSec WG
   ... perhaps that can be done while I am away

AOB

   AB: Marcos will be offline for the next three weeks; not online
   again until 21 November
   ... I will decide on Tues or Wedn of the next 3 weeks if we will
   have a voice conf on Thursdays - or not

   <arve> Have a nice trip, marcos

   AB: meeting adjourned

   <marcos> Thanks!

   RRRSAgent, make minutes

Summary of Action Items

   [End of minutes]