- From: Francois Daoust <fd@w3.org>
- Date: Tue, 12 May 2009 17:27:45 +0200
- To: Mobile Web Best Practices Working Group WG <public-bpwg@w3.org>
Hi,
The minutes of today's call are available at:
http://www.w3.org/2009/05/12-bpwg-minutes.html
... and copied as text below.
Resolutions taken during the call:
- Add an erratum to mobileOK Basic Tests, at the right time, to point to
the edited version of XHTML Basic 1.1 including the lang attribute
- on MWABP, remove second para of 3.2.1.2 and make no reference to
efficiency
Francois.
-----
12 May 2009
[2]Agenda
[2] http://lists.w3.org/Archives/Public/public-bpwg/2009May/0012.html
See also: [3]IRC log
[3] http://www.w3.org/2009/05/12-bpwg-irc
Attendees
Present
tomhume, jo, Francois, rob, EdC, adam, BruceL, SeanP, yeliz
Regrets
jeffs, kai, abel, miguel, nacho
Chair
jo
Scribe
Tom
Contents
* [4]Topics
1. [5]FYI - lang attribute back in XHTML Basic 1.1
2. [6]MWABP - new draft out. Feedback needed. Next steps?
3. [7]Mobile/Accessibility document - EOWG decision?
4. [8]Addendum to BP - next editorial session?
5. [9]CT - Abstract
6. [10]CT - Heuristics - mobileOptimized
7. [11]CT - Heuristics - rel="stylesheet"
8. [12]AOB
* [13]Summary of Action Items
_________________________________________________________
FYI - lang attribute back in XHTML Basic 1.1
francois: Not much to say on this, except we were talking about this
a few months ago and the XHTMLWG integrated the lang attribute into
XHTML Basic 1.1. The second edition is a "proposed edit
recommendation edition", close to replacing the first edition. When
it's done we can update the DTD in the mobileOK checker to have the
LANG attribute back.
... This is good news in that we're consistent with the I18N group.
jo: is there a DTD we could use if we want to?
<francois> [14]XHTML Basic 1.1 PER
[14] http://www.w3.org/TR/2009/PER-xhtml-basic-20090507/
francois: there's a schema for those who prefer it.
<francois> [15]DTD in the spec
[15] http://www.w3.org/TR/2009/PER-xhtml-basic-20090507/#a_dtd
francois: this section contains links to the actual files - the
version associated with the document and that evolving over time.
jo: when should we edit the checker?
francois: wait for the document to become a recommendation, and then
update the checker
jo: do we refer to a specific version in the checker?
francois: we're not talking about editions in the document
jo: we don't say anything specific at all, if we're talking about
mobileOK basic
francois: all we have is a link to XHTML Basic 1.1, the dated
version of the first edition. This will link to the new edition.
... mobileOK basic targets the first edition. We don't say anything
specific about DTDs, and I don't think we need to do anything to
clarify this. We could an erratum to the spec if we need to, to
explain what we mean by the XHTML Basic 1.1 DTD... that it follows
the evolution of XHTML Basic 1.1 recommendation (or not).
jo: but it's not an erratum, it's a clarification...
francois: if moving to the new DTD is a normative change (I think
it's a correction of something wrong), we can't just produce an
erratum. If we feel it's a useful clarification we can add an
erratum - that's the only way to add such comments.
... let's wait for XHTML Basic to become a recommendation.
<jo> PROPOSED RESOLUTION: Add an erratum to mobileOK Basic Tests, at
the right time, to point to the edited version of XHTML Basic 1.1
including the lang attribute
<brucel> agree
<rob> +1
<francois> +1
RESOLUTION: Add an erratum to mobileOK Basic Tests, at the right
time, to point to the edited version of XHTML Basic 1.1 including
the lang attribute
<EdC> +1
jo: francois, can you enact this pls?
<jo> ACTION: daoust to enact the resolution on XHTML Basic 1.1
revision - when it reaches rec [recorded in
[16]http://www.w3.org/2009/05/12-bpwg-minutes.html#action01]
<trackbot> Created ACTION-959 - Enact the resolution on XHTML Basic
1.1 revision - when it reaches rec [on François Daoust - due
2009-05-19].
<yeliz> +1
MWABP - new draft out. Feedback needed. Next steps?
<francois> [17]new draft of MWABP
[17] http://www.w3.org/TR/2009/WD-mwabp-20090507/
jo: we've had no feedback on this document as yet.
adam: I've had some feedback on appcache, an HTML5 feature for
caching web apps locally. It'd be good to have a BP on this subject
and discuss what form this should take. It's very HTML5-specific,
that's all the feedback I've had so far.
<EdC> What about the pending feedback from Francois and myself?
jo: How do you tell that feature is present?
adam: It's supported based on the target platform. Good question.
jo: do we want to discuss the merits of the proposal on this call,
or shall we gather some of these to discuss in a more consolidated
way once we've had feedback? The latter, I suggest
adam: agree
<jo> ISSUE: Should we have a BP on appcache?
<trackbot> Created ISSUE-297 - Should we have a BP on appcache? ;
please complete additional details at
[18]http://www.w3.org/2005/MWI/BPWG/Group/track/issues/297/edit .
[18] http://www.w3.org/2005/MWI/BPWG/Group/track/issues/297/edit
adam: re pending feedback from eduardo... the outstanding issue is
with 3.6.1 and 3.6.2. Agree with your comments, as they're
formulated right now they're a bit mickey mouse and don't reflect
reality. Not sure how to make them better - best solution might be
to make the language more woolly, talk about preferring server-side
XXX. Or if you have specific things you'd like to see in there, feel
free to propose them.
edc: I suggest you write a proposal and I'll comment on it again.
... The other issue is on sprites.
<francois> [19]3.4.6 CSS Sprites
[19] http://www.w3.org/TR/2009/WD-mwabp-20090507/#d1e8981
<jo> ACTION: adam to write a proposal in answer to EdC's comments on
3.6.1 and 3.6.2 [recorded in
[20]http://www.w3.org/2009/05/12-bpwg-minutes.html#action02]
<trackbot> Created ACTION-960 - Write a proposal in answer to EdC's
comments on 3.6.1 and 3.6.2 [on Adam Connors - due 2009-05-19].
adam: barring opinions from others, I don't have any more I can
extract through inspection. 3.4.6 and 3.4.7 are the sections; the
issue is technical and advanced, they're dependent on device support
- we have proposed icons to represent the fact that you need certain
features in the browser for them to make sense as recommendations.
There's a general issue of whether they make sense as
recommendations, or if they're "advanced tweaks"
edc: I'm reminded of multipart-mixed. if its supported (blackberry
and openwave browsers should be ok, safari has lousy support for
it), then it solves all your problems with icons and provides
greater benefits. I wonder what kind of best practice we want to put
into this document.
adam: can someone take an action to investigate multipart-mixed?
<jo> ACTION: Tom to investiagate multipart-mixed in the context of
3.4.6 and 3.4.7 of MWABP [recorded in
[21]http://www.w3.org/2009/05/12-bpwg-minutes.html#action03]
<trackbot> Created ACTION-961 - Investiagate multipart-mixed in the
context of 3.4.6 and 3.4.7 of MWABP [on Tom Hume - due 2009-05-19].
francois: I sent a comment about the security stuff (2.1 - do not
execute untrusted Javascript). Final paragraph needs rewriting.
<francois> [22]3.2.1 JSON
[22] http://www.w3.org/TR/2009/WD-mwabp-20090507/#d1e542
francois: again, it's how to find a balance between best practice
and what's acceptable. it makes sense in 99% of cases, there's 1%
where security can be impacted.
... so maybe it's not a good idea there.
adam: one option would be to remove that paragraph. Or we could
reword this paragraph to ensure that correct escaping is used with
eval(), and that this might even then not be secure.
francois: it's about not having access to sensitive data when you do
it.
adam: I'll have a crack at rewording and send to the list.
jo: wouldn't it be better for the sake of simplicity to remove the
paragraph?
francois: I'd prefer it removed
<jo> PROPOSED RESOLUTION: Remove second para of 3.2.1.2 and make no
reference to efficiency
+1
<rob> +1
<yeliz> +1
RESOLUTION: Remove second para of 3.2.1.2 and make no reference to
efficiency
<francois> +1
RESOLUTION: Remove second para of 3.2.1.2 and make no reference to
efficiency
<adam> +1
<EdC> +1
francois: one thing I can do is to write a post on the BPWG blog to
trigger reactions.
<jo> ACTION: Francois to reach out for comments on MWABP via the
BPBlog [recorded in
[23]http://www.w3.org/2009/05/12-bpwg-minutes.html#action04]
<trackbot> Created ACTION-962 - Reach out for comments on MWABP via
the BPBlog [on François Daoust - due 2009-05-19].
francois: could you reach your respective developer communities too?
I suspect we won't get much feedback, it's a working draft and not a
last call (last calls tend to trigger reactions)
jo: any idea of groups we should particularly outreach to?
<EdC> Suggestion: if anybody blogs, try to participate in the
Carnival of the mobilists: [24]http://mobili.st.
[24] http://mobili.st/
adam: feels like a dearth of feedback. Been pushing hard at work for
this.
... I'm inclined to sit on it for a week or two whilst I make
outstanding changes and wait for feedback internally. That draft can
be our LC candidate.
jo: so LC will be end this month/early june
Mobile/Accessibility document - EOWG decision?
yeliz: as far as I know there hasn't been any agreement to proceed
with the publication of the draft.
jo: any action required from our side?
francois: i've not checked emails of the education/outreach group.
We took a resolution 2 weeks ago to publish the draft as soon as
they're ok with it.
jo: so we've prior approved it... we can go ahead when ready.
Addendum to BP - next editorial session?
jo: I'm the blocker here, haven't added some further comments. We'll
need an editorial session following that.
... it'll be a couple of weeks before I get a chance to do more on
this.
CT - Abstract
<adam> +1
<brucel> bruce test
jo: on action 929, appreciate your input here Eduardo - I do have
some comments. Apologies for not replying.
... would rather make them on-list than now.
<EdC> Let us defer then.
jo: so let's defer that discussion. Apologies for not making
progress, it'll be another couple of weeks before I can address it.
... anything else on CT from anyone?
CT - Heuristics - mobileOptimized
jo: the problem with recommending specific vendor things is obvious,
but if they're prevalent in the marketplace it's equally problematic
not to mention them.
ed: MS has faced the problem of having devices that are more capable
and less capable and having several ways of viewing pages - fit to
screen, view as is, try to do something clever, etc. - and they
realised there was a need to let developers state explicitly that
the content was optimised, there's no need to try and do something
clever with it, it will display OK on mobile, etc.... so ignore the
default browser settings and do what the content provider intended.
... so it's been there for some time. It's documented in the MS
mobile windows site. Its importance is that it's a rare indication
that HTML content has been developed for mobile.
<brucel> does the metatag say *which* mobile browser/phone its
already optimised for?
ed: usually all the rules we've examined have had to do with content
types, markers, which were almost-exclusively used by mobile devices
and not by PC browsers... and this is the only indication in HTML
content that is unambiguously indicative that this is HTML content
optimised for smartphones.
... It is attached to windows mobile devices, whatever their
market-share is at present.
bruce: does it say which browser/phone the content is optimised for?
Or that I as a developer am certain this is lean and mean? I'm not
sure what the tag means.
ed: it's an indication that is meant for internet explorer
<francois> [25]Layout meta tag in MSDN
[25] http://msdn.microsoft.com/en-us/library/ms890014.aspx
<francois> [[ Web developers use the MOBILEOPTIMIZED meta tag to
control the Internet Explorer Mobile layout ]]
ed: it'll be ignored by any other browser.
bruce: so are we saying if this tag exists it must be obeyed and no
transformation must be done by any browser... or not by IE?
jo: it is conclusive evidence that the site has been created for
mobile
... the question it raises to me is that if we see evidence in the
markup that specific devices are being targeted... is it equally
conclusive evidence that its' ready for the device we're targeting
to?
... if you want to show evidence your markup is targeting mobile
there are lots of other ways of doing it.
bruce: i agree with you that we shouldn't crown any vendor-specific
marker.
<yeliz> I agree with Bruce as well
ed: there are lots of other ways of doing it. i'd like to know what
they are, given that we're restricting them.
... this is the only markup I know that will work for HTML (as
opposed to XHTML, WML, etc)
jo: you could use rel="handheld" with a self-referential link
... my preference is that vendor-specific things irrespective of
whether the device belongs to that vendor is a dangerous path.
<Zakim> tomhume, you wanted to wonder if it's evidence the site is
made-for-mobile or made-for-mobile-IE?
<brucel> I'm not against vendor-specific stuff per se, and IE can
obey its own metatags, but noone else should be bound by them
tom: Agree, this is saying IE Mobile shouldn't transform, but not
sure that this could be extended to other browsers not transforming.
<SeanP> I agree that we should avoid vendor markup
<jo> PORPOSED RESOLUTION: Adopt the MS specific <meta
name="MobileOptimized" content="NNN"> as mandatory evidence of
mobile content
<jo> -1
<rob> 0
-1
<EdC> 0
<francois> -1
<SeanP> -1
<yeliz> -1
<brucel> -`1
CT - Heuristics - rel="stylesheet"
<francois> [26]EdC's email
[26] http://lists.w3.org/Archives/Public/public-bpwg/2009May/0011.html
ed: this relates to XHTML stylesheets. You can, with some
transcoders, mark stylesheets and control how much is filtered out
by transcoders. The proposal is "if you see the XHTML stylesheet
which is marked as mobile-ready, in principle they shouldn't be
taken out". External stylesheets marked as "all" mean "good for
every device". Here the point is that there are some browsers (newer
ones especially) that will not consider ALT stylesheets but take
desktop and ALL
... since ALL covers both categories (mobile and full web), the
second part of the proposal is to say "ALL means ALL" and it
shouldn't be touched.
<Zakim> francois, you wanted to wonder about what should not be
touched
francois: what should not be touched? The XHTML content, the HTML,
the CSS?
ed: the CSS
francois: it only makes sense if you don't touch the XHTML. In many
cases you could make changes in the HTML, then CSS changes are
required... as you change the structure of the HTML page some
directives are no longer valid and don't make sense. If you change
the HTML you may need to change the CSS - not that CT proxies
necessarily do this properly...
... so we can't prevent it in the guidelines.
jo: this sounds like the discussion on the transformability or
otherwise of images. is there a parallel worth considering?
ed: you might have XHTML which does links without native for
handheld/desktop stylesheets. You might change the HTML but not
stylesheets.
<SeanP> Francois is correct; if you change the markup, you probably
will need to change the CSS
jo: sean/rob? any comment?
sean: agree with francois. If the markup is changed, good chance you
need to change the CSS too.
ed: if you don't change the XHTML, does it make sense to change the
CSS? No.
jo: this is even more similar to "if there's a no-transform on the
HTML, does this have implications to included parts". We decided
not, on images. Does the same argument apply here?
ed: The same would apply but here we're explicitly saying "yes, if
it's for a mobile device".
sean: if you have a no-transform or some other content type, it
should apply to the CSS.
jo: a derived no-transform rather than an explicit one.
<Zakim> rob, you wanted to say seems sensible to either transform
both HTML and CSS together or change neither
<brucel> I need more time to think thru after the explanations ...
rob: agree with ed and sean. typically you'd change both HTML and
CSS or neither. can't see any reason to just change CSS.
<jo> PROPSOED RESOLUTION: If the HTML is being treated as "no
transform" then external stylesheets retrieves as a consequence of
retrieving the HTML should be treated the same
<jo> PROPOSED RESOLUTION: If the HTML is being treated transparently
then external stylesheets retrieved as a consequence of retrieving
the HTML should be treated the same
<Zakim> tomhume, you wanted to wonder if we argued against "derived
transformation" as it imposed a page model on HTTP where none
existed before.
ed: at the time, we were imposing that model on a specific feature
(the no-transform directive). Here we're not saying it's linked to
that...
francois: i'm confused... I can see examples where we might want to
change CSS (e.g. absolute positions of resources).
... we're talking about external stylesheets in general, not
handheld ones.
jo: what about recursively referenced stylesheets?
ed: ALL or "handheld" would be the one to work on
jo: how far should this go?
ed: as far as the recursion goes.
jo: what about stylesheets that are referenced from stylesheets that
are themselves references as "all" or "handheld" - do they inherit
properties?
ed: logically, yes.
jo: I'd like us not to take a resolution on this without considering
our previous decision. There are caching implications here too, and
indefinite numbers of recursively referenced stylesheets.
+1
<francois> +1
<brucel> +1
<yeliz> +1
<SeanP> It does sound like we need to think about this a bit.
<SeanP> I think that is correct.
francois: another question on handheld and all. One part is about
having the "link alternate"... we agreed not to have "all" mentioned
in the list of mandatory heuristics (for reasons I can't remember).
... The same reasons should apply here as well.
... we should dig into the archives.
<EdC> Is there also a recursion problem with alternate links?
<jo> ISSUE: with reference to Eduardo's point about linked
stylesheets,
[27]http://lists.w3.org/Archives/Public/public-bpwg/2009May/0011.htm
l, we need to review in the light of an earlier decision on images
and possibly aslo in light of a recursion problem with link
rel="alternate" (per discussion of meeting on 12th May)
[27] http://lists.w3.org/Archives/Public/public-bpwg/2009May/0011.html
<trackbot> Created ISSUE-298 - With reference to Eduardo's point
about linked stylesheets,
[28]http://lists.w3.org/Archives/Public/public-bpwg/2009May/0011.htm
l, we need to review in the light of an earlier decision on images
and possibly aslo in light of a recursion problem with link
rel="alternate" (per discussion of meeting on 12th May) ; please
complete additional details at
[29]http://www.w3.org/2005/MWI/BPWG/Group/track/issues/298/edit .
[28] http://lists.w3.org/Archives/Public/public-bpwg/2009May/0011.html
[29] http://www.w3.org/2005/MWI/BPWG/Group/track/issues/298/edit
AOB
jo: anything?
<brucel> nah
<yeliz> no from me as well
<brucel> Bye, kisses all
<jo> [bye]
<EdC> bye
Summary of Action Items
[NEW] ACTION: adam to write a proposal in answer to EdC's comments
on 3.6.1 and 3.6.2 [recorded in
[30]http://www.w3.org/2009/05/12-bpwg-minutes.html#action02]
[NEW] ACTION: daoust to enact the resolution on XHTML Basic 1.1
revision - when it reaches rec [recorded in
[31]http://www.w3.org/2009/05/12-bpwg-minutes.html#action01]
[NEW] ACTION: Francois to reach out for comments on MWABP via the
BPBlog [recorded in
[32]http://www.w3.org/2009/05/12-bpwg-minutes.html#action04]
[NEW] ACTION: Tom to investiagate multipart-mixed in the context of
3.4.6 and 3.4.7 of MWABP [recorded in
[33]http://www.w3.org/2009/05/12-bpwg-minutes.html#action03]
[End of minutes]
Received on Tuesday, 12 May 2009 15:28:20 UTC