See also: IRC log
scribenic
<scribe> scribenick: slightlyoff
annevk: 0824
<noah> date: 18 April 2013
are those nicks or names?
<trackbot> Meeting: Technical Architecture Group Teleconference
<trackbot> Date: 18 April 2013
thanks
<noah> Marcos will scribe on 2 May
<noah> RESOLUTION: Minutes of 18-20 March 2013 F2F at http://www.w3.org/2001/tag/2013/03/18-agenda are approved
<noah> http://www.w3.org/2001/tag/2013/04/04-minutes
<noah> RESOLUTION: Minutes of 4 April at http://www.w3.org/2001/tag/2013/04/04-minutes are approved
<noah> TPAC 2013
<noah> 11-15 November 2013
<noah> Shenzhen, China
<JeniT> (TimBL isn't on the call, just in IRC)
thanks...is there a way to minus?
noah: we haven't talked much about TPAC, but it's a productive opportunity to talk with other groups and unless there are objections, will ask for a room
<noah> ACTION-790?
<trackbot> ACTION-790 -- Jeni Tennison to do new Editor's Draft of fragids spec for approval to publish as CR -- due 2013-04-16 -- OPEN
<trackbot> http://www.w3.org/2001/tag/group/track/actions/790
<noah> http://www.w3.org/2001/tag/doc/mimeTypesAndFragids-2013-04-04.html
<wycats_> I can commit to reviewing it for the next meeting
noah: we have a draft from Jeni, comments were received, and the new draft is http://www.w3.org/2001/tag/doc/mimeTypesAndFragids-2013-04-04.html
... current status is review to go to CR
<JeniT> wycats_, I did email it to you privately too
noah: do we need to review anything in detail?
JeniT: lets review the changes and then we can decide what to go over in detail
<JeniT> http://lists.w3.org/Archives/Public/www-tag/2013Apr/0033.html
<JeniT> http://www.w3.org/2001/tag/doc/mimeTypesAndFragids-2013-04-04.html#publishers
JeniT: have added a note to emphasize that if you have conneg, not all versions must resolve all identifiers
<wycats_> JeniT: I somehow missed the email (I see it now)
<noah> Note
<noah> Note that this best practice does not imply that every structure that is addressable within one content-negotiated representation must have an equivalent structure addressable by the same fragid in all other content-negotiated representations. It is likely that some fragids will have meaning only in one of the content-negotiated representations, for example because they are interpreted by a
<noah> script within an HTML representation but not in any others.
<wycats_> JeniT: I'm sure it's fine
<noah> Looks very good to me.
<JeniT> Scripts can also be used to map a fragid on a document that contains an embedded resource (such as an image or video) into a fragid that applies to that embedded resource. Scripts can use the fragid of the location navigated to within an iframe to alter the display of the embedding document.
JeniT: second point: pages that embed images/video and use fragids in the outer page to resolve to the sub-resource content is addressed just following the previous note (see link above)
<JeniT> http://www.w3.org/2001/tag/doc/mimeTypesAndFragids-2013-04-04.html#fragid-structures
JeniT: final major change is the addition of an appendix: http://www.w3.org/2001/tag/doc/mimeTypesAndFragids-2013-04-04.html#fragid-structures
<wycats_> if I recall correctly, my concern was just to make sure that people knew to try to create their own key/values for fragments that they didn't expect to be interoperable with other mime types
JeniT: this goes thorough many formats that define fragids and outlines them
<wycats_> looks like Best Practice 8 covers this
<wycats_> 8 and 9 I should say
JeniT: the point is to make authors aware of the other prevalant formats to avoid overlap
noah: the TAG has published a finding, and there's specific practices around hashbang...it's addressed in B7, but not mentioned elsewhere and the finding isn't referenced...should it be?
JeniT: it's referenced but not linked
wycats_: we're revisiting a lot of this now
... the outcome will be that people won't do this any more
<wycats_> specifically, insofar as we're telling people not to conflict with existing practices for other MIME types, the fact that we're calling out the use of #! on the web means that people won't use it in other contexts
<noah> I'm fine with going to CR
JeniT: the other change is the exit criteria
wycats_: what will we do about the potential JSON-pointer conflict?
JeniT: it says this is a fragid syntax you *can* use with JSON, it's not a "must"
... will expand that section in the appendix and make it explicit that if you use it as-is, it may conflict with thigns that use #/ in a different way
wycats_: if you conneg this way, yes, this is probably a bad idea
JeniT: if they're declaring conneg'd content with pointers that will map to both, the advice will be to make up your own pointer syntax in the fragid
wycats_: I care about this because I have systems that do this
JeniT: the other conflict we're likely to see will be the "+json" mimetype registration
wycats_: will take this offline and reply to your email
JeniT: if people can get back to me within the week, I'll work around wrapping up the exit criteria and JSON-pointer points and get this to Yves next week for publication
<wycats_> I'm fine with provisional approval
noah: this is about provisional approval for this...objections?
<JeniT> Yves: we agreed exit criteria at F2F, I just somehow didn't include them in the draft from what I can see
<wycats_> slightlyoff: provisional approval to go ahead with the CR with the document as-is
<noah> ACTION-790
<trackbot> ACTION-790 -- Jeni Tennison to do new Editor's Draft of fragids spec for approval to publish as CR -- due 2013-04-16 -- OPEN
<trackbot> http://www.w3.org/2001/tag/group/track/actions/790
no objection from me. Thanks JeniT for getting this done!
<noah> ACTION-790
<trackbot> ACTION-790 -- Jeni Tennison to do new Editor's Draft of fragids spec for approval to publish as CR -- due 2013-04-25 -- OPEN
<trackbot> http://www.w3.org/2001/tag/group/track/actions/790
<JeniT> Yves, oh yes, they are there!
<noah> . RESOLUTION: We will publish Best Practices for Fragment Identifiers and Media Type Definitions
<noah> W3C Editor's Draft 04 April 2013 www.w3.org/2001/tag/doc/mimeTypesAndFragids-2013-04-04.html (with minor mods at Jeni's discretion) as a CR draft, unless objections received by 23 April 2013
<JeniT> just in the wrong place
<Yves> I was looking for then in the SoTD, but here is fine
<Yves> I'll add a pointer from the SoTD when publishing
<noah> W3C Editor's Draft 04 April 2013 www.w3.org/2001/tag/doc/mimeTypesAndFragids-2013-04-04.html (with minor mods at Jeni's discretion) as a CR draft, unless objections received by 25 April 2013
<wycats_> no objection here
<noah> RESOLUTION: We will publish W3C Editor's Draft 04 April 2013 www.w3.org/2001/tag/doc/mimeTypesAndFragids-2013-04-04.html (with minor mods at Jeni's discretion) as a CR draft, unless objections received by 25 April 2013
RESOLUTION: We will publish W3C Editor's Draft 04 April 2013 www.w3.org/2001/tag/doc/mimeTypesAndFragids-2013-04-04.html (with minor mods at Jeni's discretion) as a CR draft, unless objections received by 25 April 2013
noah: anything else to do on this? don't think so
<noah> http://www.w3.org/2001/tag/products/fragids.html
noah: we have a product page for this work...thought we had a plan to update it
do product pages see much traffic?
<noah> ACTION: Jeni to tweak product page for Fragment ids - Due 2013-04-30 [recorded in http://www.w3.org/2013/04/18-tagmem-irc]
<trackbot> Created ACTION-797 - tweak product page for Fragment ids [on Jeni Tennison - due 2013-04-30].
<JeniT> slightlyoff, maybe Yves can find out
Yves: is that possible? do we have analytics on our pages?
<noah> ACTION-793?
<trackbot> ACTION-793 -- Anne van Kesteren to draft disclaimer text for Authoritative Metadata -- due 2013-04-05 -- PENDINGREVIEW
<trackbot> http://www.w3.org/2001/tag/group/track/actions/793
<noah> Finding http://www.w3.org/2001/tag/doc/mime-respect-20060412
<Yves> a bit difficult, but I can ask for some stats
<noah> ACTION-793?
<trackbot> ACTION-793 -- Anne van Kesteren to draft disclaimer text for Authoritative Metadata -- due 2013-04-05 -- PENDINGREVIEW
<trackbot> http://www.w3.org/2001/tag/group/track/actions/793
noah: annevk to review and draft disclaimer text because there's controversy about sniffing.
<noah> Anne's email: http://lists.w3.org/Archives/Public/www-tag/2013Apr/0007.html
noah: annevk's mail has generated some discussion and noah likes it
annevk: no comments for now about the proposal or responses
wycats_: seems fine
<JeniT> happy with it
I'm alright with it
<Yves> same
noah: are you proposing that the typography is the same as in the draft?
<wycats_> that is how I would prefer the typography
annevk: hoping the typography would be pretty obvious
<noah> Take a look at the proposed formatting: http://www.w3.org/2001/tag/doc/mime-respect-20130405.html
noah: what do we think?
I'm fine wit hthat
<wycats_> question: is Content-Type the only header for which this is true?
<wycats_> btw: Chrome has started allowing servers to opt into authoritative metadata
<noah> We don't have much precedent for doing things this way. I don't think this point is more important than many others that are made in the finding.
<annevk> wycats_: afaik
<wycats_> and Github has started using that
JeniT: is there any standard way to put notices in specs?
Yves: ?
JeniT: is there a standard way to put in notices for things like version succession
<annevk> http://www.w3.org/TR/2008/REC-CSS2-20080411/ <- has a similar warning
<noah> I would prefer adding it in two places:
<noah> in ordinary text as a 2nd para in the abstract
<annevk> http://www.w3.org/TR/CSS1/
annevk: we've put warning in REC's, e.g. CSS1
<noah> Following point 4 in the summary at the top
<wycats_> X-Content-Type-Options:nosniff
<wycats_> is what Chrome and IE support
annevk: it's been a big red box
noah: I think this is important, but not more important than other things the finding says
<annevk> wycats_: I believe that doesn't actually remove all sniffing
noah: counter is to put it in 2 place: 2nd paragraph of abstract and below/near summary of key points
... won't stand in the way, but it doesn't feel like an emergency
<annevk> wycats_: e.g. <img> will still sniff, it'll only check if Content-Type mentions an image type of sorts if that header is present
noah: feeling that the right way to fix this in the long run is to fix 2616
wycats_: don't think that's true. Recollection is that it tells you to do it, not based on "do what the specs say"
... IE + chrome have a header for authoritative metadata
... github is using it to prevent hot-linking
I'll just note that the idea of metadata opting into authoritative metadata is HALARIOUS
wycats_: the desire for a big-red-box is to help people who are in the mind of thinking about sniffing to see the point
... should we be investigating the state of nosniff?
noah: suggesting that we deal with formatting first, then nosniff in a sec
... there's no publication process for TAG findings
<wycats_> http://msdn.microsoft.com/en-us/library/ie/gg622941(v=vs.85).aspx
<wycats_> https://bugzilla.mozilla.org/show_bug.cgi?id=471020
noah: we just put it up and send a couple of emails
<wycats_> http://philip.html5.org/tests/ie8/cases/content-type-nosniff.html
noah: the way to do this might be to just re-publish the finding
<annevk> wycats_: I think it might be interesting to look into the specifics of that if we think we can make it work
<wycats_> ^^ some nosniff links
noah: objections? (none heard)
<noah> RESOLTUION: The TAG will republish the authoritative metadata finding to include the content and formatting proposed in http://www.w3.org/2001/tag/doc/mime-respect-20130405.html
<annevk> wycats_: and maybe define it as part of http://mimesniff.spec.whatwg.org/ if it actually does what it suggests (which I'm doubting to some extent)
RESOLUTION: The TAG will republish the authoritative metadata finding to include the content and formatting proposed in http://www.w3.org/2001/tag/doc/mime-respect-20130405.html
<scribe> ACTION: noah will republish [recorded in http://www.w3.org/2013/04/18-tagmem-irc]
<trackbot> Created ACTION-798 - Will republish [on Noah Mendelsohn - due 2013-04-25].
<noah> ACTION: Noah to republish authoritative metadata finding [recorded in http://www.w3.org/2013/04/18-tagmem-irc]
<trackbot> Created ACTION-799 - Republish authoritative metadata finding [on Noah Mendelsohn - due 2013-04-25].
oh balls
<wycats_> "This is not needed because Gecko doesn't violate the Http RFC unlike IE and doesn't do content-sniffing at all if a content-type is given."
<wycats_> I suspect that that is incorrect
I think I just did something automated and incorrect
<noah> ACTION-793?
<trackbot> ACTION-793 -- Anne van Kesteren to draft disclaimer text for Authoritative Metadata -- due 2013-04-05 -- PENDINGREVIEW
<trackbot> http://www.w3.org/2001/tag/group/track/actions/793
<noah> close ACTION-793
<trackbot> Closed ACTION-793 draft disclaimer text for Authoritative Metadata.
<annevk> noah: fwiw, smaller type as in http://www.w3.org/TR/CSS1/ would work for me
me? or wycats_ ?
<annevk> noah: it being super clear is what's most important
wycats_: it seems like Chrome has adopted it, there are publishers using it, and FF is loooking to adopt
<noah> Can we get a link to the spec for this header?
wycats_: we might want to revisit the conclusions in light of the world changing
noah: is there a spec?
wycats_: linked to http://msdn.microsoft.com/en-us/library/ie/gg622941(v=vs.85).aspx
<noah> http://msdn.microsoft.com/en-us/library/ie/gg622941%28v=vs.85%29.aspx
<noah> MIME-Handling Change: X-Content-Type-Options: nosniff
wycats_: annevk has raised a legit question: does it do what it says? or are there caveats?
noah: ugg...X-?
wycats_: IE created, others followed
<annevk> Seems Gecko is awaiting a spec
noah: Yves, do you have a sense for who we might coordinate with?
<wycats_> https://github.com/mozilla-services/cornice/issues/102
Yves: there's no need to register experimental headers
<wycats_> apparently the sniffing can trigger XSS attacks
<annevk> (from skimming the bug, which matches what I'd like us to do)
noah: was talking about something else: some future change to the flag might make it legit, and there's a coordination question around who owns "content-type"
<wycats_> everyone always says the "I really mean it" thing, but it never actually turns out to be true
noah: what's the best role for the TAG?
<annevk> I think http://mimesniff.spec.whatwg.org/ should own this.
<wycats_> it's usually just unfortunate legacy behavior
<wycats_> see: <!doctype html>
<annevk> It's an explicit opt out for the thing defined there...
Yves: it was discussed in the HTTP WG so there's not a ton of need to cooridnate
wycats_: used to be sympathetic to the "I really mean it" argument, but as I've seen it play out over time, you can imagine that people MIGHT misuse, but I don't observe that
<noah> To be clear, I was raising a point relating to coordination across groups, not the "I really mean it" technical point
<annevk> I filed a bug on MIME Sniffing: https://www.w3.org/Bugs/Public/show_bug.cgi?id=21742
wycats_: happy to investigate with annevk and slightlyoff
... if annevk is correct in that it doesn't actually mean what it says, we might come to different conclusions
noah: a group to collect and report seems good
yes
doing it now
<Yves> relevant thread: http://lists.w3.org/Archives/Public/ietf-http-wg/2009AprJun/0104.html
<scribe> ACTION: wycats, slightlyoff, and annevk to investigate nosniff current status [recorded in http://www.w3.org/2013/04/18-tagmem-irc]
<trackbot> Error finding 'wycats,'. You can review and register nicknames at <http://www.w3.org/2001/tag/group/track/users>.
sorry
I'm not great at this = \
due before F2F?
would prefer that
<annevk> Seems I filed a duplicate :-)
<annevk> https://www.w3.org/Bugs/Public/show_bug.cgi?id=19865 is the bug
<noah> ACTION: yehuda with help from slightlyoff, and annevk to investigate nosniff current status - Due 2013-05-14 [recorded in http://www.w3.org/2013/04/18-tagmem-irc]
<trackbot> Created ACTION-800 - with help from slightlyoff, and annevk to investigate nosniff current status [on Yehuda Katz - due 2013-05-14].
<noah> Note: this may be a topic for F2F, we will decide after telcon on 16 May
<noah> F2F disussion: http://www.w3.org/2001/tag/2013/03/19-minutes#item02
noah: my recollection is that we'll put out P&L as a NOTE
... couldn't find a formal decision to do that
... there was an action for Ashok to update it, and he has done it
... can others confirm that that's true?
annevk: yes, to be published as a note and then perhaps go forward from there
<noah> http://www.w3.org/2001/tag/doc/publishingAndLinkingOnTheWeb-2013-04-08
<wycats_> no objections
noah: with that clarification, latest draft is above, and the question is: are we ready to publish as a NOTE?
<wycats_> would love to get this off our plate
no objections
<noah> . RESOLUTION: We will publish http://www.w3.org/2001/tag/doc/publishingAndLinkingOnTheWeb-2013-04-08 as a W3C Note. The TAG does not commit to further work.
<annevk> wycats_: might still get updated though ;)
<noah> We note that Ashok may choose to propose some updates.
<JeniT> +1
<noah> RESOLUTION: We will publish http://www.w3.org/2001/tag/doc/publishingAndLinkingOnTheWeb-2013-04-08 as a W3C Note. The TAG does not commit to further work.
RESOLUTION: We will publish http://www.w3.org/2001/tag/doc/publishingAndLinkingOnTheWeb-2013-04-08 as a W3C Note. The TAG does not commit to further work.
<noah> ACTION: Yves to publish http://www.w3.org/2001/tag/doc/publishingAndLinkingOnTheWeb-2013-04-08 as a W3C Note and announce - Due 2013-04-30 [recorded in http://www.w3.org/2013/04/18-tagmem-irc]
<trackbot> Created ACTION-801 - publish http://www.w3.org/2001/tag/doc/publishingAndLinkingOnTheWeb-2013-04-08 as a W3C Note and announce [on Yves Lafon - due 2013-04-30].
<noah> ACTION-776?
<trackbot> ACTION-776 -- Ashok Malhotra to with help from Larry to redraft Publishing and Linking, responding to comments, for review at F2F -- due 2013-04-02 -- OPEN
<trackbot> http://www.w3.org/2001/tag/group/track/actions/776
<noah> close ACTION-776?
<trackbot> Closed ACTION-776 With help from Larry to redraft Publishing and Linking, responding to comments, for review at F2F.
noah: nothing else formal on the agenda, so inclination is to close the call
... adjourned! call in 2 weeks and Marcos will scribe