Web Application Security Working Group Teleconference

15 Dec 2014


See also: IRC log


mkwst, +1.206.753.aaaa, +1.418.907.aabb, Wendy, +1.415.736.aacc, +1.503.712.aadd, terri, gmaone, bhill2, jww, francois, dveditz, ckerschb, [Microsoft], +1.310.597.aaee, +1.310.597.aaff, +1.415.857.aagg, +1.415.426.aahh
bhill2, dveditz


<trackbot> Date: 15 December 2014

<bhill2__> zakim aaaa is bhill2

<francois> zakim aabb is francois

<jww> Zakim aacc is jww

<bhill2__> zakim insists on correct punctuation

<bhill2__> scribe volunteer?

<bhill2__> Meeting: WebAppSec Teleconference 15-Dec-2014

<bhill2__> Chairs, Bhill2, Dveditz

<bhill2__> Scribenick: bhill2_

Minutes Approval

<bhill2__> http://www.w3.org/2014/11/17-webappsec-minutes.html

<bhill2__> No objection to unanimous approval of prior minutes.

Mixed Content ends Last Call

<bhill2__> http://lists.w3.org/Archives/Public/public-webappsec/2014Dec/0056.html

<wseltzer> +1

<bhill2__> whoever just joined from aaee can you please mute?

<bhill2__> bhill: any interest from others in adding a strict mode for subresources to mixed content?

<wseltzer> [publishing moratorium: 19 December through 5 January 2015 ]

<bhill2__> ... twitter and facebook would like this

<bhill2__> mkwst: we have time given publication break to work on this, so no objections

<bhill2__> dwalp: we have folks on vacation, can we respond first week of 2015

<bhill2__> wseltzer: W3C publication break is 19-Dec to 5-Jan

<bhill2__> ... and +1 to the feature

<bhill2__> mkwst: work trying to hammer this out, many people are out of office, target mid-January for a CR seems reasonable

<bhill2__> wseltzer: where are we in process transition?

<bhill2__> bhill: on linst we seemed to agree on continuing to have an informal LC process, but it isn't part of W3C formal steps anymore regardless

FPWD of Requirements for Powerful Features

<bhill2__> http://www.w3.org/TR/powerful-features/

<dveditz> bhill2__: this is a new spec we’re taking on, but pretty much all the features were formerly in MIX

<dveditz> bhill2__: new working draft triggers a new call for exclusions (see group calendar on our homepage)

Proposed new Charter

<bhill2__> https://w3c.github.io/webappsec/admin/webappsec-charter-2015.html

<dveditz> … we’ve had a new charter up for review for a few weeks now

<bhill2__> bhill: feedback from lawyers at Facebook that this was not a bad thing but was outside the narrow IPR scope of this group so far

<bhill2__> bhill: proposed a compromise to accept it, but keep it tightly scoped to JS API only

<bhill2__> dveditz: ambivalent, agree it is useful but somewhat outside our group's current scope, happy to accept it with proposed changes

<bhill2__> dveditz: hearing no objections...

<bhill2__> dev: seems like a very long list

<bhill2__> dveditz: some of those pieces were broken out of or sprung from existing specs

<bhill2__> mkwst: this takes us through July 2016, it is long, but we have time and new people coming in to do this

<bhill2__> ... not a bad thing to bring in new things with people to do them

<bhill2__> bhill: yes, part of earlier consideration was that there are editors to own this new work

<bhill2__> dveditz: some of the dates for new work are fictional

<bhill2__> mkwst: have we hit any dates ever?

<bhill2__> bhill: I pulled those dates out of a hat, if we don't try to hit a date, we'll never make progress

<bhill2__> dveditz: there are some with schedule risk because they are not well defined or have sample implementations yet

<bhill2__> wseltzer: next steps are taking this to W3C management which should go quickly, then polling AC reps to indicate support

<bhill2__> ... and commit resources, this will help us gauge interest

<wseltzer> ACTION: wseltzer to take charter to w3m for review [recorded in http://www.w3.org/2014/12/15-webappsec-minutes.html#action01]

<trackbot> Created ACTION-208 - Take charter to w3m for review [on Wendy Seltzer - due 2014-12-22].

<bhill2__> jww moves to adopt the draft charter

<bhill2__> dev seconds

<bhill2__> dveditz: with no objections, WG unanimously decides to send the charter for w3c management approval

[POWER] New vs Legacy functionality (Re: "Requirements for Powerful Features" strawman.)

<bhill2__> http://lists.w3.org/Archives/Public/public-webappsec/2014Dec/0028.html

<bhill2__> terri: have we talked to the Web of Things CG about this?

<bhill2__> mkwst: we have generally punted IoT concerns out of this group

<bhill2__> bhill: Geolocation group is doing this themselves, we're not driving it

<bhill2__> mkwst: not actually clear that they're doing that - there is a thread, but it is unresolved

<bhill2__> dveditz: some concerns at mozilla if it is mandating or using as examples features that have had their own intense debates about this

<bhill2__> ... to the extent that those groups have made their own decisions, they don't want us to reverse them

<bhill2__> mkwst: we are trying to lay out a framework for deciding what a powerful feature is

<bhill2__> ... personally I feel some groups have made some poor decisions about this

<bhill2__> ... I don't believe that this spec is mandating that specific features be restricted to HTTPS, but that a category be restricted

<bhill2__> ... we can have healthy debate whether a feature falls into that category or not

<bhill2__> bhill: there is also a split in responsibility here between WebAppSec and TAG

<bhill2__> mkwst: we are defining the contours here of what is powerful, would be sad if we didn't try to outline when we think the algorithm should be applied

<bhill2__> mkwst: a good place to focus discussion is on section 3: "Is feature powerful?"

[REFERRER][CSP] Improving the Web Platform's Referrer Policy

<bhill2__> mkwst: there is still ongoing discussion, and we should put together a new draft in January / February that includes some of these ideas, like subresource vs. navigation policies

<bhill2__> ... this is going to take some time, still some big topics to discuss and decisions to make

[SRI] Towards a LCWD in January...

<bhill2__> http://lists.w3.org/Archives/Public/public-webappsec/2014Dec/0045.html

<bhill2__> bhill: wanted noncanonical-src but ok with a shim

<bhill2__> dev: yes, as version 1, but may still be good in future version

<bhill2__> jww: +1, freddyb is also +1

<francois> also +1 from me

<bhill2__> dev: javascript reporting is painful and not as useful as CSP-style implementation

<bhill2__> bhill: important for an experiment like this to get data from operators and not just implementers

<bhill2__> dveditz: like the idea of a unified w3c reporting uri spec

<bhill2__> ... from a website's perspective, if you are listening and start getting new reports you weren't expecting, that is bad

<bhill2__> bhill: nice for supporting policy evolution to associate each policy with its own endpoint

<bhill2__> dev: but would be great to filter reports in JS

<bhill2__> mkwst: error reporting in the DOM is shippping already

<bhill2__> bhill: do we expose SRI errors in the DOM?

<bhill2__> dev: shouldn't be a blocker for v1

<bhill2__> dveditz: most resources support onError and onLoad events

<bhill2__> mkwst: we did some work in order to avoid information disclosure for hash-guessing

<bhill2__> bhill: retaining room for spec authors and UA implementers to manage information flow here is good defense against malicious resource authors

HTTP/HTTPS for SRI (stolen from a smart thread on chromium security-dev)

<bhill2__> https://groups.google.com/a/chromium.org/forum/#!topic/security-dev/hJwYZOOPaH4

<bhill2__> jww: some disagreement on whether this is for 3rd-party integrity or integrity without confidentiality

<bhill2__> ... and sub-camps, given one of those threat models, do you care if it's over http or not

<bhill2__> ... would like to hear from application developers on this

<bhill2__> bhill: set of customers who want to use this but have parent resource insecure is small, perhaps null

Clarifying how CSP sandboxing applies to Workers, ServiceWorkers

<bhill2__> mkwst: seems a reasonable intrepretation

<bhill2__> ... hope we can have non-same-origin workers in future and language also makes sense ther

Summary of Action Items

[NEW] ACTION: wseltzer to take charter to w3m for review [recorded in http://www.w3.org/2014/12/15-webappsec-minutes.html#action01]
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.140 (CVS log)
$Date: 2017/02/15 22:32:50 $