Web Cryptography Working Group Teleconference

21 Mar 2016

See also: IRC log


hhalpin, wseltzer, jyates, jimsc, ttaubert, RobTrace
Virginie, markw


<jyates> Two of us are on the phone

<hhalpin> Just to note, the W3C AC meeting that both Virginie and Wendy are currently at is ongoing, and so I suspect they may be late. wseltzer?

the AC meeting is about to hit a break, but Virginie is on a panel right after the break

I believe she was not planning to join

<hhalpin> OK, I am happy to chair as stated on the phone.

hhalpin: review quickly with implementors, implementation schedule

<jimsc> Jim Schaad on the phone

Updated schedule

hhalpin: WebCrypto charter was renewed for a 6mo extension
... check in on scheduling
... To get to Recommendation, we need to move from CR to PR, Rec
... show that we have two interoperable implementations of each feature
... So we need to gather implementation status over the next 2 weeks
... 2-3 weeks of work on tests
... ~ May, review status of test suite and test results
... end of May, determine whehter to keep/cancel each feature
... June/July, CfC to Proposed Rec

-> https://lists.w3.org/Archives/Public/public-webcrypto/2016Mar/0048.html Virginie's email re schedule

RobTrace: When we talk about features and cutting features
... that doesn't account for an implementor who wants features but doesn't have current resources for them

hhalpin: that's a scheduel proposed by Virginie

RobTrace: Do we have to change the schedule, or if all algorithms are listed as optional,
... with algorithms that can be implement, can we finalize without having 2 implementations of each

hhalpin: there are different kinds of features

wseltzer: we need to be able to move forward; we could do that by showing progress, by doing extensions, or by cutting

RobTrace: lookng at media, codecs, we had some options, let the market decide
... we had some items that we didn't implement at first, then when the market showed interest, implemented later

wseltzer: are you asking for extension points, where we finalize the core, and then extend?

hhalpin: @@

RobTrace: RSA-PSS is the example we'd ask for

hhalpin: that will remain in the final spec, Firefox implemented
... the resolution we had to stabilize the specwas to put removed algorithms into a "proposed" status, then move them to the spec when implemented

<hhalpin> RSA-PSS will be kept

<hhalpin> The implementation report needs to be updated to take into account Firefox's latest work

<hhalpin> Microsoft will not change, but wants to keep RSA-PSS

RobTrace: MS implementation won't change in next 6mo, but we have desire to implement

<hhalpin> I believe Chrome wants to keep RSA-PSS

RobTrace: we want to complete implementation, support; just can't at the moment
... Are we making everything mandatory?

hhalpin: No. nothing is mandatory

<ttaubert> RSA-PSS is currently available in Firefox Beta

hhalpin: but we could note in test suite, "browser profile" what algorithms are implemented everywhere
... not cutting things out for the sake of process, but recognizing what has actual interop
... others that are specd but not implemnted stick around for potential later use.
... So everyone -> "browser profile"
... 2 implementations stay in the spec
... one or no implementations, go into "proposed" for later Rec

<hhalpin> AES-CMAC, AES-CFB, CONCAT, DH are the ones at risk to be moving in a different section.

hhalpin: other issues, key format, service workers

<jimsc> CONCAT has already been killed from the spec

<hhalpin> Sorry missed that jimsc

ttaubert: I'm working on getting webcrypto supported in workers
... going to turn to service workers after

RobTrace: I don't know that we've fully internalized SW crypto
... no official roadmap

hhalpin: I think it's in chrome
... so we can keep that part of the spec
... Then key format issues remain
... everyone agrees JWKs work fine, but cross-browser and in-browser issues with other formats
... there's desire from users to have PKCS

jimsc: 2 ways to deal with asn1 imports
... no longer tie algorithms hard to keys
... pretty simple in current spec
... versus getting people to implement what's in current spec is hard, not doable in time-frame

RobTrace: Not something we've had deep discussions on

ttaubert: same
... can someone summarize on ml the problems iwth key imports

jimsc: I'll send something on the ML

hhalpin: concenrn that dealing with asn1 lots of work
... markw willing to do the work
... testing training, 11 am Friday re TTWF

<jimsc> This is still a clear time for me

RobTrace: I could track that

ttaubert: I couldn't participate Friday night
... but would track ML

[calendar syncing]

hhalpin: one training, with Mike Smith, then work moving forward

ttaubert: this Friday doesn't work for me

<jimsc> Harry, we should try and record it since it is on webexo

hhalpin: I'll share the info
... any other questions?

<hhalpin> http://testthewebforward.org/

<hhalpin> We'll go through this agenda

jimsc: note, IETF meeting in 2 weeks, I can't attend then

hhalpin: I'll check with Virginie
... Rob and Tim, how much time do you want for review of Jim's proposal

ttaubert: propose 3 weeks

hhalpin: I'll follow up with Virginie


<hhalpin> trackbot, end meeting

Summary of Action Items

Summary of Resolutions

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.144 (CVS log)
$Date: 2016/03/21 20:45:35 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.144  of Date: 2015/11/17 08:39:34  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Succeeded: s/schedul/schedule/
Succeeded: s/interes/interest, implemented later/
Succeeded: s/afgter/after/
Succeeded: s/web crypt/webex/
No ScribeNick specified.  Guessing ScribeNick: wseltzer
Inferring Scribes: wseltzer
Default Present: wseltzer, virginie, hhalpin, jyates, Charles_Engelke, jimsch, timeless, kodonog, markw, jmann, ttaubert, RobTrace
Present: hhalpin wseltzer jyates jimsc ttaubert RobTrace
Regrets: Virginie markw

WARNING: No meeting chair found!
You should specify the meeting chair like this:
<dbooth> Chair: dbooth

Found Date: 21 Mar 2016
Guessing minutes URL: http://www.w3.org/2016/03/21-crypto-minutes.html
People with action items: 

[End of scribe.perl diagnostic output]