Internet Architecture Board


IETF/W3C Coordination Call Minutes

IETF/W3C Coordination Call Minutes

Time: Thursday, October 14, 2004, at 21:00 UTC

Attendees: Leslie Daigle (LD, IAB), Ted Hardie (TH, IESG),
            Scott Hollenbeck (SH, IESG), Tim Berners-Lee (TBL, W3C),
            Dan Conolly (DC, W3C), Martin Duerst (MD, W3C)

Scribe: Tim Berners-Lee (irc), Martin Duerst (cleanup)
Chair:  Martin Duerst


Review of open action items

P3P header registration

Report/discussion on Marid WG / spam measures

Status of URI and IRI drafts

URI Activity, IG, testing WG

Status of draft-phillips-langtags-06.txt

Heads up for new work:
   - on the W3C side
   - on the IETF side

Date/time of next meeting

Internet draft persistence/redirection and IETF restructuring

Actions from the meeting

Review of open action items

Martin Duerst:  update the XSLT working group on progress related to 

Ted Hardie:     confirm location of published IESG procedures
                for registration of mime types from standards organizations.
Dan Connoly:    to estimate the approximate completion time of the URI
                specification update.
                CLOSED as obsolete.

P3P header registration
(Macchiori) (carry-over from last time)

   with a note (by Ted, some time early this year):
   Have asked for a review by an HTTP expert before a Last Call -
   the plan is to not delay much longer -  we will get back to authors
   about the currency of this spec and privacy issues before next
   IETF-wc3 call.

MD: draft-marchiori-w3c-p3p-header-01.txt has been around for a long time.
TH: The last time I tried to contact these people, the note went to
     me via Alison, and I tried to get in touch with the authors,
     but have not heard back.
MD: I got mail back from Massimo Macchiori
TH: Please have Massimo contact me.
ACTION Martin: Get Massimo to contact TH about P3P header registry.
TH: Is there any sense that anyone is using this in conflicting ways?
... Massimo says this realy neeeds to move forward.

Report/discussion on Marid WG / spam measures

WG Action: Conclusion of MTA Authorization Records in DNS (marid) WG
TH: We have concluded the MARID WG. The WG was never able to come to consensus
     on what identity it was woking on.
     The chairs declared consensus on a way forward but a subgroup ended up
     just continually challenging the work on ways which were not answerable.
     We didn't think we could arrive at consensus by any definition, so we
     decioded to publish them as experimental.
     Some of the people in the WG didn't understand deeply the way email
     systems are deployed, and so may not see things as problems which in
     reality would be.
     We will revisit the standards question later.
DC: By "identities" you mean?
TH: The envelope sender and the RFC2822 "From" and "Related" senders.
     envelope sender = Mail From (rfc2822), i.e. where to send bounces.
     There is a whole process laid out about how mail moved through many
     stages, forwarded etc.
     The microsoft draft rewrote it as an algorithm to use when you receive
     mail. The patent issues was not the whole issue.
     The "RF-RAND" terms didn't meet open source.
     And they never agreed that they were trying to establish the PRA.
     They even said they wanted to check the "Hello".
MD: I have discussed this with W3C team members and we have disagreed on
this too.
     W3C has this [faked origin,spam] problem itself, and we are constantly looking
     for a solution.
     Also we are looking as a fellow stds org is there anythying we could help with?
DC: We have been using SPF for a while, and turned it on sept 2004, block 13k
     messages per day with it. (see
     But also there is stuff about squatting on TXT records, ...
TH: There is a draft on how to expand into the DNS space making new records ...
     there are 64k available types, we don't ahve a problem saying yes burn
a few of these.
     We are though very concerned about overloading TXT.
LD: in IRC notes
DC: BTW, is it possible to get persistent links for internet drafts?
MD: Would you say that using SPF was wrong?
TH: It is useful for you to be part of the experimental community
reporting on succes.
     Unintented consequences tend to have to do with forwarding but not mail lists.
TBL: Would a clear definition of RF a la W3C have helped in that group?
TH: Not really - could have solved patents if agreement had been found on the
     problem to be solved.
     It was clear that they said they were trying to give it away.
     A previous agreement about how to give it away would have cut out some of that.
     A general lack of collegiality though was evident on the mailing list.
     In the IP messageing workshop we were discussing  the whole way this set up,
     and we agreed that 1-many or many-1 communication is architected on top of 1-1
     We asked about how we can control things given that ...
     wondered about moving to a rendezvous like protocol for IM
     to set up the context and call of the form.
     We discussed authentication, as there are propsoals in front of the IETF where
     where they separte problem with people pretending to be other people and being
     themselves and doing bad things.
     Looking of things like SSH's "leap of faith" which allow better security
     after the initial risky assumption.
DC: good to know that there will be a write-up of this IAB workshop eventually.
LD: There is a fruit for a lot of throught around how to get
     out of this hole with IM systems.
MD: At the messaging workshop, any discussion of web services?
LD: Not really
    Some people thought discussions were close to what Liberty is trying to
do with SAML.
DC: We have this success disaster ... like with email...
     "anybody can talk to anybody". leads to spam.
DC: There is an Atom working group... have they started to talk
     about Comment protocols
MD: When the comment spam problem came up, the list seemed to put
     the issue out of topic.

Status of URI and IRI drafts

TH: The URI draft was approved this morning by the IESG.
     The message should go out in the next couple of days.
     Just one tiny editorial note in section 6.
     Two words "and protocols". This will be a full standard.
DC: Not draft standard?
TH: No, full standard.
MD: What about IRI draft?
TH: Russ was not able to complete his review by today,
     will be considered in two weeks time.
     1738 will NOT be obsoleted.
     It has a bunch of small old URI schemes.
     There has been talk of making small documents to get the small
     schemes out of that document and remove the last references to it.
DC: Larry Masinter asked the TAG to look at RFCs 2717 and 2718
TH: Ian King was the first person to write a draft on it, Larry took it,
     difficult to get consensus.
TH: We wanted to avoid collision, and we wanted to avoid people minting
     new schemes when they didn't reall need one. But he process was so
     hard that people just minted them, without registration, like "mms"
     which has a collision.
DC: [did I note this collision? ah.. yes...]
TH: There seemed to be a fair amount of sensin the room in San Diego
     that we could avoid collisions by having a simple registry, and later
     on have a review of the spec.
DC: The TAG spent a certian amount fo energy saying one should not make new
     new URI schemes.
MD: This is still a critical point. We just can't itms: is just http: for
TBL: And webcal too
TH: Agree still a critical poinjt, just not solved by registration hurdles.
TBL: It looks as though it is a problem with extensibility APIs in the Apple
     architecture: If so they should change the API so they can dispatch
     these things on mime type.

URI Activity, IG, testing WG
[see also]

DC: W3C has a URI activity -- once a WG, which prodcued a document, then it was
     a coordination group which didn't produce much tangible, and whose
charter expired.
     It didn't seem right to just erase the group, so I have drafted the
charter for an
     Interest Group.
     On the mailing list, sometimes people share test cases. That could be
a justification.
     Now it is worth discssing there as if they end up make test suites,
     then the tests are in a different org to the spec, which can be a problem.
TH: Sounds like not a lot of energy.,
DC: Maybe one meeting every 2 weeks (teleconf)
TH: The URI specs are done.
LH: The standard is supposed to have been done, in an ideal world.
MD: There was testing, and Roy noted it.
DC: It is about making better quality products .... making it easier to
test them.
TH: Scheme specific stuff?
DC: Not much
MD: For file:// there is a need or more convergence.
     and also quite a need to work on tests for IRIs. I might do some of that.
DC: yes, tests should be in same place ...including the conection between
them etc.

Status of draft-phillips-langtags-06.txt

MD: We are planning to charter some work on Web Services
internationalization (i18n).
     It would be useful to know the outcome fo that effort, blocking timelines etc.
DC: datatracker shows activity as recent as 2004-10-08
MD: Not very clear whether this will go to draft and another last call,
     or whether it is more or less there.
TH: [...] Have talked to IANA. Problem is that the last call produced enough
     comments that we can't say that this is clear ... but for the comments
     the authors belive they have answers. So it looks like we will need
     another last call.
MD: A lot of things have changed since the first Last Call. Another LC
seems good.
TH: Is there a specific W3C group we should notify on any new LC?
MD: The Internationalization WG, but its chair is an author on the draft,
     so no need for much notification.

Heads-up for new work

TH: SIEVE. procmail-independent mail rules

DC: A  W3C Mobile Web Initiative workshop.
     This is cool ... about people making godo standard-compliant
TBL: a response to .mobi ... let's not make a separate web,
      but use existing technologies...
      perhaps more profiling than we normally do.

TH: Open Pluggable Edge Services
TH: Securty related things
DC: I gather OPES _was_ working on standardized "content distribution networks",
     ala Akamai
somebody: OPES was more about filtering content, based on having
     filtering agents at agreed upon points in the stream
TH: BOF on "MAF" ... relat to antispam work from marid wg, but using crypto.
     we are a bit sceptical given the area it is working in

TH: Kitten WG is a daugher of the CAT working group which develpoped the GSS API
     [CAT = Common Authentication Technology]
     Generaliezed security api
LH: Certs for dummies work too.. not in Kitten.
     "GSSAPI" is an acronym; it stands for Generic Security Services Application
DC: Programming Interface." --

Date/time of next meeting

MD: We have reached the end of the agenda except for date & time of next meeting.
TBL: This day and time seems to work.
DC: next IETF is 6-11Mar2005 (says Leslie)
     PROPOSED: 27 Jan 2005
TBL: Thursday, January 27th, Same time i.e. 17:00 ET

Internet draft persistence/redirection and IETF restructuring

DC: The IETF knwos that the 00 draft and the 01 is available.
     Why not have a URI which works for the latest version.
LD: There is a lot of mirroring. Changing data is not things people want to try.
DC: What about having a message with a suggstion in the 404 message?
LD: High cost, we are just in the middle of administrative restructuring
     the other issue is why we don't keep an archive.
     they don't really evapourate , why should we make it difficult?
     They may not be ephemeral for ever.
     RFC3716 givs an overveiw of issues with the IETF admin.
    [RFC 3716 - The IETF in the Large: Administration and Execution]
DC: hmm... tasty... "3.2.2. Persistence and Accessibility of Records"
LD: The IETF grew up organically.
DC: wow! "Proposed: (Single) Formalized IETF Organizational Entity "
LD: There are all kinds of different bits.
     Meeting fees go to CNRI, while ISOC funds the RFC editor.
     It is difficult to plan changes.
     We would like a single adminisitartive point of control,
     for example about meeting fee decisions.
     A Mallamud draft suggested some structures. [draft-malamud-consultant-report]
     Check out the archives, google "Secenario O"
DC: RFC 3716 dates from March 2004
MD: ACTION MD to clean up the minutes, next week, and circulate them to
     the appropriate places.

Actions from the meeting

Ted Hardie:    confirm location of published IESG procedures
                for registration of mime types from standards organizations.

Martin Duerst: clean up minutes next week and circulate them to
                the appropriate places.

Martin Duerst: Get Massimo to contact Ted Hardie about P3P header registry.