<Frederick Hirsch> Scribe: Konrad Lanz

Administrative

can someone please post the agenda url please

<EdS> http://lists.w3.org/Archives/Public/public-xmlsec-maintwg/2007Sep/0011.html

<Frederick Hirsch> agenda http://lists.w3.org/Archives/Public/public-xmlsec-maintwg/2007Sep/0011.html

1) Administrivia

I dial in again

<Frederick Hirsch> ack ??P8

<Frederick Hirsch> mute ??p8

1a) Konrad Lanz is scheduled to scribe

Juan Carlos Cruellas is next scribe

<scribe> chair: november planary

<scribe> chair: issue with minutes

<Frederick Hirsch> http://www.w3.org/2007/09/04-xmlsec-minutes

2) Review and approval of last meeting's minutes

RESOLUTION: Last meeting's minutes approved

3) Action item review

ACTION 26 open

ACTION 71 open

ACTION 74 open

ACTION 81 open

ACTION 82 open

ACTION 83 open

ACTION 87 closed, http://www.w3.org/2007/xmlsec/interop/xmlsig-interop-doc/testcases.html#directory-structure

ACTION 88 closed

ACTION 89 closed

4) XML Signature Draft Status Update

<Frederick Hirsch> editors draft updated for examples for digest and sig values with ...

5) Workshop

<Frederick Hirsch> http://www.w3.org/2007/xmlsec/ws/agenda.html

<scribe> chair: we have an updated program

Frederick Hirsch reflects change we had before with a tweak
... combining some slight rearrangements
... people should have a look at it

5a) Program Revision

Thomas Roessler: all the speakers are confermed except for ???

<Frederick Hirsch> all speakers confirmed apart from Chris Techter

<Frederick Hirsch> need to determine who is speaking for Oracle

Thomas Roessler: who will present for oracle ...

<scribe> chair: contacted oracle

<Frederick Hirsch> new paper from MITRE

Frederick Hirsch: another paper submitted

<Frederick Hirsch> http://lists.w3.org/Archives/Public/public-xmlsec-maintwg/2007Sep/0009.html

6) Interop

Scribe: Frederick Hirsch

Frederick Hirsch: sent an email

<Frederick Hirsch> http://lists.w3.org/Archives/Public/public-xmlsec-maintwg/2007Sep/0012.html

Frederick Hirsch: Sean Mullan added to the editors ...

Konrad Lanz: +1

<Sean Mullan> fine with me

<Juan Carlos Cruellas> Juan Carlos Cruellas +1

<Frederick Hirsch> http://www.w3.org/2007/xmlsec/interop/xmlsig-interop-doc/testcases.html#Introduction.RecommendationRefs

<Frederick Hirsch> Why do we have two directory structures that appear to both have c14n11 test cases?

Frederick Hirsch: shouldn't we have just have one c14n11 dir

<scribe> Konrad Lanz: ok to test c14n11 alone as well as c14n11 tests with dsig

<scribe> ... standalone optional, sig mandatory

<scribe> ... found some divergence of input files

<scribe> Sean Mullan: merlin always showed intermediate c14n result in those tests

Sean Mullan: merlin's tests showed the intermediate results

<EdS> I can hear everyone fine.

<scribe> thomas: +1 to Sean Mullan

<scribe> Sean Mullan: we need to freeze test cases today, although ok to get it right now

<Sean Mullan> more or less by end of week or so

Konrad Lanz:+1 to end of the week

<Frederick Hirsch> proposed change: combine the c14n11 test cases into one directory with inputs and outputs (sig) as well as intermediate results

Konrad Lanz:-1 to this proposal

<scribe> thomas: needs to be done today or tomorrow

<Frederick Hirsch> ACTION: Konrad Lanz to ensure that the inputs for both c14n11 test directories are the same [recorded in http://www.w3.org/2007/09/11-xmlsec-minutes.html#action04]

<trackbot-ng> Created ACTION-90 - Ensure that the inputs for both c14n11 test directories are the same [on Konrad Lanz - due 2007-09-18].

<Frederick Hirsch> What is the status of the following incomplete sections of the document, and what is the plan for completing?

<Phillip Hallam-Baker> sorry I am late.

6b) Review and next steps for test cases document

<Frederick Hirsch> Section 3.3, Implicit/Explicit rules , XPointers DNs

<scribe> Juan Carlos Cruellas: need to work with Konrad on these

Juan Carlos Cruellas: Sean Mullan, Juan Carlos Cruellas and Konrad Lanz should work together on document
... currently only description
... no actual input at the moment

Frederick Hirsch: are these harder

?

Frederick Hirsch: just a matter of time or something else?

Juan Carlos Cruellas: just a matter of time

Frederick Hirsch: are you able to contribute after looking at cvs ?

Bruce Rich: focused on implementation ..
... bring up things later

Konrad Lanz: Juan Carlos Cruellas +1

<scribe> Konrad Lanz: +1 to Juan Carlos Cruellas

<scribe> ... need teleconference for Sean Mullan, Juan Carlos Cruellas, Konrad Lanz separate call

Juan Carlos Cruellas: propose an Action on Konrad Lanz, Juan Carlos Cruellas and Sean Mullan to distribute the rest of the work
... by email
... have the call then

<scribe> Konrad Lanz: +1 to Juan Carlos Cruellas, but wants quick call first

<scribe> ... in interest of time

<Juan Carlos Cruellas> OK

<scribe> ACTION: Juan Carlos Cruellas interop editors to plan and distribute work to complete interop tests by next week [recorded in http://www.w3.org/2007/09/11-xmlsec-minutes.html#action05]

<trackbot-ng> Created ACTION-91 - Interop editors to plan and distribute work to complete interop tests by next week [on Juan Carlos Cruellas - due 2007-09-18].

Thomas Roessler: two technical points ...

stay on the call later

<scribe> ... can stay on bridge later, or get reservation for later through thomas

Juan Carlos Cruellas: +1 to have call today

<scribe> Juan Carlos Cruellas: stay on this call for a few minutes at end for planning

Konrad Lanz: let's try to stay on this call and see how it goes

<Frederick Hirsch> What is the status of Appendix A testing? Section 3.2.4.2

7) C14N11 - Appendix A

<Frederick Hirsch> http://www.w3.org/2007/xmlsec/interop/xmlsig-interop-doc/testcases.html#XMLBASE_ANNEXA

http://www.w3.org/2007/xmlsec/interop/c14n11/appendixa/

Sean Mullan: appendix

<scribe> Sean Mullan: working with amended appendix

Sean Mullan: hard to test
... has to be a signature ...
... we should get this into a signature
... maybe hard to test
... did code modifiction

<scribe> ... had to put special hooks into code to enable use of input

Bruce Rich: not a lot of clarity
... simpler algorithm

<scribe> Bruce Rich: had issues with nested if s and use of english

Bruce Rich: something missing
... when is appendix A when normal uri processing to be used

<scribe> ... what is missing is guidance of when to use normal uri processing or appendix a, e.g. #53, //..

Bruce Rich: unclear what happens to absolute uris

<scribe> ... unclear when using absolute uris - when to collapse multiple /'s

< message 50 in the xml core June archive, Konrad Lanz:> However I agree that a complete rewording might simplify the processing as I have informally mentioned already in http://lists.w3.org/Archives/Public/public-xmlsec-maintwg/2007Jun/0031.html.
I'm unsure about the process if we'd really like to change this and hence I sticked as "close" as possible to RFC 3986.
5.2.4. "Remove Dot Segments" is modified to keep leading "../" segments and to prevent the erroneous creation of an output that looks like a net path. (seg/.././/pseudo-netpath/seg/file.ext)
* several changes as in "Remove Dot Segments" ... (see Apendix) http://www.w3.org/TR/2006/WD-xml-c14n11-20061220/#DocSubsets

Frederick Hirsch: is there a show stopper ...

Bruce Rich: when to follow uri rules

<scribe> Bruce Rich: need to be clearer which part of normal processing to follow

Juan Carlos Cruellas: all test cases by Konrad Lanz have been tested by our implementor

<scribe> Juan Carlos Cruellas: our implementation includes appendix A, all cases have been tested. Do not have RFC 4514 implementation to string conversion

<scribe> Konrad Lanz: agrees language is not clear, ok with changing language but by joining XML Core WG and doing it there

<scribe> ... was looking for smallest change to what was there already

<scribe> ... need to see support for making change in XML Core WG

<scribe> ... do not agree that not clear when appendix A applies

<scribe> (see quoted material above in chat)

Sean Mullan: hard to resolve

Frederick Hirsch: some people having problem with material

Thomas Roessler: please show examples for ambiguity
... didn't get where the ambiguity was
... tried a few lines of pearl

<Bruce Rich> ambiguity starts with "relative URI" in section 2.4

Thomas Roessler: trade off close to RFC vs. clarity

cannot hear Thomas Roessler, sonds like soap bubbles

Frederick Hirsch, please minute Thomas Roessler

<scribe> Thomas Roessler: asks Bruce to contribute simpler algorithm to XML core, also to put message on list to describe ambiguity

@Bruce Rich: maybe this helps http://lists.w3.org/Archives/Public/public-xmlsec-maintwg/2007Jun/0031.html

<Bruce Rich> e.g., is //.. a relative URI? do I apply slash-reduction to http://foo?

<scribe> Bruce Rich: need to know what to do, direction for interop in two weeks

Frederick Hirsch: if do not have the simpler algo on the list
... no one can see it
... we really need it on the list

Thomas Roessler: quick clarification, please contribute to xml-core

Frederick Hirsch, please minute Thomas Roessler, can't hear

<scribe> Frederick Hirsch: please contribute simpler algorithm to xmlsec so we can consider in time for interop

<scribe> Thomas Roessler: first step is to get to our wg, can deal with XML Core later

Konrad Lanz:+1 to this proposal

Bruce Rich: will post issues this afternoon

Konrad Lanz:I was kicked out, dialing in again

<Frederick Hirsch> Frederick Hirsch: proposal - Bruce put proposal on XMLSec WG list today, interop implementers/editors review discuss on call (including Bruce, Sean Mullan, Juan Carlos, Konrad)

<Frederick Hirsch> ... put resolution on list for group and also update document

<Frederick Hirsch> ... part of discussion is impact on implementations, feasibility etc

<Frederick Hirsch> jh: proposal - Bruce put proposal on XMLSec WG list today, interop implementers/editors review discuss on call (including Bruce, Sean Mullan, Juan Carlos, Konrad)

<Frederick Hirsch> ... put resolution on list for group and also update document

<Frederick Hirsch> .. part of discussion is impact on implementations, feasibility etc

<scribe> Sean Mullan: 1. is algorithm incorrect, or is it just simplification of wording. If latter, won't impact intetop testing

<scribe> ... if simply matter of wording then can achieve common interpretation

Konrad Lanz:+1 to Sean Mullan

Konrad Lanz:we have had this rewording discussion starting here http://lists.w3.org/Archives/Public/public-xmlsec-maintwg/2007May/0044.html and it ended here http://lists.w3.org/Archives/Public/public-xmlsec-maintwg/2007Jun/0073.html.

<scribe> Thomas Roessler: need one normative algorithm, that is understood

<scribe> ACTION: Thomas Roessler to send note to XML Core indicating potential issue with Appendix A [recorded in http://www.w3.org/2007/09/11-xmlsec-minutes.html#action06]

<trackbot-ng> Created ACTION-92 - Send note to XML Core indicating potential issue with Appendix A [on Thomas Roessler - due 2007-09-18].

<scribe> Thomas Roessler: must have only one normative algorithm, simple language would be good

Konrad Lanz:+1 to Thomas Roessler

<scribe> Konrad Lanz: majority of xml core wg wanted to stay close to RFC language, hence some of this issue

Konrad Lanz:xml core had the perception that stying close to RFC language would avoid making unwanted mistakes, I didn't share that preception, "a complete rewording might simplify the processing"

<scribe> Bruce Rich: cannot re-write simpler language in the RFC language

<Frederick Hirsch> Frederick Hirsch: propose sharing as it is, without changing

<scribe> Bruce Rich: can do

Thomas Roessler: some elaboration to the minutes required on this

Frederick Hirsch: Bruce Rich to supply easier language for appendix a

<Frederick Hirsch> Frederick Hirsch: share with xml core that there is potential issue with Appendix A and possible simplification proposal

Frederick Hirsch: Thomas Roessler take this and go to xml core with it

<scribe> Thomas Roessler: with pointer to minutes and maybe some detail

<Frederick Hirsch> Are the xml:space, xml:lang, xml:id and xml:base tests complete and what is confidence in correctness of tests and ability to implement?

Status of complete interop tests

Scribe: Konrad Lanz

<Frederick Hirsch> ACTION: Konrad Lanz to look at xml core list for additional xml:base test cases [recorded in http://www.w3.org/2007/09/11-xmlsec-minutes.html#action07]

<trackbot-ng> Created ACTION-93 - Look at xml core list for additional xml:base test cases [on Konrad Lanz - due 2007-09-18].

cannot hear Thomas Roessler

Thomas Roessler: not conformance testing

Frederick Hirsch: we test the spec not the implemenations

Sean Mullan: only issue brougt up a month ago, but resolved

Bruce Rich: not yet there time will tell

@Bruce Rich: the sooner time tells the better ;-)

<Bruce Rich> will post to list if i find issues in id, lang, space

Juan Carlos Cruellas: brought up a page by rigo
... to do anything else then reference the old test
... regression tests are of less importance given the time pressure

<scribe> Juan Carlos Cruellas: on topic regression tests, not time to update the doc, but we need to get the tests content determined

<Juan Carlos Cruellas> http://www.w3.org/Signature/2001/04/05-xmldsig-interop.html

Juan Carlos Cruellas: regression tests are mentioined in section 3.1

Frederick Hirsch: everything in tar/zip files
... we should figure out what we actually want to run

Juan Carlos Cruellas: it could be done if someone has the time

Frederick Hirsch: what are we planning to do ...

Juan Carlos Cruellas: focus on issues we discovered now

<Sean Mullan> +1

<Bruce Rich> will not have time for much/any of regression tests by interop time

<Frederick Hirsch> propose we run regression tests on '#xpointer(/)', '#xpointer(id(...))'

Frederick Hirsch: we will need regression for the xpointers

<scribe> Konrad Lanz: issues with dereferencing absolute URIs possibly

<scribe> ... everyone with implementation run all tests see what can do

<scribe> +1, lower priority, some never are passed, better to share where we are at

<scribe> Konrad Lanz: will full web access be available at interop event?

<Robert Miller> I have to go. I have another meeting to attend.

http://www.w3.org/2007/xmlsec/Group/Overview.html