W3C

- DRAFT -

MLW-LT WG

07 Jan 2013

Agenda

See also: IRC log

Attendees

Present
felix, kfritsche, dF, philr, Ankit, tadej, Yves, Naoto, dave, dom, leroy, shaunm, Jirka, joerg, chriLi, Milan, Clemens, fredrik
Regrets
Chair
felix
Scribe
Yves, fsasaki

Contents


roll call

<fsasaki> scribe: Yves_

<fsasaki> http://lists.w3.org/Archives/Public/public-multilingualweb-lt/2013Jan/0001.html

call time

Felix: proposal is Thursday 4pm UTC
... that's 5pm CET
... that's 9am MST

<Clemens_> happy new year together

all ok with the new day

Jirka: will have conflict

Felix: will do a Doodle for the time of the call

LC feedback

<fsasaki> http://tinyurl.com/its20-comments-handling

Felix: several comments, but none from outside the WG
... can't move forward in that case

Jorg: maybe the comment period was a bad one

Felix: agree
... should extend until Jan-18

all: ok with that

Felix: please nudge your contacts
... we will get comments from the i18n WG
... for sure. but we need more

Dave: can people comment that they have no comments?

Felix: yes
... endorsment is fine

2.1) New value for localization quality type "conformance"

Felix: brought up by Phil before the Holidays
... no disagreement to add the new value
... no additional support either

DF: seems to be a normative change, no?

Felix: possibly
... need a strong reason to do the change
... any supporter in addition to Phil?

DF: should try to avoid normative changes
... wait if it's really needed

Phil: being additional value makes it easier
... no back compatibility issue, etc.
... we have a definite need for it.

Felix: but it's a borderline change
... would have to justify the rational
... would drop it now, but keep it open
... you could try to fine supporters

2.2) Removing disambigClassPointer

Felix: other tool to consume/generate that data

<fsasaki> http://lists.w3.org/Archives/Public/public-multilingualweb-lt-tests/2012Dec/0170.html

<fsasaki> http://lists.w3.org/Archives/Public/public-multilingualweb-lt-tests/2012Dec/0180.html

Felix: a change in the Disambiguation data category
... 0180 email is the right link
... this is dropping an attribute
... but ok because to test or processing is changed by it

DF: looks like an omission correction

Felix: will do the edit after the call

2.3) Need to clarify global provenance

<fsasaki> http://lists.w3.org/Archives/Public/public-multilingualweb-lt-tests/2012Dec/0170.html

Felix: this one is to clarify the global provenance

<fsasaki> The attribute provenanceRecordsRefPointer does not apply to HTML as local markup is provided for direct annotation in HTML. Because that information is the only one in the provRule element, the global rule does not apply to HTML.

Felix: it explains why the global rule does not apply to HTML

Dave: sounds like a clarification, not a substantive change
... original wording was close to the LQI text
... maybe we need to look at it there too

Felix: for LQI the requirements are different
... that's why we have different cases
... best practices will be important

2.4) Location of its:version

<fsasaki> http://lists.w3.org/Archives/Public/public-multilingualweb-lt-tests/2012Dec/0202.html

<fsasaki> "If there is no rules element in an XML document, a prefixed ITS version attribute (e.g. its:version) MUST be provided at the root element of the

Felix: this one is to allow more flexible location of the version

<fsasaki> document."

Felix: not only in the root, but anywhere in the upper part of the tree of a given node
... not a sever change
... just change checking before you do the process

Shaun: would require to change implementation though
... not too bad, but it's mean changing the code

Yves: agree

Felix: means a difference between 1.0 and 2.0

Yves: not sure there is a problem

Shaun: if it's in the root, and it's an ITS 1.0 processor it's invalid

DF: 1.0 requires it in the root
... but what is the gaol
... what is compliant to our principles?

<fsasaki> yves_: 1.0 processor can't check 2.0 subset

shaun: 2.0 processor should process 1.0, but not the reverse
... and here 1.0 processor can know the case

Felix: need to have note in the spc

<scribe> ACTION: Yves to propose notes for the version location change [recorded in http://www.w3.org/2013/01/07-mlw-lt-minutes.html#action02]

<trackbot> Created ACTION-382 - Propose notes for the version location change [on Yves Savourel - due 2013-01-14].

2.5) Regular expression change

<daveL> FYI: TCD has rejoined the internet

<fsasaki> scribe: fsasaki

yves: I re-raised issue I raised earlier
... about using XSD regex for allowed characters
... earlier we started with a subset of that
... we then moved to the XSD character class
... I didn't disagree strongly at that point so that other implmeneters can work on this
... I think it's an interop issue
... it would be better for the allowedchararacter class to be processed with other engines than xsd
... we can have a subset that works with dotnet, java, xsd etc.
... the only difference is that the user might have to change the way to write a regex
... but that's not a big deal, the interop we gain is more important

shaun: you are going to say that you will have the subset
... but tools need to be required to check that they conform from to the subset

yves: the schema would ensure that the expression is valid or not

phil: agree with Yves' argument
... having to figure out equivalent regex syntax can be hard

shaun: in general I agree with Yves
... just worried about the verification of the subset

felix: would we have a volunteer to help defining the regex subset?

shaun: I could do that

<scribe> scribe: Yves

Felix: this change may be seen as normative

<scribe> ACTION: Felix to give heads up to W3C about the possible change [recorded in http://www.w3.org/2013/01/07-mlw-lt-minutes.html#action03]

<trackbot> Created ACTION-383 - Give heads up to W3C about the possible change [on Felix Sasaki - due 2013-01-14].

DF: is it absolute that one such change would force a new revision
... is it a problem to have another LC?

Felix: there is a difference between extending the LC period and doing a new publication

DF: agree, this change seems big and warrant a new publication
... would hope we can do without
... because it brings additional feature change often

Jirka: this change is not so big, each XSD-based implementation will be compatible
... with the old syntax
... using a sub-set is compatible

DF: so we would not break anything?

<chriLi> +q

Jirka: only issue would be if current instances use special XSD faeture
... but that's not the case

Jorg: with sub-set an XSD RE would be ok
... need to explain what kind of RE engine are supported by the sub-set

Christian: does the sub-set has the same expressive power?

Yves: from examples, yes.

DF: a formal proof would be an agrument to not do a new LC
... even if we have a formal proof that the change does not affect the capability, the spec needs a new LC?

Felix: yes
... e.g using the Unicode blocks would be 'broken'
... basically it's a matter of negoiation with the w3c mgmt

DF: seems the discussion go back and forth
... so what is the new evidence is the question.
... seems people have different view at different time

felix: from implementer it's unlikely to go back to XSD RE
... no-one is pushing in that direction

felix: so we wait to see the regex from Shaun/yves

3.4) Test suite dashboard

<fsasaki> http://lists.w3.org/Archives/Public/public-multilingualweb-lt-tests/2012Dec/0190.html

Felix: worked on the test suite dashboard
... did you have a time to look at it?
... was proposing to drop the Google doc and replace it with an XML doc

<fsasaki> http://lists.w3.org/Archives/Public/public-multilingualweb-lt-tests/2012Dec/att-0190/TestStuiteMaster.xml

Felix: automatically generated from the test suite results

<fsasaki> http://lists.w3.org/Archives/Public/public-multilingualweb-lt-tests/2012Dec/att-0190/testsuiteDashboard.html

Felix: and the HTML page would be generated automatically
... this way a simple stylesheet can re-generate the output from the developer outputs

Dave: appreciate the work

leroy: agree, looks fine
... need to finish the corrections from Jirka
... add a few test files
... and after that I can look at it
... should be able to work on it by end of week

Felix: so by next monday we can deprecate the Google doc and use the XML output
... it would be on Github as well

leroy: agree

<fsasaki> http://lists.w3.org/Archives/Public/public-multilingualweb-lt/2012Dec/0087.html

Felix: need input test cases for some data categories
... see the list
... M3 deadline is at the end of January

<fsasaki> http://lists.w3.org/Archives/Public/public-multilingualweb-lt/2013Jan/0001.html

Felix: changes in the suite are documented in the above link

5) XML Prague demo session

<fsasaki> http://lists.w3.org/Archives/Public/public-multilingualweb-lt/2012Dec/0094.html

Felix: we'll have confirmation from fredrik (ENLASO) that he'll be there
... Dave as well?

Dave: leroy should be able to go
... are we presenting this as a MLW event?

<Jirka> http://www.xmlprague.cz/pre-conference-friday/

Felix: pre-conference day, 1.5h slot
... Karl also said he would be available
... need to distribute the work between those people

<scribe> ACTION: Felix to make agenda proposal for Prague meeting about the XML prague day [recorded in http://www.w3.org/2013/01/07-mlw-lt-minutes.html#action04]

<trackbot> Created ACTION-384 - Make agenda proposal for Prague meeting about the XML prague day [on Felix Sasaki - due 2013-01-14].

Karl: do i need to register?

Jirka: yes
... fee is small

felix: this is ok to use the LT-Web budget for this

<Jirka> Registration for XML Prague http://www.xmlprague.cz/conference-registration/

Clemens: can we access the full conference for 40 Euro?

Jirka: yes, but not to lunches and social events
... for that it's 100 Euros

Felix: anyone else?

DF: will try see if it's needed

Felix: this is about demoing implementations
... good opportunity to bridge to the content world

DF: yes, good way to contact XML people

Felix: Prague F2F is soon
... let's talk about the EC-specific parts.

8) EC-project internal

Felix: need input from Ankit

Felix: MT harvesting tool taks: any update?

Ankit: started to work with Cocomore on this

Felix: talked to Clemens about this, but no info on the tool

Ankit: we need to data to make the tool
... by end of January we should have prototype

Karl: Not quite sure

Felix: Ankit let me know if you don't get info from Clemens
... david: any update on roundtipping with XLIFF?

DF: yes, should be able to provide that Wednesday

Phil: what info will you have on Wednesday?
... does it includes the mapping?

DF: we have a working prototype
... part of it is the mapping, but not frozen

Felix: the task is about a prototype

DF: Q1 is the time when we'll work on this
... we have work in progress

Felix: anything else?

meeting adjourned

Summary of Action Items

[NEW] ACTION: Felix to give heads up to W3C about the possible change [recorded in http://www.w3.org/2013/01/07-mlw-lt-minutes.html#action03]
[NEW] ACTION: Felix to make agenda proposal for Prague meeting about the XML prague day [recorded in http://www.w3.org/2013/01/07-mlw-lt-minutes.html#action04]
[NEW] ACTION: Yves to propose notes for the version location change [recorded in http://www.w3.org/2013/01/07-mlw-lt-minutes.html#action02]
 
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.137 (CVS log)
$Date: 2013/01/08 16:54:54 $