W3C

- DRAFT -

mlw-lt wg

17 Apr 2013

Agenda

See also: IRC log

Attendees

Present
kfritsche, dF, philr, Karl, pnietoca, Jirka, Des, shaunm, mdelolmo, leroy, Ankit, fsasaki(onlyIRC), Yves, chriLi, tadej
Regrets
Jörg,Dom
Chair
david
Scribe
daveL

Contents


1) Review of issues 70 - 92

<dF> https://www.w3.org/International/multilingualweb/lt/track/actions/465

<dF> https://www.w3.org/International/multilingualweb/lt/track/issues/pendingreview

2) LC issues in general see http://tinyurl.com/its20-comments-handling

df: dave to check a close remaining issues offline, checking for response timeouts

df: last four issues remaining at:

http://www.w3.org/TR/2013/WD-its20-20130411/#unresolved-issues

dF: asks if there are any other issues apart from those four?

none raised

3) Change of regex for allowed characters https://www.w3.org/International/multilingualweb/lt/track/issues/67

pablo: has produced ABNF to go into specification, and waiting on schema to validate regex

dF: asks jirka is he is OK to make changes to schema

jirka: we first need to agree grammar
... allow a few extra days so everyone is happy before addressing schema for regEx

<fsasaki> hi dF, all, just for the record, I am happy to close the issue-67 without the schema change

jirka: some points to clarify
... one issue is encoding beyond unicode
... BMP support, e.g. this was motivation for compatibility with some existing regex engine that don't have complete unicode support

jirka: so people who raised this issue should check that they are now happy with regEx proposal from Pablo in this regard

dF: felix happy to proceed without schema change, but have we stable regEx for schema

<fsasaki> hi dF, just a clarifiation: we have to have stable ABNF (not regex)

Jirka: agrees, doesn't need change to schema, but need to tie down option for encoding so that regex will work with these other engines

df: feels we should stick to unicode
... asks WG if we are OK with Pablo to go ahead with suggested grammar

pablo: but there are still some issues, related to allowed characters

<pnietoca> it's issue 67, Yves raised it

<Yves_> yes I'm here.

<Yves_> yes

dF: asks Yves as he raised topic originally

<fsasaki> hi all, the issue is the dependency on XML Schema - issue-67 is about removing that dependency. Having the abnf specified is a way to resolve this

<Yves_> sure

dF: can pablo go ahead and raise regex given syntax for grammar agreed on the mailing list
... no objections just clarification, so Pablo please go ahead.
... with aim to close action for approval by next week

4) HTML translate

<dF> https://www.w3.org/International/multilingualweb/lt/track/issues/97

df: should this be handled separately to HTML5 defaults issue?
... asks Jirka to report on development of HTML translate in HTML5 WG

Jirka: nothing to report, things can be changed at any time at this point as it is still at CR stage
... so its hard to build on top of it right now

dF: several options available.
... one is to drop it as no viable to nail down
... second is to wait for the HTML5 to stabalise
... third, set our own defaults
... Yves proposed the last

jirka: support this as waiting for HTML5 could take a long time

<Ankit> 3rd option

dF: propose to issue on pending review, with resolution not to wait for HTML5 but make issue related to HTML5 defaults (up next)

<pnietoca> I'm in favour of merging both issue

<fsasaki> for the record, I would agree with our own defaults, but a question is what they relate to: all content labelled as HTML5 (there is no such label, btw), or each HTML fragment in the HTML namespace, or also including HTML legacy content without namespace?

5) HTML defaults https://www.w3.org/International/multilingualweb/lt/track/issues/118

<Yves_> http://www.w3.org/International/multilingualweb/lt/wiki/HTML5_Defaults

dF: asks Yves to post link to proposed defaults

<dF> http://lists.w3.org/Archives/Public/public-multilingualweb-lt/2013Apr/0088.html

<Yves_> not stable yet

<Yves_> probably

df: asks if content of defaults page is upto date
... propose making this page stable and then referring to this page

<dF> http://www.w3.org/International/multilingualweb/lt/wiki/HTML5_Defaults

dF: i.e. to implementing these stable defaults in the spec

<Yves_> Is it fine to use a wiki page as a normative source?

dF: asks everyone to review this and raise any objections
... better to have reference on wiki page as issue is complex

<fsasaki> Hi Yves, yes, but we should have the wiki page in the ITS IG so that the update is easier, and a description in the ITS2 spec how the update of the wiki page should work

dF: so we keep issue open and aim to resolve via wiki page by next call

<fsasaki> +1

6) Data categories that need mappings in the wiki See http://lists.w3.org/Archives/Public/public-multilingualweb-lt/2013Apr/0087.html

<fsasaki> dF, can you discuss also the mail from jirka?

<fsasaki> see http://lists.w3.org/Archives/Public/public-multilingualweb-lt/2013Apr/0110.html

dF: move mapping to IG wiki
... but aim to continue work within MLW-LT and within its timeline over the remaining months
... ask felix if that is OK from a MLW-LT workgroup proceedure/

christian: raises that version control for mapping is needed, specifically for quality mapping to specific LQI tools

<chriLi> http://www.w3.org/International/its/ig/its20-tool-specific-mappings.html

<fsasaki> hi dF, Dave, not sure what "that" refers to. For a topic like the mapping it is ok to move to the ITS IG wiki now and also include input from outside.

<fsasaki> hi christian, just to emphasize: the quality mapping is non-normative. Version control makes sense, but it won't influence the ITS2 spec - rather the way how the info is presented in the wiki

felix, I mean that MLW-LT can continue develop mapping directly in IG wiki - so i guess yes

<fsasaki> yes, thanks, daveL

df: asks how this table in particular is version managed
... who developed the table

<Yves_> Arle created the table

<fsasaki> hi dF, the table is not version managed. it was developed by arle.

<Yves_> (I think)

<fsasaki> if people agree I'm ok to add a version info to http://www.w3.org/International/its/ig/its20-tool-specific-mappings.html - do you want to give me an action item?

<scribe> ACTION: fsasaki to add tool version info to http://www.w3.org/International/its/ig/its20-tool-specific-mappings.html [recorded in http://www.w3.org/2013/04/17-mlw-lt-minutes.html#action01]

<trackbot> Created ACTION-492 - Add tool version info to http://www.w3.org/International/its/ig/its20-tool-specific-mappings.html [on Felix Sasaki - due 2013-04-24].

<fsasaki> thanks

christian: also not sure about the effect of this table - would it discourage implementation of ITS if the mapping existing
... suggest to include explicit statement to encourage implementation of ITS, rather than rely on mapping

dF: implementation typically use such a mapping internally

<fsasaki> hi christian, I think it depends on how you present the table, not on the table itself. I'm happy to add such an explicit statement to http://www.w3.org/International/multilingualweb/lt/drafts/its20/its20.html#lqissue-typevalues if you give me an action

<scribe> ACTION: fsasaki to add comment to LQI tool mapping hihglighting its informative nature and the separate need to conformance testing of ITs impementations [recorded in http://www.w3.org/2013/04/17-mlw-lt-minutes.html#action02]

<trackbot> Created ACTION-493 - Add comment to LQI tool mapping hihglighting its informative nature and the separate need to conformance testing of ITs impementations [on Felix Sasaki - due 2013-04-24].

7) Importance of normalization See i18n WG discussion at https://lists.w3.org/Archives/Member/member-i18n-core/2013Apr/0021.html

<fsasaki> hi all, my *tentative* proposal to resolve this is here http://www.w3.org/International/multilingualweb/lt/drafts/its20/its20.html#its2-and-unicode-normalization

<fsasaki> I sent that link also to the i18n wg, waiting for feedback - if you have feedback here please let me know

<fsasaki> see my mail to i18n wg asking for feedback on "normalization" section at https://lists.w3.org/Archives/Member/member-i18n-core/2013Apr/0024.html

dF: suggest non-normative text suggesting implementors include a normaliser implemenitng unicode nomalisation form C, see post form felix

<fsasaki> (only visible to w3c members)

dF: the aim is to encourge migration by implementors from legacy encoding to unicode

<pnietoca> It seems good to me

dF: supports this proposed solution
... other comments

christian: clarifies, if a text analytics tool provides annotations it should use such a normaliser?

<philr> When we say Unicode - we mean all encodings: utf-8, utf-16, etc.?

dF: this is a non-normative piece of implementation guidance

<fsasaki> hi christian, yes, like all other implementations of ITS2 - the normalization section doesn't differentate between what piece of the spec is implemented

<fsasaki> hi philr - yes, all unicode, encouraging normalization form C

christian: is this normalisation requirement similar to those in XLIFF, TMX etc?
... if not does it make ITS implementation seem more onerous than these others in comparison

<fsasaki> hi christian, do you have pointers to the requirements in XLIFF, TMX etc.?

dF: reports there are some normalisation statements in XLIFF in places, e.g. for content comparison

<fsasaki> (to be able to answer your question)

<fsasaki> answering the question: no. The requirement in ITS is a general one: when you process content on the web, assure that you have unicode in normalizaiton form C

dF: this is the default behaviour in several places

<fsasaki> (that is at least the purpose of the section). The requirement is not specific to selected processes (NIF, XPath execution, ...)

dF: so the comparison to XLIFF is not to the detriment of ITS2.0 adoption

<fsasaki> hi dF, daveL, what is the outcome of the normalization section discussion - any feedback for me to discuss with the i18n WG? Or can the i18n folks just look at http://www.w3.org/International/multilingualweb/lt/drafts/its20/its20.html#its2-and-unicode-normalization and give comments?

Summary of Action Items

[NEW] ACTION: fsasaki to add comment to LQI tool mapping hihglighting its informative nature and the separate need to conformance testing of ITs impementations [recorded in http://www.w3.org/2013/04/17-mlw-lt-minutes.html#action02]
[NEW] ACTION: fsasaki to add tool version info to http://www.w3.org/International/its/ig/its20-tool-specific-mappings.html [recorded in http://www.w3.org/2013/04/17-mlw-lt-minutes.html#action01]
 
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.137 (CVS log)
$Date: 2013-04-17 17:28:04 $