ITS IG -- 29 Sep 2014

yves: nothing to add on the mapping at the moment
... did not have time to follow up, will try to work on that in the coming months
... not sure if we will have time to implement everything before the end of the year
... phil and others are working on using it and implementing that in ocelot
... using e.g. lqi and provenance
... they will be working on it also on the end of the year rather than now

felix: is that an issue process wise in OASIS? the OASIS xliff 2.1 timing and the mapping progress?

david: xliff tc decided that we want to have yearly releases
... two related things: 1) transfering the mapping into the OASIS specification format
... I am starting that work this week
... I am a bit worried
... if work continues in the wiki we may end up with scattered information
... like to have a simple rule: don't touch ... in the wiki so that I can transfer that to the OASIS template
... the OASIS draft will be public, only OASIS TC members will be able to edit
... must happen sooner or later
... we are kind of late
... implementations by the end of the year is fine
... question is: will we have approved spec in the committee specification phase
... I would take all categogries marked for review and leave others for now until next call

felix: next call timing?

david: before loc world

<yves_> Oct-20 would work for me.

<scribe> ACTION: felix to announce next call 20 october, to prepare for loc world [recorded in http://www.w3.org/2014/09/29-i18nits-minutes.html#action01]

<trackbot> Created ACTION-51 - Announce next call 20 october, to prepare for loc world [on Felix Sasaki - due 2014-10-06].

david: felix had good input, also summarizing issues - some are not even issues, some categories are not metadata, this is not really relevant
... we should have a feeling how it goes in three weeks

jörg: maybe inbetween david can give feedback on the work in progress

david: sure, I will send a link to the xliff 2.1 draft

serge: we (logrus) want to come back on XLIFF 2.0 mapping issue, for sure
... I understand that xliff 2.1 is going to incooperate ITS 2.0
... so we would like to get back to everything we discovered

felix: would be great to get some feedback in the next three weeks

serge: ok, I'll look into that

http://lists.w3.org/Archives/Public/public-i18n-its-ig/2014Aug/0000.html

decide about XLIFF testing design

http://lists.w3.org/Archives/Public/public-i18n-its-ig/2014Sep/0008.html

david: not sure how the ITS2 test site worked - in OASIS we don't have application hosting, just simple SVN
... we have wiki and SVN
... that's it
... does not seem to me that we would be able to replicate W3C testing structure
... we will have reference files that we had for XLIFF 2.0
... XLIFF files with ITS extensions or mappings
... how they are supposed to look before and after
... test automation is not a requirement for OASIS
... in W3C you need to present test suite results to progress
... in OASIS you need to present statement of use
... advanced validation was also approved for XLIFF 2.1
... we will be able to test a lot using schematron rules, without having a specific test suite

serge:

<dF> https://www.oasis-open.org/committees/download.php/54178/ballot_2674.html

<dF> XLIFF TC approval of the ITS feature for 2.1

df: )

[discussion on ITS2 / XLIFF 2.1 outreach]

felix: testing the mapping - would we need input files in a given format, e.g. HTML5?

david: how the extraction is performed is normally out of scope of xliff - this may be an issue with ITS now
... for ITS we may have more stages even than input - output
... not sure about step between original format and XLIFF - it is probably not going to be normative
... we can have best practices - the xliff tc had reference guides but never talked about how the extraction is done

other agenda items

rdf-its ontology - postponed

"open data management for public automated translation services"

https://www.w3.org/International/its/wiki/Open_Data_Management_for_Public_Automated_Translation_Services

serge: please keep this on the agenda
... this topic will develop further
... if we don't have substantial revision of the document, it is time to get through the channels we have and try to push this throuhg
... I will try to propose this as a presentation topic at the lt-innovate summit
... we can try to email this to various people in this theme

<scribe> ACTION: felix to announce work for data managemnt document [recorded in http://www.w3.org/2014/09/29-i18nits-minutes.html#action02]

<trackbot> Created ACTION-52 - Announce work for data managemnt document [on Felix Sasaki - due 2014-10-06].

other open items

http://www.dercom.de/en/projects

"An interface COTI between authoring and translation memory systems is planned. The specification may be downloaded from this page and commented."

felix: any comments on this? should we contact the people?

david: the people do what they think they have to do
... I went through the doc today
... this has an overlap with linport
... approach of spec is similar to linport and tip approach - sending a zip package that can contain project files
... not sure how the project files are supposed to be structured - the doc seems to define a soap envelope
... which is not in conflict with what we are doing
... this is a kind of web services interface
... they define an XML wrapper based on SOAP
... there is not necessarily a conflict
... but I'm a bit worried that there was no interest in contact - there is a potential for mismatch
... but at the same time I think there is not a clash - one could propose xliff 2.0 as a payload for their soap envelope
... xliff does not say anything about the web service should work
... there may be a conflict with the object model and API that people now want to see for XLIFF 2.0
... above feedback is based on me looking into this before today's call
... behind this there is a trade org of german content producers
... the spec does not seem to follow some current practices, e.g. it has its own definitions of keywords "must, can, ..."
... they are referencing xml and soap, and they are referencing some standards

jörg: there is a huge gap in communication between different communities - it seems they don't know about xliff and so on

david: hard to achieve things via mail

felix: will be at tekom again, will try again to contact them and see what happens

david: maybe alan melby could be helpful - linport is in clahs with this

jörg: not really

scribe: it is a little bit from linport and tipp, and then their own way

david: xliff can be payload of anything

jörg: I can check with Alan to see if he has talked to these guys

felix: thanks

aob

http://lists.w3.org/Archives/Public/public-i18n-its-ig/2014Sep/0007.html

adjourned, by to all! next call in three weeks

- DRAFT -

ITS IG

29 Sep 2014

Attendees

Contents

roll call

action items

XLIFF 2.0 mapping

decide about XLIFF testing design

other agenda items

other open items

aob

Summary of Action Items