ITS IG -- 10 Nov 2014

<scribe> scribe: fsasaki

its in xliff module time line

yves: timeline for delivery of its module
... felix noted, we don't have much time to finish things
... we need to finish both spec + implementations by end of December
... looking at what we need to do that is very hard
... there may be three options
... 1) postpone release
... 2) not put ITS in that release
... 3) have only parts of ITS in that module

david: we should have a concrete spec by Christmas
... implementations can come later
... during February
... if you start public reviews in January, you have at least 60 days to provide implementations

yves: our schedule is based on a single review?

david: yes
... if we need more than one review it will automatically postpone anyway

yves: a really hard schedule

david: I will have now more time and transfer categories, lot of this is putting attributes in module

yves: there are data categories that we have not even discussed yet

david: right
... one is the note, but that is done, right?
... it is not on the wiki, but I made a draft based on 1.2
... there might be roadblocks, but I should manage to transfer stuff by christmas
... postponing module for next release is not really an option, this is the main feature for xliff 2.1
... can work on this during christmas break
... we can still send spec for review during january

felix: if you need help with any editing please let me know (no need to be an editor)

david: good to know that you are ready to help, need to see myself if there is stuff to give to you etc.
... work on docbook template is straightforward

progress and next call

next call 8 december, we need to make progress via mail, then status check on 8 december

<scribe> ACTION: david to make status check on 8 december call [recorded in http://www.w3.org/2014/11/10-i18nits-minutes.html#action01]

<trackbot> 'david' is an ambiguous username. Please try a different identifier, such as family name or username (e.g., dfilip, dlewis6).

<scribe> ACTION: dfilip to make status check on 8 december call [recorded in http://www.w3.org/2014/11/10-i18nits-minutes.html#action02]

<trackbot> Created ACTION-55 - Make status check on 8 december call [on David Filip - due 2014-11-17].

http://lists.w3.org/Archives/Public/public-i18n-its-ig/2014Nov/0020.html

ITS module namespace

http://lists.w3.org/Archives/Public/public-i18n-its-ig/2014Nov/0013.html

yves: we will use a single namespace for the module
... the URI is made from the other (xliff) modules
... the question is arising for the detault prefix for that namespace
... by convention we use the prefix for the same identifier
... david things we should use ITS, I think we should use something else because ITS is taken

<dF> urn:oasis:names:tc:xliff:its:2.1

david: good summary
... above is URI that is on the draft
... argument to use something else than ITS is not to confuse things
... prefix appears on a few places

<dF> its:

david: can also appear in values as an authorative namespace

<dF> its:

<dF> mtc:

david: we can have values from various modules

<dF> msft:

<dF> subState

david: all occurences should use the same short cut
... question is: is ITS taken because it is used by w3c rec?
... or is it ok to use it in xliff?
... not sure if it matters - what is the danger?
... your extractor needs to follow the mapping anyway

http://lists.w3.org/Archives/Public/public-i18n-its-ig/2014Nov/0023.html

felix: see above mail for technical reaons where prefix matters

http://www.w3.org/International/its/itstagset/verbatim.xsl

yves: ITS is default prefix for ITS in general
... we should respect the convention
... there is nothing that prevents us for using a different namespace

david: xits has a private connotation
... the other namespaces are not based on a w3c namespace
... itsm is a good idea, it indicates "its module"

felix: how to move forward with discussion?

david: yves copied xliff TC on discussion, they will decide before next call
... xslt example is probably the most convincing
... as yves said there are many options, main are: the same "its" or something else, something short like itsm - for some reasons itsx or xits are not good
... I could live with its + s.t. else

XLIFF test outupt

<dF> itsm

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

<dF> meaning its module

http://lists.w3.org/Archives/Public/public-i18n-its-ig/2014Nov/0016.html

http://okapi.opentag.com/snapshots/okapi-xliffLib_all-platforms_1.1-SNAPSHOT.zip

felix: any open questions related to the output?

yves: need to define the line break type + encoding
... having the same one would be useful
... need to start producing the output
... what we need to produce to make progress:
... 1) ITS rules for the mapping
... 2) write algorithm that felix, fredrick etc. talked about, how to convert an original file to work with ITS proceccing, e.g. deleting markers etc.
... 3) after that it is matter of processing things

<scribe> ACTION: felix to do write up of processing ITS+XLIFF files [recorded in http://www.w3.org/2014/11/10-i18nits-minutes.html#action03]

<trackbot> Created ACTION-56 - Do write up of processing its+xliff files [on Felix Sasaki - due 2014-11-17].

david: need to define generic rules file, and 2) the algorithm
... the generic file that would go for all generic ITS processors
... makes sense to have rules file associated with an XLIFF file that does not have the module

yves: annotators reference - we could have them for all data categories
... in some case they are mandatory
... should we put them if they are existing, in addition to data category info?

felix: if avail., should put them in the output

yves: should we just put information for that data category

felix: +1

yves: and shows that the tool know inheritance etc.

ITS preserve space and language information

http://lists.w3.org/Archives/Public/public-i18n-its-ig/2014Oct/0039.html

http://lists.w3.org/Archives/Public/public-i18n-its-ig/2014Oct/0045.html

felix: looking in the archive I was not sure if there is a clear outcome of the discussion

yves: latest was: two additional attributes - that are part of the ITS module
... its:space and its:lang

felix: itsm:space or itsm:lang ?

yves: yes, if we use that prefix

david: pretty much clear and resolved for space
... the lang topic: I have not done a draft yet for this one
... there are two options: say it is not possible to express inline in xliff
... or to introduce itsm: attribute
... if we use itsm: prefix
... the main use case of xliff is bilingual - source and target
... is this strictly analogical to preserve space or different?

yves: you could have a sentence with a citation
... spell checking could help if language information is inline

david: it would be non translatable with translation one language > other
... trying to thing about use cases

discussion on citations

yves: you don't make segementation based on lang info
... but if you have s.t. quoted in a language you don't segment it
... I am trying to say: language information is independent of segmentration, translate etc.
... useful e.g. for spell checker - we should try to provide it

david: technically it can be done like space
... fine to handle it the same way as space
... enough info for doing the editing in the spec

felix: great

Provenance and change track module

http://lists.w3.org/Archives/Public/public-i18n-its-ig/2014Oct/0034.html

yves: looking at ITS provenance, wondering how that relates to change tracking
... some if the info is not the same
... there is an action on dave lewis to look into that
... I tried to look at it - I don't use change tracking currently

david: owner in xliff tc is ryan
... dave + ryan would be best team to move this forward
... change tracking addresses only a specific use case
... you can record provenance also on things that have not been changed
... we should look into mapping between them - there should be information from change tracking mapped via ITS rules into ITS provenance

felix: not sure if it is feasible to always do that mapping

david: there will be some 1:1 relationships between attributes

action-9?

<trackbot> action-9 -- David Lewis to Look at the XLIFF 2.0 change tracking module for provenance -- due 2014-05-30 -- OPEN

<trackbot> http://www.w3.org/International/its/ig/track/actions/9

felix: will check with dave

loc world

david: we can be at locworld with feisgillt again, in berlin, will be week of june 1st, feisgillt then june 2-3
... reserve the dates asap

General: example XLIFF file with ITS2 metadata

http://lists.w3.org/Archives/Public/public-i18n-its-ig/2014Oct/0033.html

http://lists.w3.org/Archives/Public/public-i18n-its-ig/2014Oct/att-0033/ITSInXLIFFtests.zip

yves: will update the example with invalid files as well, so that people can test their validation
... any additioal examples are welcome

aob

<yves_> bye

next call 8 december, let's progress via mail before

adjourned

- DRAFT -

ITS IG

10 Nov 2014

Attendees

Contents