Description:
See the private action item for dial in details
https://www.oasis-open.org/apps/org/workgroup/xliff-omos/members/action_item.php?action_item_id=3822
==========
Agenda:
A. Admin
1- Roll call
[7 voters in total]
2- Minutes from Nov 8 https://lists.oasis-open.org/archives/xliff-omos/201611/msg00009.html
B. Material
1- XLIFF OM
dF:
Grandfathering the provisional directory: https://github.com/DavidFatDavidF/XLIFFOM/blob/master/model/Class_Diagram.PNG.
Content transferred to the new OASIS directory, at Repository URL: https://github.com/oasis-tcs/xliff-omos-om
Primary Maintainer: David Filip
Backup Maintainer: Felix Sasaki
2- JLIFF
Repository URL;: https://github.com/oasis-tcs/xliff-omos-jliff
Primary Maintainer: Robert van Engelen
Backup Maintainer: Chase Tingley
AI Robert to add readme info / instructions – using the JLIFF repo “Wiki” page if possible, or as part of the schema itself
dF:
Feel free to use all available Repo features, including Issues and Wiki
Check content of 3 OASIS provided .md files for high level instructions / guidance regarding the use of publicly accessible GitHub repository for TC chartered work
AI David to update TC page on OASIS site so it points to new GitHub repository
Most committees members have GitHub accounts =>
AI All to follow the repositories (From main repository page, click on “Watch” icon to follow what happens in the repository).
3- TBX Mapping
Continue work on SVN: https://tools.oasis-open.org/version-control/browse/wsvn/xliff-omos/trunk/XLIFF-TBX/xliff-tbx-10.xml
Link to latest Editors' Draft of TBX mapping: https://tools.oasis-open.org/version-control/browse/wsvn/xliff-omos/trunk/XLIFF-TBX/xliff-tbx-v1.0.pdf
AI James and dF to schedule 1:1 editorial call for follow-up. Good progress.
C. Other Topics
1- Date of next meeting?
· Normally on Dec 27, 2016, but ..
· Next meeting scheduled for Jan 10 (2nd Tuesday, as usual).
2- Liaisons
dF/James:
TBX Steering Committee / ISO TC37/SC3
dF:
XLIFF TC - XLIFF 2.1 CSPRD01 launched on 26th October, will end on 25th November
Review closed, significant cahnges under way:
https://issues.oasis-open.org/browse/XLIFF/?selectedTab=com.atlassian.jira.jira-projects-plugin:summary-panel
Do subscribe to the comment list https://www.oasis-open.org/committees/comments/index.php?wg_abbrev=xliff
if you're not a member of the XLIFF TC you can only comment through the comment list (if you have a subscription)
3- Promotion
4- AOB
==========
Minutes:
A. Admin
1- Roll call
dF: 4 voters present, we have quorum
[7 voters in total]
2- Minutes from Nov 8 https://lists.oasis-open.org/archives/xliff-omos/201611/msg00009.html
dF: I move these are meetings of record, do I have a second?
[Robert seconds]
dF: I hear no dissent.. Minutes are accepted
B. Material
1- XLIFF OM
dF:
Grandfathering the provisional directory: https://github.com/DavidFatDavidF/XLIFFOM/blob/master/model/Class_Diagram.PNG.
Content transferred to the new OASIS directory, at Repository URL: https://github.com/oasis-tcs/xliff-omos-om
Primary Maintainer: David Filip
Backup Maintainer: Felix Sasaki
dF: Felix and I had direct write access and discussed with Robin Cover about who else could have direct write access. We can have more maintainers but primarily there are only two and others should contribute via pul requests through the primary maintainers. Everyone can edit directly Issues and Wiki though
If you do a full checkout of the Repo you can edit it in Papyrus as described by Chase on the XLIFF OMOS Wiki
https://wiki.oasis-open.org/xliff-omos/WorkingWithPapyrus
The reason to set it up on GitHub with Papyrus is to allow several people to edit the diagram without conflict.
The grandfathered repo is available for record.
Does anyone have input about the OM Repo? If not we can progress to the JLIFF Repo.
2- JLIFF
Repository URL;: https://github.com/oasis-tcs/xliff-omos-jliff
Primary Maintainer: Robert van Engelen
Backup Maintainer: Chase Tingley
AI Robert to add readme info / instructions – using the JLIFF repo “Wiki” page if possible, or as part of the schema itself
DONE
dF:
Feel free to use all available Repo features, including Issues and Wiki
Check content of 3 OASIS provided .md files for high level instructions / guidance regarding the use of publicly accessible GitHub repository for TC chartered work
AI David to update TC page on OASIS site so it points to new GitHub repository
DONE
Most committees members have GitHub accounts =>
AI All to follow the repositories (From main repository page, click on “Watch” icon to follow what happens in the repository).
ONGOING
dF: Robert, could you provide a summary of JLIFF repo progess
Robert: Yes, it is bare-bones but the schema and examples are there. There are three validators (links on repo) to check the examples against schemas. If I find more I will add them.
I will also add more explanation in the Wiki.
I will add the information to the Wiki because it is easier to edit than the README.
dF: So the validator http://jsonschemalint.com/#/version/draft-04/markup/json needs two inputs: JSON schema and an example.
Robert: Yes, it is possibly easier to have two screens open to do this. It is not very user friendly..
dF: If someone has an automation idea that would be good.
Robert: I spent most time on the comparison to XLIFF OM and the XLIFF schema v2. As you see there are elements which are not in the OM Diagram yet. The reason I did that is so we can look at an XML version so we can compare back and forth. I feel conversion is theoretically possible.
dF: It is good that you have overtaken the OM diagram.
Robert: [discussion of schema]
dF: Chase do you have any input?
Chase: This seems like a good starting point.
Robert: Let’s look at Example 1, which answers your question about the usage of kind to define tags in JSON.
There’s an array of optional properties and the required “kind” tells us what kind of information this is.
dF: So it basically enforces the data model?
Robert: Yes. I recommend we look at what additional properties are allowed. The validator would ignore additional properties
dF: We are going step-by-step which is good, but we need to support core first. Modules are extensions that are sanctioned by the committee, so there are module attributes allowed on the inline codes but other attr. are not allowed. Markers allow extensibility. So we may have issues with markers because you can’t forbid additional attributes.
Robert: Do you want me to start adding those in so there are constraints that allow extensibility?
dF: I think we can, for now, try to forbid the extensibility on the codes. “sc, ec, pc, ph, cp”
“sm, em, mrk” are the markers that should stay extensible
Robert: I will do that in the schema and describe it in the wiki. I would like if we had more eyeballs to see if this is acceptable.
dF: Thank you Robert
Felix: I have a question about JSON capabiblity to reference IDs
Robert: In JSON there is no equivalent ID/IDREF mechanism. It is up to the data model to represent IDs and references. The validator does not validate IDREF consisitency. There are no constraints as far as I know to manage that. We can enforce the existence of the attribute, but not the IDREF.
dF: There will probably be need to develop custom validation methods, but it is interesting to see how far we can get with the transparent schema-based method.
Robert: Yes, that makes sense, no comments.
dF: There will be no way in JLIFF schema to say whether the id referenced from an ec startRef actually exists.
Robert: Yes, the concept of validation is different in JSON, you can add constraints as listed values.
dF: Constraints by enumerations. Are there any mismatches between XML and JSON in the basic schema data types?
Robert: There are discussions on the best way to do this There is still work being done on JSON Schema because it is hard to create extensible schemas due to strict handling of properties. It is a work-in-progress. V5 is the current WIP.
dF: This seems quite popular
Robert: Yes, but I have worries about JSON Schema V5. Progress is slow. I am uneasy about certain things happening. So I would be uneasy about the JSON Schema v5 draft. V4 seems safe.
dF: They have all versions on the one repo?
Robert: Yes, it seems JSON Schema isn’t used often because most JSON validation is done in code rather than with JSON Schema.
Also the cost of validation with JSON sechema may be high, it may grow exponencially rather than linearly.. but I made sure our is not so hungry..
dF: The use-case is different, in JLIFF we want to focus on the unit fragment, since you would not likely transfer the entire XLIFF in JLIFF, but only unit level exchange.
Robert: In JSON the entire element must be read into memory before you can validate certain things.
dF: same in XML, people are not required to use our artifacts, but the artifacts are good because they are transparent. People are still free not to use them as long as their custom code produces the same results..
Robert: I think there are additional constraints that are not able to be shown in JSON, such as IDRef or type. This schema is relatively efficient and doesn’t involve backtracking in any major way.
dF: Great, let’s move on.
3- TBX Mapping
Continue work on SVN: https://tools.oasis-open.org/version-control/browse/wsvn/xliff-omos/trunk/XLIFF-TBX/xliff-tbx-10.xml
Link to latest Editors' Draft of TBX mapping: https://tools.oasis-open.org/version-control/browse/wsvn/xliff-omos/trunk/XLIFF-TBX/xliff-tbx-v1.0.pdf
AI James and dF to schedule 1:1 editorial call for follow-up. Good progress.
James: I made a few more updates to the file
dF: Great, please send me an email so we can meet before Christmas.
C. Other Topics
1- Date of next meeting?
· Normally on Dec 27, 2016, but ..
· Next meeting scheduled for Jan 10 (2nd Tuesday, as usual).
dF: I will cancel Dec. 27 meeting and schdule the regular for Jan. 10
Will everyone be back by 10th Jan?
Chase: seems fine..
2- Liaisons
dF/James:
TBX Steering Committee / ISO TC37/SC3
James: I don’t have anything specific to report
dF: Ok, we will see tomorrow.
dF:
XLIFF TC - XLIFF 2.1 CSPRD01 launched on 26th October, will end on 25th November
Review closed, significant cahnges under way:
https://issues.oasis-open.org/browse/XLIFF/?selectedTab=com.atlassian.jira.jira-projects-plugin:summary-panel
Do subscribe to the comment list https://www.oasis-open.org/committees/comments/index.php?wg_abbrev=xliff
if you're not a member of the XLIFF TC you can only comment through the comment list (if you have a subscription)
dF: Felix, do you want to report on 2.1 progress? Our official liaison Yves is not here today.
Felix: progress is going well, only change tracking has still some discussions
dF: I will post the link you can use if you want to follow the progress. https://issues.oasis-open.org/browse/XLIFF
dF: Chase did you follow the changes with change tracking?
Chase: Yves pointed out all the changes I had concern with.
dF: I think it’s on good track
You may follow the CTR discussion here https://issues.oasis-open.org/browse/XLIFF-4
3- Promotion
dF: I will talk on 19. December at DCU about the multilingualism stnadards
https://www.eventbrite.ie/e/expert-talk-multilingual-content-interoperability-dr-david-filip-tickets-29793020727
and I spoke at an EC workshop last week in Gdansk. It the Translating Eirope series.. http://ec.europa.eu/dgs/translation/programmes/translating_europe/index_en.htm
Anyone else have any promotion stuff to report?
Felix: FREME is closing end of January. FREME will be continued by Adapt centre. see http://adaptcentre.ie/ezine/Nov16/FREME/ . We are doing XLIFF - linked data processing in FREME. that is now part of FREME native - the Dublin June demo was still in the browser
dF: maybe it will be a good venue to publish the XLIFF 2.1 schematrons
Felix: schematron publishing is possible. sure the peopel in ADAPT doing that are Dave Lewis and Kevin Koidl. I will continue to contribute from outside
4- AOB
dF: Any agenda item proposals? No
I encourage people to follow XLIFF 2.1 development on the TC JIRAand the GitHub repos to see the changes there.
XLIFF OMOS list will be notified on teh 2nd public review
Meeting adjourned
==========
Attendance:
Meeting Statistics |
Quorum rule |
51% of voting members
|
Achieved quorum |
yes |
Individual Attendance |
Contributing Members: 5 of 18 (27%) Voting Members: 4 of 7 (57%) (used for quorum calculation)
|
Company Attendance |
Contributing Companies: 5 of 14 (35%) Voting Companies: 4 of 7 (57%)
|