JLReq TF meeting notes 2019 June 12th

JLReq TF meeting notes 2019 June 12th

Attendee
Kobayashi, Tatsuo
Kobayashi, Toshi (Bin-sensei)
McCully, Nat (online)
Murakami, Shinyu
Shimono, Atsushi
Tajima, Jun
Takase, Hiroshi
Kida, Yasuo

Meeting notes

Status of JLReq update (Shimono-san)
Status
There are a few remaining open errata: #10 <https://github.com/w3c/jlreq/issues/10>, #15 <https://github.com/w3c/jlreq/issues/15>, #18 <https://github.com/w3c/jlreq/issues/18>, #53 <https://github.com/w3c/jlreq/issues/53>, #56 <https://github.com/w3c/jlreq/issues/56>
Consensus
Will fix discrepancies between English and Japanese (like #53 <https://github.com/w3c/jlreq/issues/53>, #56 <https://github.com/w3c/jlreq/issues/56>) as they are reported.
Regarding the glossary (#60 <https://github.com/w3c/jlreq/issues/60>), the agreement is to fix the links.
Actions
Archive images on git-lfs (#63 <https://github.com/w3c/jlreq/issues/63>). Thank you Xue Fuquiao for suggestions. (Shimono-san)
Fix the links in the glossary (Shimono-san. The GA report is higher priority.)

Gap Analysis: analysis of the work (Murakami-san)
Murakami-san reported the overview of GA tasks based on his chart that describes availability of corresponding CSS and tests for each section of JLReq. Link to the chart <https://docs.google.com/spreadsheets/d/18igcZxU7LErg2bn9qfjJe40gqAkx_lvt059exqsX2ZQ>.
Discussions
There are two different cases for “css exists? = no”. One is that the feature require CSS but they have not been defined yet. The other case is that the feature does not require CSS because it should always work in a certain way. They still need tests. We need to separate these two. → Murakami-san to reflect it in the action column.
We want rough priority on each section to organise the work. → Bin-sensei to assign three level prioritisation to the chart.
Concerns are raised against the current approach of analysing and organising the work. They are all good points. We will discuss them at the next meeting.
There are features that are left unsaid in JLReq because they are so basic and natural to native speakers but they are new concepts to English speakers. → We want to capture them in a section of the GA report, as well as in the future version of JLReq.
There are features that are achieved by combinations of CSS, e.g. section 2 “Kihon-hanmen”. A concern is that they are not properly tested with approach of developing tests for individual CSS or for particular aspect of a feature. (This might be similar to the concept of unit tests vs integration tests)
Tests should be developed in a way that they try to reproduce good results that JLReq expects, rather than to test CSS to see if they work in a way that they are designed for.
Actions
Add an “Action” column (Murakami-san)
Assign rough priority (Bin-sensei. Kida will send a chart to him)
Will discuss more about the concerns raised at the next meeting (Kida / Nat)

Gap Analysis: framework of the work (Kida)
We are almost ready to start the GA work. Agreed on the following assignments:
Developing missing i18n tests: Murakami-san, Takase-san, Tajima-san with help from Shimono-san
Writing GA reports:  Shimono-san
Kobayashi-san will cover topics that do not fit well with the structure of the report.
Priorities: Bin-sensei
Available resources (other than i18n tests and wpt)
We can take advantage of tests that JAGAT developed to test EPUB.
Example text used in JLReq can be freely used for tests. If necessary Bin-sensei can generate example texts.
Question to Richard
Which GitHub issues should be used to track the work of developing i18n tests?

User Feedback (Tajima)
Shimono-san extracted user feedback in a spread sheet (34 items).
Action
Add comments to each item (Bin-sensei)

The next meeting (Kida)
The next meeting will be F2F. Kida to find a date.
The rough agenda will be:
Status of JLReq update
Status of Gap Analysis work. is it ramping up?
Discuss how we address concerns raised on the current approach of analysing and organising the work
any other?
//