Re: Analysis of TTML2 2nd Ed and test suite

On Thu, Feb 6, 2020 at 11:44 AM Glenn Adams <glenn@skynav.com> wrote:

>
>
> On Thu, Feb 6, 2020 at 12:36 PM Cyril Concolato <cconcolato@netflix.com>
> wrote:
>
>> Hi all,
>>
>> I had a look at the list of changes in TTML2 2nd Ed (
>> https://www.w3.org/TR/2020/CR-ttml2-20200128/ttml2-changes.html). We
>> have 13 syntax changes and 14 semantic only changes, for a total of 27.
>> Thanks to Glenn's work it's fairly easy to track the associated tests. Out
>> of these, 14 already have tests (corresponding to 50 new files, thanks to
>> Glenn again). 4 are marked as 'untestable'. We are left with 9 changes that
>> need review. Here they are:
>>
>> Syntax Changes
>> * Add #presentation-audio feature designation (w3c/ttml2#1097).
>> * Add IMSC profile designators (w3c/ttml2#1123).
>> * Permit audio|image in body content model (w3c/ttml2#1125).
>>
>> Semantic (Only) Changes
>> * Ignore text children of ruby container spans (w3c/ttml2#1077).
>> * Define shear value to degree mapping (w3c/ttml2#1094).
>> * Clarify media type parameters (w3c/ttml2#1099).
>> * Define font selection strategy used by line height algorithm
>> (w3c/ttml2#1106).
>> * Remove error condition from scaling procedure (w3c/ttml2#1109).
>> * Correct #extent-root-version-2 designation (w3c/ttml2#1110).
>>
>> Glenn, does this match your analysis?
>>
>> We could start populating the Implementation report with the 50 rows for
>> the 50 test files currently available, to determine the gap in terms of
>> implementation support.
>>
>
> I presume you are counting the new entries marked "2e" in the JSON files?
>
No, I looked at all test PRs associated with the changes and counted all
new files. See this spreadsheet
<https://docs.google.com/spreadsheets/d/1oo9DtHBBn_t0Nhba4thDASibS0G5fZaJ9HkvlRXdTqo/edit?usp=sharing>
.


> I need to review these to check their mapping/association with the PRs
> listed in the change document to verify whether (and which)
>
>    - "2e" test files are misassociated with a 2e substantive PR
>
> I didn't check that specifically.

>
>    - "2e" test files are missing, i.e., a 2e substantive PR exists but no
>    "2e" test file is associated with it
>
> I think these are the 9 bullet points above.


>> Glenn, is there a script to generate these rows from the JSON "since:
>> "2e"" attribute?
>>
>
> no
>
Do you mind if I go ahead and update the IR? or do you prefer to wait until
we have all the files?

Cyril

>
>
>>
>> Nigel, I noticed that we have 4 open PR on the test repo that we should
>> probably review and possibly merge or discuss at the next meeting.
>>
>> HTH,
>> Cyril
>>
>