RE: Minutes for Teleconference on 24 July 2014

Automated WCAG Monitoring Community Group Teleconference 2014-07-24, 16:00 -
17:00 h (CET)
Attendees:
- Annika Nietzio
- Birkir Gunnarsson
- Kamyar Rasta
- Wilco Fiers

scribe: Birkir

Congrats to Kamyar who just got married!

item 1 - Action items
Birkir: Will complete 2.4.2 test specification next week.
Have been waiting for a couple of sample wiki entries from Anika, and now
they are here.
Annika: Has added 3.1.1/3.1.2 to the wiki.
John: Has uploaded his sample html files to the wiki.
Annika: has completed creating templates and uploaded to wiki.
Template question:
What about assigning unique IDs to each test?
Make sure every test has a unique idea, suffix test id pattern with a number
starting with 1 for first test for a given sc.
Wilco: There is a field for unique IDs in the template.
Action item for Wilco, create a sample of how to generate a unique ID using
a consistent pattern, put on wiki.

item #2: tests for success criteria 3.1.1 & 3.1.2 (Annika)
3.1.1:
- Discussed combining steps 2 and 3 for SC 3.1.1 into a single test.
- Run test on files where lang definition in xml is different from html lang
definition to see which one assistive technologies pick. Test with various
screen readers (Annika/Birkir).
- Modify the interactive step in the test for 3.1.1 so that the application
retrieves the language code from the page header an asks the user whether
that language is appropriate for the page (make it into a "yes or no"
question).

3.1.2 -
- Should we retreive every page element with a lang definition different
from language of page and ask user to verify them individually, or should we
rely on the application to highlight the sections with a language definition
in a page overlay and ask users about whether any of them is inappropriate?
If we present each piece of content with a language definition assigned we
could overload the user with tests.
If we do the same via a GUI, we risk a confusing and inconsistent UI for
tools. 
No decision made as-of-yet, need to sleep on it, for a week.
- What about alt attributes for images? Often bi-lingual websites have a
single field for image alt text (not one per language). Also, since alt text
is invisible, it may easily get missed by the developers. 
- Should we display all alt texts and ask if they are written in the
appropriate language?
- Should we display one alt text at random for the user to evaluate?
- Should this be part of testing for 1.1.1 (stretches the definition of
"appropriate alt content" but it could apply).
- How do we determine what constitutes a "phrase" per definition of 3.1.2?
This is a difficult one to determine, in most cases the problem is that
people forget to specify a language, so if we see lang definition in page
content attributes it is overwhelmingly likely they are correct.

item 3: Review:
http://lists.w3.org/Archives/Public/public-auto-wcag/2014Jul/0005.html

Wilco: Please review the tests linked to from this page and letīs discuss
next week, it is important that everyone has read and formed an opinion
rather than doing it on the fly during the meeting.
- If you have comments on the tests, please add them to the wiki and send a
notification email out to the list.
- If no comments are received, the assumption is that everybody approves.

item 4: Test case Status and versioning  
We will have the following 3 statuses for each item.
"draft" 
"review" 
"completed"
Once tests are completed we need to introduce versions.
- Annika will review template and add status/version info.

item 5: Assignment for next iteration
Birkir: 2.1.2
Annika: 2.4.3, various others.
Everyone feel free to pick a success criterion or two.

Final comments:
Wilco: We need to get the tests implemented in code as we are developing
them, when could this happen?
Kamyar: not sure, just got back from vacation. Will have more information
next week.

Received on Friday, 25 July 2014 04:04:46 UTC