<scribe> scribe: CarlosD
Wilco: there are 4 PR in Call for Review
Jean-Yves: Hidden State ends today and I'll probably merge it this afternoon
Wilco: PR 1669 and 1671 fix issues related to semantic roles, more editorial
Wilco: we are still in the
process of having the designs approved
... waiting on Shawn to approve them
... we still have to decide on the rules having a table of
contents
... the template includes implementation tracking tables
... and they are in a new design
... test cases' links will open in a new page
... Another update: the TF has gone through all of the
rules
... in theory every rule can be moved to the WAI website
... and we started surveying again
... there are 3 rules related to the letter-spacing and
text-spacing are being reviewed
... if there are no issues they will be sent to the AG
<Wilco> https://github.com/act-rules/act-rules.github.io/issues/assigned/WilcoFiers
Wilco: I haven't looked at my
issues, since I've been working on the migration
... I'll pick up 1482 next
... and give 459 another pass following a recent PR related to
it
<Wilco> https://github.com/act-rules/act-rules.github.io/issues/assigned/Jym77
Jean-Yves: many of my issues have PR open, I'm moving forward
<Wilco> https://github.com/act-rules/act-rules.github.io/issues/assigned/carlosapaduarte
Wilco: Anne has no assigned issues
<Wilco> Carlos: Haven't looked during my vacation. I have a few open PRs.
<Wilco> ... Will work on 1445 and 1446 next.
Wilco: Aron prepared a proposal
for test instructions
... any comments
Jean-Yves: Having test
instructions before the test case will not be picked-up by test
scripts running automated checks of the test cases
... we will end up forget about them
... and that will force us to update our tools to place the
page in the required state for those specific cases
Wilco: this is not going to be on
a lot of rules
... just the ones where it is essential to have some
interaction before running the test
Jean-Yves: I see the need for
this
... but I don't think the current proposal is scalable
... if I'm an end-user choosing tools, when I run the tool
agains the test cases it will miss those cases if the end-user
doesn't place the page in the required state
... but it might be good for now, since we have few test cases
that need it
Wilco: we might constraint it a
bit
... if we have two instructions that are the same, they should
programmatically match
dmontalvo: we could have a list of instructions that we use
<dmontalvo> Carlos: I don't think this will address Jean-Yves concerns, we would still need to have some kind of metadata
Wilco: should we have regular expressions for each instruction that would allow access to the variables that need to be input?
Jean-Yves: we could link the test
case to the test instructions so they could be automatically
parsed
... and implementation checks could make use of the test
instructions
Wilco: I will work on this with
Aron
... Should we switch the order and have the description before
the test instructions?
CarlosD: That makes sense to me
Jean-Yves: The description could start with something "After running the test instructions ..."
anne_thyme: The instructions
should explicitly state when the test should be run
... there might be instances where the test needs to be run
more than once
<dmontalvo> Carlos: Test seems ambiguous. whole process versus each of the parts
Wilco: These tests need to
compare a before and after states
... should the instructions make that explicit
... or do we leave to the tester/tool knowing what are the
states to compare?
... that are not many tools that can track state now, but there
are some
anne_thyme: That should be covered by the applicability and the expectation because those should specify what needs to be tested
Wilco: a tester is responsible
for generating the states
... WCAG doesn't tell you how to get to the different
states
... what we are trying to do here is including the exploration
part in the test cases in some scenarios to make sure the test
case works
... we can't just assume that everyone gets that right
anne_thyme: I still think it
would be nice to state when to run the test
... The wording "Test instructions" might be misleading...
these are not instructions to test, they are instructions to
get to the point where the test is to be executed
Jean-Yves: Perhaps use "Setup
instructions"
... replace "Start the test when the window is loaded" with
"Wait until the window is loaded"
Wilco: But what matters is the transition when we "click the button", so these are test instructions
anne_thyme: But when does it end?
<Wilco> "End the test when the status message is updated"
dmontalvo: But what happens when there is no aria... that's an error
Wilco: But in this case there
will be error - the test case is written that way
... we're just giving instructions to trigger the message
anne_thyme: I would be more comfortable if we state what should happen in the test case, so we know when to end it
aron: This is the test description
anne_thyme: But I would like to have it also in the test instructions
Jean-Yves: I agree. We can have a
test where there is a delay between clicking the button and
something happening
... we need to know how long to wait
anne_thyme: Sometimes we have to wait before several resources are loaded before running the test and we have to know that
Jean-Yves: We now what is going to happen in the test case, so we can be very precise in describing it
Wilco: I propose a final bullet in the instructions with a description of what needs to happen before completing the test
Jean-Yves: I don't really like
the name "Hidden state"
... I like Wilco's suggestion of "Programatically hidden"
dmontalvo: I agree with the suggestion
<dmontalvo> Carlos: Agree with the suggestion, can see the rationale behind,
Jean-Yves: I'll update it
Wilco: ARIA 1.2 is in CR
... it should be moving out of CR soon
... we should adapt to 1.2
... is there any reason for us not to move?
aron: shouldn't we do it when it's actually a recommendation?
Wilco: it's pretty close... it's
safe to assume it's not going to change anymore
... and it's a lot of work to move it
... when we end up the migration effort it should be a
recommendation
aron: can we have PR for that, which we merge only it becomes a recommendation
Wilco: that's a good idea
anne_thyme: will tool vendors update?
aron: I'm happy to grab a rule each week and update it to 1.2
Wilco: the required owned
elements and roles have test cases that need to be
updated
... and the definition of WAI-ARIA specifications that has all
the ARIA specs in it
<Wilco> https://act-rules.github.io/glossary/#wai-aria-specifications
Wilco: final thoughts
<Wilco> https://github.com/act-rules/act-rules.github.io/pull/1590
This is scribe.perl Revision VERSION of 2020-12-31 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: Irssi_ISO8601_Log_Text_Format (score 1.00) Default Present: CarlosD, Jean-Yves, anne_thyme, aron Present: CarlosD, Jean-Yves, anne_thyme, aron Found Scribe: CarlosD Inferring ScribeNick: CarlosD WARNING: No meeting chair found! You should specify the meeting chair like this: <dbooth> Chair: dbooth WARNING: No date found! Assuming today. (Hint: Specify the W3C IRC log URL, and the date will be determined from that.) Or specify the date like this: <dbooth> Date: 12 Sep 2002 People with action items: WARNING: IRC log location not specified! (You can ignore this warning if you do not want the generated minutes to contain a link to the original IRC log.)[End of scribe.perl diagnostic output]