See also: IRC log
<trackbot> Date: 04 June 2015
<MichaelC> meeting: ARIA
<scribe> scribe: fesch
zakim item 1
<scribe> POSTPONED
<clown> matt's email: https://lists.w3.org/Archives/Public/public-pfwg/2015Jun/0007.html
<joanie> https://lists.w3.org/Archives/Public/public-pfwg/2015Jun/0009.html
mc: I understand there are two branches a Matt branch and Joanmarie branch
<clown> joanie's email: https://lists.w3.org/Archives/Public/public-pfwg/2015Jun/0010.html
<joanie> joanie's email with details: https://lists.w3.org/Archives/Public/public-pfwg/2015Jun/0009.html
mk: after studying it much more,
on the side to NOT use a property to turn a static table into
an interactive grid
... primary reason, state and properties associated with an
interactive grid are different than states/properties for
tables which can cause mass confusion.
... anyone disagree?
... a concrete example - aria-selected could be used with grid
but not with table - how would we make that clear in the
spec?
<clown> http://w3c.github.io/aria/aria/aria.html#aria-selected
<Zakim> LJWatson, you wanted to agree separating design patterns makes sense.
mk: if you have a grid that could be a grid (active true) or a table (active false) how do you work with properties
lw: talked with developers - all
a table is a static thing and interactive thing are different
use cases, separating them is logical thing to do
... a simple set of roles for tables seem like the soln
mc: Matt you are OK with either approach?
mk: no, I only like approach of
separate roles (not interactive property)
... I wouldn't be comfortable if we implemented both... (only
separate roles)
<LJWatson> s/a simple set of roles for tables seem like the soln/Based on conversations with developers, a separate set of roles for simple/static tables as opposed to interactive/grid tables makes most sense./
mc: Joanie has branch with separate roles...
<joanie> https://lists.w3.org/Archives/Public/public-pfwg/2015Jun/0009.html
jN: what do we mean by separate roles? Just table vs grid or header roles as well
<clown> https://rawgit.com/w3c/aria/table-and-cell/aria/aria.html#table
<clown> https://rawgit.com/w3c/aria/table-and-cell/aria/aria.html#cell
jd: wanted to keep it separate -
two roles table and cell
... reads her email...
... read about cell
... reads about table...
mk: first question on colheader
and rowheader - sometimes colheader is a widget and sometimes
not... properties not appropriate for table
... aria-readonly - should never appear on static table,
aria-require... are not global if context is table but would be
for grid
jn: not sure how readonly applies to a grid?
mk: editable column and header
could be in a spreadsheet....
... could mark readonly if user can't change name
cs: or in a pivot table, but I
don't think it is a common situatuion
... something about an old COBAL table - not editiable
jd: how the spec pulls it in and
how we rely on inherited properties and states, doesn't apply
here but a general problem
... ... expandable rows, sloppy....
mk: aria-expanded....
<joanie> In a treegrid, authors may mark rows as expandable, using the aria-expanded attribute to indicate the present status. This is not the case for an ordinary table or grid, in which the aria-expanded attribute is not present.
<clown> https://rawgit.com/w3c/aria/table-and-cell/aria/aria.html#aria-expanded
<clown> https://rawgit.com/w3c/aria/master/aria/aria.html#aria-expanded
bg: why isn't it OK for aira-expanded to be used on a table, suppose a cell has a link and it expands a row
mk: what does it mean for a row
to be collasped
... what do you show in the table?
<jamesn> http://www.codeproject.com/KB/aspnet/Expandabe-DataTable-MVC/expandableDataTables-table.png
bg: no, a medical service
provider has a pharmacy table.... also links which you can
expand it and it gives instructions for the
medication....
... aria-expanded just lets you know if it is hidden
jn: took a screenshot - company name column, .... when you expand the row, you get all the employees... and this is pretty common
mk: it is hard to claim you are expanding the row... you are showing something hidden
jd: I brought up that thing, is
that sometimes it doesn't make sense, so don't do it.
... what about a sortable table by column?
<clown> http://w3c.github.io/aria/aria/aria.html#aria-sort
mk: aria-sort the spec says that
applies both to column headers in both table and grid, but
there is no way to support that in mappings, aria-sort is not
global -
... IBM checker will put an error on a spec with it, but it is
not correct
<jamesn> another example
mc: seems we have a emerging consensus for Joanie's path,
<jamesn> (notes we often add that info in the summary attribute)
mk: when I approach a table I want to know who the table is sorted without reading the data
bg: need a better mapping... but sort works for that information
mc: seems like we are OK with sort
mk: aria-required is the other
one I am concerned about
... it is weird to hear on a static table, just noise
cs: sounds like a developer problem, not spec problem
jn: if someone puts an * on a column header, to tell the user a column is required, why isn't that valid?
mk: you put required on the input not the header
mc: sounds like nothing needs to
be done with required
... what aria-readonly and aria-required should be done in the
spec
... seems like consensus with Joaine's table branch
+1
mk: the required context role was
missing from the table but mentioned in the text
... trying to find....
mc: easy fix to do offline
<MichaelC> proposed resolution: Accept the new table roles and taxonomy changes modulo notes about aria-readonly and aria-required (Joanie to merge that branch, Matt to park the alternate one)
<LJWatson> +1
+1
<clown> Matt to park the aria-interactive branch for tables.
RESOLUTION: Accept the new table roles and taxonomy changes modulo notes about aria-readonly and aria-required (Joanie to merge that branch, Matt to park the aria -interactive branch)
<MichaelC> action-1505?
<trackbot> action-1505 -- Matthew King to Create an aria-interactive property for issue 633 -- due 2015-05-08 -- OPEN
<trackbot> https://www.w3.org/WAI/PF/Group/track/actions/1505
<MichaelC> close action-1505
s/-nter/-inter/
<trackbot> Closed action-1505.
<MichaelC> action-1505: dropped this proposal in favor of a separate set of table roles
<trackbot> Notes added to action-1505 Create an aria-interactive property for issue 633.
jd: what about the HTML discussion about aria-interaction?
mk: related to action ...
... aria-interactive could come back.
<MichaelC> issue-633: closed action-1505 regarding tables but the issue still needs to be addressed for other use cases
<trackbot> Sorry, but I think you meant to close issue-633.
<MichaelC> issue-633: we closed action-1505 regarding tables but the issue still needs to be addressed for other use cases
<trackbot> Notes added to issue-633 listbox and tree may contain only static items; badly need interactive widgets that can contain interactive typed items.
<clown> https://www.w3.org/Bugs/Public/show_bug.cgi?id=27866
<joanie> I created action-1646 on the readonly and required.
<joanie> action-1646
<trackbot> action-1646 -- Joanmarie Diggs to Add clarifying text to readonly and required properties -- due 2015-06-11 -- OPEN
<trackbot> https://www.w3.org/WAI/PF/Group/track/actions/1646
<MichaelC> issue-633: https://www.w3.org/Bugs/Public/show_bug.cgi?id=27866
<trackbot> Notes added to issue-633 listbox and tree may contain only static items; badly need interactive widgets that can contain interactive typed items.
mc: when we file issues with user agents, that we want to track, we want a keyword in the bug so we can track it
cs: can filter on accessibility
mc: may want something more specific... and add a note on the page, then we can go through the bugs...
jd: make a metabug - for example there is an orca metabug --- I make a lot of bugs.... that block the meta bug....
js: so JD is proposing a ARIA meta bug
cs: could do #aria... or...
js: tried aria lower case picks up lots of stuff
<clown> for example, https://connect.microsoft.com/IE/SearchResults.aspx?SearchQuery=aria
mc: sounds like we need different
practices - for the different trackers (no suprise)
... need to document the practices for the 4 trackers, do we
like the idea of meta-bugs?
js: yes
mc: want meta-bugs for the ones where we can use them...
jd: in chromium you can block bugs
need to expand bug tracking page, then create meta bugs and document in this page
zakim: next topic
mc: does it make sense for a
checkbox to be read only?
... currently checkbox doesn't have readonly should it?
mk: why do you need readonly if you have disabled?
jn: may need readonly if the value is important to the user, disabled the value shouldn't be important to user
cs: often used for choices... with dependencies or you don't have access to change them
jn: preference settings, where administrators change stuff, so you are trying to convey that there are changeable options but you may not have privileges
mk: what is the visible difference?
jn: may see the check but no border, you want the consistency of the check...
cs: we often use disabled...
jn: using disabled - the disabled color contrast isn't sufficient to read
mk: can some User agents take disabled stuff out of the keyboard sequence...
mc: Alexanders noticed different behaviors between browsers... not that he was supporting one way or another
jn; a reason to make custom checkboxes - because you want it to be readonly instead of disabled
<clown> http://rawgit.com/w3c/aria/master/aria/aria.html#aria-readonly
mk: if we want to do this we
should need to reread readonly and disabled
... it is not global so you can't put on a div
<clown> http://rawgit.com/w3c/aria/master/aria/aria.html#aria-disabled
jn: ---- audio connections webex --- eaten by webex...
<clown> from html 5.1 nightly: The difference between disabled and readonly is that read-only controls are still focusable, so the user can still select the text and interact with it, whereas disabled controls are entirely non-interactive. (For this reason, only text controls can be made read-only: it wouldn't make sense for checkboxes or buttons, for instances.
jn: HTML checkbox behavior is
weird, if you put readonly on it - it renders differently, only
keeps it from changing the value, not whether the checkbox
checks or not!
... that is an application developer does that, what we are
doing for AT...
bg: what happens when someone puts readonly on the checkbox
mk: need an author should statement
<Zakim> clown, you wanted to point out what HTML5 says about readonly checkboxes.
js: reads what he pasted in...
cs: we should follow HTML
mk: do we always follow HTML? no
<mattking> This the current description of aria-readonly: Indicates that the element is not editable, but is otherwise operable.
cs: read a real benefit to NOT follow HTML...
jn: a disabled value means - the value in unimportant, readonly means it is important but not changable
mk: is our description the same as HTML?
js: depends on what operable means
<clown> http://rawgit.com/w3c/aria/master/aria/aria.html#dfn-operable
lw: are we trying to fix a problem?
jn: we have 3 states with checkboxes...
lw: I've not come across this as an issue
mc: so James you have use cases for readonly use cases?
jn: yes, but if everyone else disagrees, I will drop it
cs: having it behave different than HTML will cause more problems...
js: this sounds like we disagree with HTML on what readonly means
mc: we have identified use cases but question how widespread they are
mk: a screen reader doesn't care which you use
jn: as a screen reader user, if you hear that an item is disabled do you think the value is important?
mk: yes
lw: an example is a disabled submit button
cs: that has been a paradigm on windows for a long time
lw: suggested a WCAG technique for this
jn: we could do that
mc: makes me wonder if it is
really a checkbox, but is really a different role...
... do we want to continue this discussion?
<MichaelC> ACTION: nurthen to document use cases for readonly checkboxes [recorded in http://www.w3.org/2015/06/04-aria-minutes.html#action01]
<trackbot> Created ACTION-1647 - Document use cases for readonly checkboxes [on James Nurthen - due 2015-06-11].
jn: if I can come up with use cases I will send them out...
<MichaelC> trackbot, associate action-1647 with issue-712
<trackbot> action-1647 (Document use cases for readonly checkboxes) associated with issue-712.
<MichaelC> action-1647 due 3 weeks
<trackbot> Set action-1647 Document use cases for readonly checkboxes due date to 2015-06-25.
<MichaelC> issue-712: http://www.w3.org/2015/06/04-aria-minutes.html
<trackbot> Notes added to issue-712 Should checkbox role support aria-readonly?.
This is scribe.perl Revision: 1.140 of Date: 2014-11-06 18:16:30 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Succeeded: s/Joniemarie/Joanmarie/ WARNING: Bad s/// command: s/a simple set of roles for tables seem like the soln/Based on conversations with developers, a separate set of roles for simple/static tables as opposed to interactive/grid tables makes most sense./ Succeeded: s/::/:/ Succeeded: s/-nter/-inter/ FAILED: s/-nter/-inter/ Succeeded: s/js/mc/ Found Scribe: fesch Inferring ScribeNick: fesch Present: Fred_Esch Joseph_Scheuhammer Léonie_Watson Michael_Cooper Joanmarie_Diggs Matt_King James_Nurthen Bryan_Garaventa Cynthia_Shelly Janina_Sajka Regrets: Rich_Schwerdtfeger Stefan_Schnabel Agenda: https://lists.w3.org/Archives/Public/public-pfwg/2015Jun/0005.html Found Date: 04 Jun 2015 Guessing minutes URL: http://www.w3.org/2015/06/04-aria-minutes.html People with action items: nurthen WARNING: Input appears to use implicit continuation lines. You may need the "-implicitContinuations" option.[End of scribe.perl diagnostic output]
