See also: IRC log
<trackbot> Date: 22 October 2013
<Kathy> scribe: Kathy
<AWK> https://www.w3.org/2002/09/wbs/35422/20131008comments/
<AWK> https://www.w3.org/2002/09/wbs/35422/20131008comments/
<AWK> https://www.w3.org/2002/09/wbs/35422/20131008comments/results
Kathy: Josh was going to put together an example for this one
AWK: what is the update on this
David: there is a few things in the testing. This is about scope on the th. Testing on two tables one with scope and the other without
<David> http://davidmacd.com/test/sailesh-table-test.html
David: Test results for complex
table and simple table. Same behavior in JAWS; any th to the
left of the column header is read as a row header
... the issues is localized to JAWS. So is this a bug in JAWS
or expected behavior?
<Zakim> AWK, you wanted to ask what is read when querying headings for a single cell, not using table navigation commands?
AWK: how are you testing? Are you using the table navigation keys?
David: Yes, using table navigation keys.
<AWK> RESOLUTION: leave open
AWK: the scope issue is #9 in the survey
David: scope fixes the issues in JAWS. Do we consider this as a bug in JAWS or is it expected behavior
AWK: If you move the virtual focus, in JAWS you can read the table headers for a cell. If you don't use the scope does this read?
David: Yes, it reads in td but
not in th
... JAWS does read in a different voice the header cells
... It does this because it is looking to the left and above
for headers. Since there is no scope, it thinks it is a row
heading
AWK: Does the read marked heading change the way that it reads?
David: that does not change the behavior
AWK: Bruce what advice do you give?
Bruce: It depends on the complexity of the table. For simple tables, don't need scope since it is implied.
<Joshue> have we a URI for the table?
<Joshue> +q
Bruce: This example has two levels of headers so you need scope
<AWK> http://davidmacd.com/test/sailesh-table-test.html
http://davidmacd.com/test/sailesh-table-test.html
David: do you really need headers and id for the tables ... but this is another conversation. For simple tables the JAWS has the same behavior
AWK: At the bottom of the page, there are two small examples. David's test results show that the JAWS reads all heading to the left of the header
David: I have seen this behavior and so has Sailesh
Josh: The conversation last week was about whether scope should be added. When you have table group headings, Sailesh was using H3 headings. The question is whether to add scope to the example
David: Plays JAWS example
... it is a good thing to do (adding scope) but this is
specific to JAWS
Bruce: one area where you need scope is on row 1, column 1. Maybe this is causing the buggy behavior
<David> http://davidmacd.com/test/sailesh-table-test.html
Josh: does it change if you just add column headers does that change the results
AWK: There is more AT that we need more information
David: tested with JAWS, NVDA,
VoiceOver in IE, FF and Safari
... Also looked at it with aViewer
<Joshue> +q
David: the screen readers need to
make their own decision regarding headers
... we should test with no row headers
AWK: JAWS is making the best
guess based on the tables that are most common. I am not sure
JAWS would count this as a bug. If you want the table to read
correctly you need to add scope.
... May be better to have scope but it is difficult to get
developers to do this
Josh: Headers/id is more robust but it is harder to code. The scope method is easier. If it is a good thing to do, then we should support this.
David: finished testing, it has the same behavior without the row headers
AWK: Josh tell me what you said
Josh: It is part of good development practice. Sounds like it should be suggesting scope
Katie: don't have a problem supporting it. The author should have options though
<Joshue> +q
<AWK> We do say "Note 1: For simple tables that have the headers in the first row or column then it is sufficient to simply use the TH elements without scope." in h63 at present.
Katie: continue support headers/id
AWK: The crux of the questions, should we allow table without scope
Katie: yes, it should be ok
... it does read the header... it just reads more
information
David: we need to encourage people to use scope since it is leading screen reader and it is not just a bug. In our examples we should have scope but we should not have failure if they don't have a scope
Josh: I agree with David. This came up because Sailesh has example where this does not work. Scope is an easier method to do this; we should support this because it increases user experience.
David: This goes back to different versions in JAWS. Headers/id and scope are two different issues. The heuristics on the the table are pretty good so don't always need headers/id
<Joshue> no
AWK: #9 is just talking about scope always being required. Hearing from a few of you that scope should not always should be required
<Alan_Smith> Team, sorry I had to take another call and missed the conversation up to now and need to drop the call.
AWK: we should demonstrate good practices. Should we fail if there is no scope
Bruce: only for the corner cell
<Joshue> h
AWK: H63 says they can use just th for simple table. Beyond that something is required
<Joshue> http://www.w3.org/TR/WCAG20-TECHS/H63.html
<Joshue> +q
David: we should tell people that we recommend it strongly to use scope
<Alan_Smith> Thank you. Regrets that I need to drop. Tables is a challenge for many.
<Joshue> -q
AWK: modify note 1
<AWK> current note 1: Note 1: For simple tables that have the headers in the first row or column then it is sufficient to simply use the TH elements without scope.
<Alan_Smith> We are putting in scope in our's.
<Alan_Smith> I'll check emails for updates. Thank you.
David: say that is recommended to add scope
Josh: will suggest text
AWK: is there anyone that wants
to speak against not changing h63 except for adding language to
encourage scope
... Ok, so we will suggest language
<Joshue> For simple tables that have table headers in the first row or column then it is sufficient to use only the TH elements. Otherwise, it is advisable to use the scope attribute to provide information to a screen reader about col and row relationships to these headers.
<AWK> Note 1: For simple tables that have the headers in the first row or column then it is recommended to use scope but sufficient to simply use the TH elements without scope.
<David> switch without scope. to use the TH elements alone
<AWK> Note 1: For simple tables that have the headers in the first row or column then it is recommended to use scope but sufficient to simply use the TH elements.
<Zakim> bbailey, you wanted to say simple tables that ONLY have the headers in the first row or column
<Joshue> + 1 For simple tables that have the headers in the first row or column then it is recommended to use scope but sufficient to simply use the TH elements
<AWK> Note 1: For simple tables that only have the headers in the first row or column then it is recommended to use scope but sufficient to simply use the TH elements.
David: the example has the scope in the second column. Add a new example of a simple table that shows good form
AWK: May not be for this one
David: This example would have scope, but would be in the first column instead of the second column for the row headers
Josh: what is wrong with example 1
David: The row headers are in the second column not in the far left column
AWK: this does not meet what we are saying in note 1
David: I can modify the example to meet that
AWK: let's get back to the issue Sailesh raised. We got the general action. Who would like to write up the official response? I will do it
<AWK> ACTION on AWK to add h63 issue to tracker and add response, based on working work disagreeing that scope is always required.
<trackbot> Error finding 'on'. You can review and register nicknames at <http://www.w3.org/WAI/GL/track/users>.
<MichaelC> ACTION: AWK to add h63 issue to tracker and add response, based on working work disagreeing that scope is always required. [recorded in http://www.w3.org/2013/10/22-wai-wcag-minutes.html#action01]
<trackbot> Created ACTION-226 - Add h63 issue to tracker and add response, based on working work disagreeing that scope is always required. [on Andrew Kirkpatrick - due 2013-10-29].
RESOLUTION: return for future work; see action
<Peter_Thiessen> noted :)
<Joshue> http://davidmacd.com/test/sailesh-table-test.html
<David> http://davidmacd.com/test/sailesh-table-test.html
AWK: you can look at the table
example from Josh; the question - is it acceptable to have the
sub-headings that represent a series of cells below them?
Sailesh reports it works well with JAWS. Loretta thinks it is
stretch
... Does anyone think that we should suggest this?
David: it does work in JAWS
Josh: it does not provide the information to let screen reader users to know they are in a data table
<Zakim> MichaelC, you wanted to say there´s a difference between what works and what the WG might want to ¨bless¨
Michael: regardless of whether or not it works, we don't need to add it. Can encourage others to add it as a technique
David: it does validate
AWK: It does not validate under HTML5
<Joshue> http://dev.w3.org/html5/spec-preview/headings-and-sections.html
<Joshue> “The first element of heading content in an element of sectioning content represents the heading for that section.”
<Joshue> “Certain elements are said to be sectioning roots, including blockquote and td elements. These elements can have their own outlines, but the sections and headings inside these elements do not contribute to the outlines of their ancestors.”
<Joshue> + 1 to AWK, this method doesn't indicate context for the user or relationships between a document heading with a table and its surrounding content.
AWK: my concern is in line with
Josh. The headings within the table will not be conveyed as a
header when in a data cell. It is not strong enough
... we should respond as Michael suggested
<bbailey> +1 to not blessing as technique
RESOLUTION: do not accept
Josh: we would like to get
feedback on how we can improve it
... the comment often has raw code and it is hard to associate
back to the technique
Bruce: until someone puts the example on a live page, it is hard to discuss. Can we have HTML code rendered as part of the submission?
David: talk about moving to Github
Michael: can do poll request after people have changed the technique. There may be issues with quality
Josh: not sure what the features are on Github. Is there a big win to move to this? What does it above what we can do today?
<Joshue> Can anyone speak to that, what are the benefits of Github that will help us to improve the submission and commenting process?
Michael: Github is a platform that many are familiar with. If someone copies the HTML and makes edit, it tracks the changes and allows you to accept or reject them
<kathleen> what is the url of the submission form?
Michael: could create a template on Github but may not address the quality issue. May need public training, helping people with submissions or editing them ourselves
<Peter_Thiessen> +1 github
<David> +1
<Peter_Thiessen> (use it daily)
AWK: Who knows Github>
... using it, increases your familiarity with it. There is a
bit of a learning curve
Peter: Very familiar with Github
Josh: If you are familiar with it, do you think Github would be a win for moving to it
<MichaelC> Note we set up space for WCAG on Github, but haven´t done anything with it: https://github.com/w3c/wcag
Peter: good for version control; users can comment on versions and branching to add sub-work. The history is tracked. The only resolution is that it does have a steep learning curve and is mainly used for developers
AWK: it is worth some
investigation. One questions - when you do something in Github
and change the html, do you need to take it down?
... can you render it in the browser?
Peter: you need to take it down and make the edits and upload it.
<David> See https://github.com/chaals
AWK: you can view the source but then you need to check it out and then launch it from the browser
<David> https://github.com/chaals/longdesc-tests
David: The example shows the examples and information for longdesc discussion that was going on
<Joshue> s// \
Peter: you can add structured content
David: the links goes to pages that are on Github
<bbailey> here are wet-boew examples: http://wet-boew.github.io/wet-boew/demos/index-en.html
Bruce: here are some live examples of Github
<AWK> Scribe: AWKl
<AWK> JO: anything that people think would improve submission and comment process?
<AWK> JO: ie. actual improvements?
<Peter_Thiessen> ps. you can point to resources from github and for example host pages using (http://pages.github.com/) but as far as within a repository the view is only RAW (Source)
<AWK> AWK: we are _highly_ receptive to suggestions for improvements in this area. Please keep that in mind and let the chairs know if you have any ideas for greater efficiency.
<AWK> David: thinks use of GitHub will help
<Peter_Thiessen> +1 Github is really useful (we could keep it simple)
<AWK> David: may be able to help with some training and repository setup
<AWK> AWK: we will talk about at editor's meeting tomorrow, follow up with WG about trying it out.
This is scribe.perl Revision: 1.138 of Date: 2013-04-25 13:59:11 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Succeeded: s/JOSH Note: Does this mean that a heading used within a data table would NOT be able to relate to any following <td> cells? Looks like it// Succeeded: s/rssagent, draft minutes// FAILED: s/rssagent, draft minutes/ \/ Succeeded: s/rssagent, draft minutes// Found Scribe: Kathy Inferring ScribeNick: Kathy Found Scribe: AWKl Scribes: Kathy, AWKl Default Present: David_MacDonald, Kathleen_Anderson, Kathy_Wahlbin, AWK, Michael_Cooper, Bruce_Bailey, Peter_Thiessen, Marc_Johlic, Alan_Smith, Katie_Haritos-Shea, Joshue_O_Connor Present: David_MacDonald Kathleen_Anderson Kathy_Wahlbin AWK Michael_Cooper Bruce_Bailey Peter_Thiessen Marc_Johlic Alan_Smith Katie_Haritos-Shea Joshue_O_Connor WARNING: Replacing previous Regrets list. (Old list: Korn) Use 'Regrets+ ... ' if you meant to add people without replacing the list, such as: <dbooth> Regrets+ Kerstin_Probiesch, James_Nurthen, Peter_Korn Regrets: Kerstin_Probiesch James_Nurthen Peter_Korn Agenda: http://lists.w3.org/Archives/Public/w3c-wai-gl/2013OctDec/0052.html Found Date: 22 Oct 2013 Guessing minutes URL: http://www.w3.org/2013/10/22-wai-wcag-minutes.html People with action items: awk[End of scribe.perl diagnostic output]