IRC log of wai-wcag on 2004-06-02

Timestamps are in UTC.

13:56:42 [RRSAgent]
RRSAgent has joined #wai-wcag
13:56:54 [sh1m]
RRSAgent, make visible world
13:56:54 [sh1m]
I'm logging. I don't understand 'make visible world', sh1m. Try /msg RRSAgent help
13:57:17 [sh1m]
Zakim, this will be WAI_WCAG
13:57:17 [Zakim]
ok, sh1m, I see WAI_WCAG(techniques)10:00AM already started
13:57:34 [MichaelC]
rrsagent, make log world-visible
13:57:49 [sh1m]
Thats the one. :)
13:58:01 [MichaelC]
13:58:04 [sh1m]
Zakim, who's on the phone?
13:58:04 [Zakim]
On the phone I see Jim_Thatcher
13:58:09 [bcaldwell]
bcaldwell has joined #wai-wcag
13:58:11 [wendy]
wendy has joined #wai-wcag
13:58:32 [Zakim]
13:59:08 [Zakim]
13:59:25 [wendy]
zakim, who's on the phone?
13:59:25 [Zakim]
On the phone I see Jim_Thatcher, Michael_Cooper, Wendy
13:59:26 [Zakim]
13:59:33 [Zakim]
13:59:34 [wendy]
zakim, ??P6 is Tom
13:59:34 [Zakim]
+Tom; got it
13:59:46 [wendy]
zakim, [Microsoft] is Jenae
13:59:46 [Zakim]
+Jenae; got it
13:59:59 [Zakim]
14:00:10 [wendy]
zakim, ??P10 is Ben
14:00:10 [Zakim]
+Ben; got it
14:01:48 [Zakim]
14:01:58 [wendy]
regrets: chris ridpath, john slatin, don evans
14:02:06 [JimT]
JimT has left #wai-wcag
14:02:08 [wendy]
zakim, ??P12 is David_MacDonald
14:02:08 [Zakim]
+David_MacDonald; got it
14:02:20 [JimT]
JimT has joined #wai-wcag
14:03:02 [Zakim]
14:03:11 [wendy]
regrets: roberto scano
14:04:07 [wendy]
14:04:53 [Zakim]
14:05:05 [wendy]
zakim, ??P17 is Lisa
14:05:05 [Zakim]
+Lisa; got it
14:05:42 [wendy]
chris' mapping:
14:05:56 [Becky]
Becky has joined #wai-wcag
14:06:31 [wendy]
alt for area is not mentioned wrt 1.1.
14:06:33 [sh1m]
Zakim, I am Tom
14:06:33 [Zakim]
ok, sh1m, I now associate you with Tom
14:06:34 [Zakim]
14:07:37 [wendy]
longdesc for iframe as level 2 vs longdesc for object as level 1?
14:08:01 [wendy]
chris is deliberately trying to provoke discussion by assigning levels
14:08:32 [wendy]
longdesc of image as p3?
14:08:45 [wendy]
[many comments about adjusting priorities]
14:09:17 [wendy]
techs have to map to a success criterion and thus inherit the priority.
14:09:32 [wendy]
he wanted to see if the priorities make sense for the techniques.
14:09:54 [wendy]
(using "priority" as refer to levels)
14:10:30 [wendy]
perhaps assigning some to level 2if they effect presentation.
14:11:00 [wendy]
primary focus was likely impact on accessibility to assign levels. but we have to go per the definition of success criteria.
14:11:22 [wendy]
d-links, since deprecated, definitely fall into level 3?
14:11:31 [wendy]
should deprecation effect level assignment?
14:11:42 [wendy]
could be level 1 and deprecated?
14:11:53 [wendy]
however, effects presentation, thus map to level 3.
14:11:58 [wendy]
however - there is not a level 3 criterion.
14:12:16 [wendy]
questions of using longdesc instead of d-link?
14:12:22 [wendy]
14:13:32 [wendy]
we're providing transition from 1.0 to 2.0 techniques. want to ack d-links (since disucssed in 1.0) and want to give guidance about moving to longdesc.
14:13:52 [wendy]
however, even if use deprecated technique can still conform.
14:15:02 [wendy]
when there are not success criterion should we expect that box to be empty in the table?
14:15:31 [wendy]
the box should be empty, but in some cases, chris found that problematic.
14:16:00 [wendy]
he's created 3 levels of the guidelines w/out considering the existing success criteria
14:16:10 [wendy]
summaries for tables as level 3 - is that appropriate?
14:18:20 [wendy]
3.1 level 3 techniques are not relevant.
14:19:17 [wendy]
since chris is gone for two weeks, should someone else take action item to repropose?
14:20:04 [wendy]
this helps us determine where we are missing techniques. maps to success criteria.
14:20:16 [wendy]
traffic cop has mapping
14:20:48 [wendy]
ben writing xslt that will generate traffic cop automatically
14:20:51 [wendy]
14:22:47 [wendy]
action: michael propose update to that ties to specific success criteria. expect to have done by 4 June 2004 for discussion at next week's telecon.
14:30:24 [wendy]
14:30:58 [wendy]
ben describes xslt created to generate "big picture" of guidelines/techniques. similar to traffic cop.
14:36:12 [wendy]
5 columns: level, success criteria, gateway, html, css
14:36:32 [wendy]
a step towards the traffic cop
14:36:59 [wendy]
trying to generate complex table layout (row/colspan)
14:37:26 [wendy]
via xslt was bit of bear (col/rowspan ala david's traffic cop proposal)
14:37:33 [wendy]
with tweaking could become the traffic cop
14:37:57 [wendy]
tweaking table layout is not high priority - top priority it to get all the info into this table
14:38:19 [wendy]
chris' mappings would magically appear (once have all that data in the xml)
14:38:28 [wendy]
ready to do this mapping in gateway techniques?
14:39:02 [wendy]
1/3 of the criteria don't have html techniques - because they are gateway techniques?
14:39:08 [wendy]
don't need html-specific?
14:40:39 [wendy]
(in some cases, don't need an html-specific technique if have gateway)
14:42:06 [wendy]
renaming gateway? common, general,
14:42:30 [wendy]
common techniques, general techniques, core techniques, technology-independent (no-no b/c of acronym)
14:43:00 [wendy]
traffic cop becomes "gateway"
14:44:59 [wendy]
technology-independent examples and techniques
14:45:36 [wendy]
would then rename others? e.g., html examples and techniques for wcag 2.0
14:46:15 [wendy]
based on usability study, "techniques" didn't have any "scent" people were looking for "examples"
14:46:31 [wendy]
cross-technology techniques
14:46:58 [wendy]
does that mean multiple technologies?
14:47:09 [wendy]
"General" good term b/c says how to accomplish generally.
14:47:21 [wendy]
to do something, will need to use some technology.
14:47:27 [wendy]
they are the principles that underly the techniques
14:48:02 [wendy]
Techniques Principles and Examples
14:48:57 [bcaldwell]
history on naming -
14:49:39 [wendy]
move forward with "general" and "gateway" - let ferment for a while. this was good brainstorming session, but let's move forward for now. make decision before publish drafts in july.
14:51:14 [wendy]
14:51:23 [wendy]
14:51:45 [wendy]
complex table
14:53:02 [wendy]
layered tables
14:53:14 [wendy]
"irregular" tables
14:54:59 [wendy]
table headers instead of putting inline
14:55:28 [wendy]
14:56:55 [sh1m]
Zakim, who's on the phone?
14:56:55 [Zakim]
On the phone I see Jim_Thatcher, Michael_Cooper, Wendy, Tom, Jenae, Ben, David_MacDonald, Sailesh_Panchang, Lisa, Becky_Gibson
14:57:30 [wendy]
classes of table for which to require different aspects of markup.
14:57:35 [wendy]
e.g., if complex, require a summary.
14:58:01 [wendy]
agreed upon classes of tables and what is required for them.
14:58:16 [wendy]
seem to have agreement (with minor objections) that all data tables should have th element.
14:58:22 [wendy]
complex tables should have additional markup for headers.
14:58:31 [wendy]
that's when decided need for defn of complex tables.
14:58:41 [wendy]
think we decided that complex tables should have a caption
14:58:51 [wendy]
14:59:18 [wendy]
no agreement about caption? required on all data tables? only complex data tables? optional?
15:00:07 [wendy]
irregular (headings are not in row or col), layered (more than one row and/or col of headings), simple (one of each)
15:00:13 [wendy]
a table could be irregular and layered.
15:01:39 [wendy]
in large part, these are working definitions for us.
15:01:46 [wendy]
if we don't define, will be difficult to write test cases for them.
15:03:09 [wendy]
defns at: but replace "complex" with "irregular"
15:03:23 [wendy]
irregular that uses rowspan? just irregular
15:03:42 [wendy]
if it meets of defn of irregular, then irregular regardless of use of spanning
15:04:02 [wendy]
layout talbes are not part of this discussion.
15:04:33 [wendy]
2 types of tables 1. data (with 3 classes: irregular, layered, simple) 2. layout
15:05:11 [wendy]
think have agreement w/th. not agreement: extended header markup, summary and caption
15:05:17 [JimT]
Irregular tables have cells for which header cells are not in the same row
15:05:17 [JimT]
or same column as the cell.
15:05:17 [wendy]
those issues dependent on these definitions
15:05:45 [wendy]
in html spec, says, "if cell is both header and data it should be marked up with td"
15:05:52 [JimT]
Layered tables have more than one column of row headers and/or more than one row of column headers and header cells are in the same row and/or column as the data cell
15:06:12 [JimT]
Simple tables have at most one row and at most one column of headings and
15:06:12 [JimT]
header cells are in the same row and/or column as the data cell
15:07:44 [MichaelC]
Possible TH rule: simple data tables must have TH
15:08:01 [MichaelC]
layered and irregular must have TH and/or extended header markup
15:08:38 [MichaelC]
irregular must use extended header markup
15:09:15 [wendy]
<!-- TH is for headers, TD for data, but for cells acting as both use TD -->
15:09:28 [wendy]
15:10:40 [wendy]
when simple, use scope
15:10:58 [wendy]
we'll need a technique for each of them
15:11:03 [wendy]
we'll log user agent issues with each
15:11:13 [wendy]
have tech for scope, but document the UA issues
15:12:11 [wendy]
col headings use th, row headings use td w/scope=row
15:13:19 [wendy]
15:15:37 [wendy]
can a cell be irregular? if header info not in the same col or row?
15:15:47 [wendy]
thus, require that markup only on that cell?
15:16:17 [wendy]
then defn of irregular table is one that contains irregular cells
15:18:51 [wendy]
[scribe misses discussion due to reading xhtml 2.0 spec -]
15:19:32 [wendy]
layered: no req to use extended header markup, but not forbidding it. if use it, don't need th.
15:19:42 [sh1m]
15:19:47 [wendy]
say same for simple?
15:20:05 [wendy]
to support current AT you should... [insert jim's comment here]
15:20:25 [wendy]
we're requiring some semantic to identify header info
15:20:42 [wendy]
if th element is present, here is the algorithm for which it is a header.
15:20:52 [JimT]
To support current AT, you should use headers/id markup on layered tables
15:20:54 [wendy]
if that algorithm works in your table, if not use extended header markup
15:20:59 [wendy]
(thanks jim)
15:21:18 [wendy]
rule is, "headers must be identified" but how you do that depends on your table
15:22:01 [wendy]
that (headers must be identified) is the success criterion. these are all techniques that support it.
15:25:59 [wendy]
take issue to PF? - if cell is data and header, make td (is only a comment in html 4.01 spec). since dtd not compmlete for xhtml 2.0 - not there yet.
15:26:08 [wendy]
suggest that if a cell is both th and td, should be th?
15:26:22 [wendy]
if style is an issue, use style.
15:27:58 [wendy]
encourage html wg to make a clearer distinction about layout tables.
15:28:07 [wendy]
also, ask for defn of "simple"
15:29:35 [wendy]
we could advise pfwg on our use of terms and definitions and use that in response to html wg about xhtml 2.0.
15:30:15 [wendy]
somewhere are we saying that a table can only be one of the 3 classes?
15:30:20 [Zakim]
15:31:00 [wendy]
action: wendy summarize what we would like to take to pfwg as advice for html wg.
15:31:30 [JimT]
Phone went out!
15:32:57 [MichaelC]
That sucks, we're talking about summary now
15:33:44 [JimT]
"The conference is restricted"
15:33:50 [wendy]
15:34:05 [wendy]
ben or tom - could you take minutes for a couple minutes?
15:34:54 [sh1m]
15:35:19 [JimT]
Reason - schedule for 90 minutes
15:35:20 [MichaelC]
summary to describe how to navigate table, caption to identify purpose of table?
15:35:26 [Zakim]
15:35:33 [wendy]
jim are you here now?
15:35:44 [JimT]
15:36:15 [wendy]
great. yes, the issue seems to be that we are "over". (i'm off to send an email to our phone admin to formally extend time from now on)
15:37:10 [sh1m]
Summary was not designed to be a visible interface, it is designed to describe the design of the table for people using AT or audio interfaces
15:39:39 [sh1m]
Caption provides a title for the table
15:39:50 [sh1m]
There should always be a caption?
15:40:48 [sh1m]
If there is a title it should be implemented as a captioin
15:40:53 [sh1m]
15:42:12 [wendy]
we need to poll screen reader users about use of summary.
15:42:22 [sh1m]
Need to talk to screen reader users so that we properly understand before we talk to PF/HTML WG
15:42:26 [wendy]
summary might say, "rainfall decreased" - like longdesc of a picture.
15:42:48 [sh1m]
hehe ok
15:43:09 [sh1m]
summary should describe and should be required on iregular and layered tables
15:43:18 [sh1m]
describe layout
15:44:24 [sh1m]
shouldnt misuse summary, should leave it according to spec even though there may be other needs for a 'summary' of the table
15:44:35 [sh1m]
a summary of the content
15:45:42 [MichaelC]
15:46:34 [sh1m]
never seen a summary used to do that
15:50:16 [Zakim]
15:52:02 [wendy]
need common terminology?
15:52:18 [wendy]
examples of how summaries are written
15:52:52 [wendy]
agree that summary required on x types of tables, but open space about how describe what should go into summary?
15:53:15 [wendy]
expect people will continue to put unuseful in summary
15:53:39 [wendy]
think we're leaning towards: meaningful summary required on irregular and layered
15:53:43 [wendy]
optional on simple data tables?
15:53:49 [wendy]
later, determine what the content should be?
15:54:13 [wendy]
steer us away from someone using, "this is a layout table" since should not be used on layout table
15:54:44 [wendy]
action: tom and sailesh write techniques for writing effective data table summaries
15:55:04 [JimT]
Real life Example: summary="The rows represent non-stop, 1 stop, and 2+ stops (a single row is provided for Multi-Destination itineraries) itineraries. The columns represent your choice of airlines. Clicking on the price shown in each cell will redisplay this page with only flights matching the airline and number of flights selected. Clicking on a stop category or airline will display all flights within that category."
15:55:11 [wendy]
summary: absent or null on data. may be used on simple. must be used on irregular and layered.
15:55:21 [wendy]
15:55:33 [wendy]
if table is title, the title should be implemented via caption
15:55:51 [wendy]
talk to pf that should have something else to deal with summaries?
15:56:32 [wendy]
summary should be limited to purpose and layout
15:57:03 [wendy]
lisa's question about summary that is more than title, but interpret content of table
15:57:12 [wendy]
perhaps another tag?
15:59:12 [wendy]
purpose: does that mean a summary (an interpretation of the data, the conclusion)...
15:59:17 [wendy]
however, purpose gets in way of layout
16:01:38 [wendy]
summary: is it for summary of data or of layout? if author marks up table correctly, shoiuldn't asst. tech be able to generate layout summary?
16:01:57 [wendy]
the asst. tech summary would be very general. perhaps not very useful.
16:02:21 [wendy]
however, it could read the header information as associated via markup to be less general.
16:03:01 [wendy]
jim's example - asst. tech not know "delta" is an "airline"...unless use axis attribute.
16:04:37 [wendy]
today's tech: they are not reliably or consistently implementing markup. generating summaries is a long ways off.
16:04:43 [Zakim]
16:04:59 [wendy]
getting author's to implement that informatio is a long ways off.
16:05:10 [wendy]
if it's required by wcag 2.0, will get more.
16:05:24 [wendy]
if humans can do it, perhaps then asst. tech more likely to do automatically.
16:05:48 [wendy]
put issue on hold. if sailesh and tom come up with techs, if looks like easy for UA, then do that.
16:06:09 [wendy]
in addition, come up with something concrete, may seem more reasonable for humans to do.
16:06:46 [wendy]
16:06:59 [wendy]
next week: pick up on summary discussion
16:07:03 [wendy]
also revised mapping
16:07:18 [wendy]
and css techs
16:07:23 [wendy]
action wendy: send review of css techniques
16:07:48 [Zakim]
16:07:49 [Zakim]
16:07:49 [Zakim]
16:07:50 [Zakim]
16:07:50 [Zakim]
16:07:51 [Zakim]
16:07:52 [Zakim]
16:08:02 [Zakim]
16:08:03 [wendy]
RRSAgent, make log world-visible
16:08:03 [Zakim]
WAI_WCAG(techniques)10:00AM has ended
16:08:04 [Zakim]
Attendees were Jim_Thatcher, Michael_Cooper, Wendy, Tom, Jenae, Ben, David_MacDonald, Sailesh_Panchang, Lisa, Becky_Gibson
16:09:17 [sh1m]
im off home
16:09:18 [sh1m]
16:09:20 [sh1m]
back soon
16:10:28 [wendy]
zakim, bye
16:10:28 [Zakim]
Zakim has left #wai-wcag
16:10:31 [wendy]
RRSAgent, bye
16:10:31 [RRSAgent]
I see 4 open action items:
16:10:31 [RRSAgent]
ACTION: michael propose update to that ties to specific success criteria. expect to have done by 4 June 2004 for discussion at next week's telecon. [1]
16:10:31 [RRSAgent]
recorded in
16:10:31 [RRSAgent]
ACTION: wendy summarize what we would like to take to pfwg as advice for html wg. [2]
16:10:31 [RRSAgent]
recorded in
16:10:31 [RRSAgent]
ACTION: tom and sailesh write techniques for writing effective data table summaries [3]
16:10:31 [RRSAgent]
recorded in
16:10:31 [RRSAgent]
ACTION: wendy to send review of css techniques [4]
16:10:31 [RRSAgent]
recorded in