IRC log of wai-wcag on 2004-01-21

Timestamps are in UTC.

00:01:53 [Zakim]
00:02:04 [Zakim]
00:02:05 [Zakim]
00:02:06 [Zakim]
WAI_WCAG(plan)6:00PM has ended
00:02:07 [Zakim]
Attendees were JasonWhite, Ben, Wendy, Gregg_Vanderheiden, JasonWhite
00:09:09 [wendy]
zakim, bye
00:09:09 [Zakim]
Zakim has left #wai-wcag
00:09:12 [wendy]
RRSAgent, bye
00:09:12 [RRSAgent]
I see 2 open action items:
00:09:12 [RRSAgent]
ACTION: wendy call people who aren't responding to email about action items. [1]
00:09:12 [RRSAgent]
recorded in
00:09:12 [RRSAgent]
ACTION: wendy ask judy about funding for jason for csun [2]
00:09:12 [RRSAgent]
recorded in
15:07:47 [RRSAgent]
RRSAgent has joined #wai-wcag
15:08:19 [wendy]
zakim, who's on the phone?
15:08:19 [Zakim]
On the phone I see +1.703.265.aaaa, Michael_Cooper, Dave_MacDonald, Chris_Ridpath, tom, DanBri, Charles (muted), Jonathan, Wendy
15:08:24 [ChrisR]
Here's a demo document that shows what the WCAG HTML test suite may look like:
15:08:25 [ChrisR]
15:08:35 [ben]
zakim, Charles is Ben
15:08:37 [Zakim]
+Ben; got it
15:08:43 [ChrisR]
Note: The URL of the test doc is very temporary.
15:08:45 [wendy]
zakim, +1.703.265.aaaa is Don_Evans
15:08:45 [Zakim]
+Don_Evans; got it
15:10:30 [sh1mmer]
sh1mmer has joined #wai-wcag
15:10:47 [sh1mmer]
hi can I get that url again pls?
15:10:59 [ben]
15:11:07 [sh1mmer]
Zakim, who's on the call
15:11:07 [Zakim]
I don't understand 'who's on the call', sh1mmer
15:11:32 [wendy]
have you looked at using UAAG 1.0 test files?
15:12:16 [sh1mmer]
Zakim, I am tom
15:12:16 [Zakim]
ok, sh1mmer, I now associate you with tom
15:12:52 [wendy]
chris has been doing one test per file. UAAG and HTML tests have more than one test in a file.
15:13:07 [wendy]
UAAG has examples as well as test files. examples help identify the test.
15:13:41 [wendy]
Charles has a student who is doing something similar to Chris, but using RDF instead of XML.
15:13:50 [wendy]
EuroAccessibility building a suite of tests.
15:15:13 [chaalsBRS] - EuroAccessibility's current draft list of tests
15:15:47 [chaalsBRS]
No "latest version URI" :(
15:18:07 [Zakim]
15:18:32 [MichaelC]
ack chaalsBRS
15:18:32 [Zakim]
chaalsBRS, you wanted to say that checks and techniques aren't one to one, and a mapping should be added
15:20:03 [sh1mmer]
15:20:36 [wendy]
are tests designed for automatic tools? e.g., "alt-text no longer than 150 characters"
15:20:43 [wendy]
concern about a test like that (seems ambiguous)
15:21:02 [sh1mmer]
ack tom
15:21:12 [wendy]
we should spell out as much as possible.
15:21:19 [wendy]
spell out if not spelled out in techniques?
15:21:38 [wendy]
if a long title causes an accessibility problem, should spell it out somewhere.
15:21:42 [wendy]
15:22:04 [sh1mmer]
15:22:16 [ben]
15:22:18 [wendy]
ack charles
15:22:18 [Zakim]
charles, you wanted to suggest we should be in the middle of auto-tests and manual tests
15:22:31 [wendy]
tricky for wcag to have tests that are different from techniques.
15:22:43 [wendy]
be in the middle of automatic and non-auto tests.
15:23:32 [wendy]
80 char hard limit (in a success criterion) would be wrong due to language constraints
15:23:54 [wendy]
need to build tests so that "here are tests you can do" - "this test is diagnostic aid rather than clear test"
15:24:02 [wendy]
"this test useful if trying to flag things to look at"
15:24:12 [wendy]
however, not using to say "this alt-text is appropriate or not"
15:24:45 [wendy]
checks have confidence level. e.g., title length = 150 chars is medium. i.e, likely a problem but not necessarily.
15:24:55 [wendy]
so 200 chars could be appropriate for an image.
15:25:16 [wendy]
confidence, not a good measure. it's a short-hand for something that should be more descriptive.
15:26:49 [chaalsBRS]
rather than just confidence, there should be a "what to do with the result" - maybe a consequential test, ask for user decision, etc.
15:26:57 [wendy]
ack wendy
15:28:02 [chaalsBRS]
wendy: Focus on auto-testing is a concern
15:28:38 [chaalsBRS]
... would like to clearly state assumptions for how to meet a technique, why it is there, separate from what auto tools can test.
15:29:23 [sh1mmer]
ack tom
15:29:30 [chaalsBRS]
... Don't want to see a hard limit set in documents just so it is easy to make tools. We also don't want to tie tests to specific technoologies
15:29:31 [ChrisR]
15:29:56 [sh1mmer]
am i on?
15:30:00 [sh1mmer]
i was talking
15:30:01 [wendy]
don't hear you
15:30:03 [sh1mmer]
15:30:05 [wendy]
zakim, who's muted?
15:30:05 [Zakim]
I see Ben muted
15:30:09 [wendy]
zakim, unmute Ben
15:30:09 [Zakim]
Ben should no longer be muted
15:30:35 [Zakim]
15:30:45 [wendy]
rather than defining values, but thinking behind it is important.
15:31:17 [wendy]
we'll ask when tom takes a breathe
15:32:17 [wendy]
zakim, ++P29 is Lisa_Seeman
15:32:18 [Zakim]
sorry, wendy, I do not recognize a party named '++P29'
15:32:36 [wendy]
zakim, ??P29 is Lisa_Seeman
15:32:36 [Zakim]
+Lisa_Seeman; got it
15:32:45 [wendy]
ack David
15:33:09 [wendy]
title element: describes document, not site structure. should be name of the resource.
15:34:35 [wendy]
15:35:14 [wendy]
regardless of length, title of document is title element.
15:35:36 [wendy]
what is the restriction based on?
15:35:53 [wendy]
is it an accessibility issue?
15:35:59 [wendy]
zakim, who's muted?
15:35:59 [Zakim]
I see ben, Lisa_Seeman muted
15:36:30 [Zakim]
15:36:33 [ben]
can you hear me?
15:36:35 [sh1mmer]
unmute ben
15:36:44 [ben]
still talking
15:36:46 [sh1mmer]
Zakim, unmute tom
15:36:46 [Zakim]
tom was not muted, sh1mmer
15:36:46 [wendy]
ack ben
15:36:49 [sh1mmer]
Zakim, unmute ben
15:36:50 [Zakim]
ben was not muted, sh1mmer
15:37:08 [sh1mmer]
Zakim, I am tom
15:37:08 [Zakim]
ok, sh1mmer, I now associate you with tom
15:37:10 [wendy]
in looking at checks, need a better job of differentiating them. confidence is an interesting method.
15:37:32 [wendy]
some checks will be sufficient for meeting a technique. others are good ideas or optional or helpful in automated testing, but need clear lines between them.
15:37:48 [wendy]
while they could all be part of test suite, we should be most concerned with those that are sufficient for techniques.
15:38:06 [wendy]
e.g., technique that says "use title element" minimum test is to determine if it is there.
15:38:13 [wendy]
other checks are optional.
15:38:28 [wendy]
up to the author to decide if want to use in their development scenario
15:38:44 [wendy]
include rationale about why excessively long title might cause issues, but why it is not a problem.
15:38:50 [wendy]
ack ChrisR
15:39:04 [wendy]
these tests are not just automatic tests, some can not be automated (like longdesc).
15:39:39 [wendy]
these are meant to define what the techniques are. if say something in techniques, it needs to be defined.
15:39:46 [wendy]
if we say "short" then it needs to be defined.
15:40:40 [wendy]
q+ to say shouldn't techniques have testable criteria that are not ambiguous? (part of checklists)
15:41:12 [wendy]
this is just a draft, not carved in stone. beginning point for discussion.
15:42:23 [wendy]
should it be part of technique or not? are there other checks that could be included?
15:42:27 [wendy]
ack don
15:43:13 [wendy]
have 100s of pages to test every day. if don't have specific tests, won't do them.
15:43:17 [sh1mmer]
15:43:36 [wendy]
ack Lisa
15:44:50 [wendy]
two kinds of things: some are testable (alt text exists?) others are not (alt text length).
15:45:05 [wendy]
"short" depends on context.
15:45:26 [wendy]
e.g., mission statement in a picture. alt-text should be all of the text.
15:45:58 [wendy]
putting all of that text into the alt attribute doesn't make my site inaccessible (if it is longer than the length specified in some check).
15:48:42 [wendy]
ack wendy
15:48:42 [Zakim]
wendy, you wanted to say shouldn't techniques have testable criteria that are not ambiguous? (part of checklists)
15:52:19 [chaalsBRS]
wendy: think it is important to have auto tests. Also important that a technique contains everything you need to know if you meet the requirement. These might not be the same.
15:52:40 [chaalsBRS]
... intersted in types of tests, and where information needs to go.
15:52:53 [wendy]
let's talk about framework rather than specific tests.
15:53:08 [wendy]
ack charles
15:53:08 [Zakim]
charles, you wanted to say that auto tests are important, but we need a layer of indirection between them and the guidelines requirements
15:53:40 [wendy]
automatic tests are good to have, different from techniques (not always one-to-one mapping). ought to have mapping: this tests belongs to this technique.
15:53:43 [wendy]
ack tom
15:53:43 [sh1mmer]
ack tom
15:54:07 [chaalsBRS]
where a test can falg an issue for several techniques, or where one technique implies several different tests.
15:54:13 [wendy]
a framework which helps specify how the tests should be rather than the actual value is important for us to go for.
15:54:35 [wendy]
ack david
15:54:54 [wendy]
developing techniques based on accessibility issue. guideline, technique, test.
15:55:09 [wendy]
whether address in automatic or not, but if can't, it doesn't make the accessibility issue go away.
15:55:56 [wendy]
should manual checks be identified, and here's technique to address it.
15:56:12 [wendy]
guidelines and techniques should be agnostic (as to whether automated test or not).
15:56:20 [wendy]
manual checks are equally as important.
15:56:48 [wendy]
challenge: independent of a lot of assumptions around testing. we need to test guidelines.
15:57:07 [wendy]
are test documents will or have to be used by evaluation tool or authoring tool developers.
15:57:17 [sh1mmer]
15:57:17 [wendy]
validate techniques and guidelines.
15:57:40 [wendy]
useful to dive deep, but bring back up to ask, "what will our test suite look like?"
15:57:55 [wendy]
ack tom
15:58:09 [wendy]
personas: have been working on some more.
15:58:15 [wendy]
building use cases.
15:58:23 [wendy]
how does that fit into techniques?
15:59:09 [wendy]
ack Lisa
15:59:37 [wendy]
q+ to say let's take one technique and follow it through (from guideline/success criteria, to task to checks)
15:59:51 [wendy]
(perhaps title, since already had so much discussion about today)
16:00:07 [sh1mmer]
16:00:10 [wendy]
(assemble all of the pieces and see how works. think about checklist)
16:00:28 [wendy]
(personas fit in to help us think about different paths and pieces of content that are needed)
16:00:57 [sh1mmer]
ack tom
16:02:13 [sh1mmer]
16:02:22 [chaalsBRS]
ACTION charles: look at the chain from a guideline to a particular test.
16:02:51 [sh1mmer]
ack tom
16:02:53 [wendy]
ack wendy
16:02:53 [Zakim]
wendy, you wanted to say let's take one technique and follow it through (from guideline/success criteria, to task to checks)
16:03:00 [wendy]
group action?
16:03:38 [wendy]
part of agenda for today is to talk about f2f. could also do this exercise at that mtg.
16:03:53 [wendy]
16:04:23 [ChrisR]
16:04:34 [wendy]
charles send something to the list week after next.
16:05:03 [wendy]
diff people take diff techniques/tests to get different takes.
16:05:43 [wendy]
action: david - title element
16:05:51 [wendy]
action: charles alternatives for images
16:05:56 [wendy]
ack ChrisR
16:08:23 [wendy]
action: wendy pick guideline and work way from top to bottom
16:09:37 [wendy]
action: tom offer use cases (we're doing this technique because it helps x, y, z)
16:09:47 [wendy]
continuing discussion next week, primary part of agenda.
16:10:12 [wendy]
review one or more than one of analyses (results of action items)
16:10:44 [wendy]
when reviewing QA documents, wondering about QA process.
16:10:52 [wendy]
get someone from QA to walk us through it?
16:11:31 [wendy]
the QA Process. how are the docs interrelated?
16:11:36 [wendy]
are the test files too detailed?
16:13:17 [wendy]
action: chris ask olivier to review tests and ask questions for.
16:13:24 [wendy]
another thing to tech plen
16:14:24 [wendy]
michael, tom, wendy, charles, lisa, ben (will be at t.p.)
16:14:46 [wendy]
agenda for t.p.:
16:14:52 [wendy]
16:15:32 [wendy]
testing, coord w/euroaccessibility
16:15:49 [wendy]
(cmn student - modeling tests. run tests)
16:16:33 [wendy]
rdf techniques, svg techniques
16:17:46 [wendy]
20-30 minutes of svg time on friday
16:18:10 [wendy]
user agent imp issues, content writing/authoring tool
16:18:32 [wendy]
QAWG to talk about process, testing
16:18:38 [wendy]
html and css techniques
16:18:47 [wendy]
techniques gateway
16:19:17 [wendy]
plus, meeting 2 weeks after at csun
16:19:54 [Zakim]
16:20:12 [wendy]
zakim, ??P4 is Jonathan_Chetwynd
16:20:12 [Zakim]
+Jonathan_Chetwynd; got it
16:20:36 [Zakim]
16:20:40 [ben]
16:20:54 [wendy]
ack ben
16:21:35 [wendy]
talking about the chain, what do things look like start to finish - could we spend some time on a draft that helps us figure out what each piece of the chain looks like. snippets as roadmap
16:22:23 [wendy]
guideline, techniques gateway, checklists, test suite. links to each piece look like they should. get an idea of how it work and fits together.
16:23:03 [wendy]
charles, david, and wendy's action items feed into this process.
16:25:29 [wendy]
action: wendy publish latest draft of RDF Techs from lisa.
16:26:14 [wendy]
rdf techniques perhaps next week, more likely week after
16:26:45 [ChrisR]
ChrisR has left #wai-wcag
16:26:53 [Zakim]
16:26:54 [Zakim]
16:26:55 [Zakim]
16:26:56 [Zakim]
16:26:57 [Zakim]
16:26:58 [Zakim]
16:26:59 [Zakim]
16:27:00 [Zakim]
16:28:48 [ben]
ben has left #wai-wcag
16:30:54 [Zakim]
16:31:03 [Zakim]
16:31:06 [Zakim]
WAI_WCAG(techniques)10:00AM has ended
16:31:07 [Zakim]
Attendees were Michael_Cooper, +1.703.265.aaaa, Dave_MacDonald, Chris_Ridpath, Jonathan, Wendy, Don_Evans, David_D, charles, ben, tom, Lisa_Seeman, Jonathan_Chetwynd
16:37:50 [wendy]
zakim, bye
16:37:50 [Zakim]
Zakim has left #wai-wcag
16:37:53 [wendy]
RRSAgent, bye
16:37:53 [RRSAgent]
I see 7 open action items:
16:37:53 [RRSAgent]
ACTION: charles to look at the chain from a guideline to a particular test. [1]
16:37:53 [RRSAgent]
recorded in
16:37:53 [RRSAgent]
ACTION: david - title element [2]
16:37:53 [RRSAgent]
recorded in
16:37:53 [RRSAgent]
ACTION: charles alternatives for images [3]
16:37:53 [RRSAgent]
recorded in
16:37:53 [RRSAgent]
ACTION: wendy pick guideline and work way from top to bottom [4]
16:37:53 [RRSAgent]
recorded in
16:37:53 [RRSAgent]
ACTION: tom offer use cases (we're doing this technique because it helps x, y, z) [5]
16:37:53 [RRSAgent]
recorded in
16:37:53 [RRSAgent]
ACTION: chris ask olivier to review tests and ask questions for. [6]
16:37:53 [RRSAgent]
recorded in
16:37:53 [RRSAgent]
ACTION: wendy publish latest draft of RDF Techs from lisa. [7]
16:37:53 [RRSAgent]
recorded in