10 May 2001 WCAG WG minutes

http://www.w3.org/WAI/GL/2001/05/10-minutes.html

10 May 2001 WCAG WG minutes
Summary of action items and resolutions
·       Action KHS: help finish introduction - link to info about 
assumptions and info from EO work and UAAG1.0. Plan to have something to 
the group by next week.
·       Resolved: process for sending open issues to techniques subgroups: 
subgroup gets assigned open issue, up to them to resolve. best to do it in 
context of techniques document. editor of the techniques doc takes the 
action item.
·       Action MM: resolve issue of absolute units for table layout within 
the HTML techniques.
·       Resolved (at least for today): criteria for determining equivalency 
is based on purpose and task. caveat: translations - purpose may be to gain 
full meaning of text. where that is the purpose, simplifying text is not 
equivalent.
·       Action MM: think about this resolution and respond.
·       Action everyone: review PDF techniques.
·       Resolved: PDF techniques on the agenda for next week. KHS says, 
"ignore the numbers and think about which checkpoints they fit under." Some 
feel like they can fit under more than one checkpoint.

Participants
·       Jason White
·       Gregory Rosmaita
·       William Loughborough
·       Katie Haritos-Shea
·       Cynthia Shelly
·       Wendy Chisholm
·       Paul Bohman
·       Jeff Isom
·       Marti McCuller
·       Matt May

Introduction to WCAG 2.0
WL Looked good last time I read it.
JW Last year we noted that assumptions re: user agents, authors, users, and 
assistive technologies should be documented.
KHS Know where to steal it from - stuff that Judy did. Also, nice piece in 
UAAG1.0
WC Point to or summarize? Don't want to duplicate.
Action KHS: help finish introduction - link to info about assumptions and 
info from EO work and UAAG1.0. Plan to have something to the group by next 
week.

Bidirectional languages
WCAG 2.0 Issue #11
WC More than a bidirectional issue, I've heard it said about HTML tables 
for layout in general.
JW Regardless, goes to HTML techniques group, does not go to the higher level.
WC Process then is that subgroup takes action, sends proposal to WG once 
decided.
JW Yes, subgroup gets assigned open issue, up to them to resolve. best to 
do it in context of techniques document.
Resolved: process for sending open issues to techniques subgroups: subgroup 
gets assigned open issue, up to them to resolve. best to do it in context 
of techniques document. editor of the techniques doc takes the action item.
Action MM: resolve issue of absolute units for table layout within the HTML 
techniques.

Server-side issues
WL It could be a glossary rathole. Equivalents, even at fundamental level, 
has spawned huge thread about alt-text and what does equivalent mean for that.
JW We need to make sure it doesn't go on forever.
WL Depends on what equivalent is equivalent to.
CS Something along the lines of: you should be able to accomplish the same 
task in about the same amount of time. If the task is, "read content" then 
using a different version you should be able to get the content. If 
downloading a pen, or buying a pen...you should be able to do these things 
in any of the versions.
PB I have developed a system where you take database content and put into 
different templates. You're talking about the content itself, which is 
different. e.g. you have a graphical version and one that replaces graphics 
with text. The content doesn't change. YOu are talking about changing the 
content.
CS Talking about both.
JW Continuim between changing structure, content, and just presentation. If 
you remove all audio files and does not show that were there, then 
disagreement.
CS Len mentioned idea of "essential purpose." Can you accomplish what the 
site is intended to accomplish. A simplified version of text might do in 
some cases, not in others.
WC Where is UAAG at on this?
GR Idea of "conditional content." Incorporates switch, alt, title, 
longdesc. Wrestling with how do you alert user that there is conditional 
content and does it need to be synchronized. Want to step through 
conditional content. Can configure order to step through or to step through 
at all. UAAG does not address "what can you do with what is served to you."
WC Assumes no processing on server-side and that all info passed to the client.
GR Since UAAG 1.0 does not deal with server-side. Encourage you to raise it 
as an issue with UA WG. Probably end up on open issues for UAAG 2.0.
JW Some people who would say that they have the right to know that the 
conditional content is there. others don't care.
CS Others confused or harmed by it in some way.
JW Certainly, a divide of opinion.
CS Fair amount of work in limiting choices to make interface easier to use.
JW We need to decide what are the criteria for determining if content is 
equivalent. e.g. provide text equivalents in every version or providing 
only in one.
CS There will be cases where that has to happen. e.g., the wap or any thin 
client case where you optimize for a small device. you have to leave some 
things out. There is an argument to be made for optimized versions. My 
opinion - as long as you can switch and that the version w/alternatives is 
equivalent (a diff. conversation).
PB Common that the main version should have WCAG 1.0 A conformance.
CS Assuming that there is a main version.
GR Just b/c serve something, assume 99% use all senses together, no reason 
not to include accessibility features that are built into the language. 
Can't assume that someone can access a proxy. If serving images have to 
serve conditional context.
CS Unless the user says that they don't want alt-text.
GR Should be done through something like CC/PP.
CS Or some cookie-based mechanism or whatever. Could definitely be abused, 
but don't make it against the rules if it can be done well.
WC We need to document those assumptions. Gets back to baseline discussion 
that is not resolved.
JW To be resolved at the technology level.
CS Many of these things will be covered in the server-side techniques.
JW When do two versions count as equivalents? If there is a clear task to 
be performed with the content, then one can use that as the criteria, ala 
Cynthia's proposal. It can get complicated if you want you want to use it 
for is different from what the creator intended. Or if there is not a clear 
purpose. Then, perhaps a rewrite is necessary. Also raised in relation to 
cognitive discussions a few weeks ago.
PB Where in the guidelines do we reference server-side?
CS There will be server-side techniques. I believe there should be some 
normative aspects in this document.
PB There are things that shouldn't apply to every page.
JW Statement of conformance requirements.
WC Also hooks to server-side techniques from various checkpoints. What 
about making some of the checkpoint solutions normative, as cynthia suggested?
JW We'll figure that out when we get there.
CS It does sound like a conformance issue. Only when testing things does it 
become a concern. Not a guideline.
WC Could someone take an action to document so that we can refer to this 
discussion later when we get to the point when we are focusing on conformance.
WL How do we currently define content?
CS Include tags or not.
JW Refer to meaning rather than structure or presentation.
KHS No definition in WCAG 2.0, UAAG "3 ways. ..." @@link.
JW Clear for UAAG purposes. Even tho used for multiple purposes can make 
contextual distinctions.
WL How far back to we reach to find what the content is that we'll have 
guidelines for. Do we reach to the server? Do we go into the mind of the 
author?
MM Already separate content from presentaiton. Do we tell people to present 
same content regardless of presentation. I have issue with using different 
verbiage for different audiences. As long as you say, "if you can present 
the same words and ideas on 2 different types of presentations, then we can 
consider that accessible."
CS Can present same idea with different words.
MM Yes, but 2 audiences who are drawing different conclusions. When you 
separate the idea from the words and communicate in different ways, people 
will absorb in different ways.
CS I agree but if one group draws zero inferences, then how do you handle it?
WL No translation versus bad translation problem.
CS Adjusting things to reading levels. Common in educational material. 
e.g., 6th grade math leaves a lot out. Have to do that to move on.
MM Also say write clearly and simply for your audience. Go too far if you 
have a technique that says, "if you take this and boil it down to a certain 
level - change wording, structure, etc. you make it inaccessible." You will 
leave things out.
CS I don't want to see 2 things happen: 1 all content written to reading 
level X, 2 prevent people from providing content at a lower reading level.
MM The issue is, if you say that a company can rewrite something in a lower 
level. When someone reads the reduced content and is only able to use that 
content, then reduced for all content.
WC How do you see that happening?
MM If someone has a laundry list, will apply it all to an "accessible version."
WC I hear you saying that we have to be very careful about how we define 
the laundry list, but I do not think that we should avoid helping people 
make easy to read versions of content in fear that that will be only 
accessible version of content.
JW In guidelines we don't actually deal with translation, closest to 
cognitive case. Translation is not always equivalent to the original. 
Whatever we decide to do, we should not try to stretch the term equivalent 
to cover both of them.
CS The statement that you made about expressing different things to 
different audiences can apply to images and alt-text.
GR Functional equivalents and contextual equivalents.
JW An alt attribute is equivalent to image for purpose of understanding 
content, they are not equivalent. Could get info from one that can't get 
from the other. For the purpose of understanding, they are equivalent.
GR Functional equivalent of image is alt, contextual is longdesc. Image 
serves a function in context of the page. The description of the image is a 
contextual thing.
JW If I thought I could convey function of image in alt-text, not describe it.
GR If I have a the W3C "valid HTML" image, alt-text would be "valid HTML." 
the function is, this content has been checked. in longdesc dcribe image.
JW Always differentiated between function and description.
JW Back to issue, cognitive equivalents seems to be stickiest aspect. My 
feeling is to treat more as translation rather than equivalent.
GR It is an alternative not equivalent.
WL Not congruent.
JW If use Cynthia's definition, in some cases count as equivalents 
sometimes not.
GR Function of image is to show has been validated. That it is a hyperlink, 
belongs to link not image.
WL Important to have "reasonable" and "where appropriate" is important.
JW We seem to agree that functional equivalence is the ultimate criteria. 
Separately, have translations or rewrites for cognitive purposes where the 
same function might not be fulfilled but it may partially overlap.
CS There are cases where the goal is to "buy a pencil" rather than "convey 
info" that you can simplify user interface and language so that person with 
CD/LD can use it.
MM That falls under write simpley and clearly.
CS May not sell as many pencils without the pretty words.
Resolved (at least for today): criteria for determining equivalency is 
based on purpose and task. caveat: translations - purpose may be to gain 
full meaning of text. where that is the purpose, simplifying text is not 
equivalent.
Action MM: think about this resolution and respond.
JW suspect that with right qualifications we can make this definition work.
PB Server-side ideas and conformance: if we are going to have a caveat that 
if served out on some page some where in an accessible form, then can call 
the site rather than the page accessible.
CS Don't think we've quite reached agreement. I will draft server-side 
techniques in next couple weeks, that will help. Does that work for everyone?
JW Yes, gets down to conformance and priority issues. We decided not to 
discuss until had substantially complete draft of the 3 levels of the document.
KHS Please look at the PDF techniques. Not sure which techniques fit under 
which checkpoints.
Action everyone: review PDF techniques.
Resolved: PDF techniques on the agenda for next week. KHS says, "ignore the 
numbers and think about which checkpoints they fit under." Some feel like 
they can fit under more than one checkpoint.

$Date: 2001/05/10 21:36:56 $ Wendy Chisholm

--
wendy a chisholm
world wide web consortium
web accessibility initiative
seattle, wa usa
tel: +1 206.706.5263
/--

Received on Thursday, 10 May 2001 17:38:10 UTC