20000120 WCAG telecon

Summary of action items

Carryover from last week

Agenda items not discussed



This agenda was posted to the list on 18 January 2000.

Action items from last week

"One size fits all" thread

GV don't want to take it off the list, since we'll be seeing more of it. If people are not interested then they don't have to read. As long as they keep the thread title.

JW I was proposing that it ought to become related to the deliverables of the group.

GV yes, it does look more like an IG topic.

IJ JW's point is well taken. It is reasonable to tease out a proposal since this is working group bandwidth.

@@GV suggest move "one size fits all" thread to IG.

WL questions on IG re: priority of validation

JW how move discussions between IG and WCAG WG?

@@ JW propose way for handling mechanism for moving threads between IG and WG.

ETA issues

IJ who is "working group" that can add issue?

JW those who have completed "call for participation."

IJ but might want subset of "members in good standing."

GV yes, should revisit periodically.

JW starts from scratch w/call from participation?

@@JW track incoming requests for participation.

Clarification of what is meant by a "linearized" table

JW gregg submitted proposal that WCAG defn should be definitive. I agree. Concerned that agreed as co-chair.

IJ why?

JW b/c as co-chair concerned about agreeing w/other chair and influencing others.

GV you are a member of the working group and should have a vote.

IJ acknowledged art of acting with chair hat on or off. yes, you don't want to bias people. you will find a groove for how to contribute. i can assure you that anything you say will not have an influence on me.

/* laughter */

JW proposes that techniques should be updated to be in synch with guidelines.

GV Discuss implications in techniques document. You can't declare in order by column, since table markup is row based. Then if unroll by column, that is o.k. unless use machinations.

JW another suggestion is that if the table is used for layout then TH should not be used. "scope and headers" still should be used. but best to avoid since it might assume it is a genuine table.

@@WC update the discussion of table linearization in the Techniques.

@@WC take resolution to ER (defn in WCAG is the one to use).

WL moving towards not using tables for layout?

WC yes. once style sheets supported.

JW in the near future.

IJ MS still supports IE3, but won't drop support for 4 or 5 for a long time. realistically, we won't get rid of tables any time soon.

WL If CSS2 supported, that's a step in that direciton.

IJ but until IE4 and 5 disappear, designers will use tables. this is based on my experience from talking with the MSN folks. Money drives the design by pixel not by structure.

WL ought to read Jakob Nielsen's new book. it makes the point that people are more concerned about appearance than function. it is very compelling.

@@WL give electronic reference to Jakob Nielsen's new book.

Conflicts between checkpoints for layout

3.3 Use style sheets to control layout and presentation. [Priority 2]

5.3 Do not use tables for layout unless the table makes sense when linearized. Otherwise, if the table does not make sense, provide an alternative equivalent (which may be a linearized version). [Priority 2] Note. Once user agents support style sheet positioning, tables should not be used for layout. Refer also to checkpoint 3.3.

IJ it holds together but is not explicit.

GV you can not say "use something." like, "drink poison" but "don't drink posion if it's going to kill you."

IJ but there is an until user agents clause. also 11.1. 11.1 ought to be explained as an umbrella.

WL someone who reads this but not has read 11.1 is in the dark.

IJ whether the first time it appears in the document is a different question. perhaps it should.

JW Ian had a proposal which would have modified checkpoint 11.1 to eliminate some of the UUA clauses in the document. Gregg replied by asking if conformance claim holds at time it is made or if it changes.

IJ thus you date your conformance claims. you can't claim into the future.

WL nor can you require it in the past. grandfathering would then be alive and well.

GV every day you have your server up you have the info available. if it is accessible today, then it is not in the past.

WL that if they stamp w/our icon?

JW it is up to them to determine if they conform. does ian's unified conformance proposal (dating of conformance claims) meet the umbrella technology exception in a revision of checkpoint 11.1?

IJ I can repropose the idea that 11.1 should be an umbrella.

GV I worry about dating conformance claims someone can say, "i conformed years ago."

IJ e.g. we publish a document and never touch it again. we can't go in and change that icon. you can't punish us for that document b/c our policy does not allow us to change it. we won't change that document.

WL but it says "until" and when that is passed, then what?

IJ someone can say, "you are inaccessible. you need to republish." it's like old literature that was racist. you don't delete it, but you supercede it with new thinking.

JW this is an interesting issue as to how one treats a relationship between the unsupported technology and the question of conformance. any existing document that conformed at one time, remains conformant but when it is revised those then need to conform w/guidelines at the time the new doc is published.

WL like architecture. when modify buildling have to comply.

IJ old RFC's still available but marked obsolete.

WL yes.

GV yes. if you state as a preference not a requirement.

IJ as soon as you republish, the date of your conformance claim changes.

GV it is preferable that whenever the guidelines are updated that you bring your material up to date with the guidelines.

IJ a huge burden. good idea, but...

JW the proposal to generalize 11.1 as an unsupported technology to qualification to the guidelines. make clear that conformance claims may change as technologies evolve.

IJ resubmit proposal with tie to new unified conformance statement.

GV each of the guidelines can be cited out of context. therefore, a blanket statement could get lost.

IJ "Once user agents..."

WL "after style sheets are supported..."

JW I don't like the out of context requirement at all. There are many interdependencies between checkpoints. Reading a couple out of context does not resolve that problem. We can provide cross-references.

GV that's fine, if it says "see also.." There's no way that a person can say this is it because the other piece is linked right there.

JW whether a cross reference is in the document, one could distinguish between those that are informative and those that are intended to qualify the statement. for example, you would have a statement below each checkpoint. "Qualifications: ... see also [these checkpoint]."

GV which fall into qualification category?

IJ we do this in the HTML Rec. "check out ..." vs "this requires..."

JW under 3.3 say "qualifications: see 11.1 [revised appropriately], 5.3" if you want to add other references they could be informative.

GV under 3.1, it says refer to guidelines 6 and 11. where's a requirement?

JW under 3.3

WC can someone take an action item to submit a proposal?

JW there is a general understanding that changes in errata make changes if the document were revised. therefore, you mentally substitute errata in document. Therefore, needs to be as close as possible as to what put in next version. The relationships ought to be considered before errata is proposed. IJ willing to resubmit proposal for 11.1 and consider GV's issue with timing issues with unified conformance statement. We ought to revisit the time of our calls. It should be 1 1/2 hours.

@@IJ resubmit proposal for 11.1. Refer to Ian's original proposal from 28 July.

GV guidelines will be used more and more in legal context.

IJ that is fairly unsupportable. when you take it so far out of context.

GV this is political. it has to be easy to defend. only one case someplace that proves this doesn't have to be followed will have ripple effect and no one will support. It won't matter what you say, it will matter what they think.

IJ when GV raised issue re: dating conformance claim what should I try to capture?

GV frequently used materials should be kept up to date. e.g., if you don't update the DOS manual in 10 years but they still sell it it ought to be up to date.

@@JW proposal for dealing with relationships between 3.3, 5.3, and 11.1.

JW move to 90 minute calls?

IJ lenght of call should be in the call for participation.

GV, IJ, WC suggest keeping it at 60 minutes.

$Date: 2000/11/08 08:30:14 $ Wendy Chisholm