Minutes November 8, 1999 telecon


Summary of action items and consensus

AC presentation

/* wendy describes AC meeting */

Slides from the WAI technical highlight.

Report tool

DD several comments on the list. to do's: wording, development, restructuring of the 3rd step (making it simpler). We need to decide what to do first and who is going to do it.

MC how many sites do not have an e-mail address to send comments to?

DD we don't have a measurement. today, there is nothing else being done. The tool doesn't access the page.

MC my impression that many pages do not allow any feedback.

LK Finding it more difficult, especially with larger sites, to find how to give feedback.

HB some way to encourage companies to get feedback?

DD basic meta data convention that people have been using (meta author=). there is no requirement and no widespread RDF notation. so have to look for a mailto: in the page . this is not currently handled by the tool.

LK what about the InterNIC database as a means of contacts? so, given a domain name, who owns it and who the contact is.

DD that is one way. for large sites, you get the person who is officially responsible. for example, for w3.org you get Tim Berners-Lee. If the e-mail name is valid, then send something there.

LK so send to webmaster and see if get a message that it bounced?

DD more of a technical problem. the main issue is the usability of the tool. there is the issue of covering all of the checkpoints, at leas the p1s or some subset. The more pages we have the harder it is for people to understand the tool.

LK also discussion about symptoms as opposed to the code itself. e.g., instead of asking, "are they using accessible javascript" versus, "i clicked on this and nothing happened." the average user doesnt know what's going on.

DD I've been driving the comments. I can continue to do that, but I'm busy this week and next week.. If someone wants to take responsibility of the tool, or i can with some delays.wendy do you want to?

WC long term? would i have access to students?

DD yes and yes. although guess it doesn't need to be done in the next 2 weeks.

WC can definitely look at symptoms vs all of the checkpoints.

DD regardless of impementation, we need some consensus.

WC @@ put together a proposal send to list as to how to proceed.

WL is it too premature to enlist EO?

DD we got some feedback saying it is too complicated. the risk is that people say it is garbage and they won't use it again.

WL we could use some publicity to get feedback.

WC i think we should deal with the feedback we have alread


LK go ahead and discuss since Chris isn't here?

BM looking through for "must" then we don't need much discussion. e.g. 1.1.J what about an empty script?

MC using active scripts, can certainly happen.

BM then does it need a NOSCRIPT?

LK could have the situation like purely decorative images, where blank alt-text is allowable.

MC it's hard to determine relationship between script and NOscript. several scripts are purely decorative. "there are scripts on the page...this is what they do"

BM Brian implies that NOSCRIPT must follow the script.

MC required by HTML syntax? therefore, we dont want to be flagging errors that are valid HTML. sometimes NOSCRIPT is appropriately placed elsewhere.do we need to discuss?

BM therefore his wording may need to be cleaned up.

HB if arbitrarily placed should be a way to couple them.

MC there isnt any in the definitions.

WL there better be!

BM problem with the script tag.

LK so you could have 6 scripts adn one noscript.

WL dont even have to have one noscript.

WC more like longdesc than alt.

LK does that check also have a provision for viewing the noscript contents so that person may compare. so if you want to expose the noscripts just so one can see. its pretty much a global comment. there are so many things that require human intervention. the tool should make it easy for the human to make judgements. for me, its more important to do that than the automatic stuff.

MC chris and i have talked about. not sure that it has migrated to the document. seems to be outside of discussion about regular techniques.

LK show the person the document with the alternative content visible, where it falls, or you could links to the alternative document, or side by side.

WC already in authoring tool guidelines. we don't need to define here. let's just focus on what needs to be checked and what can be repaired, not how and when.

LK then need a general note?

MC maybe a suggestion that is made with ERT doc. it is a guide to implementation for automatic tools. the rest of it should be in preamble or in separate note.

LK that's a major disconnect. i thought the ERT doc was both to manual and automatic checking.

BM it does, but more along the idea of the checkbox, "i've checked and I comply."

LK the alt-text acurrately describes each picture, or each picture should show alt-text. and not just show in a list, but with the image in some way. am i the only one who has thought that?

WL AU has a lot of that.

MC these types of things be suggestions. i see it primarily as the algorithms for checking. the interface are a different part of it.

WL both are necessary. need to do as much as a validator.

WC people go through AU to make sure 2 docs fit together well?

LK if ERT is stand alone. then we can point to them in AU. but if we have a general statement. therefore, best case is that they have the same words, stuck here for your convienience. or just have a bunch of links.

WC yes, links. we dont want this document to balloon.

LK in general the ERT doc has links to AU features that are required or that should be in document.

WC if they have a techniques, then we can just link to that technique.

HB no guarantee they have an authoring tool. some depend on an authoring environment.

LK consensus that we need to have pointers to AU doc for all features that be part of ERT tool.

WC propose process: for each checkpoint see if technique or relevant info in AU and link to.

LK group agrees this will work.

next meeting is nov 22