04 April 2000 ER WG/IG telecon

Summary of action items and resolutions



What should we be doing wrt implementations of ERT

LK emphasis is on "we" what should we be doing? already we're working on ERT and the list of tools.

WC we need to bring ERT techniques into reality.

BM unusual situations for almost all of them.

DB easier for us to do for us.

LK there are lots of authoring tools...

DB but not enough that conform.

LK there are some w3c groups where recommendations are only put out after there are only one reference implementation of the whole thing.

WL it's being done with SVG, SMIL, CSS, etc.

LK a whole implementation before Rec?

WL have to show it works before you write the book. we're in a similar boat. my question: which recommendations do our stuff refer to? ATAG?

LK CMN's e-mail considers repair tool a special case of an authoring tool. in a way LYNX is an accessibility evalution tool. a sighted person can use w/out speech to get some idea of what will happen. it's not a bad rule of thumb to use. are there tools that come close to conforming to ATAG?

WL some on AU would say yes. Microsoft is listening and talking, must be doing something.

DB yes, not sure what. not being ignored.

WL any movement is a step forward. should we investigate if any ERT stuff being included?

LK we're referenced in the techniques of ATAG, not in Rec itself.

WC work with AU to find out what conforms or not.

LK list of conformance somewhere?

WC generate a checklist of ERT to see who does what.

LK systematic work going on in AU to list what various tools are doing?

WL no.

WC in ATAG techniques, has examples of tools that do x, y, z. by technique not by tool. techniques for 4.1 - says that amaya is working on it.

Action Wendy: invite amaya folks to ER/AU meeting.

BM HTML and CSS validator are listed on our page.

WC Gerald (maintainer of HTML validator) will be at our face2face.

WL what about MathML? SMIL?

WC an XML validator exists that should cover. not sure if it is public.

LK there is one that exists elsewhere. I will add to tool list.

WC think we should include in the exercise of adding tools to the exisitng list. create a checklist and check "supported" or "not supported." this list does not overlap with the tools that AU is evaluating.

MC there are techniques that we have refrained from suggesting, because it is hard to do.

LK hope we are not. that's a separate issue. let's add them.

WL how far do you go?

MC you could assume a neural net processor.

LK sure, a single sentence. right now, if don't mention OCR on a picture, we could say, "it is desirable to read the text with OCR." even if we can't define it. there will be things that we haven't.

WC we need to make a difference which means we need to bring this into being. find the holes and either find something that fills them or try to fill them ourselves.

WL have we gone through to determine which exist? possible? unlikely? I think it is unlikely that something will exist that writes alt-text.

LK there are things that write alt-text for pictures - like the Altifier. instead of saying "likely or unlikely" i'd prefer to say, "easy or difficult."

WL exists or doesn't exist.

LK needs research or not.

WL huge thing in cognitive world: illustrating text. this will take a tool.

Action: WC generate checklist.

Action: MC go through checklist for Bobby.

Action: CR go through checklist for A-Prompt.

Action: LK go through checklist for WAVE.

LK also need an "other category."

WC yes, by checkpoint here are the techniques, and here's a box for "other ways to satisfy this checkpoints that we have not thought of." then an overall "other."

LK Lynx can help you understand how a page looks without JavaScript. it doesn't say "you've got it."

MC Lynx vs ERT - if author uses Lynx they will know right away if javascript works. in ERT, we don't address.

LK Lynx or listening to the page with a screen reader are ways to test usability that are not captured by ERT techniques. when we evaluate the tools, although we check it off against the checklist, but don't stop there. look for other things that are missing. not just things that could be added as a checkpoint, but not part of the ERT paradigm at all.

WC should we add it? could be under "ensure that new technologies tranform gracefully."

LK we can see when those comments come up, if we should add to ERT or not. a big "other" category.

WL for example, using a black and white monitor falls under checkpoint 2 (uses of color).

LK this is a little off-topic, looking at something in black and white is usually not a good technique.

WL it does check if you can use the page w/out color.

WC other people to review tools?

DB front page does some testing. I'll talk with Heather.

Action DB to talk with Heather Swayne to review FrontPage as a testing tool.

DB Roger Hill from MSN is testing the MSN site.

Action DB talk with Roger Hill about testing methods used on MSN.

WL testing techniques as important as testing tools.

LK cynthia brought up interesting points: what happen when try to scale these to sites with lots of pages (>1,000).

DB there are hundreds of thousands of pages on Microsoft site. MSN also lots of pages. almost has to be automated at that size.

LK perhaps do a statistical sample and do manual on a few pages.

Action WC review a few tools, get Gerald to review HTML Validator.

LK quick way for tool provider to fill out their own checklist. need a simple form. encourage people to review their own tools and get included on our list. gets them thinking about this.

WL therefore, contact Allaire. seems positive.

MC we should continue to list tools, publisher not responded. people may not respond. some may not feel they need to spend time on. helpful as service of people we are serving, "here's our brief analysis" then "full analysis" for people who responded.

WL we can try! "what does your tool do to check for these matters."

MC coordinate with authoring tools.

WC Charles working on relationships with developers: he contacts them then shows up on their doorstep. the intersection with AU (between the tools they are investigating and those we have listed) is small. that part needs to be coordinated. the rest, we need to track down.

WL ERT tools part of other tools rather than stand alone.

LK disagree that the utlimate goal is to get these tools integrated in with other tools. there is still the need for stand alone tools.

DB you can consider UA's testing tools. in the future they may be doing more.

LK the ideal: techniques part of authoring tools and browsers.

WL can we suggest something to UA for UA's to perform validity checks for the user?

LK i don't think that has been part of the UA philosophy.

Action DB take issue to UA - should they be considering ERT techniques in their guidelines? have they already? include in techniques document?

WL if you can be warned that you have the wrong browser because it is going to be ugly, that would be great.

ERT going public

WC one comment, about null alt-text. need results from survey. working with kitch barnicle on survey. jason reminded WCAG group. think we can go public this weekend. put it on this weeks WCAG agenda.

Resolved: assuming no more comments will release draft publicly this weekend.

$Date: 2000/11/08 08:17:43 $ Wendy Chisholm