WAI Authoring Tool Guidelines Working Group
Chair: Jutta Treviranus
Date: Wednesday 16 February 2000
Time: 4:00pm - 5:30pm Boston time (1900Z - 2030Z)
Phone number: Tobin Bridge, +1 (617) 252 7000
The Latest Draft is the Recommendation dated 3 February, available at http://www.w3.org/TR/2000/REC-ATAG10-20000203.
JT: Time to get together at CSUN (informally)?
WL: Sunday March 26 is AU meeting
JT: Phil leaving on Saturday
WL: other WAI meetings Saturday
CMN: encourage people to go to IG session if don't have anything else to do
WL: is EO on Sat
CMN: no on Sunday -- it is a split day;
WL: Saturday is EO and prior Sunday is GL?
CMN: yes
DB: will be there from 19 to 26
JT: proposal do it Saturday night
CMN: conflicts abound
WL Friday night?
DB: steering committee stuff possibly on Friday night
CMN: is everyone going to be at CSUN before the meeting?
KB, WL, DB, JT, GJR: yes
JT: Tuesday is out
CMN: let's decide via email
JT: should we go onto agenda yet?
CMN: let's keep the meeting stuff for later; please register -- there is a link to registration form off AU home page
CMN: rather address "where do we go from here first"
JT: reads from charter -- finish Techniques as Note, other sub-tasks, evals, implementation reviews, etc.
CMN: what still needs to be done: published first draft of Techniques -- basically hints and tips; provides tips for pretty much every checkpoint; some checkpoints extremely straightforward; some of them -- especially relative priority checkpoints -- don't have a whole lot of info in them; only 7.1 (a special case) has a fair bit of info; should put in Techniques for how to do WCAG dependent things; what it means to implement a WCAG checkpoint in an Authoring Tool; need to sit down with relative checkpoints and go through them one-by-one, build a matrix, insert WCAG checklist -- put in techniques for each item in checklist -- for example, checking errors -- what sort of things can be done, what might be difficult; is possible to check readability level (Word has 3 different ways of doing this, HotDog does none; small programs not likely to integrate a monstrous piece of code -- have to push back to the author
WL; so is that P4?
CMN: no, P3 under WCAG
JT: example implementations important; rather than say tool X does this and that, have real life examples of how tools have complied;
CMN: related to that is what we've been talking about on the list; actual test cases -- if can describe what it is when test for conformance, that will often give developers a much better idea of what need to provide; in HTML, can you add ALT text and a LONGDESC to an image? -- developer then sits down, says "Ah, that's the test!" and then checks tool in development; simple thing that tells them explicitly what they need; another thing I've been working on and will send notes to list when get into readable form, is the RDF (metadata) scheme to describe conformance to any of the WAI guidelines; been building a tool -- take a checklist and adds a bunch of information through an RDF store -- essentially a database -- so that can query -- show user tools that comply with criteria defined by user - find tools, show me comments on how satisfied
WL: how does it work?
CMN: at this point is nasty ugly code
WL: what about word processor that allows to save-as HTML
CMN: won't -- tool for users to help users to gain info about conformance
WL: for developers?
JT: for anyone; efficient way of storing and retrieving information; not a checking tool, but a retrieval tool
WL: if I am purchasing agent for a county office and want to find a conformant tool,
CMN: could go to list and be disappointed
WL: if build database, will eventually find list
CMN: not definitive be-all end-all info -- contains fields such as name of reviewer, hardware used in eval, AT used in eval (if any); don't imagine it as more than a rough tool with which to check implementations; not for use by major decision making bodies -- could be used to collect a set of tools for further evaluation; kind of like Bobby -- if people get a Bobby report with no hats, but with a long list of manual checks, they are likely to stick a Bobby approved icon on their site; can't stop stupidity
JT: timeline for tool?
CMN: just hacking away in spare time; have basics in place, need to sit down with Dan Berkley or someone; hoping to do in next couple of weeks; one thing need to do is assessment of "where do we go from here?"
JT: that is derived from WAI charter
CMN: right; assuming that WAI activity will be onging, this summer, we need to assess our opinion on the guidelines; do they need revision in the next 6 months; what do we have in terms of conformance evals, implementations, implementation commitments; what kinds of resources do we still need; know we need to provide Techniques; need much more information about how to meet relative priority checkpoints; aim to have new version of Techniques Note (document published with a date and W3C guarantees that that document will be available at that URI forever; no guarantee that it is not garbage or is exhaustive; can publish new Note fairly easily
WL: what does this Note cover?
CMN: Techniques document is a Note; can update pretty straightforwardly as we do WG drafts in AU space; would like to have a new version of the Techniques document published as a note by CSUN
JT: will require quite a bit of work; gathering of examples, implementation commitments; how should we recruit volunteers?
WL: rather than can be implemented, has been implemented?
JT: purpose of conformance doc -- describe how has been implemented; where no implementation, suggested implementations; those are the 2 things need to work out; need also to look at overall structure of Techniques document -- is there a better way to index or organize within checkpoints; Ian created categories --- 3 different areas
CMN: links, references, suggestions, sample implementations
JT: think is very useful; doubt if large laundry list would be as effective; that's the type of organization I'm talking about -- how to make more usable; would also be interesting to query Authoring Tool developers who are committing to implementations to contribute how they plan to implement them
CMN: Non-disclosure and rivalry keeps companies from doing that; need perhaps, more vague ideas from developers
JT: ER group will spin off of our work and vice versa; Chris Ridpath who is editor of ERT document, could be recruited, perhaps
WL: at least as liaison
CMN: skipping back-and-forth on agenda; the ERT document is something that each of us should try and sit down and read; will be new draft of it any day; encourage people to grab latest copy at earliest possible convenience and spit their comments out onto the AU mailing list; new IBM Guidelines out -- assuming that everyone's read the Guidelines referenced in Techniques for 7.1; good reading; encourage people to read a couple of those things; haven't read the IBM Guidelines yet, myself
WL: been reading these type of GLs for 5 years, and most of it is pie-in-the-sky;
CMN: yeah, but much which used to be pie-in-the-sky aren't anymore; as historian trained to read documents and comparatively analyze them, so for me (and maybe GJR) it is a rewarding and enjoyable experience
JT: what is best mechanism to populate Techniques document? import from ERT?
WL: suggest that whoever wrote up notes at IBM be contacted to fill in some blanks; same at Microsoft, Allaire, etc.
DB: could you repeat -- there was someone at the door
WL: need people who are writing guidelines for their company, show real life examples of how ATAG satisfied by company's own projects
CMN: 2 approaches: do a conformance eval by sitting down with tool and describing how it happens -- not a simple Yes or No, but how it does it or how it fails to do it, and/or how it could do it consistent with the "look-and-feel" of the tool; other thing is to sit down and run threads on the AU list checkpoint by checkpoint -- would need to keep subject lines -- reply to message, so that they are automatically threaded in mail archive -- makes tracking easier
WL: and vice versa -- if have new topic, create new subject line
CMN: yeah; running threads -- 4.2 WCAG 1.1, ATAG 4.2 WCAG 1.2; etc;
DB, JR, KB, WL, GJR: cool-o-matic
CMN: to return to the agenda or some semblance thereof -- our goals are to have a new Techniques document published by the Wednesday before CSUN starts -- 17 or 18 March -- means a bit of fast work; if aim to have new version, containing as much as we can do before F2F would be good; then can discuss implementation -- what kind of support materials are still needed, and the guidelines -- should we be looking at a revision within the next 6 months, or do they look like they'll stand the test of time (at least for 6 months)
JT: id gaps in Techniques, what do we mean by implementation, assess whether or not ATAG doc needs to be updated in next 6 month
WL: are we writing a new charter
JT: will be in April
CMN: part of the purpose of addressing these questions;
WL: conformance certification could take forever
CMN: question of what is conformance? don't do any formal certification -- do we want to? my suggestion - at this stage -- is probably not; do we want to say informally that the following conformance evals are reliable -- that might be very helpful; if developer comes up with tool, can sit down with checklist and techniques to find out "what does that mean for my type of tool"; next question is "how do I check this", which is where conformance evals come in -- show me how tools satisfy checkpoints X, Y, and Z
WL: someone has to be person called on by the judge
CMN: that is judge's responsibility; yes, important question, but well outside WG's goals; WG could act as an expert witness, but should set itself up as one; shouldn't be the sentencing judge; want to set put documents that do as much of that as possible
WL: sounds good; each time a new ML comes up, same thing happens all over again
JT: anyone want to add anything to that?
// NO //
CMN: question we skimmed is WL's question -- organize techniques by other than checkpoint; big meta-question that don't formally address very often; my feeling on it with WCAG and UAAG, is no -- doing by checkp9oint best
JT: could link it to other organizations of material -- by type of tool, such as conversion tools -- could be crosslinked; need enough material to class it by tool type
WL: even if classed by tool type or ML, still going to be in checkpoint order
JT: core document, then P1 compliance stuff
CMN: checklist ordered by priority
WL: but within that is ordered by applicable checkpoint
CMN: think about it as sit down to build techiques or perform conformance evals -- is there a way to better organize this info
JT: anything else to discuss today?
WL: gives everyone enough to do for next 6 months
JT: CMN and I will start posting checkpoints to the list, ideas about organization, etc.
WL: give me an example -- start with 1.1 -- what are you going to tell us?
CMN: first, allow author to create anything in the ML
WL: should have a source editing mode
CMN: don't need source editing mode, but ensure that author can do it; if a WYSIWYG editor, way to satisfy is have standard user interface method as way of getting LONGDESC into image
// unminuted comments from GJR about support for tags that aren't currently supported by UAs -- approaching problem backwards -- needs to be emphasized to developers //
WL: still not sure what "ensure" means
CMN: simply that it is possible; no requirement in ATAG for author to be able to see the markup; in Word, for example, couldn't see the markup until very recent releases; not a problem or a bad thing
WL: could view in plain text editor
DB: produces HTML file
CMN: point is, not necessary for author to view source code; no requirement for that on purpose -- not a requirement, a technique
WL: question for GJR -- should this be a requirement
GJR: no -- average user doesn't want, nor should be forced, to look at doc source or learn intricacies of all MLs supported by tool; if ATAG is to have the effect we want tools need to be
JT: allow author to verify or check; has or hasn't been done, but not how
CMN: send pointer to HTML slide presentation used in Japan for W3C staff;
WL: without revealing state secrets, are you encouraged that developers are trying to do this stuff?
CMN: more optimistic than I would have expected a year ago; there is a lot of work for people to do, but pleased by how much is being addressed and how well being addressed; good tools can change quickly
WL: completely new tools, or revising tools?
CMN: mostly existing tool developers
WL: talked to Sausage?
CMN: yeah;
WL: sent newsletter, but no pointer to URI for downloading beta; only pointed you to comments form!
CMN: will bug them on that; anymore?
WL: I've had enough
// laughter //
JT: ok -- let's take it up on list
WL: badger EO group to have same PR push as given WCAG
CMN: need more focused push; average person doesn't build an authoring tool
WL: author needs to know what to expect, what to buy, when to buy
CMN: needs more targeting than WCAG
WL: meeting next week?
// YES, SAME TIME SAME STATION //
Last Modified $Date: 2000/11/08 08:13:13 $