WAI Authoring Tool Guidelines Working Group
Chair: Jutta treviranus, <jutta.treviranus@utoronto.ca>
Date: Wednesday 18 August 1999
Time: 3.30pm - 4:30pm Boston time (1930Z - 2030Z)
Phone number: Tobin Bridge, +1 (617) 252 7000
The latest draft is http://www.w3.org/WAI/AU/WAI-AUTOOLS-19990811
CMN The wording I have included allows you to show what is bad, but requires that the example show how to make it good.
GR I don' understand why we can' say "o not use inaccesible examples - there are already lots of places to leran how to do it wrong
JR I see Charles'problem - you can say here is a problem, if you say how to fix it
GR I don' see why we should allow tools to use inaccessible markup
WL That isn' use that is demonstrate
GR Kynn was saying tools would show how to do things -0 he wanted t otag that this is an inaccessible practise. I hanve no problem with negative examples (actually I do a bit). I don't understand why it' ok to use inaccessible examples
JT The point was raised that there will be tools which allow you to do inaccessible things, and the docuemntation needs to be there
CMN My wording will allow people to build documentation for inaccessible examples, so long as they then demonstrate how to make it accessible
JT An example is developing a server side image map. They are going to have to include an example
IJ That is an example where you include what is necessary to make a server side image map accessible
CMN An example may be a multi-part example...
JT Maybe we should clarify that
IJ As a person who writes examples, some of this is difficult. Writing SMIL examples I was pulled up for leaving out accessible markup. That can lengthen the example considerably, which can be counter-productive - maybe we should be able to link the partial example to a full accessibility example.
JR The number of times when I use a reference, and just cut and paste an example, makes me think that that is a bad idea
GR If we want to have a checkpoint about using accessible markup it should be restricted to something about proprietary markup which does not allow accessiblity features - you must note the limitations of such markup
JT some kind of access aert?
GR Right, but I would restrict it
IJ I wouldn'
CMN Nor would I
WL Is 7.3 currently considered a draconian checkpoint?
CMN Yes, but it is necessary. And it would be nice to phrase it positively. There is a need to be able to take a small piece of an example, rather than have to make every example complete in itself - maybe linking is part of that. I think there needs to be a note in the checkpoint
JT We want to alert the author if an example is accessible and link the example to an accessible version of the same.
WC That makes sense - we have an example of how to do ASCII art in the WCAG.
WL There is a situation where you say something like "don't use click here" - that contains the inaccessible example
WC Right - we have a similar multipart example in the techniques for WCAG. As long as information is provided you can make inaccessible examples triple-A compliant
JR If the whole thing is considered to be one example with three sub-parts.
IJ Is this about documentation of the product?
general: Yes
IJ Then it should be moved to the User Agent guidelines.
JT This is about how the author uses the tool.
IJ I am hearing two things - this is about examples in foobar, and this is about when you are using foobar to produce documentation. These checkpoints are equally applicable to user agent guidelines
JT No they aren't. In an authoring tool there is documentation about how to do authoring.
IJ Why doesn' that apply to user agents as well.
JR The problem is that it can be taken in two ways. One - the example is inaccessible - the person cannot read it. In the way we are trying to use it we mean - do not provide an example of how to do something inaccessible. That is applicable to our guideilnes.
IJ I retract my proposal
CMN The checkpoints need to be in here to describe how to use authoring tools.
GR I think the checkpoint should be "Ensure that all documentation examples conform to WCAG [P1] Clearly label as inaccessible any practices which are inaccessible"
CMN I think that waters down the proposed text, in a bad way.
GR How?
CMN I think we want to say something stronger than label inaccessible examples - we want to say and show how to fix them.
GR Fine
JT Label inaccessible examples and provide a repair strategy
/* Phill Jenkins joins
JR You can give instances of an inaccessible examples, so long as it is accompanied by a note and an accessible version. I would like to add the word instace, to disambiguate examples and sub-parts
IJ I think that can be understood from context. I don't know if Chartles wanted this to be done so it could be verified by machine...
JR In our document, because we have already run into some problems, it could be helpful...
IJ In WCAG, when we talk about alternative pages, does that mean that the primary page can be made inaccessible? People felt that it could be understood what the whole and the parts were
JR We would like to say do not use inaccessible markup in examples. We would like to temper that and say do not use inaccessible markup alone in an example
IJ I think that is too low-level
JT We seem to be saying the same things. Does everyone agree that we have the requirement for triple-A, with a note saying that if ther is inaccessiblity there needs to be a note saying that and a link to an accessible version.
JR I would like to note that we mean the process being illustrated must be accessible.
JT maybe we need to add something to the introduction to this guideline.
JR Yeah, I agree
GR I proposed "do not use inaccessible markup - that is why the word is there.
IJ If I were documenting an authoring tool i would not show markup...
CMN/JT That won' work for popular markup editors
JT Does anyone disagree with the rewording.
CMN Ensure that examples comply triple-A to the WCAG. 2 Notes: This refers to the process being illustrated producing accessible content, and partial examples which are not themselves accessible should be noteed and linked to full example.
PJ The original proposal to have relative priority. You can' be a P2 tool unless all your examples are triple-A? That doesn't make sense
CMN Yes.
PJ I have a problem with that.
IJ I do too.
PJ We are being too prescriptive by that. GR, you want to know whether the examples are compliant or not. You can't make it a P1 to make examples triple-A
WL The examples can be required to be P1.
PJ You give no option to be P2 level
IJ One would not be able to produce P2-compliant examples. That's a ghood thing
PJ But too restrictive
JT Is it not the intention that notes would say the examples do not need to be triple-A
IJ That is for me the equivalent of making the whole thing triple-A
PJ So I can have an example of a javascript that is not triple-A but I need to have an exxample of a link to have a triple-A alternative.
CMN. Yes. The checkpoint as proposed would require you to demonstrate how all examples can be triple-A
PJ The problem I have is that this is P1 not relative. I could be P1 without showing how to do the P3 requirement of acronym expansion.
CMN I think we all understand - should this be a P1 straight, or should it be relative?
PJ Right.
JR I think I agree with Phill. For example a tool can't do ABBR expansion. But if you don't include that in your examples you can't get a P1
GR It makes more sense to say ensure that examples conform to WCAG. You can address the sliding requirement in techniques. Then it is a no-brainer for a developer.
CMN We need to say what level conformance we require, and what the priority is.
JT I may want to give a complex example, with different levels of priority. It will be hard to determine that.
PJ Each example should be veriiable to a given level of WCAG
JT I think we are confusing the purpose.
PJ Gregory, aren't you saying that the example
CMN The question is that when you follow the example being discussed, what level will your work conform to WCAG. SHould it be P1 that your work is trilpe-A, or should it be reltive priorities
IJ It is not always possible to provide an accessible counter-example
JR It makes sense to have a relative priority
CMN I am happy with the relative priority
WC Part of what the checkpoints are intended to do is to make content accessible. You don't necessarily want to make content accessible.
JR I didn't mean that
WC The text within the example could be gibberish.
PJ If we could simple the definiiton that all examples are web content, that would be it, right?
WC The emphasis is the gist is accessible, not necessarily the content of the example. There are two things here - we want to promote ABBR etc, but I don't know that that will always be accessible.
PJ 7.2 is the other one. We want accessible examples of the accessibilty features.
GR We want the accessiblity features included in whatever we want to show. I agree with relative priority.
CMN I think if you want to include something, you must show how to make it accessible.
WC Using ABBR is only required once - if you take the example out of context, what do you do? I think that supports relative priority.
CMN Proposal. Ensure that examples show how to conform to WCAG. P1 for level-A P2 for level double-A P3 for triple-A NOTE: An example may be built from several parts, some of which are not themselves conformant, so long as those parts are identified and linked to a conforming example.
Resolved: 7.3 to read Ensure that examples show how to conform to WCAG. P1 for level-A P2 for level double-A P3 for triple-A NOTE: An example may be built from several parts, some of which are not themselves conformant, so long as those parts are identified and linked to a conforming example.
resolved: Adopt proposed wording.
JR I can't make a meeting for the moment.
CMN nobody else has a problem with the date. The one possible conflict we identified was ATIA.
CMN I would like to check the list of groups for last call. Within W3C we have identified HTML, MathML, SMIL, SVG, Amaya, WCAG, WAI IG, WAI EO, WAI PF, User Agent.
WL CSS, XML. Wendy had a listing of TopStyle.
WL It would be helpful to have an authoring tool list from EO
CMN I have started one. It is intensely painful to gather. We have to identify W3C groups - asking developers is for our own benefit.
PJ Go through W3C member list
Action CMN: Do this
JT I have a list from a year ago
Action JT: Send list.
PJ guideline 6 - allow users to control the nature, provide a summary of documentation. We are being too prescriptive - wanting that to be configurable is too prescriptive. How can we ask for function without describing how it will work. We want users to be alerted, but we want to give them some cntrol over that. At what level? We were struggling with this in Home Page Reader. How much configuration is really useful? Some of that is part of the tool. We want something like "allow the user to control some of, or the important, or ..." Here we appear to be saying allow the user to control all.. I propose allowing the author to control important features.
JT You don't need to configure for every type of alert, but there needs to be some choice of how they're going to be alerted
WL But only when the function is available
PJ You're forcing it on everything
JT We're not asking for a huge dialogue box
CMN Can we character this as "what is the scope of configuration requirements?"
PJ That's the issue. Another one - when we read "double-A" we have to spell it out? Why is this.
GR That's for the purpose of the document
PJ So we are trying to make this a triple-A document so we are trying to make them accessible. That's fair. Overall the person thought this was a reasonable set of requirements and would be putting it into the docuement.
PJ Here's one that comes from website developers. We found on a site we were doing that it would be cheaper to have a text alternative than to have a graceful degradation.
JT How does that have to do here with authoring tool guidelines?
PJ He was surprised that there was not much emphasis on how to manage text alternative sites?
WL So if you must use alternative pages you should talk about managin these things
PJ Yes. So it makes it easy. We just need to mention it more. He was really thinking that we have a tool that helps manage the text.
JT We should have this in the techiniques
PJ Right. He was looking for more mention of this issue.
CMN I will write up my long technique for such a tool - Phill, if you could get this guy to review it that would be good.
PJ OK.
Quote was Overall I think the requirements make sense and are possible to achieve. [Ed: roughly]
PJ I am encouraged.
JT Yes. It is encouraging.
PJ It shows that if you get it to the right guy who understands the tool and can understand the requirements we can get good comments.
Action PJ: Send review points to the list.
Last Modified $Date: 2000/11/08 08:11:51 $ by Charles McCathieNevile