Meeting minutes
Jeanne: Last week we had 3 ideas about what direction to go.
Makoto's ideas about accessibility supported
Makoto: I was facilitating the acc support subgroup last year.
… Reading the definition of acc support (WCAG 2.1).
Makoto: My idea, NVDA and VoiceOver can be the baseline
… These are almost globally available for free.
… Maybe I don't have to stick with PC-Talker (80% in Japan)
… But about 55% use NVDA, many use multiple screen readers.
… There is an a11y committee, and a working group in charge of maintaining test files for testing screen readers.
… We have to think about other countries / languages.
+1 that data would be very useful from that organization testing interoperability/accessibility in Japan
Makoto: When I joined WCAG WG in 2005, the concept of a baseline was considered.
… Unfortunately, there were no UAAG compliant user agents, so this couldn't be used as a baseline.
… So instead techniques was used. Anything could be a sufficient technique if it was supported by UA, and relied on authors to decide if it was supported
… Screen reader techniques were based on support by JAWS
… John Slatin who was the chair understood how important this was.
… Techniques were based on JAWS, but this created problems in Japan where JAWS was not as popular.
… For example 2.4.4 link purpose was an issue.
… If a link is in a data cell, in a table, and there is a table header cell, then link text can meet the success criterion.
… These techniques were based on JAWS support, which allowed users to check the context, but PC-Talker does not have these shortcuts
… Japan we had a hard time to address the criterion.
… Another example was bypass block. There's a technique for providing heading elements.
… When WCAG 2.0 was published in 2008, PC-Talker did not have heading-jump functionality, so we had to rely on skip links at first
… The concept of accessibility support has been flexible, but at the same time it brought big challenges
… We had to create test files, which we tested with Japanese browsers & screen readers. This took much time, and we needed many resources to do the testing
… But if we go to the definition, it says "if it does not cost" and is "easy to obtain" it should be okay.
… So the question is if NVDA & VO can be the baseline.
… If it works in NVDA it may be okay, but at the same time can we use PC-Talker users?
<Zakim> jeanne, you wanted to ask if Makoto knows why multiple screen reader users don't use NVDA all the time? Do they still need PCTalker for some applications?
Jeanne: Do you know why people in Japan would use multiple screen readers if they have NVDA?
… I think we have two directions. One is, we should do more to support screen readers like PC-Talker to bring them up to standard.
… The other is that we could use the idea of a baseline. Although if we say what the baseline is, are we hurting users?
Wilco: I think what is great about how Accessibility Supported has worked is it's flexibility.
… if an org wants all major screenreaders to work in all major browsers, they know their goal
… others may choose to narrow it to a single screen reader and browser
… we should not underestimate the problems caused by the lack of standards for screenreaders and what features to support -- probably not in AG
<Zakim> Makoto, you wanted to answer Jeanne's questions
Makoto: PC-Talker has been popular for a long time. The company developing PC-Talker also provides other software for VIPs.
… When the Japanese version of NVDA became available, it worked better, especially for web applications.
… People who only use PC Talker, started using NVDA. The number of people who use multiple screen readers increased.
… When PC Talker doesn't work well they switch to NVDA, but long-time PC Talker users stick with it.
<Zakim> Azlan, you wanted to ask I recall a list shared of tools for testing accessibility with capabilities and limitations - is there an equivalent for screen readers?
Azlan: Last year a list was shared where vendors could share the capabilities of the products?
<Zakim> janina, you wanted to ask whether PC-Talker implements ARIA?
Janina: For Linux there is excersizer. On Linux there is Orca. NVDA and Orca are do similar things.
… In general, does PC Talker implements ARIA? ARIA is so screen reader TTS specific that the next version of ARIA needs to do more for braille
… This is one of the reason APA has Adapt.
<Zakim> Makoto, you wanted to answer Janina's question
Makoto: PC Talker has limited support for WAI-ARIA.
… I contacted the developer and asked him to implement the easier attributes, which they did quickly
… PC Talker is developed by one person, and he needs to also support things like e-mail clients.
… A web browser is just one program for him, and he doesn't speak English. Technical specifications are written in English, which made things harder.
… Once he had the information, he can do it
Janina: It sounds to me like a good next step could be if someone could sponsor a translation of ARIA into Japanese.
Makoto: PC Talker is watching NVDA. They can see how it supports ARIA and can follow that.
<Zakim> jeanne, you wanted to say that setting a baseline of NVDA would solve the problem of the JAWS bugs, but raises the problem of W3C officially endorsing a product.
Jeanne: Setting a baseline of NVDA solves the problem of people knocking themselves out over a JAWS bug.
… NVDA is a more standard-compliant screen reader, although I don't see how W3C could name a specific product. That could be a problem for endorsement.
… Perhaps we can work out the qualities of NVDA & VO, and define the baseline by those qualities
Janina: I'm not aware of W3C translations other than WCAG. I think that's a fundamental problem.
… It seems there should be a way to do more. ARIA is established enough to look at where resourcing for that would come from.
Makoto: Browsers have automated translations. The quality of that is getting better. Even if there's no translation, we can read automatically translated specifications
… We don't have to name NVDA specifically, it might be possible to say if a free and easy to obtain screen reader supports the technique it is supported
MJM: I wonder if anyone's explored WAI-ARIA becoming an ISO standard. They get translated.
… I thought the plan was to get WCAG 2.2 updated as an ISO standard.
Janina: We have WCAG, we have ARIA, we have other stuff in the pipeline.
… I'm not seeing a path where we need acc support going forward.
… Just the fact that we need established standards that benefit, not having translations is an obvious first problem.
… maybe we still need work-arounds.
Jeanne: I think we do for new technologies.
… There's very much a place for if you have something new, it needs to work.
Janina: If our poster-child is screen reader, and the main problem is that the standard's not available in the language of the engineer, that's not a new technology problem.
Jeanne: Kim Dirks used to talk about the problems her company had, advanced application not in HTML, where they ran into a JAWS bug. They could not get it working, and because of their interpretation of accessibility support they felt the product was inaccessible.
… That's another part of this we should discuss. Accessibility support and screen reader bugs.
… We could make things simpler if we said that you need to follow the standard, instead of forcing it to follow screen readers and break the standard.
Makoto: On the other hand, local government website will not be able to ignore PC Talker users. We'll need test files.
… I see great code examples in ACT documents. If we used code examples from ACT rules as screen reader test files it may reduce the burden of creating test files.
Wilco: ACT has been exploring use cases where different assistive technologies implement features differently
… accessibility is a long-tail problem. The people who are not standard users that have different needs than other users. It's easy to say we don't care about PCTalker, it will be aproblem for many people. I see where it is coming from, but it seems to contradict what we are hoping to achieve
Makoto: I came up with two points. I'm going to talk to PC Talker dev to check what they're challenges are.
… I'll also need to meet with the Japanese working group to discuss what a better way is to do accessibility support. This will take some time, but I'll share what I get.
Janina: It's possible that a curated overlay could be part of an acc supported solution
-1
accessibility at the edge community group
<Makoto> Link to my slides https://