[Accessibility] Revised minutes for June 1 teleconference
john goldthwaite
jgoldthwaite at yahoo.com
Fri Jun 17 05:03:58 PDT 2005
These have been revised using the recording. Its
still not an exact transcript.
June 1, 2005 Accessibility meeting
Attendees:
Gunnar Schmidt
Olaf Schmidt
Bill Hanaman
Janina Sajka
Matthew Wilcox
John Goldthwaite
Larry Weiss
George Kraft
Pete Burnet
Andreas Gonzales
Earl Johnson
Recording the meetings starting today. The
transcriber will have a password protected audio file
to use. Please still speak freely. We will test it
out today.
George posted minutes to the main list this morning.
Are there any additional corrections?
Andreas- I haven’t had a chance to read them yet.
Janina- We will hold approval until all have had
opportunity to review them. If no objections within
48 hours, they will be considered approved and will be
posted. We had a good session last week. I’m sorry we
didn’t have our recording technology working last
week. Any announcements from anyone? I have one,
there is activity on the ISO working group which as
proposed a gap analysis. That was a bit of a
surprise. I have posted the documents on our site and
there are additional documents on their website. If
you need that url let me know.
Bill- do we have a contact with that workgroup?
Janina- I am the formal contact and it is open for any
of us to participate. I believe I will attend their
meeting in Toronto in July.
Bill- the current state of things is not documented
well enough for anyone outside the immediate developer
community to do a gap analysis at least on Gnome
technologies, the free desktop technology, right now.
Especially if they do it without contacting us. I;d
have concerns that they may be missing things or
making assumptions.
Janina- I will do my best to forward the
communications I get from them. There has not been
much the last two weeks. They have set up a website
where they are posting their documents. There is a new
document app. I will make sure that message gets
carried through and they don’t come up with some
standards without consulting with us.
Bill- even if they come up with a gap analysis without
consulting us, it will be at best confusing. If they
are in agreement you could draw some confidence from
that but if they divergent it could be confusing.
Janina- If they are divergent there needs to be some
resolution and we need to be part of that resolution.
The outcome needs to be something we can support in
free and open environments.
Earl- is this the INCITS accessibility study group
headed by Connie Myers?
Janina- Not INCITS, it is ISO JTC 1. It is the one
headed by Connie Myers.
Earl- I think Peter may be attending those
occasionally, you may want to check in with Peter to
see.
Bill- Peter isn’t with us tonight is he? I think he is
nearly in my time zone, wasn’t sure he’d dial in..
George- Who in the FSG is in a national body to be in
this conversation? if you are in the national body
you can review and comment.
Janina- asside from us?
George- who from FSG is participating and which
national body are they in?
Janina- I am not aware that there was anyone from FSG
that was at the Sheffield meeting other than myself.
George- so when they come up with their gap analysis
at the next meetings, if you are in a national body
you can interject comments that will be put into the
record.
Janina- we can do that directly as FSG. we can comment
as FSG. After the Singapore meeting, FSG can propose
standards directly to JTC1 for ISO certification That
was what Scott McNeil was doing before he left FSG. I
was told I could come to the meeting as a member of a
national body or as FSG and I chose FSG
.
Bill- if we are productive in our current gap analysis
activity, we should let ISO know we are carrying on a
gap analysis on existing APIs and feed that
information into their process. to JTC.
Olaf - is that gap analysis on the APIs or is it on
what is available for free desktop.
Earl- I thought it was a gap analysis about
accessibility standards and not API
Olaf- there is also a lack of information about which
applications available etc. Sometime people see only
AT written with a particular tootkit. While, in fact
the different solutions are in fact complementary.
Good idea to document that. If they are looking at
API that’s different.
Janina- I believe they are at a level above API at
this point. They are more at the level of Section 508
lets you have scripted web content while the W3C is
noticably silent and that is a gap and we need some
consensus.
Bill- then it’s a harmonization activity?
Janina- yes, from a desire to make sure we don’t have
multiple standards. It’s to make sure that
accessibility is planet wide and not different in
various political areas. Industry doesn’t want to
meet different spec’s in each political region. How
far they go with that is that is another matter. They
may want to adopt some of our spec’s as a model. The
group is new, it has only had one meeting. The next
will be at the University of Ontario, Jutta Treviranus
will be hosting.
Back to the political point of being a member of a
national body. There is substantial value to attending
other than in a national body. Thanks to Scott for
getting FSG into the ISO track. We can attend as FSG
and not as part of a national body. You are limited as
a member of a national body as to when you can speak
and about what you can say. There is assumption that
the members of a national body will come to a
consensus before speaking in the international
meeting. So we have a great deal more freedom if we
attend as FSG. I’d invite anyone from the group to
attend. Any other news?
Earl- On the keyboard spec the goal was to have it
completed in June. I’m working on adding the test
assertions and reformating. The June date won’t be
met but we should be complete by the end of July.
We’re looking at July for the finalization of the
spec. I’ll be on vacation from next week until July
11. I’ll be working on it but its taking longer than
I anticipated.
Janina- better to get it right than to meet a
particular date. We have some outstanding issues - we
owe NSF a report and we still need to complete the
transcripts. We hope to have the transcripts in June.
We may be able to spend some additional funds to
attend meetings.
We had two deep dives with Bill last week and the week
before on AT SPI and the support of complex rich
documents. What are the next steps? We are going to
have more complex doc’s and ever richer media as we
move into the future and we need them to be
accessible. Its better to build it in than to add it
on as Peter Korn said. Where do we stand with that,
what are people’s reactions? Olaf I know you had some
questions that you posted on the list. Anyone want to
take a stab at it?
Bill- If it is the case that we have visited a
significant part of the issues around documents, we
could propose a strawman solution for the issue in the
aT-SPI context. I’d be willing to propose a set of
extensions. Not necessarily to advocate them but to
use to discuss them as to whether they are adequate or
necessary. It might not be too early to start
drafting a counter proposal. Working from there we
could see whether to move forward and on what time
frame.
Olaf- I think it is also a good idea. Just a couple
of points, we have what might be useful to be
extented. We can discuss whether it needs to be
extended.
Bill- I everyone agrees, I would take an action to
create an extension to AT SPI to address some of the
issues that have arisen. .. Will address as many as
possible, not decide which issues are valid but to get
them started for discussion.
Janina- good, what Time frame?
Bill- I think I could have a draft proposal by our
next meeting. I’m not envisioning it being large or
complex.
Janina- since we are in agreement, do we have an idea
of how we will test whether this meets the needs for
complex documents? We might want to test that this
really meets the needs for complex documents. It
seems reasonable that we would want to test before we
commit to a standard How do we test, we’d need to
find several sample complex documents. Need to look
at them from several view points, from several user
needs, kinds of technologies involved. See it these
are working well enough with AT, that we’d have
reasonable expectation that it is going to work.
Bill- In the case of capabilities, if something is
expressible as an API, and how expressible. we may
have to do that as a paper exercise, may not
reasonable to construct software, to demonstrate it.
Where there are performance issues, we will have no
choice but to do some benchmarking. We have to
demonstrate that there is a performance problem and
then quantify it. On other fronts, it could be a
paper exercise.
Olaf- I very much agree with that. If we are looking
at what needs to be extended, then we don’t want to
lose these points in the discussion. its nothing we
need to have working solution to have a complete
solution.
Janina- will come back to this topic with Bill’s straw
man. If takes too much time we may need to have a
separate call. I’ll be on a plane at this time next
week
Bill- this may be best for the mail list for review
and discussion. For problem solving, its nice to be in
realtime but for this the mail list is okay.
Janina- We will be grappling with certification, will
need to have some comments about the same time we
release the keyboard spec. Maybe not then, but
certainly by the time of the second spec.
Bill- when we publish the first draft, we should
provide some guidance for readers.
Janina- Earl- can you see who as been cc’d on the FSG
email. We’ve had a response from the FSG board
regarding our request for guidance. They are not first
time problems, they were encountered by LSB and
internationalization. We were looking for guidance so
we could respond in similar way. It turns out that
while it has been addressed in the past, they are
currently reviewing it. It is an issue that needs to
be carefully stated- We’re providing a level of
service for accessibility, we’d add to it as will work
on the roadmap. Jim Zemlin has returned from China and
has requested that Janet Sun set up a meeting
including Janina, Earl, Janet Sun,
Earl- Arthur Tide, Amanda McPherson are the other
people
Janina- Amanda is a PR person. Does any know him?
Matthew- I remember Mr. Tide from Linux Care? and I’m
unhappy that FSG has chosen to hire him.
Janina- The topic is open, date of meeting will be at
earliest next Friday. It seemed most meaningful to
start here. We won’t take any action without
discussion with the group. Would like full consensus
on this topic. Do we want to brainstorm on this? We
need to have the meeting and have approval for this
language from beyond this group.
Earl- Did you say that they don’t have any precedent
for this situation?
Janina- I was surprised to read that. They said that
they were re-writing their processes so they don’t
have one right now. We need some rough text. That
puts in the situation of who gets to write first. I’m
will to take a first stab. It seemed useful to have
some ideas from FSG. What does it mean to have a
keyboard standard that ISV’s can test against.
Earl- that requires manual as opposed to automatic
testing.
Bill- key thing is why are we doing; what is implied
by conformance. Our mission gives use a broad idea of
what we are doing and why. But we have always had a
idea of being circumscribed by issues of practicality.
We need to justify the trade off and make the
limitation of the specification explicit.
Janina- we want to be clear for people that read it so
that they don’t get unrealistic expectations. We
don’t want to demean or sell short what the standards
actually does provide. But it is critical that users
understand the what is necessary for accessibility in
the OS and application.
Bill- One of concerns raised in the keyboard
accessibility group when we have made decisions that
have limited our scope or opted to not provide a
particular thing, the concern is that by certifying to
a low standard, a higher standard will be impaired or
prevented.
Earl- the implementer will stop at the lower standard.
it’s a minimum but the state of the art is further
along.
Bill- state of the art is too strong, best practice is
possibly even too strong. Don’t know how it works with
FSG but one thing we want to prevent is the abuse of
the certification process by vendors overstating their
case. Because we have certified to FSG 1.0, we provide
complete keyboard accessability. It may be that part
of the conditions for certification would include
limitations on the language that could be used about
certification. Condition that you don’t make invalid
claims based on the certification. There should be
limits on what the vendors can say based on
certification.
Janina- the service provided are by conforming the
certification are essential to keyboard accessibility
but do not guareenty that any particular application
on this distribution utilizes it appropriately.
Bill- If we are concerned about abuse or
misinterpretation, it would be worth it to see how the
FSG handles this in other areas. Are there limits to
what the vendors can claim based on the certification.
Are there ways to limit vendors over excessive
self-aggrandizement?
Janina- W3C is an example with 3 levels. Most sites
are going with level 1 and Section 508 and not going
on to higher levels.
Bill- these are moving targets. If we as a group try
to specify end to end accessibility, I don’t think
we’ll publish a spec or validation suite.
Earl- there is missing technology that will need to be
developed.. Achievability is an aspect of what we have
been doing
Bill- If we plan for the spec to grow, Version 2.0
will require more features. There is the potential
for providing continually improving accessability.
George- example- FSG will say we are creating an
access standard - Keyboard access 1.0. We as a
workgroup define what it means to be keyboard
accessible 1.0, what it doesn’t mean. If some
marketing person says xxx and gets it wrong, you send
them a polite note, it doesn’t mean that. If they
continue you slash dot them. One xx some marketing
person said y and we slash dot’d them.
Bill- Whether slash dotting it is possibly the most
effective thing, I don’t know.
George- We did that in the LSB. We were trying to get
out the 1.0 and we said that noone can claim
conformance until these criteria are met. Some
marketing person jumped the gun and we made them
retract their statement. And said if you don’t
retract your statement we’re going to slash dot you.
It was in the best interest of their product to
retract the statement.
Bill- What do you think about my suggestion as a final
recourse? That one’s certification gets pulled for
misuse. Is that a notion that has any precedence?
George- Currently, the certification gets pulled for
non-compliance. If you don’t pass a test or someone
finds you don’t meet the spec that wasn’t in the test
case. You have some many days to fix the problem or
to challenge the specification. There is a process to
do that. I haven’t seen anything about a
certification being pulled for someone making extra
claims.
Bill- does anyone think this is useful to explore?
Janina- I think you are getting to the right stuff,
Bill.
George- When you get to the claims, you’d have to be
very explicit about what they can say in claims; these
are the marketing claims you can make, if they make
other claims you’d have recourse against them. If you
did not explicitly say what the claims were, legally
they would not. If you iterated all the claims, you
could get it into the certification. I haven;t seen
it happen before
Bill- I know it would be an additional complication
for us.
Earl- is certification a black and white thing or are
there levels like W3C?
George- it is black and white; if there is a spec
either you meet it or you don’t. If you had
sub-levels, they would have to be different product
standards. It costs the FSG to administrate each
product. The board would not want to have multiple
levels since it would have more administrative costs.
The board would frown on it. In Posix testing, there
things, if there are things that don’t exist in the
system you can’t test but you can still pass. I think
the FSG can handle that, said I’m compliant as long as
its on the system. In my presentation in Hawaii, I
showed the 9 different failure methods. Its up to the
workgroup and FSG as to whether to pass it.
Earl - is there a risk then. I thought we could
specify more than might be supported by current
technology. If they say they don’t have the
technology on the platform, they could say the we
don’t have a keyboard on our product therefore we
don’t have to support XKB. Because its not there they
can pass the certification.
Gunnar- it would pass if your system has no support
for keyboard. But, once you have a keyboard, you have
to do the testing.
Janina- that is an interesting question, a cell phone
is going to have keyboard with maybe 12 keys. But it
has no shift keys that could be used press 5 times to
turn on sticky keys. Yet you might use some other
combination of keystokes to activate sticky keys. Do
we accommodate that and let it pass?
Earl - no, at the moment we don’t accommodate that
Bill- at the moment we accommodate systems that don’t
pointers or mice. But that’s the only explicit thing
we make conditional. It’s the only place where the
system provide bla bla
Olaf- it would be possible to change the keyboard spec
so that it would be valid for a PDA or phone if people
can use a combination of keys to turn on keys, then
there must be a key for turning on sticky keys.
Bill- if we did that, we would in effect weaken the
spec for desktops. It means that the familiarity of
the current gestures wouldn’t be guarreented which is
important for users. There is always a trade off.
Olaf- if it was a standard keyboard you could say, ..
Bill- if a system has a shift key then hitting it 5
times has to starts sticky keys
George- That’s true when you come up with another
product standard and you could call it imbedded
accessability, you could define a subset for imbedded
accessibility for devices that have
Janina- I hope we get to imbedded, that is something
we should address. There is a lot of need there.
Based on what we’ve said, will this model serve us
with other spec’s in our pipeline. Will these
principles work for AT-SPI or accessible doc?
Bill- I’m not sure that it would. It’s a different
kind of a spec, that’s an API ABI spec. This is a
functional spec (partly at least)
Janina - so I’m looking for the wrong template?
Bill- the model where allowing conformance where
certain parts of the spec aren’t relevant to the
product is valid for functional specs. I’m not sure
how valid they will be to ABI specs. Part of the
purpose of ABI spec is to insure that clients can
work. If someone wrote a AT that required a
particular API library to be present, and you removed
the library or modified it in some way, then the app
wouldn’t run so you lose the benefit. Certification
doesn’t mean your app will run. I think the intent of
ABI certification is that an app written to the ABI
will run.
Janina- that’s enough to get started. I’ll volunteer
to start on that with Earl. We will let you know when
the FSG meeting occurs. Can we skip next week? Okay
our next meeting will be on the 15th. Thanks very
much everyone.
____________________________________________________
Yahoo! Sports
Rekindle the Rivalries. Sign up for Fantasy Football
http://football.fantasysports.yahoo.com
-------------- next part --------------
June 1, 2005 Accessibility meeting
Attendees:
Gunnar Schmidt
Olaf Schmidt
Bill Hanaman
Janina Sajka
Matthew Wilcox
John Goldthwaite
Larry Weiss
George Kraft
Pete Burnet
Andreas Gonzales
Earl Johnson
Recording the meetings starting today. The transcriber will have a password protected audio file
to use. Please still speak freely. We will test it out today.
George posted minutes to the main list this morning. Are there any additional corrections?
Andreas- I haven't had a chance to read them yet.
Janina- We will hold approval until all have had opportunity to review them. If no objections
within 48 hours, they will be considered approved and will be posted. We had a good session
last week. I'm sorry we didn't have our recording technology working last week. Any
announcements from anyone? I have one, there is activity on the ISO working group which as
proposed a gap analysis. That was a bit of a surprise. I have posted the documents on our site
and there are additional documents on their website. If you need that url let me know.
Bill- do we have a contact with that workgroup?
Janina- I am the formal contact and it is open for any of us to participate. I believe I will attend
their meeting in Toronto in July.
Bill- the current state of things is not documented well enough for anyone outside the immediate
developer community to do a gap analysis at least on Gnome technologies, the free desktop
technology, right now. Especially if they do it without contacting us. I;d have concerns that they
may be missing things or making assumptions.
Janina- I will do my best to forward the communications I get from them. There has not been
much the last two weeks. They have set up a website where they are posting their documents.
There is a new document app. I will make sure that message gets carried through and they don't
come up with some standards without consulting with us.
Bill- even if they come up with a gap analysis without consulting us, it will be at best confusing.
If they are in agreement you could draw some confidence from that but if they divergent it could
be confusing.
Janina- If they are divergent there needs to be some resolution and we need to be part of that
resolution. The outcome needs to be something we can support in free and open environments.
Earl- is this the INCITS accessibility study group headed by Connie Myers?
Janina- Not INCITS, it is ISO JTC 1. It is the one headed by Connie Myers.
Earl- I think Peter may be attending those occasionally, you may want to check in with Peter to
see.
Bill- Peter isn't with us tonight is he? I think he is nearly in my time zone, wasn't sure he'd dial
in..
George- Who in the FSG is in a national body to be in this conversation? if you are in the
national body you can review and comment.
Janina- asside from us?
George- who from FSG is participating and which national body are they in?
Janina- I am not aware that there was anyone from FSG that was at the Sheffield meeting other
than myself.
George- so when they come up with their gap analysis at the next meetings, if you are in a
national body you can interject comments that will be put into the record.
Janina- we can do that directly as FSG. we can comment as FSG. After the Singapore meeting,
FSG can propose standards directly to JTC1 for ISO certification That was what Scott McNeil
was doing before he left FSG. I was told I could come to the meeting as a member of a national
body or as FSG and I chose FSG
.
Bill- if we are productive in our current gap analysis activity, we should let ISO know we are
carrying on a gap analysis on existing APIs and feed that information into their process. to JTC.
Olaf - is that gap analysis on the APIs or is it on what is available for free desktop.
Earl- I thought it was a gap analysis about accessibility standards and not API
Olaf- there is also a lack of information about which applications available etc. Sometime people
see only AT written with a particular tootkit. While, in fact the different solutions are in fact
complementary. Good idea to document that. If they are looking at API that's different.
Janina- I believe they are at a level above API at this point. They are more at the level of Section
508 lets you have scripted web content while the W3C is noticably silent and that is a gap and we
need some consensus.
Bill- then it's a harmonization activity?
Janina- yes, from a desire to make sure we don't have multiple standards. It's to make sure that
accessibility is planet wide and not different in various political areas. Industry doesn't want to
meet different spec's in each political region. How far they go with that is that is another matter.
They may want to adopt some of our spec's as a model. The group is new, it has only had one
meeting. The next will be at the University of Ontario, Jutta Treviranus will be hosting.
Back to the political point of being a member of a national body. There is substantial value to
attending other than in a national body. Thanks to Scott for getting FSG into the ISO track. We
can attend as FSG and not as part of a national body. You are limited as a member of a national
body as to when you can speak and about what you can say. There is assumption that the
members of a national body will come to a consensus before speaking in the international
meeting. So we have a great deal more freedom if we attend as FSG. I'd invite anyone from the
group to attend. Any other news?
Earl- On the keyboard spec the goal was to have it completed in June. I'm working on adding
the test assertions and reformating. The June date won't be met but we should be complete by
the end of July. We're looking at July for the finalization of the spec. I'll be on vacation from
next week until July 11. I'll be working on it but its taking longer than I anticipated.
Janina- better to get it right than to meet a particular date. We have some outstanding issues - we
owe NSF a report and we still need to complete the transcripts. We hope to have the transcripts
in June. We may be able to spend some additional funds to attend meetings.
We had two deep dives with Bill last week and the week before on AT SPI and the support of
complex rich documents. What are the next steps? We are going to have more complex doc's
and ever richer media as we move into the future and we need them to be accessible. Its better to
build it in than to add it on as Peter Korn said. Where do we stand with that, what are people's
reactions? Olaf I know you had some questions that you posted on the list. Anyone want to take
a stab at it?
Bill- If it is the case that we have visited a significant part of the issues around documents, we
could propose a strawman solution for the issue in the aT-SPI context. I'd be willing to propose a
set of extensions. Not necessarily to advocate them but to use to discuss them as to whether they
are adequate or necessary. It might not be too early to start drafting a counter proposal. Working
from there we could see whether to move forward and on what time frame.
Olaf- I think it is also a good idea. Just a couple of points, we have what might be useful to be
extented. We can discuss whether it needs to be extended.
Bill- I everyone agrees, I would take an action to create an extension to AT SPI to address some
of the issues that have arisen. .. Will address as many as possible, not decide which issues are
valid but to get them started for discussion.
Janina- good, what Time frame?
Bill- I think I could have a draft proposal by our next meeting. I'm not envisioning it being large
or complex.
Janina- since we are in agreement, do we have an idea of how we will test whether this meets the
needs for complex documents? We might want to test that this really meets the needs for
complex documents. It seems reasonable that we would want to test before we commit to a
standard How do we test, we'd need to find several sample complex documents. Need to look at
them from several view points, from several user needs, kinds of technologies involved. See it
these are working well enough with AT, that we'd have reasonable expectation that it is going to
work.
Bill- In the case of capabilities, if something is expressible as an API, and how expressible. we
may have to do that as a paper exercise, may not reasonable to construct software, to demonstrate
it. Where there are performance issues, we will have no choice but to do some benchmarking.
We have to demonstrate that there is a performance problem and then quantify it. On other
fronts, it could be a paper exercise.
Olaf- I very much agree with that. If we are looking at what needs to be extended, then we don't
want to lose these points in the discussion. its nothing we need to have working solution to have
a complete solution.
Janina- will come back to this topic with Bill's straw man. If takes too much time we may need
to have a separate call. I'll be on a plane at this time next week
Bill- this may be best for the mail list for review and discussion. For problem solving, its nice to
be in realtime but for this the mail list is okay.
Janina- We will be grappling with certification, will need to have some comments about the same
time we release the keyboard spec. Maybe not then, but certainly by the time of the second spec.
Bill- when we publish the first draft, we should provide some guidance for readers.
Janina- Earl- can you see who as been cc'd on the FSG email. We've had a response from the
FSG board regarding our request for guidance. They are not first time problems, they were
encountered by LSB and internationalization. We were looking for guidance so we could respond
in similar way. It turns out that while it has been addressed in the past, they are currently
reviewing it. It is an issue that needs to be carefully stated- We're providing a level of service
for accessibility, we'd add to it as will work on the roadmap. Jim Zemlin has returned from
China and has requested that Janet Sun set up a meeting including Janina, Earl, Janet Sun,
Earl- Arthur Tide, Amanda McPherson are the other people
Janina- Amanda is a PR person. Does any know him?
Matthew- I remember Mr. Tide from Linux Care? and I'm unhappy that FSG has chosen to hire
him.
Janina- The topic is open, date of meeting will be at earliest next Friday. It seemed most
meaningful to start here. We won't take any action without discussion with the group. Would
like full consensus on this topic. Do we want to brainstorm on this? We need to have the
meeting and have approval for this language from beyond this group.
Earl- Did you say that they don't have any precedent for this situation?
Janina- I was surprised to read that. They said that they were re-writing their processes so they
don't have one right now. We need some rough text. That puts in the situation of who gets to
write first. I'm will to take a first stab. It seemed useful to have some ideas from FSG. What
does it mean to have a keyboard standard that ISV's can test against.
Earl- that requires manual as opposed to automatic testing.
Bill- key thing is why are we doing; what is implied by conformance. Our mission gives use a
broad idea of what we are doing and why. But we have always had a idea of being circumscribed
by issues of practicality. We need to justify the trade off and make the limitation of the
specification explicit.
Janina- we want to be clear for people that read it so that they don't get unrealistic expectations.
We don't want to demean or sell short what the standards actually does provide. But it is critical
that users understand the what is necessary for accessibility in the OS and application.
Bill- One of concerns raised in the keyboard accessibility group when we have made decisions
that have limited our scope or opted to not provide a particular thing, the concern is that by
certifying to a low standard, a higher standard will be impaired or prevented.
Earl- the implementer will stop at the lower standard. it's a minimum but the state of the art is
further along.
Bill- state of the art is too strong, best practice is possibly even too strong. Don't know how it
works with FSG but one thing we want to prevent is the abuse of the certification process by
vendors overstating their case. Because we have certified to FSG 1.0, we provide complete
keyboard accessability. It may be that part of the conditions for certification would include
limitations on the language that could be used about certification. Condition that you don't make
invalid claims based on the certification. There should be limits on what the vendors can say
based on certification.
Janina- the service provided are by conforming the certification are essential to keyboard
accessibility but do not guareenty that any particular application on this distribution utilizes it
appropriately.
Bill- If we are concerned about abuse or misinterpretation, it would be worth it to see how the
FSG handles this in other areas. Are there limits to what the vendors can claim based on the
certification. Are there ways to limit vendors over excessive self-aggrandizement?
Janina- W3C is an example with 3 levels. Most sites are going with level 1 and Section 508 and
not going on to higher levels.
Bill- these are moving targets. If we as a group try to specify end to end accessibility, I don't
think we'll publish a spec or validation suite.
Earl- there is missing technology that will need to be developed.. Achievability is an aspect of
what we have been doing
Bill- If we plan for the spec to grow, Version 2.0 will require more features. There is the
potential for providing continually improving accessability.
George- example- FSG will say we are creating an access standard - Keyboard access 1.0. We as
a workgroup define what it means to be keyboard accessible 1.0, what it doesn't mean. If some
marketing person says xxx and gets it wrong, you send them a polite note, it doesn't mean that.
If they continue you slash dot them. One xx some marketing person said y and we slash dot'd
them.
Bill- Whether slash dotting it is possibly the most effective thing, I don't know.
George- We did that in the LSB. We were trying to get out the 1.0 and we said that noone can
claim conformance until these criteria are met. Some marketing person jumped the gun and we
made them retract their statement. And said if you don't retract your statement we're going to
slash dot you. It was in the best interest of their product to retract the statement.
Bill- What do you think about my suggestion as a final recourse? That one's certification gets
pulled for misuse. Is that a notion that has any precedence?
George- Currently, the certification gets pulled for non-compliance. If you don't pass a test or
someone finds you don't meet the spec that wasn't in the test case. You have some many days to
fix the problem or to challenge the specification. There is a process to do that. I haven't seen
anything about a certification being pulled for someone making extra claims.
Bill- does anyone think this is useful to explore?
Janina- I think you are getting to the right stuff, Bill.
George- When you get to the claims, you'd have to be very explicit about what they can say in
claims; these are the marketing claims you can make, if they make other claims you'd have
recourse against them. If you did not explicitly say what the claims were, legally they would not.
If you iterated all the claims, you could get it into the certification. I haven;t seen it happen
before
Bill- I know it would be an additional complication for us.
Earl- is certification a black and white thing or are there levels like W3C?
George- it is black and white; if there is a spec either you meet it or you don't. If you had sub-
levels, they would have to be different product standards. It costs the FSG to administrate each
product. The board would not want to have multiple levels since it would have more
administrative costs. The board would frown on it. In Posix testing, there things, if there are
things that don't exist in the system you can't test but you can still pass. I think the FSG can
handle that, said I'm compliant as long as its on the system. In my presentation in Hawaii, I
showed the 9 different failure methods. Its up to the workgroup and FSG as to whether to pass it.
Earl - is there a risk then. I thought we could specify more than might be supported by current
technology. If they say they don't have the technology on the platform, they could say the we
don't have a keyboard on our product therefore we don't have to support XKB. Because its not
there they can pass the certification.
Gunnar- it would pass if your system has no support for keyboard. But, once you have a
keyboard, you have to do the testing.
Janina- that is an interesting question, a cell phone is going to have keyboard with maybe 12
keys. But it has no shift keys that could be used press 5 times to turn on sticky keys. Yet you
might use some other combination of keystokes to activate sticky keys. Do we accommodate
that and let it pass?
Earl - no, at the moment we don't accommodate that
Bill- at the moment we accommodate systems that don't pointers or mice. But that's the only
explicit thing we make conditional. It's the only place where the system provide bla bla
Olaf- it would be possible to change the keyboard spec so that it would be valid for a PDA or
phone if people can use a combination of keys to turn on keys, then there must be a key for
turning on sticky keys.
Bill- if we did that, we would in effect weaken the spec for desktops. It means that the familiarity
of the current gestures wouldn't be guarreented which is important for users. There is always a
trade off.
Olaf- if it was a standard keyboard you could say, ..
Bill- if a system has a shift key then hitting it 5 times has to starts sticky keys
George- That's true when you come up with another product standard and you could call it
imbedded accessability, you could define a subset for imbedded accessibility for devices that
have
Janina- I hope we get to imbedded, that is something we should address. There is a lot of need
there. Based on what we've said, will this model serve us with other spec's in our pipeline. Will
these principles work for AT-SPI or accessible doc?
Bill- I'm not sure that it would. It's a different kind of a spec, that's an API ABI spec. This is a
functional spec (partly at least)
Janina - so I'm looking for the wrong template?
Bill- the model where allowing conformance where certain parts of the spec aren't relevant to the
product is valid for functional specs. I'm not sure how valid they will be to ABI specs. Part of
the purpose of ABI spec is to insure that clients can work. If someone wrote a AT that required
a particular API library to be present, and you removed the library or modified it in some way,
then the app wouldn't run so you lose the benefit. Certification doesn't mean your app will run. I
think the intent of ABI certification is that an app written to the ABI will run.
Janina- that's enough to get started. I'll volunteer to start on that with Earl. We will let you
know when the FSG meeting occurs. Can we skip next week? Okay our next meeting will be on
the 15th. Thanks very much everyone.
More information about the Accessibility
mailing list