- From: Ben Caldwell <caldwell@trace.wisc.edu>
- Date: Tue, 24 Oct 2006 09:38:03 -0500
- To: public-wcag-teama@w3.org, public-wcag-teamb@w3.org, public-wcag-teamc@w3.org
- Message-ID: <453E254B.9020801@trace.wisc.edu>
There are 5 issues related to aggregation. Many of them have draft
responses, but require a decision from the group about whether (and if
so, how) conformance claims can be made for content that includes
aggregate parts or user-contributed content.
Here are the issues:
*Comment LC-464*
*Sort Terms:* AGGREGATED - should be exempt
*Document:* WCAG 2.0 Guidelines
*Submitter:* Jason White <jasonw@ariel.its.unimelb.edu.au>
*Comment Type:* substantive
*Location:*
*Comment:*
Item Number: Related Documents
Part of Item:
Comment Type: TE
Comment (Including rationale for any proposed change):
Conformance: Aggregated Content.
If content contains authored units that do not themselves carry any
conformance claims, and those authored units are modified or substituted
as a result of an aggregation process, then the conformance status of
those authored units is unknown at any point in time unless individual
assessments are carried out. Such assessments may be impractical, for
example on sites that collect comments from the public, maintain e-mail
archives, etc.
As the guidelines are currently drafted, the conformance of any Web unit
containing such authored units depends in turn on the conformance of
those authored units, which may vary over time. In order to avoid making
false conformance claims, the operator of such a Web site would,
presumably, have to exclude such Web units from the scope of any
conformance claim, in accordance with the scoping provisions of the
conformance section. I think this consequence needs to be clarified and
stated explicitly.
Alternatively, the scoping provisions could be modified to allow
individual authored units to be excluded from the ambit of a claim, but
in that case it is by no means clear how the "authored units" could be
precisely identified and specified in the claim.
Proposed Change:
Clarify that if it is unknown whether an authored unit participating in
aggregation conforms to WCAG 2.0, or which level of conformance is
achieved, then it is likewise unknown what, if any, level of conformance
is attained by Web units in which it appears. Implementors should be
advised to exclude Web units containing such "unknown" authored units
from the scope of any conformance claim in accordance with the "scoping"
provisions of the conformance section of WCAG 2.0.
Note that by controlling what may appear in authored units participating
in the aggregation process, through technical or other means, it may be
possible to ensure that a given level of conformance is always
satisfied. Under these circumstances (where the conformance of resulting
Web units is guaranteed), conformance claims with respect to such
aggregated content may reliably be made.
*Status:* open
*Working Group Notes:* [TEAMA] [HOLD]
Discussed in the 16 May 2006 Team A call, proposal sent to list:
<http://lists.w3.org/Archives/Public/w3c-wai-gl/2006AprJun/0189.html>
<http://lists.w3.org/Archives/Public/w3c-wai-gl/2006AprJun/0189.html%3E>;
suggestions from WG Call
Author can also choose to provide "fail-safe" alternative content for
included content of unknown conformance, just in case it doesn't comply.
But overall, yes, authors should choose to include content of known or
knowable conformance if they wish to claim conformance for a page.
Also, this overlaps with user input issue and needs to be reworded so
that it doesnt cover that by mistake.
OLD PROPOSAL {needs rewriting - see above} (do not accept)
This comment is based on supposition that if content includes parts that
have unknown accessibility, the author should not be responsible. This
is not correct. The page either conforms or doesn't. The fact that parts
change means the authors need to ensure that changing parts are
accessible if they want to claim the web unit is accessible, not that
they could scope them out and still claim access. The guidelines are
written to reflect this, and that is the intent of the working group.
5/25/06 telcon
additional comments to consider in resolving this item (from survey)
- We may want to be sure that our conformance section makes this clear.
- I believe this is the issue that Alex Li raised a couple of weeks ago.
Philosophically, I agree that a page cannot claim conformance if it
contains authored units for which the conformance level is not known.
However, this would mean that any Web page that contains a text entry
field (such as this one) would not be able to claim conformance. In
particular, 1.3.1, 2.4.4, and 3.1.2 would be difficult, if not
impossible to meet.
We can't just say you just don't claim conformance for those pages. For
one thing, we want developers to make the part they do control conform
but they may not be motivated to do that if they can't, in the end,
claim conformance. For another thing, if a government has a policy
requiring conformance with WCAG Levels 1 and 2, does that mean that
government websites cannot have pages with user contributed content such
as this one?
- I think the other side of ths issue is that if we cave into the
rational that corporations will not claim conformance, we may also
notice that other competing standards are already emerging (ie European
Union and the UK crowd) and are being adopted (or being considered)
because they *do* require acessibility. I think various filtering
techniques and algorithmes could overcome the issue of text based
submitted comments like this page. I don't think this kind of text based
filtering and fixing would be that difficult for large companies and
governments that are on the cutting edge of technology. Perhaps even the
rules for "text based documents" could be applied. If we let this go I
think we are going to get scorched by the disability community and by
the accessibility community and I am concerned our GL will be supplanted
by other standards that at much more demandng than ours. I think we've
all worked too hard and long to allow for that to happen.
*Resolution Working Notes - Unapproved:*
{parital accept}
@@ (same action for 1222) Add the following paragraph to the Section on
Aggregated content in Conformance:
By controlling what may appear in aggregated authored units, it may be
possible to ensure that a given level of conformance is always
satisfied. Under these circumstances (where the conformance of resulting
Web units is guaranteed), conformance claims with respect to such
aggregated content may reliably be made.
@@ respond with
We have added the following paragraph to the section on aggregated content:
"By controlling what may appear in aggregated authored units, it may be
possible to ensure that a given level of conformance is always
satisfied. Under these circumstances (where the conformance of resulting
Web units is guaranteed), conformance claims with respect to such
aggregated content may reliably be made."
However, this comment seems to be based on supposition that if content
includes parts that have unknown accessibility, then the author should
not be responsible. This is not the position of the Working Group.
Content either conforms or doesn't. The fact that parts change means the
authors need to ensure that changing parts are accessible after they
change if they want to claim the Web unit is accessible (note that they
could scope them out and still claim access). The guidelines are written
to reflect this.
Authors may also choose to moderate or otherwise monitor aggregated or
contributed content for conformance to WCAG to ensure conformance.
*Related Issues:*
1222 <http://trace.wisc.edu/bugzilla_wcag/show_bug.cgi?id=1222>
*Assigned To:* Nobody
*Last Edited:* 2006-08-08 20:16:11
**
------------------------------------------------------------------------
*Comment LC-1168*
*Sort Terms:* Blinking aggregated
*Document:* WCAG 2.0 Guidelines
*Submitter:* Greg Lowney
<gcl-0039@access-research.org> *Affiliation:* Lowney Access
Research, LLC
*Comment Type:* substantive
*Location:* time-limits-blink
<http://www.w3.org/TR/WCAG20/complete.html#time-limits-blink> (Description)
*Comment:*
2.2.2 requires content to not blink for more than three seconds or a
method be available to stop all blinking in the Web unit or authored
component. Many Web sites display GIF images that are provided by a
third-party (e.g. advertisements, or user-contributed photos); are such
sites required to ensure that none of those are animated GIFs, in case
some blink? Is it sufficient for the authors to define a baseline that
includes user agents that allow the user to stop blinking on the current
Web unit (e.g. pressing ESC)?
*Status:* open
*Working Group Notes:* [TEAMC] []
baselines are technologies not user agents.
The comment about third-party ads seems related to LC-1318.
*Resolution Working Notes - Unapproved:*
*Related Issues:*
1318 <http://trace.wisc.edu/bugzilla_wcag/show_bug.cgi?id=1318>
*Assigned To:* Nobody
*Last Edited:* 2006-09-14 05:14:39
**
------------------------------------------------------------------------
*Comment LC-1222*
*Sort Terms:* AGGREGATED CONFORMANCE
*Document:* WCAG 2.0 Guidelines
*Submitter:* Masafumi Nakane <max@wide.ad.jp> *Affiliation:* Keio
University
*Comment Type:* general comment
*Location:* conformance
<http://www.w3.org/TR/WCAG20/complete.html#conformance> (Aggregated Content)
*Comment:*
Will these examples conform to WCAG? 1. A perfectly conforming blogsite,
with comment form, and comments posted by someone breaks theconformance;
2. The page is perfectly conforming, except for a fragmentof code which
was provided by affiliate program of a shopping site, andthe user is not
allowed to modify the code either for technical reasonsor for legal
reasons. From the current wording of WCAG, these exampleslook to be not
conforming to WCAG. This can limit the creativity andvariety in ideas
for Web content if authors are keen to meet the WCAG,or else, let many
authors decide not to care too much about theconformance.
Proposed Change:
Limit the responsibility of the content creator to those Web units
thatare authored by the creators themselves.
*Status:* open
*Working Group Notes:* [TEAMA] [HOLD]
*Resolution Working Notes - Unapproved:*
{not accept}
@@ see 1222 for action items
@@ Respond with:
We have added the following paragraph to the section on aggregated content:
"By controlling what may appear in aggregated authored units, it may be
possible to ensure that a given level of conformance is always
satisfied. Under these circumstances (where the conformance of resulting
Web units is guaranteed), conformance claims with respect to such
aggregated content may reliably be made."
However, the suggestion that if content includes parts that have unknown
accessibility, then the author should not be responsible is not the
position of the Working Group. Content either conforms or doesn't. The
fact that parts change means the authors need to ensure that changing
parts are accessible after they change if they want to claim the Web
unit is accessible (note that they could scope them out and still claim
access). The guidelines are written to reflect this.
Authors may also choose to moderate or otherwise monitor aggregated or
contributed content for conformance to WCAG to ensure conformance.
*Related Issues:*
464 <http://trace.wisc.edu/bugzilla_wcag/show_bug.cgi?id=464>
*Assigned To:* Nobody
*Last Edited:* 2006-08-10 22:58:06
**
------------------------------------------------------------------------
*Comment LC-1318*
*Sort Terms:* AGGREGATED
*Document:* WCAG 2.0 Guidelines
*Submitter:* Takayuki Watanabe, Makoto Ueki, and Masahiro Umegaki
<nabe@lab.twcu.ac.jp> *Affiliation:* JIS WG2
*Comment Type:* general comment
*Location:* (Aggregated Content)
*Comment:*
Comment: We think it is a good idea for WCAG 2.0 explaining how
aggregated contents conform to WCAG because of their popularity.
Aggregated contents must be considered carefully because such kinds of
content have been increasing on the web. This paragraph, however, is
difficult to understand: This paragraph deals with aggregated content,
Web unit, authored units, and aggregated (authored) units, which terms
and their differences are difficult. It is difficult to understand what
'aggregated content' means. Thus, Good examples of aggregated content,
Web unit, and authored units are needed.
In addition to that we can not understand the responsibility of Web
authors and aggregated contents. We also can not understand how authors
make a conformance claim to aggregated content.
*Status:* open
*Working Group Notes:* [TEAMA] [HOLD]
Discussed in the 10 August 2006 telecon:
Resolution: send 1318 back to Team A
http://www.w3.org/WAI/GL/2006/08/10-wai-wcag-minutes.htm
10 August survey results:
http://www.w3.org/2002/09/wbs/35422/20060810teama/results#x1318
*Resolution Working Notes - Unapproved:*
@@ add links to "Web unit," and "authored unit" in aggregation section
Respond with:
We have added the following paragraph to the section on aggregated content:
"By controlling what may appear in aggregated authored units, it may be
possible to ensure that a given level of conformance is always
satisfied. Under these circumstances (where the conformance of resulting
Web units is guaranteed), conformance claims with respect to such
aggregated content may reliably be made."
The Working Group plans to develop additional materials related to
conformance in conjunction with the education and outreach working group
in the future.
*Related Issues:*
*Assigned To:* Nobody
*Last Edited:* 2006-08-10 22:59:20
**
------------------------------------------------------------------------
*Comment LC-1332*
*Sort Terms:* LEVEL-CHANGE CAPTIONING AGGREGATED
*Document:* WCAG 2.0 Guidelines
*Submitter:* Arun Ranganathan <arunranga@aol.com> *Affiliation:*
Advisory Committee Representative to the W3C for AOL LLC
*Comment Type:* general comment
*Location:* media-equiv
<http://www.w3.org/TR/WCAG20/complete.html#media-equiv> (Guideline 1.2
Provide synchronized)
*Comment:*
Comment (Including rationale for any proposed change):
The purpose of the following LC Comment is to highlight issues that we
believe warrant further consideration by the Web Content Accessibility
working group before assigning a Level 1 requirement to Guideline 1.2.1:
"captions are provided for prerecorded multimedia," particularly as the
WAI's Web Content Accessibility guidelines are used as the benchmark for
web accessibility by government and other policy-making bodies.
AOL LLC fully understands and supports the need for captioned
multimedia, and we do provide the same on several of our highly
trafficked areas. AOL was the first commercial Internet Service Provider
to offer captioned video. Today, we provide captions for two cartoon
series "Princess Natasha" and "SKWOD" on KOL, AOL's online channel for
kids ages 6-12, and on video help tutorials developed for the AOL 9.0
software. Additionally, we are currently testing delivery of captioned
news and entertainment content through our video portal. We continue to
work hard at this area, and plan on announcing further such developments
as they take place.
Technologies such as SMIL, Microsoft's SAMI and Apple's QuickTime all
enable display of closed captions on multimedia, and tools like Caption
Keeper from the WGBH Media Access Group can be used to repurpose Line 21
television caption data. However, AOL's research to date shows that the
acquisition process and production model for the majority of video
content distributed by commercial Internet portals such as AOL LLC does
not support cost-effective and efficient processes for delivery of
closed captions in a timely manner. A collaborative effort between the
Internet industry, content producers and web accessibility experts is
required to develop solutions before commercial web portals can fully
conform to this Level 1 requirement to caption prerecorded multimedia.
Guideline 1.2.1 assumes that the web site displaying the multimedia
content is the producer of the content. What is not considered is the
barriers created by the process of acquiring repurposed third party
content, or who is responsible for captioning content produced by a
third party and distributed via multiple web sites/services. While AOL
LLC has made substantial progress towards captioning of our video
content, there are three barriers inhibiting AOL LLC's goal of complete
conformance to a Level 1 success criteria:
i. Internet production units of broadcast networks prepare the content
for streaming before the content is captioned, usually in real time. For
example, field packages produced for TV networks' nightly newscasts are
often streamed before they air. As a result, Internet portals receive
the video asset too far up stream in the content production workflow.
This presents two possible scenarios:
- A content aggregator (Internet portal such as AOL's) needs to manually
caption a video stream produced and owned by a separate content
provider. Neither is this scalable, nor are vendor solutions robust
enough at this point (e.g leveraging a programmed transcript which only
provides the text of the audio, and excludes ambient sounds and time
stamp data).
- Captions are added to the streaming video long after it has been
published to the web site assuming the portal and partner repurpose the
captions originally created during the TV broadcast. This is problematic
as some videos have a very short shelf life.
ii. Lack of information on the whereabouts of existing caption files
when broadcast content is repurposed for the Internet. There is an
increasing amount of "video on demand" products online that allow people
to view archives of current or old TV series, movies, music videos,
short films, etc. It is very likely that most of the content has been
captioned. Unfortunately there isn't a central database that Internet
portals or content partners can search to locate the caption agency who
captioned a particular season of a show. It is important to note that
the content provider to the portal may not always be the content
producer or the entity responsible for captioning the content for
television.
iii. Need for a common delivery protocol. Commercial Internet portals
receive video from many of the same content providers (broadcast
networks, etc.). Internet production units are generally very small in
terms of staff so delivering multiple text formats to multiple portals
is not feasible. Solutions are required to ensure content providers can
deliver caption data in an efficient, cost-effective manner. This is a
solvable problem, but identifying solutions will require cooperation
from many players.
AOL LLC proposes changing the Level 1 Success Criteria for Guideline
1.2, namely 1.2.1 and 1.2.2, to Level 2 Success Criteria. This change
reflects the ground realities of being a content aggregator on the Web.
This proposal is necessitated by the current reality that content
aggregators on the Web partner with multiple content providers. Issues
such as who is responsible for producing captions, delivery of caption
text files and other barriers described above must be addressed before
policy-making bodies can effectively leverage this guideline.
Alternatively, we recommend adding language which recognizes the current
barriers to wide scale availability of captions for prerecorded
multimedia, and encourages development of solutions to resolve them.
*Status:* open
*Working Group Notes:* [TEAMA] [HOLD]
BBC: I don't think this effects anything in the Aggregated category.
*Resolution Working Notes - Unapproved:*
*Related Issues:*
*Assigned To:* Nobody
*Last Edited:* 2006-08-08 16:32:16
--
Ben Caldwell | <caldwell@trace.wisc.edu>
Trace Research and Development Center <http://trace.wisc.edu>
Received on Tuesday, 24 October 2006 14:38:50 UTC