2008-148 Minutes
PDF-UA 2008-147
November 5, 2008
PDF/UA Working Group Teleconference
PDF/UA (Universal Accessibility) Working Group Draft Meeting Agenda
Wednesday, November 5, 2008, 3:00 P.M. - 5:00 P.M. EST
Conference Call Information
International Dial-in: +1...
U.S. Dial-in: 1-888-742-8686
Conference ID: 7083945
all participating Committee members to edit the Wiki during the meeting.
Roll Call
Duff Johnson, DJ Chair Appligent Document Solutions
Peter Abrahams, PA Bloor Research
Joe Clark, JC Independent Accessibility Consultant
Cherie Ekholm, CE Microsoft
Mark Gavin, MG Appligent Document Solutions
Greg Pisocky, GP Adobe Systems
Leonard Rosenthol, LR Adobe Systems
Neil Soiffer, NS Design Science
Review PDF/UA 2008-147 Agenda
This represents a modified agenda due to conversations with Betsy Fanning.
Gavin / Soiffer
Approval of pdfua-2008-10-15-minutes
Gavin / Soiffer
Review Cherie's report from Beijing
Cherie attended the meetings in Beijing. An email was sent reporting the results.
Question raised by Neil Soiffer, all of Cherie's recommendations were approved with exception of table algorithm. 14.8.1.
Issue does not warrant it's own separate standard according to other member participants
Comments related too lack of partipation by non-English languages and some additional technical requirements.
Duff indicates that the document must be prepared to such an extent that a technical editor can work wtih it. The group is facing a deadline of November 11 for timely submission to the meeting in Germany in March.
JC proposal cannot create an interchangeable specification.
Cherie also forwarded an email from Leonard
DJ: PDF/UA will not be disassembled or become a part of 32000
Continue work
Time was alotted for a discussion of the issues raised by the ISO 32000 technical committee and detailed in Cherie's email to the listserv (see below).
The chair closed discussion further discussion concerning the issues the ISO 32000 committee had concerning the PDF/UA specification submitted for ISO-32000 consideration. Joe Clark objected to closing the discussion,
Deferring the discussion of Section on 14.8.1 after November 11, 2008
Plan required follow-up meeting in order to get PDF/UA to the Technical Editor by 11/11/08
Table Algorithm issues must now be resolved. The section had not been submitted to ISO 32000
Reviewed the section on Conformance. LR identified issues around scripts.
JC recommends review section on scripts from WAI ARIA as a reference for PDF/UA
Playback Controls should be perceivable and accessible
Wrapup
Extra meeting added on November 6, 2008 in order to meet the November 11 deadline.
Adjourn
4:50 Eastern
Gavin / Soiffer
Addendum
Cherie Ekholm's summary of the discussion of PDF/UA at the ISO 32000 meeting in Beijing
The following is feedback from the PDF 32000-2 meetings in Beijing last week. Please pardon any language oddities as I’ve been linguistically contused since arriving in Beijing on a Japan Air flight that went through customs at the same time as an Air France flight, and spent a week hanging out with Brits and a German. These are quick notes I forwarded to Duff, who requested I forward them for thought prior to Wednesday’s PDF/UA meeting.
1. Consider rolling the entirety of PDF/UA in 32000-2, or at least all except conforming reader/writer, or all except conforming writer. We should discuss in the PDF/UA committee whether there are large swaths of the PDF/UA spec that can and should be rolled into 32000-2. This would, of course, change the scope of what is being worked on by the committee and require the PDF/UA committee to revisit its charter. There are a several thoughts behind the idea of rolling most of PDF/UA into 32000-2:
a. Conforming reader requirements are growing from all the various sub-standards. There is concern that this will cause a proliferation of readers that conform to just one or two of the standards and not others, leading to confusing in as to what reader you need to have installed to read any given document.
b. Most of those items which deal with tagging are already in 32000-1 or have been proposed for 32000-2. These are generally regarded as best practices for well-formed PDF, and accessibility tagging is required by PDF/A, therefore it makes sense to have these items as part of the superset of all PDF standards rather than in a standard of its own.
c. There is already confusion among the PDF consumers about what standard is best to conform to and it is not always possible to conform to two or more of the standards. Moving PDF/UA requirements and conforming reader requirements into PDF32000-2 would prevent further confusion.
d. Having the accessibility requirements an integral part of 32000-2 strengthens the standard overall.
e. It is generally understood that PDF/UA has an important role to fill. Detailing the conforming writer in a separate document is a non-trivial work item and a vital part of that roll.
f. The 32000 committee also discussed moving conforming reader statements from other subset standards into the main 32000-2 standard, particularly those from PDF/A. In general, it's believed that the alphabet soup of PDF standards is hitting a critical mass for confusion among consumers and it would be wise for the committees to work together more effectively to minimize this.
2. Line numbers – artifacts – such as line numbering for document should be covered by the PDF/UA spec – the UK delegate had a good example where difficulties with this were encountered in work that was being done on PDF/X.
3. The 32000 committee concerned about lack of international involvement in the PDF/UA work, specifically or especially from member countries whose languages are DBCS or Complex Scripts. (Side note – I sent an email to Microsoft’s accessibility standards folks to see if they could recommend anyone either within Microsoft or without who is a subject matter expert on both accessibility and either DBCS or CS languages. So far I’ve not received any recommendations.)
4. Questions on why MathML instead of using the XML itself, resolved in discussion.
5. Sent Neil a note on how Math ML handles grouped glyphs such as ¾ vs 3 / 4. Resolved a question by the German delegation.
6. Section on 14.8.1 – clean up and resubmit turning this into normative text with less explanatory. Does not match spec. Another example, also unneeded would be code samples within a PDF. The committee could not make any sense of this in context with the rest of the PDF specification and requests that PDF/UA take this proposal up again and rewrite the normative text to be more clear, to minimize the informative examples, and to conform to the style of the PDF 32000 spec. The bevy of examples and the markedly different style of this one proposal caused confusion rather than clarification. On the other hand, one member also noted that there is also another common example: sample computer language code. It was felt that the examples are such that a reasonable person working to create a compliant reader/writer/document should be able to understand what they are without this list. Also, there was some confusion caused by the formatting in the document submitted to member countries that caused the stipulation “shall declare its natural language” to look like it is part of this same proposal, and the confusion in the wording of the non-linguistic PDF section compounded that. We need to make the distinction clear, rewrite and resubmit for consideration at the next 32000 meeting. No one said no to either of these. They simply would like them rewritten before saying yes.
7. The table algorithm is listed in the 32000 proposals. However, it didn't appear in the circulated document of proposals, so was not discussed. Also, I was going to withdraw it pending further conversation, as I need to have a discussion with the PDF/UA committee about difficulties Microsoft encountered when we looked into implementing it.
