– DRAFT –
PDF-AAM Teleconference

Substructures for figure tags

Paul: The image would go inside the figure and the image gets the alt text
… But it doesn't read anything that's inside the figure text
… In PDF 2.0 there are scenarios where it'd be semantically appropriate to have substructures
… We would need both the caption and the alt text to be read
… There could be nested figures

zkinsey: We are curious how these tables in the AAM would look in these instances?

spectranaut_: You could format this info however it makes sense
… I feel that there is conceptual difficulty here, it would be nice to have examples of this
… If something has an alternative name, wouldn't it be read?

Zakim: That's AT behavior that we are observing
… We are trying to write things up to mimic current implementations

spectranaut_: In the web there are things that can be named, not ignored, in the subtrees, and there are others that cannot
… It'd be really nice to have more examples of the kind of structure trees that are causing problems for you to try to map them to similar HTML paradigms

Export 2

Zakim: We are working on this, figure is the main topic right now
… I'd like to discuss in today's call to start creating a parent issue to group these concerns for the second batch
… What tasks do we need to do to the index.html?
… We have our intro sections that shouldd be transferred to the HTML docs

<spectranaut_> https://w3c.github.io/pdf-aam/

Daniel: We should add the ED link to the PDF-AAM readme

<spectranaut_> There is a link under About: w3c/pdf-aam

zkinsey: What are next steps to do?

Daniel: You can open as many PRs as you want but they should be focused

spectranaut_: There are previews from netlify

zkinsey: Once I run my convertion from google sheets to html, then we may need feedback even before opening these PRs

spectranaut_: No worries,if you have a proposal it's easiest to open a PR

zkinsey: I'll try to push these as PRs

spectranaut_: IF you could make available some information about the problems with the naming things before next week that would be very much appreciated

zkinsey: We don't necessarily see a naming issue, it's more the parent/child relation
… What the naming convention in the AAM would look like

Paul: One is figure containing captions, that's the new rulee in PDF 2.0

zkinsey: There are disparities between 1.7 and 2.0. Which version should we be targetting for the first release?

spectranaut_: You could just have two role mapping tables and then subheadings for 1.7 and 2.0. Not everything changed

zkinsey: 75% of the structure are the same between these two versions. It's just some of them that we added in 2.0 or deprecated

Paul: I think it makes sense to have one section for 1.7 and another section for 2.0

zkinsey: Or maybe added rows

Zakim: For now let's stop ignoring 2.0 and discuss later whether we use different tables or sectionss etc

Paul: +1

Meeting duration

zkinsey: Let's first do the second export and then we may ask for an hour meeting
… There are issues for now when it's not known how certain APIs are handling the mappings

spectranaut_: We should discuss these ones with the ARIA WG, which has expertise in these APIs