Meeting minutes
Review of extended descriptions explainer:
Avneesh Singh: We are writing the explainer document to discuss with APA and TAG why we need semantics for extended descriptions.
<Avneesh Singh> https://
Avneesh Singh: added a PR to add: highlight HTML spec and ARIA principle that semantics should be able to be programmaticly determined.
… its a shortcoming in specifications from the perspective of estabilished principles.
GeorgeK: opened a PR and viewed Preview in Markdown.
… uses: mathematical notation, we should remove this.
… most cases MathML will be the way info is presented. there may be mathematical notation may be described but we want these extended descriptions, removing it should not diminish the strength of this explainer, and it will prevent a distraction.
Jonas: I need to look at the latest version, but in the last meeting this explainer only recognizes one technique for extended descriptions using a link with aria-details, but best practices for extended descriptions and the other technique having the desc. in the reading order in an aside or some other way. this explainer should acknowledge there is more than one way to add ext. desc.
Avneesh Singh: the DAISY document is focused on best practices for ext desc. This document is to highlight more of the future standards work and semantics.
Jonas: the way the future standards are formulated may be influenced by this document, and reading systems support for one model the other models may not get support that would be beneficial.
Simon_M: It is great reading this doc. real guidance and this is a logical way. any document about this but the alt retains gives enough information that the user may not need the ext. desc. the Patent example looks like its duplicated, and missing a close image tag.
<Gregorio Pellegrino> https://
GeorgeK: Purpose is to get support W3C to adding these ARIA attributes. the ext. desc. best practices which we will talk about in the next week TIES call has the aside and details approach explained.
… those should remain, and once we have a draft of ARIA with the new semantics we will update our best practices document to reflect this.
… not sure when we could add this new semantics if ARIA new semantics is only in a draft state.
Gregorio Pellegrino: there is some confusion with ext. desc. with two groups and this semantics could take years to arrive. this is not already here. but the Daisy's Best Practice document and is how to implement ext. desc. Right now with currently approved valid techniques.
Matt Garrish: Don't jump the gun. we need confirmation first in ARIA so wait until it is solid and in a Rec track.
Jonas: Good to hear if we get the semantics we can use it with different models/techniques/elements for the markup. since we use the aside model in the Nordic region.
… we missing an option to hide / skip image descriptions if we don't have these semantics.
Gregorio Pellegrino: the overall Note is fine, but there are some small things. we can go through the list? Maybe larger issues to discuss...
… should we stress more in user section managing alt text we have a standardize way but we don't have this in the ext. desc. we may stress math more, We have this concept of ext. desc. but no standardize way to manage it?
… technically we have no semantics for ext. desc. having the roles is what we need.
Avneesh Singh: this section is trying to say exactly that. After Gautier has done the next revision, you can go through the text to try to highlight this more.
GeorgeK: after we merge gautierchomel_'s PR
Jonas: useful to mention the aria-details would point directly to an extended description and not only to a link.
… which are the two main options in the best practices document.
… i think it is worth pointing out.
GeorgeK: I agree. if you are doing this on the web you will probably be pointing to a details element
… same tech. could be applied to an aside, we could use that in the DAISY world but might not be appropriate for the web.
Avneesh Singh: lets incorporate this and work on the next revision.
Work on low hanging fruits for FXL techniques
Avneesh Singh: Fixed layout doc. we broke it into 2, future and more practical work that we can do now.
… It would be good to know what we can do in the FXL Technique document, next steps.
<Avneesh Singh> https://
Gregorio Pellegrino: I read these, showed screen. looking at Single A
… I don't see anything that is problematic for FXL. there may be some more explanations needed but is not really an issue.
… the SC that we need to explain, 1.1 Text Alts and especially images split on mult. pages. and 1.3.2 meaningful sequence and we have reading order across the fold etc. and SC 2.4.2 Page Title we may need to provide guidance on how to add meaningful title of these types of pages.
… req. to page breaks/nav we may need to give guidance on how to manage page breaks in FXL and some cases they may not be needed but needed in others.
… if we look at Fixed Layout EPUB 1.1 WCAG 2.2 (Single A) not really problemantic, and we may need to add 5 additional guidance.
… but if we go to AA there are more serious concerns which will be difficult for FXL.
… there are 6 SC that are problematic, 1.3.4 Orientation (not block orientation in metadata) landscape. 1.4.4 Resize text is not possible in FXL. 1.4.5 images of text we have some info in the document, and 1.4.10 Reflow is also not possible, and 1.4.12 Text spacing is not possible in FXL, and if we want to get into 2.5.7 Dragging Movements is also an issue in FXL.
… my proposal would be to reorganize the documents those SC that are problematic without retelling common things like we don't discuss how to tag headings, tables etc.
Gregorio Pellegrino: given this, question 1: which level of WCAG for this document A and AA or just A.
… Question 2: main focus should be on content created with HTML/CSS, or SVG for Fixed Layout
<Charles: In our GCA a publisher had us review their fixed layout. Our focus was single A and we were able to get their title to pass.
<GeorgeK> The low hanging fruit would be single A
<GeorgeK> We should address AA at a future time.
Charles: GCA had a publisher Amazon who passed their Fixed Layout workflow at WCAG 2.2 A. I think we should focus on A only for this document.
Ken Jones: Target A+ close to AA as possible except for those that is impossible currently.
… Its not an end user how to do it, this is how it could be archived, this is what we want to end up with.
… I have solved some of these issues with my own software.
Wendy Reid: We are under-estimating the number of WCAG AA. Reflow and Text spacing but the rest should be able to be met. There are a few that are possibly problematic ie. orientation. Images of Text, not text contrast can be met.
… We almost want to say you can does these with the exception of these small handful of AA requirements.
Jonas: the EAA requires AA, conformance to A wouldn't help our publishers. conformance to AA is normal in the world.
Simon_M: AA is often interpreted as a regulator req. we are here to maximize the a11y here.
… paring this with a EAA req. ie. convert to Reflow.
… A+ as Ken said.
Ken Jones: What about Fixed Layout AA?
GeorgeK: WCAG folks wouldn't like it.
Avneesh Singh: We want to have techniques for A and AA and for those AA where we can't meet the success criteria we clearly say this is a challenge. lets start with Level A and then start working on the AA. Restructuring the document to align with the EPUB 1.1 techniques. we structure it in a Success Criteria kind of layout. What are the thoughts on 2nd question, should we have HTML/CSS or should also have SVG options?
GeorgeK: targeting A and the AA which can be achieved. Not sure how we label that. we should solidly reach A and most of the AA. Does this get us into Manga/Comics.
Gregorio Pellegrino: Comics and manga are only images one per page so another way which is nearly impossible to make accessible.
Avneesh Singh: Level A + HTML/CSS can be achieved in the next step. We can then step into level AA. I want to move step wise and achieve some success, I Don'T want this to take 3 years to achieve.
Ken Jones: We may also say this may be achieved in SVG but can set that for a future.
Wendy Reid: I think we could restructure that to align with the EPUB 1.1 techniques.
Avneesh Singh: the EPUB accessibility 1.1 techniques quite a lot refer to the WCAG techniques, but we may need to work on some of these for FXL because WCAG focuses on reflowable. For example, there are PDF techniques for WCAG. So we may need to add specific techniques for FXL for meeting WCAG SC (which WCAG techniques are already addressing for reflowable).
Simon_M: Do we need to worry about WCAG3?
Wendy Reid: WCAG3 and their req. and guidelines there is a new draft coming out. Nothing has changed dramatically for EPUB. lot are similar / organized differently.
Matt Garrish: they are rewriting and the conformance model is still being spec'd out. WCAG 3 is years out.
Simon_M: If we keep this tuned to focus not specific to WCAG.
Matt Garrish: Whats the best case / hope to achieve is our goal with this document.
Avneesh Singh: Wendy will start with this initial restructuring the document to align with the EPUB 1.1 techniques document and WCAG SC. We can review and provide feedback when first pass is done
… Thanks all, great start to 2026