Meeting minutes
TPAC Breakout 2025
janina: Overview
… reminder of W3C code of conduct
… link: https://
Lionel_Wolberger: Reviewed the outline structure
… and the diagram
jason: Reminded that no visible widget is needed, and source is the preferred method
… no magic
shahdi: we should think about authoring tools and user agents
… in an ideal world, the 1st cloud would be in the source
… and the 2nd cloud (user-authorized) would be in the browser
… we are not in an ideal world
… things do not happen in the authoring process, not in the browser
shahdi: do we explain that?
Shadi: IN an ideal world, anything missing in source would be added by source in a loopback
… or for user preferences, the browser (user-agent) would ask, what does the user want
jason: I pick up on the phrase you use, an ideal world
… extremely rapid world, developers not keeping up, the ease with coding up sites that look great and are not necessarily accessible
… post-source needs to be dev'd, tested and managed in the same way [and quality] of source
shadi: yardstick not only WCAG, look to ATAG and UAG
janina: We did consider linking to ATAG, we found it has not been maintained recently
… that is a normative recommendation
… UAG was never accepted normatively, it remains in published note status
… we do not get deeply into the user-agent
… top of my mind: so many things are being mashed together: payment gateway, login authentication service
… the user experience seems, to the user, as coming from a single domain, but it does not
… pages are unique
Lionel_Wolberger: How do you see monitoring of the eventual experience?
shadi: I am addressing the domain of what the user gets, the additional things that are inserted (inserted libraries)
… have the user-agent do the lifting
… this is an evergreen discussion in W3C
… UAG did not advance to normative status
… user-agents can do a lot more than they are doing today
… as a community we should think about getting more mileage out of user-agents
… "There is no right font, size or color"
… there is my particular need
… that is best done on the user side
… and most ideally done natively by the user-agent
janina: Yes, since a user wants to take the same preferences everywhere
… they may have multipel devices as well: a fire device, tablet, phone, computer
shadi: exactly
janina: We have tried, with our browsers, to address all types of disabilities
… some new approaches build for one particular audience and requirements
jamesn: user agents can handle this
… they sync preferences across devices
… [there's a limit]
… my color preferences on my phone are not necessarily what I want on my desktop
giacomo-petri: I agree, user-agent can be the heavy lifter on font size, background color, etc
… this can be done in authoring as well
jason: We would welcome user-agents "taking up the mantle"
janina: Should we reesume work on UAG?
… current link, https://
janina: How do we assert, which things can be defined, to move the industry forward
… ATAG link, https://
shadi: I did not intend for a11yedge CG to be tasked with this
… I still ahve to read the report
… I welcome this discussion
front-endian-jane: portability issue with browsers
… when an author's end installs an overlay
… is that worse?
… is there any other place where this technology can exist?
Lionel_Wolberger: State can be storedin a cookie, though that has privacy implications