Matthias Clasen has written a post on the GTK blog about recent changes in the accessibility implementation of the toolkit, as well as improvements in the tooling planned for the 4.12 release: https://blog.gtk.org/2023/06/21/evolving-accessibility/
It's important to note that more work is needed to improve the accessibility stack on Linux; if you want to contribute you can read the at-spi2-core developers documentation for an overview of the changes to the stack: https://gnome.pages.gitlab.gnome.org/at-spi2-core/devel-docs/index.html
We especially need more documentation on how ATs interpret the accessibility interfaces, and what kind of expectations should be satisfied by toolkits and apps.
@GTK I don't see any overview on this page though...
@emersion For an actual overview of the stack, the readme for at-spi2-core has a good outline with links: https://gitlab.gnome.org/GNOME/at-spi2-core/-/blob/main/README.md
> It is no secret that we would like to see some modernization of the AT-SPI D-Bus APIs.
I'm curious about what the GTK team, and other GNOME teams if applicable, would like to see here. Has the wish list been documented anywhere?
@matt @federicomena The at-spi2-core devel docs have a page on it: https://gnome.pages.gitlab.gnome.org/at-spi2-core/devel-docs/at-spi3.html
In general: more bulk operations, especially at setup time; fewer roundtrips; and fewer index-based methods. The D-Bus API doesn't use property notifications, so it's not very efficient for bulk changes. Actions should be parametrised, to eliminate a bunch of methods that are currently used for automation.
@matt @GTK https://gnome.pages.gitlab.gnome.org/at-spi2-core/devel-docs/at-spi3.html has a bit of this, but there's no "new protocol" document yet, AFAIK.
It may be useful to look at bug lists to see if there's anything worth distilling into a document. I think we can have a call-for-ideas from the people who have been working on this for a while; I have a feeling that there are a lot of good ideas in people's minds, but there has not been a push to write them down - because there was little hope of implementing them soon.