My Accessibility Stack and the future on Wayland
89 points by nocoffei
89 points by nocoffei
This big and long-winded post isn't intended to punch down-- I'm genuinely looking for the right people to talk to to solve a very big and difficult problem. If you think you might be one of those people, or know someone who is, please reach out!
Also, previously on lobste.rs: https://lobste.rs/s/o0x7rb/your_mouse_free_setups -- and I'm terribly grateful to all of you for the suggestions! You all helped me discover Talon and the Svalboard in the first place.
This has been shared around KDE chatrooms. I can't promise anything, but people are aware.
I do not work on this stack, but still.. I am very sorry about this. I will try my best to make this issue more known in our spaces, so someone with the right knowledge might be able to improve the situation.
Again, I can't promise anything, but want to let you know that we hear you and definitely want to help.
Edit: any interested parties could hop on to our matrix channels:
#kde-accessibility:kde.org#kwin:kde.orgEdit2: apparently people have tried to contact Talos development about how to get things to work on Wayland, but the Talos dev replied "Wayland is not supported."
I wonder if it'd be possible to get this article in front of someone with decision-making authority over government software procurement. An email from the right person saying "We cannot certify deployment of Wayland-based Linux desktops such as Ubuntu or Red Hat due to EN 301 549/Section 508 because assistive-input users will soon lack a supported accommodation path" might go a long way to breaking the stalemate here.
But really is any large government using Linux desktops?
Lately I think that the (very indirect) way of improving the accessibility situation on Linux is for an organization to try to make money from selling Linux desktops to governments. They would likely have to solve the accessibility issues to do that. I thought with the current world situation, someone would try to make money from making governments less dependent on Apple and Microsoft.
(My understanding is that a lot of advancements in UNIX-related accessibility came in the era when Sun was selling desktops to governments.)
But really is any large government using Linux desktops?
I believe the French civil service is moving to Linux this year, and that Germany and Denmark have been making similar moves. It would surely then be necessary for their computers to be accessible for civil servants, right?
I don't know. Most of what I see in France has been server-side and applications. (E.g. https://lasuite.numerique.gouv.fr/)
Germany has been migrating to Linux desktops for ages, but it never pans out or they roll back. For a long while I followed with interest a blog about Linux in the Munich (city) government. Lots of good work came of that; however https://en.wikipedia.org/wiki/LiMux says they went back to Windows.
What I hear from a very limited set of accessibility users in Spain is that these efforts tend up preserving Windows for people who need accessibility features.
Schleswig-Holstein has announced that they'll switch to Linux/LibreOffice semi-recently. The interesting point is that this is largely driven by Digital Sovereignty, so not just cost measures that are easily passed by some Microsoft "deals", and thus with better footing in majority parties (one issue LiMux had, although at some point in time even the Greens turned coat, and the OSS-friendly "Pirates" were a short insignificant blip anyways).
Accessibility is quite important in that space, so I guess they'll have to pay for some developers helping there.
Most of what I see in France has been server-side and applications
I'm referring to the recent announcement by the French gov's technical agency (DINUM) that they will move from Windows to Linux and that all other French govt agencies need to submit their plan to do the same by autumn (source, in French).
Germany has been migrating to Linux desktops for ages, but it never pans out or they roll back.
Yeah, they've been very annoying about it. It does tend to be only individual states or cities that try, which I guess is why it's easier for them to balk. Denmark seems to be slowly planning a government-wide transition, but it will be slow and thoughtful.
What I hear from a very limited set of accessibility users in Spain is that these efforts tend up preserving Windows for people who need accessibility features.
This is the problem, I suppose. But even if only a subset of European governments are using Linux, they surely need it to be compliant with their laws.
I, like you, was diagnosed with EDS, but at 12 years old. No one told me that it would worsen -- in fact, I was told it would just "go away with time". Fast forward, I've been losing muscle strength in my hands for the last year, to the point that it has been repeatedly confused for neurological issues.
I know my timer is ticking for configuring my system to be usable without my hands. Watching the Wayland compositor become the standard and X be put away is destroying my motivation to learn Talon, which was recommended to me by another friend with EDS. I wanted to try helping build support while I still have the ability to, but have no idea where to start, and it feels very much like few care. This whole situation sucks.
I want to thank you for raising this as it affects many.
There is one sentence I wish would have been left out:
The important thing is that absolutely no one wants to touch that garbage fire of a codebase anymore, and Wayland IS the future the FOSS desktop community has decided on.
Neither freedesktop, RedHat, GNOME, you or me can speak on behalf of the entire FOSS desktop community. It's certainly not a monolith, though the echo chambers are several. I don't believe the sentiment and buy-in are as unanimous as some will make you believe. Several major projects seem undecided and haven't communicated any commitments to going fully Wayland and dropping X. The assumed conclusion isn't written in stone.
It would be a shame to have to give up Linux for BSD, though.
There is one sentence I wish would have been left out:
The important thing is that absolutely no one wants to touch that garbage fire of a codebase anymore, and Wayland IS the future the FOSS desktop community has decided on.
Neither freedesktop, RedHat, GNOME, you or me can speak on behalf of the entire FOSS desktop community.
OP's sentence is true: the Xorg maintainers indeed "stepped away from that garbage fire of a codebase", nobody forced them to stop. This a concrete downside of FOSS' do-ocracy: if there's not enough people who want to do something (or enough people able & willing to pay for it), it won't happen.
It is true that anybody can maintain and publish a Xorg fork to keep the flame burning, but I consider that the fact that it has not happened yet does "speak on behalf of the entire FOSS desktop community.". The BSDs are also adding Wayland support and, AFAIK, no BSD project has announced a maintained fork. And no, I'm not entertaining XLibre as a viable fork.
I recognize my privilege of not suffering (yet) from significant physical handicap and empathize with OP's situation. I feel like the level of access that Talon requires might be better achieved as a KWin plugin, but Talon being closed-source unfortunately makes it hard to evaluate and would make it pretty challenging to maintain.
It is true that anybody can maintain and publish a Xorg fork to keep the flame burning, but I consider that the fact that it has not happened yet does "speak on behalf of the entire FOSS desktop community.".
Huh? There's a very prominent fork by the one maintainer that wanted to keep it going. link
It was all over the place last year, and seems to be maintained well enough that it has commits as of yesterday.
The problem seems to be that the guy who did the split has some... shall we say... unsavory political opinions.
The problem seems to be that the guy who did the split has some... shall we say... unsavory political opinions.
Is it not an unsavory political opinion to say - and more importantly, actually do things like described in the OP, refusing to engage in discussion and referring to people with real world needs as "accessibility maximalists" while actively breaking their workflows?
Is it not an unsavory political opinion to say to innumerable downstream maintainers and users that "we have no choice but to try and contribute" and forcing many to conclude that "the Linux desktop [is] a lost cause"?
i'm impressed by talon. i never knew this was possible. the situation is similar for ideographic languages. which means that wayland adoption will either hit a wall in parts of the world, or they'll have to fix this.
I was somewhat aware of the limitations of Wayland, but as a somewhat casual user I have only ran into some smaller idiosyncrasies and never had to deal with this sort of thing. I hope your post contributes to more awareness in the circles of those who can make a difference.
Are you okay with your post being reposted on other parts of the internet to try and increase reach? I'd make sure to actually include your own disclaimer about this explicitly not being intended to be a punch down and instruct people to leave the pitchforks in storage.
I'm wondering if this is something that could be discussed with the context of the European Accessibility Act. I'm not familiar with the text of the act, but I know that accessibility in operating systems is a core part of it. Maybe it could be discussed with developers and European funding agencies such as the Sovereign Tech Agency and NLNet as something that needs to be addressed for Linux to be compliant.
The US ADA is IIRC much stronger than EU accessibility regulation (and around for much longer), and that hasn't stopped Red Hat selling to the US government.
(though in the future, I may try to integrate a foot petal with Talon)
I haven't used Talon, but I've had a pretty easy time using midi devices, like piano sustain pedals, with a small forwarder to generate other kinds of input. The hardware is pretty cheap and there's quite a few options you can creatively repurpose with a pretty small bit of bridge code. Just something to consider.
Unrelated to the main subject, custom(ized) input devices might be an interesting direction for accessibility, and there's already a community around open/DIY keyboards and such. I imagine a ploopy trackball could have a foot-operated variant (larger ball, sensitivity adjustment; buttons would need a redesign, but in a simplest approach one could operate a separate button pad with the other foot, split keyboard style).
I wonder if Wayback could be a short/medium-term stopgap workaround.
Not for that use case. Wayback's use case is to run X11-only desktops (Window Maker for example) on top of Wayland, to remove the almost-unmaintained Xorg server from the software stack. It's spearheaded by Alpine folks, who want to stop packaging Xorg without dropping the window managers that have not transitioned to Wayland yet.
What the OP is writing about is Plasma and Gnome removing support for their X11 session.
Ah, right, they mention KDE explicitly.
Maybe some of the desktops that still support X11 have good accessibility, but I have some doubts (Mate would be my candidate- given it's derived from GNOME 2, which I think was also part of the Sun's accessibility push, IIRC).
Last year, the maintainer of xdotool, the defacto input automation tool for X11, investigated how to do this and walked away confused.
He tried again this year and made a lot of progress, for example https://hachyderm.io/@whack/116554328265192238 It seems he's on a related side quest of automatic testing across multiple desktop environments at the moment.