2024-04-23 20:50:30
2024-04-09 11:38:27
2024-04-09 11:38:25
4797267
Some idle musings about Linux Accessibility with no real scientific basis for my claims I'm just thinking out loud here
Sensitive content
I saw another big thread about people talking at each other concerning Linux accessibility where one side goes "no this sucks" and the other side goes "But how?", and then it's a constant back and forth with no real resolution.
I wonder if a lot of this stems from the fact that a lot of people actually putting in work for Linux accessibility don't really need it? Like, for example if I boot up a Linux distro, sometimes I'll be able to do what I want, sometimes I won't. One of the biggest issues is that Orca and the accessibility stack will randomly fall over. Like I'll be in some app like Mate software update center or whatever and then it just stops reading. Even worse if Orca was playing a progress beep at the time, which then becomes stuck. So I have a constantly beeping unresponsive system. That's anything but pleasant. But here's the thing. These systems working together are complex. You can have just a slightly different config, or sometimes even just different apps open, and you'll get a completely different experience. Sometimes the update center works fine. Sometimes it doesn't. Telling me that this is a bug with the update center doesn't help because it's the screen reader that died, and without it, I can't do anything. So that's not the solution. The solution is clearly to make the screen reader or accessibility stack somehow work despite this.
But the real thing I'm trying to get at is that we're expecting a largely sighted developer base to fix this for us. Then they do some work, it works theoretically, so they throw on Orca and mess around for a few minutes and everything seems to work fine, so issue solved right? So off goes Orca, and back to visual working it is without a screen reader.
But not much longer and someone blind tries to use it, has lots of apps open, is doing complex and elaborate work, and something falls over again. So they report it. "Wait, but this was fixed wasn't it? Let me try." So they do. And it works for them. Can't reproduce, must be user error, and Orca gets disabled again and on with normal work.
And who can blame them? I mean this makes perfect sense right? We don't really want to use something we don't need to use. There's really no reason to keep using Orca if you can read the screen just fine.
Don't take this as fact. For example I know that there are at the very least a few blind devs and devs with assistive tech background working really hard on trying to make Linux accessibility better, and I'm not sure how many. But that's probably gonna take some time. Because if all this stuff was easy, we wouldn't even have all these issues right?
Anyway if you're working on Linux accessibility, don't take it personally. This is a hard problem to solve, and it's not your fault. Neither is it ours. If something sucks, listen to the ones who're complaining. It's the same stuff over and over. :)
I wonder if a lot of this stems from the fact that a lot of people actually putting in work for Linux accessibility don't really need it? Like, for example if I boot up a Linux distro, sometimes I'll be able to do what I want, sometimes I won't. One of the biggest issues is that Orca and the accessibility stack will randomly fall over. Like I'll be in some app like Mate software update center or whatever and then it just stops reading. Even worse if Orca was playing a progress beep at the time, which then becomes stuck. So I have a constantly beeping unresponsive system. That's anything but pleasant. But here's the thing. These systems working together are complex. You can have just a slightly different config, or sometimes even just different apps open, and you'll get a completely different experience. Sometimes the update center works fine. Sometimes it doesn't. Telling me that this is a bug with the update center doesn't help because it's the screen reader that died, and without it, I can't do anything. So that's not the solution. The solution is clearly to make the screen reader or accessibility stack somehow work despite this.
But the real thing I'm trying to get at is that we're expecting a largely sighted developer base to fix this for us. Then they do some work, it works theoretically, so they throw on Orca and mess around for a few minutes and everything seems to work fine, so issue solved right? So off goes Orca, and back to visual working it is without a screen reader.
But not much longer and someone blind tries to use it, has lots of apps open, is doing complex and elaborate work, and something falls over again. So they report it. "Wait, but this was fixed wasn't it? Let me try." So they do. And it works for them. Can't reproduce, must be user error, and Orca gets disabled again and on with normal work.
And who can blame them? I mean this makes perfect sense right? We don't really want to use something we don't need to use. There's really no reason to keep using Orca if you can read the screen just fine.
Don't take this as fact. For example I know that there are at the very least a few blind devs and devs with assistive tech background working really hard on trying to make Linux accessibility better, and I'm not sure how many. But that's probably gonna take some time. Because if all this stuff was easy, we wouldn't even have all these issues right?
Anyway if you're working on Linux accessibility, don't take it personally. This is a hard problem to solve, and it's not your fault. Neither is it ours. If something sucks, listen to the ones who're complaining. It's the same stuff over and over. :)
reshared this
modulux and Matt Campbell reshared this.
Talon
in reply to Talon • • •Sensitive content
Tyler Spivey
in reply to Talon • • •Sensitive content
Talon
in reply to Tyler Spivey • • •Sensitive content
Thomas Depierre
in reply to Talon • • •Sensitive content
i mean another part of this is that "work kinda despite other parts being broken" is considered a bad thing in software, for relatively good reasons.
As such, patterns, techniques and research around things that can do so are... Highly niche or under developed.
Hell look at the whole state of parsing and how people doing language servers stuff have to reinvent the wheel
Talon
in reply to Thomas Depierre • • •Sensitive content
Thomas Depierre
in reply to Talon • • •Sensitive content
oh i definitely agree. That is part of why compilers are such an atrocious UX in general.
Thing is, I don't see a way out anytime soon, because the research exist, but we stop funding bringing research to industry decades ago.
So everything is based on old stuff and we barely can keep it running
Matt Campbell
in reply to Thomas Depierre • • •Sensitive content
A new accessibility architecture for modern free desktops – GNOME Accessibility
blogs.gnome.orgTalon
in reply to Matt Campbell • • •Sensitive content
Matt Campbell
in reply to Talon • • •Sensitive content
Talon
in reply to Matt Campbell • • •Sensitive content
Quin
in reply to Talon • • •Sensitive content
Matt Campbell
in reply to Quin • • •Sensitive content
James H
in reply to Talon • • •Sensitive content
Drew Mochak
in reply to James H • • •Sensitive content
Talon
in reply to Drew Mochak • • •Sensitive content
Drew Mochak
in reply to Talon • • •Sensitive content
I believe the folks who work on Orca are part of some official foundation or other, not sure if they get paid though. I also don't know the structure of Apple/FS but I would be very surprised if anyone there was blind who had decision-making power over the screenreader. Google hasn't for years. I wish this were otherwise, but it is the norm.
Drew Mochak
in reply to Talon • • •Sensitive content
Matt Campbell
in reply to Talon • • •Sensitive content
Matt Campbell
in reply to Matt Campbell • • •Sensitive content
Matt Campbell
in reply to Matt Campbell • • •Sensitive content
James Scholes
in reply to Matt Campbell • • •Sensitive content
Talon
in reply to Matt Campbell • • •Sensitive content
Mikołaj Hołysz
in reply to Matt Campbell • • •Sensitive content
Arnold Loubriat
in reply to Talon • • •Sensitive content
Talon
in reply to Talon • • •Sensitive content
Al Puzzuoli
in reply to Talon • • •Sensitive content
Talon
in reply to Al Puzzuoli • • •Sensitive content
Kujiu
in reply to Talon • • •Sensitive content
James Scholes
in reply to Talon • • •Sensitive content
James Scholes
in reply to James Scholes • • •Sensitive content
Talon
in reply to James Scholes • • •Sensitive content
Talon
in reply to James Scholes • • •Sensitive content
James Scholes
in reply to Talon • • •Sensitive content
Talon
in reply to James Scholes • • •Sensitive content
Drew Mochak
in reply to Talon • • •idlestate's SDF liason acct
in reply to Talon • • •I can't speak for anyone you've encountered doing this, but I know I'm not alone in having pretty intense aversion to sounds that even other sighted people don't have a problem with.
@jscholes
Drew Mochak
in reply to Talon • • •Sensitive content
Cleverson
in reply to Talon • • •WestphalDenn
in reply to Talon • • •Sensitive content
Peter Vágner
in reply to WestphalDenn • •When I am leaving my office I am often closing some 15 browser windows, some 10 terminal windows, about 5 different files open in the text editor.
Most used apps on my desktop include #Firefox #Thunderbird file manager (pcmanfm or nautilus), Gedit, VLC media player, electron based apps such as teamsforlinux, losslesscut and gnome-terminal.
Next I'm using @LibreOffice, I am also using #Emacs with #speechd-el a little and finally some other less frequently used apps.
As for the #TTS or the #audio setup I am using #RHVoice, speech-dispatcher and @PipeWire Project .
Finally with @Matt Campbell and @Lukáš Tyrychtr we do have tallented visually disabled developers dogfooding or partially dog fooding so let me finish this post by saying it really is gold era of a linux #a11y and we are looking forward for what it brings us in the future.
like this
WestphalDenn and FediVerseExplorer like this.
WestphalDenn
in reply to Peter Vágner • • •Peter Vágner
in reply to WestphalDenn • •WestphalDenn likes this.
WestphalDenn
in reply to Peter Vágner • • •Peter Vágner
in reply to WestphalDenn • •Also gnome is very likelly to pioneer #accessibility innovations in upcoming releases so for me the answer is clear. I'm on @GNOME since v3.4 and v3.6 days and I am not going to change that.
WestphalDenn
in reply to Peter Vágner • • •