Thread, more thoughts on FOSS accessibility, accountability, ETC.
So, in many FOSS programs, there's this thing that says that that program shouldn't be counted on for any purpose and all that, that it may not work and such. And devs can hide under that, saying stuff like "well you know, out of scope, and besides it's my program, I don't have to make it accessible, do it yourself." Well, those devs should be scorned and boycotted, because they don't make FOSS better. The Pinephone in one project that said something like this, putting the blame on desktop Linux, while microaggressively telling me to do it myself. When you, as a developer, put something out there on Github, you are giving the community something, and if that something is inaccessible, it's your fault, and the community suffers for it. I no longer use Linux natively because of this repeated thing. No accountability, no responsibility, and no care for anyone besides biotypical and neurotypical people.
I mean, when billion dollar companies who are supposed to be these totally depraved and evil corporations can make their stuff at least somewhat enjoyable to use, and the other billion dollar Linux companies, even million dollar ones, can't, then that's saying something. Yes, KDE continues to take little steps every month or so to be accessible. And that's wonderful for those who can mess with Linux day in and day out. Maybe they'll get there one day, a few *years* from now. But I have a job, where we have to work out butts off, and I don't have time to stress about trying to convence FOSS bros to label their controls, or to try to use GTK next time. I just don't have the spoons to do that anymore.
Also, we, as a society, me included, seem to be so tired and worn out and just, like, small issues that we face are the biggest issues. Like, how much work would it take for one physical product manufacturer to make their packaging accessible? You'd have to hire a contractor who knows Braille, another who knows sign language, and increase the package size for that and a large print version of all print. The box labels, the printed user guide and instructions if any, and maybe put a physical marker where the barcode is. Then, you'd have to recall any packages sent out because you don't want a blind customer suing you because they got an earlier revision. And then send them all out. But for software? You're dealing with pure text. Pure, freaking, text! Screen readers can speak or braille it. Maybe one day there can be something to sign it. But yet, oh no, that's just too hard and CBA and can't be bothered with all that.
Also, none of these projects even try to survey us. No one hops on irc.linux-a11y.org to ask us questions (I've not been there in a few weeks though so that theoretically could have changed), no one subscribes to the Gnome Orca mailing list and even freaking tries to talk to us. Oh no, we have to freaking ***yell*** to be heard by them over the den of other stuff. And yet, Microsoft has done *two* surveys in the past half year alone, with a more focused group doing more. One was about the voices we use with our screen readers (and then they went and bought Neuance, which makes Vocalizer (one of the most used voices for screen readers)). And then, they did one about how well screen readers work for us. I mean, if I didn't know any better, I'd think Microsoft, oh my gosh, evil empire, cross my heart and hope to die or something, is trying to *help* us. Meanwhile, Gnome 40 comes out and it's a strggle to just escape the top bar.
So, yeah, I'm happily reading Kindle books, with Kindle for PC, on Windows, because it's actually accessible on Windows. Yeah, I'm using Google Chrome on Windows, because it's not just a second class citizen next to Firefox like it is on Linux. And NVDA's web navigation is smooth and slick and dreamy compared to Orca, where, if I go into a form field and start typing, I always dread it when Orca thinks I'm done and starts using what I type as navigation keys. Because in Orca, browse and focus modes and the navigation toggle are two separate things. It's just, well, a breath of fresh, much-needed air after trying the FOSS jail. And let's be clear; when you are stuck with a small group of accessible programs because the rest is inaccessible, that's not freedom. That's being trapped, with another group of people who are able to use other stuff happily around you, reminding you that you're tied up. That you aren't good enough.
Developers may then say, "Well we didn't even know blind people exist.", like Jesus healing the blind man just healed all blind people by default or something lol. Or, "Well resources aren't enough, so we don't know what we're doing." The person who works on Orca is active on the Gnome Orca mailing list. I mean come on, devs can Google. Right? "Well we didn't know accessibility is a thing." So you didn't go through the settings of your desktop environment and see the Universal Access part, or the "turn on assistive technologies" part, or see "orca" somewhere in your packages list, which is a required dependency of Gnome? I just don't think there is any excuse for reasonable accessibility in software anymore. We yell so much, and we try so hard to not offend anyone because we don't want to alienate anyone. Well, my patience grows thin, and I'm glad to be rid of Linux at least.
@devinprater one of the most annoying things about, "it's just text!" is when i tri to copy some of that text somewhere, but i can't, because the UI at that point doesn't allow selection, and as such, doesn't allow copying
@devinprater I cant tell you how many times I was paid for a Microsoft user study or giving paid feedback on hardware. A lot more than any other company, for sure. More than 200 times in one year.
@devinprater Dear god Devin. If you really have to switch to Chromium, at least use Brave or Ungoogled Chromium or something. Behave the same way, but aren't evil!
Seriously though, while I used to dislike Brave, I've actually been using it some more recently. It's not bad at all. Not as good as FF though, of course.
@Mayana @devinprater By all means look at alternatives to Chrome, but I do recommend Chromium over Firefox on Windows. Mozilla made an architectural decision in FIrefox's accessibility implementation that seriously limits performance. It frustrates me, but I'm not sure I can fault them for it. I discussed this with Jamie Teh (Mozilla accessibility tech lead), and they were caught between a rock and a hard place. Still, pragmatically speaking, Chromium's accessibility on Windows is much better.
@matt That's odd. I haven't had any accessibility issues I can recall in the recent months. At least with NVDA (and tbh, mostly also with the newer versions of JAWS), it runs perfectly smoothly.
Also, just check out all the things you can tweak in it! So many about:config options! For example: https://privacytools.io/browsers/
And it's also just ... everything is using Chromium now. Even freaking Microsoft Edge. If even Firefox uses the userbase it has, then websites will start only accounting for Chrome(ium) and nothing else, and Google will have even more power over the market than it already does. Mozilla has its flaws -- many! -- and is currently half-owned by Google already, but still, better use that if it works.
@Mayana @devinprater You can really feel the performance difference between Firefox's and Chromium's accessibility implementations when you load a long web page. Try loading a long Wikipedia article, then press NVDA+F5 to refresh the buffer. Much faster with Chromium.
But if Firefox works for you, that's good.
@matt Hmm. I suppose it could just be that lag doesn't bother me as much. I still use an incredibly old laptop for school, so this one, even with FF, feels smooth in comparison. And it definitely does compared to my ancient Android phone. So I'm perhaps willing to sacrifice more speed for the "feel" of it, the layout I've gotten used to.
@weirdwriter Well yeah, maybe. But it's fucking Edge. I for one have too much pride to use it. There's better Chromium browsers out there. Better as in not owned by a company that doesn't respect privacy in the slightest.
However little my choice actually matters, given that -- no matter what privacy settings I might've turned on -- I'm still using Microsoft Windows in the end.
@devinprater In Edge's defense, a sound effect played by the application wouldn't help deafblind users. UI Automation does have a way for an application to send a notification with a non-localized ID (that can be matched programmatically) as well as localized text. I think Edge even uses that feature. Now we need app developers and screen reader developers to work together on standardizing some notification IDs so the screen readers can map them to sound effects.
@devinprater I think that FOSS is actually still seen as an IT stuff even if it's begin to open to other disciplines but there is still a long way to go to make the foss tool even better and more inclusive
@devinprater yep. If the goal is to specifically exclude blind people, well, that is possible. If the goal is to make money and there’s no money developing for blind people, well, too bad blind people.
But free software that crows about being free for all that excludes blind people is like a free lunch behind a locked door. The reply of “why don’t you build your own house without a locked door” is a crummy reply.
@onan I mean, Apple, Microsoft, and sometimes Google develop for blind people. So apparently there *is* money in developing for blind people, if nothing else for the good PR and being chosen for government use, by any respectable government. I guess FOSS really just doesn't care, along with the general "fuck government" mentality, and the focus on absolute autonomy but also wanting to be a community somehow.
Google, Microsoft, Apple
you mention them as if they are the opposite of FOSS somehow. Yet all rely very heavily on FOSS and participate in FOSS very widely. (Often to the consternation of some *other* FOSS folk, to be sure.)
Nevertheless maybe the question isn't why "FOSS" isn't spending money on accessibility but why those making money with FOSS and spending money on *some* FOSS aren't doing it to make FOSS more accessible.
maybe it would help in explaining my confusion to add another large company to the list that has a huge presence in FOSS:
Here is someone from its Redhat division arguing from the point of "we need to do this for accessibility" in GNOME more than 10 years ago:
@deejoe @swift You know, I find this video actually pretty entertaining, and seems to show a good bit of perfection being the enemy of the good from Wolfgang Draxinger. Wolfgang does seem to try to get all the technical things just right, while Lennart seems to be more user focused and practical. Of course, I may have gotten their names mixed, lol.
I think age is a factor here (certainly of endusers and maybe some corporate managers who decide what the priority is for what gets developed).
On top of governments legislating for disability provision and requiring their suppliers to do so,
*everyone* will get some level of eyesight deterioration as they age - so companies with an older userbase would have to account for that to keep their business... >>
GAFAM (or at least Microsoft and Apple) have a considerable number of fairly affluent customers aged 70+, even in late 1990s I remember how my grandmother in Malaysia (in her 80s and with limited eyesight due to cataracts) easily sent an email using Outlook Express as it was comparatively easy to make the fonts large enough for her to work with comfortably...
I'm so happy you are speaking on this! It's far too often that developers assume that accessibility "is not their focus."
Uhm, it's for the community. So yes, it is!
We would have far more powerful and developed resources if we thought about how all users interacted with them from the beginning.
@Lindsaythelibrarian I just wish more people would learn about this and speak about it. I mean, I can't do this alone! I mean, I know a few people on here talk about web accessibility, and that's important too. But the desktop is still here.
I'm sorry it feels like so much of the responsibility has been in your hands! I was never exposed to these concepts before I went to school - which is sad! You would think it would naturally be brought up by socializing with people. All this to say, I'm still learning too so I so appreciate being able to hear from you, even though it may be (understandably) frustrating to repeat.
@Lindsaythelibrarian @devinprater Schools of opinion are multiple in FOSS and sadly "every man on his own" (gendered language intentional) is a prominent one. So is the ideology that tech > people and code > character. While many people are in it for community, many others are in it for (a varying combination of) their legacy, their CV or their company's interests. We thus end up with a lot of half baked implementations of the same thing.
Maybe what I'll say now is a bit anachronistic but AFAICT I've kinda watched the transition from a more community oriented FOSS to a more corporate oriented FOSS, in the years around 2010, as business realised open source was actually beneficial to them. Some part of me believes, if that whole story had been a bit different, maybe we'd be at a different, better place today.
FOSS doesn’t and shouldn’t wait until it’s finished to be released. It’s a “hey, here’s what’s cooking in my kitchen right now”.
Someone hacked something up for their own use. They can keep it in their own local/bin drawer for all time, or they can upload it as is without a warranty of any kind, either expressed or implied, including, but not limited to, the implied warranties of merchantability and fitness for a particular purpose.
It’s better that they upload than not upload.
If the goal is accessible software, it’s still better that they upload than not upload. Because it’s a start.
If people are gonna get attacked for what they make and share then that’s fucked. I’m not onboard with that policy. Keep on releasing in the free world.
Now, do I flinch everytime I see someone push hobby-level phones like PinePhone (or the one I’m involved with, Mudita, which is even crappier in the accessibility department) as if they were polished and ready? Yes. It’s not ready and done until it’s accessible.
But I stand by releasing unready things.
When billion dollar companies can make something that, in some areas and aspects, outperform what indies, smaller companies, and hobbyists can make, that’s saying… that a corporation with full-time paid accessibility team can do things that a one-person show can’t.
The discourse on here is that if you release something that is bad at RTL or calligraphic text, or doesn’t interface well with screen-readers, then you’re a racist and an ableist. Sure, if the software had been locked down to only be that. Since there’s an open license, other people can build on it.
The release-early philosophy is like… say the goal is to build shelter for homeless and someone contributes a plot and foundation, and give that away without walls or roof yet. If that were here on Fedi they’d’ve gotten their head bitten off for not providing a wheelchair ramp yet.
Now, it definitively is fair to criticize charge-money-early, or market-early, or set-up-a-customer-relationship early. “Release early and share what you’re working on” is a great FOSS philosophy that in just a few decades has changed our world to be more inclusive of hobbyist level devs, people who, alone and in their free time, make and share amazing stuff not just in software but in music and literature. Unfortunately, that has opened the door to a sort of quasi-hobby, quasi-commercial world that wants the best of both worlds: all of the money but none of the responsibility. Kickstarters that release garbage after months of waiting, for example. I’m not defending that at all.
What I am saying is that I’m definitively not on board with dissing the release-early philosophy and the warranty disclaimer that enables that philosophy.
We can’t get to “done” without taking those first steps. This goes double for features like accessibilty and localization, which are specialized subfields of software design. Those features are especially helped by a larger team and the only way to build a larger team is release-early.
Saying “hey, garage dev, fuck you if you haven’t mastered feature X, Y or Z out of the gate with your first release” is the same as expecting every single dev to master all fields. Talk about gatekeeping! Except it’s probably very few devs who are both blind and masters of Arabic typesetting, right? (Not saying zero devs can do that, just trying to say that different devs can usually have different strengths.) When humans cooperate, pool their strengths, and work together, we can make things that are more then the sum of the parts. Release-early is the stepping stone towards cooperation. It’s opening the lab doors a bit and seeing if anyone in the neighborhood can help make it better. I’ll never sign off on an attack against that.
@Sandra But that's not what has happened. We haven't come together to work on accessibility and internationalization, and all the other things people need. We just keep releasing stuff and then hoping other developers will help us with it, and that's not what has happened. And then users like me come around, asking politely for accessibility, and the Gnome folks are like "Uh, do we have anything for accessibility in Gnome 40?" *crickets.* So, accountability and responsibility are important.
re: Long post
If people are gonna get scorned, attacked, held responsible and accountable for releasing stuff then I’m not onboard with that. This stuff exists because someone thought “enh, might as well upload it as not upload it”. People shouldn’t get punished for uploading, nor be roped in to be the ones responsible for every lacuna.
We agree that the Earth would be a better planet if some app, say Gnome 4, had non-crickets accessibility.
We agree that it would be good if someone added that.
I’ll even grant that since some people (for example those who made it) know the code base better than some other people (for example those who have not made it), it is often ideal if those who do know the codebase get involved witn the accessibility and internalization efforts
What I don’t sign off on is that those people should be mandated to be the ones to add those features, to be responsible for them, to be accountable for them. Releasing FOSS should not come with strings beyond “don’t be actively malicious”—don’t release viruses or revenge porn, for example.
I’m gonna call this the death clause. It has to always be legal to release software and then die. I should not be mandated to keep working on it for all time.
The problem in our community is that we say software freedom is a human right, a fundamental freedom. If that's the case, then it being usable by anyone should be the number 1 priority, and where devs don't know, they should be inviting to people with diverse needs to share their contributions, code or otherwise. Sadly, both are exception, rather than the rule.
That is the part of the talk I am specifically against.
I don’t mind criticizing Gnome, PinePhone, or Mudita for being bad at accessibility. Heck, I criticized a FOSS project myself the other day (pulseaudio).
And OMG I can’t belive I did the “if you such-and-such then you get labeled a such-and-such” Karen move. That was awful. Instant regret.
What I do wanna adamantly say though is that I stand by the “death clause”. Releasing something comes with obligation that it’s not actively malware, but not with any obligation ever that the person is gonna keep working on it or any merchantability or fitness etc.
assume perfect literacy of FOSS.
That’s why am I trying to explain the problems with the argument.
Releasing not-early should be just as allowed as releasing early. That’s fine.
Labeled buttons have two parts. The labels and the buttons. You can make them three ways:
We all agree that labeled buttons are what we want. I appreciate unlabeled buttons as a stepping stone to labeled buttons, not as an enemy of labeled buttons.
It’s like a potluck dinner. If someone brings salad and gets beaten up for not bringing any dressing then that’s not a cool party.
I respect the lack of spoons and you don’t have to reply further if you can’t. I appreciate what you’ve said so far.
This is just in case someone else is following the thread:
I believe release now and fix later is good for inclusive design. Bookwyrm, which you mentioned, is a good example since it actually did release very early. It had contributors as early as a few weeks after the initial repository, while it was still only in a skeletal state, and many of the accessibility and internationalization features were added by contributors that were able to do so because the code had been released early.
Does Linux culture have an accessibility and ableism problem? Yes, a bad one, unfortunately. But that is because of other factors, not because of release-early.
Selling early, hyping early, promising early… those are bad things. Releasing early is part of the antidote to that. That’s what happened with Bookwyrm; people could and did contribute fixes.
@Sandra Seconding @weirdwriter 's opinion that adding accessibility later is not a good idea. Ignoring the disabled users for a second, it's also worse for the developers.
I'm not as active in FOSS as Robert and @devinprater; definitely not as much as I should be. But I have been observing the accessible gaming scene and beta-testing sometimes, and pretty much all the mainstream games that are also playable for us are that way because the developers had accessibility in mind at the start. Because that way, they didn't run into "small" problems like making some buttons reachable by mouse only, and being stumped when it comes time to add keyboard support. They didn't create fancy buttons for each action, then realize that this also means that they had to edit each one rather than just the main button function. Etc.
What is a simple change in a beginning could be a huge change later on.
But of course, that's gamedev. It may not apply.
A fun, happy little Mastodon/Hometown instance. Join us by the fire and have awesome discussions about things, stuff and everything in between! Admins: @Talon and @Mayana.