I need employment, *urgently*. I am not picky. My family needs to eat, and that overrides everything else. Please, does anybody out there work at a company that has machine learning or Python programming positions? Would you be willing to recommend me? I can demonstrate my abilities to you, first, if you need convincing, or I can share git repos.
To all you creatives out there with very understandable fears about AI...create. Get creating! Don't stop! Your creations matter. The corporations running AI who want to optimise us out of existence want us to believe our creations don't matter any more, that they can't compete. they want that hopelessness. Once you start creating, feeling it, that noise dials down. Keep on keeping on!
I know that you might get tired of my words, but we are still suffering and the situation is very bad. We didn't receive any donations for three days. Please donate and share to help my family overcome this tough time
People who grew up playing Zork and other Infocom games know that chat interfaces don't have to be perfect to be magical. Conversational interfaces are awesome, and it's amazing to see them develop.
The trouble is that a lot of the hype is based around chatbots not being an interface, but a universal knower and doer, as a kind of silver bullet for automation in general, as well as a creator of code, text, images, movies, etc.
When the chatbot is an interface, the skill and effort remain with the user. They are using a conversational interface to make something worth making or to find something worth knowing, and that skill and effort is what makes it valuable.
When the idea is you can offload the knowing and doing to the AI, that the chatbot is not an interface but a skilled servant, then it's presumed that the user isn't adding any skill or effort of their own. This is where things go off the rails.
Feasibility comes into question, like can "AI" (LLMs, etc.) really know and do things without human skill and knowledge? Early results don't look good. And economic questions, like if it can be produced with no skill and effort, is it worth anything?
One is a small set of companies effectively prevents competition by hoarding capital, hardware, and network resources, effectively blocking new entrants and stifling innovation. Let's call that the #enshittification path.
The other is that companies focus on conversational interfaces, not on the agency of the chatbot, but on the agency of the people using them, whether they are shopping, creating software, writing, or whatever, focusing on maximizing the agency of the person.
While the #enshittification path depends on centralization and hoarding to create an all-knowing oracle, the #dialogic path depends on diffusion, openness, and human-centred design to create a powerful, expressive lever for human agency.
With the dialogic path, you've effectively captured something that I've been thinking about. I regularly find myself revisiting the capacities of these new AI systems. I find myself always landing on a mixture of disappointed and impressed. The impressed part is outweighed by my disappointment.
But, for example, the chat interface to OCR is pretty useful. I wanted a serial number off of an access point the other day. I snapped a picture and sent it to my desktop.
I looked at it, and I had the "I wonder" thought, and I asked for the serial number AND MAC address. The system faithfully ignored the other information and presented what I asked for. I double-checked, and everything was accurate. That was legitimately cool.
What I don't like is that I don't have a lot of control or insight over these things. Ideally, I would be able to evaluate and plug and play these with a combination of models and deterministic plugins. Starting with a canvas and building my ideal user agent from the ground up. I don't want this to be folded into a web browser because I think having them as discrete tools is useful and adds a layer of cognitive protection.
I have found a few desktop applications that sort of gesture towards this, but I'm not seeing anything that feels good in the way my IDE or web browser feel good.
Admittedly, the knowledge of how these systems operate currently sort of saps any good feeling one might have away.
Boeing does not deserve sympathy and their attempts to make it other people's fault is wrong. #AllTheCoffee - Boeing Sued for $10 Million by Captain of Alaska Airlines Flight 1282 - Business Insider
More Americans with disabilities are now eligible for ABLE accounts.
If your blindness or qualifying condition began before age 46, this could open new doors for saving and financial independence — without risking benefits.
wow I had no idea that smartmontools added json output years ago
this fixes some annoying issues I have parsing it because there are some fields that have names that are not consistent (Device Model or Model Number) which are normalized in the json version of the data as model_name
The raw SMART attributes (temperature, power-on lifetime, and so on) are stored in vendor-specific structures. Sometime these are strange. Hitachi disks (at least some of them) store power-on lifetime in minutes, rather than hours (see next question below). IBM disks (at least some of them) have three temperatures stored in the raw structure, not just one. And so on.
For anyone backing up a pleroma instance DB, if you're using an sql backup (all variants of pg_dump) - YA DON'T HAVE A WORKABLE RESTORE WITHOUT SOME MASSAGING! You MUST have indexes pre-created to load data. Post creation of indexes effectively won't work...unless your DB is small.
Soooo I think I know what causes the restores to basically never complete, maybe...I'm near the limits of my psql debugging capability because I've never hit this so I actually don't exactly know how to see what the CREATE INDEX is actively doing... But tl;dr; 1.8T rows. One Point Eight TRILLION rows. But...HOW?!
The dump+load upgrade method I tend to use causes this. In the default mode it creates the schema (sequences setup, all the functions, that sort of thing, BUT NO INDEXES), loads the data, then creates the indexes.
The index it's never finishing creating is calling this function:
CREATE OR REPLACE FUNCTION public.activity_visibility(actor character varying, recipients character varying[], data jsonb)
RETURNS character varying
LANGUAGE plpgsql
IMMUTABLE PARALLEL SAFE SECURITY DEFINER
AS $function$
DECLARE
fa varchar;
public varchar := 'https://www.w3.org/ns/activitystreams#Public';
BEGIN
SELECT COALESCE(users.follower_address, '') into fa from public.users where users.ap_id = actor;
IF data->'to' ? public THEN
RETURN 'public';
ELSIF data->'cc' ? public THEN
RETURN 'unlisted';
ELSIF ARRAY[fa] && recipients THEN
RETURN 'private';
ELSIF not(ARRAY[fa, public] && recipients) THEN
RETURN 'direct';
ELSE
RETURN 'unknown';
END IF;
END;
$function$
Well. users is ~500k rows, and this function would be called for the ~3.7M activities rows....at a point in time where users has no indexes! So each invocation of activity_visibility would scan ~500k rows, times 3.7M rows in activities, and .... that's ~1.8T rows.
Why did I footgun myself like this when pg_upgrade exists? Well, binary incompatibilities, index collation differences, in the past causing Major Heartburn. I'm not going to dive into how it might be possible to fix the dump+load procedure, since, well, yeah. But for anyone backing up a pleroma DB be warned. Either take a binary copy (f/ex pg_basebackup) or dump schema and data seperately so you can load the data with indexes. Or find another workaround!
Over the weekend, did you see the much-shared story on Reddit by a supposed whistleblower?
It started like this: "I’m a developer for a major food delivery app. The 'Priority Fee' and 'Driver Benefit Fee' go 100% to the company. The driver sees $0 of it. I’m posting this from a library Wi-Fi on a burner laptop because I am technically under a massive NDA. I don’t care anymore."
A “whistleblower” tried to corroborate his viral post with AI-generated evidence. This is how I caught him. PLUS: Grok's image-generation crisis, and the rapture over Claude Opus 4.5
Basically he advocates that a retaliation with the US fucking trade treaties that nation void the anti-circumvention laws. That's the very short version.
Learn how to install Ghost locally and start developing a custom theme from scratch.0:00 Intro0:25 Developers documentation0:45 Requirements1:14 Install Node...
Over the weekend, did you see the much-shared story on Reddit by a supposed whistleblower?
It started like this: "I’m a developer for a major food delivery app. The 'Priority Fee' and 'Driver Benefit Fee' go 100% to the company. The driver sees $0 of it. I’m posting this from a library Wi-Fi on a burner laptop because I am technically under a massive NDA. I don’t care anymore."
A “whistleblower” tried to corroborate his viral post with AI-generated evidence. This is how I caught him. PLUS: Grok's image-generation crisis, and the rapture over Claude Opus 4.5
Edit: Erledigt, danke euch! Ein lieber Mensch gibt mir seine. 🥰
😭 Noooo. Hat jemand diese Tasse von den @MRMCD 2024 im Schrank stehen und braucht sie nicht? Ich würde mich extrem freuen. Tausche z.B. auch gern gegen die von 2019 oder 2025. Es ist eine meiner Lieblingstassen gewesen, weil ich Größe, Form und Farbe so schön finde.
Edit: es hat sich jemand gemeldet, vllt wird eine aus ihrem Keller-Dasein erlöst. 🤞
For everyone who enjoys discovering new music, I have a set of tools that can help. Shaq-GUI consists of three applications, each serving a different purpose: 1. shaqgui.exe – a simple application that listens to a selected input or output of your computer and saves whatever it recognizes to a file. 2. shaqfilegui.exe – do you have old radio shows archived somewhere and want to know which songs were played in them? This app will definitely help. 3. shaqcast.exe – something for radio broadcasters who play music from CDs, vinyl, etc., and want to send the titles of the tracks they’re playing to a server. For now it supports Shoutcast; if needed, I can add support for Icecast as well. Everything is accessible with screen readers: github.com/michaldziwisz/shaq-… You can download the programs here: gitrls.com/michaldziwisz/shaq-… Feel like you could add something, improve it, or help with development? I’m open to pull requests.
Seeking recommendations for a reliable, accessible platform to host a support/discussion email list/listserv/snapshot of the 90's with around 1000 members. We're with GroupsIO right now but I'm tired of their in-browser inaccessibility. Cheap would be brilliant as my group doesn't directly make money, can pay something for solid service, accessibility and responsive support though.
I probably don't fancy self-hosting this much traffic.
@jaybird110127 Reading/searching archives on the web is grim and it seems like they're A-B testing something in the back-end most times I log in. Haven't used the app for reading firsthand but I'm seeing reports that's regressing too. Support haven't replied to anything since switching on monetization, presumably because we're there for free.
I don't know the address of this, but apparently there's a Beta group where support is active. There's also a Group Managers' Forum, but apparently no one from Groups.io is actually there, it's just for support among members, so that wouldn't be the right place to raise any concerns.
I don’t know if the Business package does mailing lists yet, but I’ve just switched accessibility online over to proton, because the web and apps on LINUX became accessible when I had my back turned.
RWP doesn't generate much traffic. My VPS had 83 Gigs of traffic last month, that is about 2 to 3 gigs a day, way less than my usual monthly traffic at home. What I want to say is, traffic for text like on RWP is not an argument to deny self-hosted services lol.
@ToniBarth Isn't email delivery fiddly to keep consistent though? Had pretty rough reliability in the past when we were on a VPS, but admittedly I wasn't hands-on with improving that because I didn't (and still wouldn't) know what I'm doing.
Runs pretty smoothly here. Its not like we need to run self-hosted, but we shouldn't stop looking into that direction, just in case something comes up that looks promising.
@diffrentcolours I use MailmanLists for the mailing lists of the social groups I organise. The pricing goes up as your list grows, but 2,000 members is less than 5 EUR/month (less if you pay annually). Haven't had any problems with deliverability or support.
@pwaring @diffrentcolours Fair warning that unless it's recently been overhauled, Mailman has a UI that seems to have been frozen some time in the 1990s. Which in some ways is a selling point in its favour, and in other ways means it doesn't make consistent use of accessibility-friendly innovations like headings, landmarks, skip links, and live regions.
"I will never use Flatpaks because I have to type `flatpak run com.example.Application`. It's insane Flatpak developers don't change it!"
Except you can add `/var/lib/flatpak/exports/bin` to your $PATH. This way, you can type `com.example.Application` in your terminal, dmenu, everywhere. Also if you aren't using completions and insist on doing everything in the terminal, you have bigger problems.
Whenever I browse through the web, I find many “tips and tricks” from various blog writers, YouTubers and others who recommend users to take steps that either they aren’t supposed to, or have better alternatives.
@bugaevc For me, it's the cherry on top of why I don't like using Flatpak, but I can see why it would be a primary reason, especially for more command-driven workflows.
Ein Sachverständiger in einem Zivilprozess fordert Vergütung für ein Gutachten. Das Landgericht lehnt ab: Das Gutachten sei KI-generiert und damit unbrauchbar.
I don't think you should be able to claim your (macos) app is accessible with a screen reader if I have to learn app specific keyboard combos to effectively navigate to read things that a sighted mouse user would just click and scroll to find.
Tonight's frustration culprit: Discord. Which to be honest I find pretty bafflingly bad visually as well.
Relatedly, the trend towards native apps which are really web apps (electron app or similar) leads to some really baffling behavior compared to real native apps (at least on MacOS). Lack of full menus, VO cursor position not being consistent after similar actions, etc. Or at least I assume the containerized-web-app is why this happens. #accessibility
Do not underestimate the power of boredom+frustration. Paperback was born after I got progressively more and more frustrated with reading on Windows, and finally one day got bored and got on TeamTalk with one of the most intelligent hackers I know, and started bouncing ideas off of him. And look where it is now.
Thanks to my recent accessibility changes, and my first contribution to an Open Source project other than my own, Auto Claude is now the most screen reader accessible front-end UI that extends Claude Code with features like a changelog generator, creating roadmaps, and feature ideation. At its core, it has tasks, spec, and review workflows built in. github.com/AndyMik90/Auto-Clau…
I've installed it and set it up, no actual project with it yet, though I have a little thing I wanna try vibe coding so I guess I'll try it out soon. Seems quite decent so far though, nothing bad which made me wanting to uninstall it directly XD
So I'm currently trying it, reauthorized my Claude account already, updated claude code but still, created a new task and every minute or so it crashes with this: RESEARCHStarting phase 6: CONTEXT DISCOVERYStarting phase 7: SPEC DOCUMENT CREATION05:00:13 PMAPI Error: 401 {"type":"error","error":{"type":"authentication_error","message":"Invalid bearer token"},"request_id":"XXX"} · Please run /login05:00:15 PMAPI Error: 401 {"type":"error","error":{"type":"authentication_error","message":"Invalid bearer token"},"request_id":"XXX"} · Please run /login05:00:17 PMAPI Error: 401 {"type":"error","error":{"type":"authentication_error","message":"Invalid bearer token"},"request_id":"XXX"} · Please run /loginPhase 'spec_writing' failed: Attempt 1: Agent did not create spec.md; Attempt 2: Agent did not create spec.md; Attempt 3: Agent did not create spec.md05:00:17 PMPhase spec_writing failed
@jonathan859 Hmm. Are you using the development version, or stable? Anthropic just killed any client not using Claude Code, but that shouldn't be the issue here.
playing christmas dash will newer be the same after this song thanks to @HKC Radio and @tunmi13. Well, I am listening some christmas shows which I recorded because of different timezones and saved for today just because christmas oficially ends in Slovakia today.
The anti-commercial bias in much countercultural and activist thought is self-defeating by design, and is actually a product of internalized bourgeois ideology.
Veblen's Theory of the Leisure Class gave us terms like conspicuous consumption and conspicuous waste. 1/5
The core idea is that elites demonstrate superiority by showing how little they work. To be high-class means never having to work, sell, or promote, because you already have everything. 2/5
Even those opposing capitalism conform to its performance, perhaps replacing wealth with ethetics, but still performing a lack of need, and seeing the outward signs of money-earning as vulgar. This serves to maintain the class structure. 4/5
i humbly disagree, at least in the u.s. context. the national framework (reinforced by media) is to pull oneself up by one's bootstraps and work 50-60-70 hours a week. be the first one in and the last one out... movies with the executive class taking phone calls at all hours in all places, lower rung employees buy the gifts for weddings and birthdays, caregiving for children and elders is for the low class, spouses are for show not caring for, marriages are not partnerships they are for decoration... the morbidly wealthy ever busy scamming each other for the next big deal.
@melioristicmarie sure, but you get sweated and pushed to your limits, but self-promotion, selling, etc, is still viewed negatively. You're meant to just suffer.
no, that's how the rich and powerful are portrayed. from what i see, counter-cultures in hollywood media are mostly portrayed as noisy and ineffective.
@melioristicmarie the negative rich and powerful, these are generally not positive figures on counterculture/activist circles, they are not to be emulated.
@melioristicmarie the virtuous ones, the ones that are to be emulated in countercultural/activst circles are the ones that don't try and make money, either because they're already rich (enough) or because they get buy with very little/asceticism.
virtuous rich people is an oxymoron, a fiction. no one in my circles believes the morbidly wealthy are to be emulated, we all see the blood on their hands.
what about virtuous rich communities tho, don't we want everyone to be rich? How do we do that?
no such thing, no, we do not.
my version of we wants:
everyone to contribute meaningfully without the unknown, unseen suffering of others. in a global context. no cheap chinese goods made by slave labor, no dumping toxic waste for indian children to sort through, no oppression anywhere for the convenience of any one, at all.
it is the work of generations. generations of colonial exploitation built and continue to reify the current unjust systems, locally and globally.
our inheritances (you and i and every human who will see this conversation) include debts to others. to those who physically construct the digital tools we use, to those who have less clean water and more polluted air to support our lifestyles, to the nonhuman living world that has been destroyed and continues to be mined and clear cut and mutilated for our daily life to continue.
honorably making amends is our entire life's work.
accepting our responsibility to work, of proactive reciprocity, and then doing that work is how we have a virtuous life.
richness in community, in living in harmony with the humans and nonhuman world locally, in the love with share, in expanding our capacity to love... in recognizing the abundance in existence, wallowing in the pleasure of service to the life all around us.
@melioristicmarie lots of great ideals there, but there does seem to be an anti-commercial bias. I'm definitely not comfortable with and do not agree with the phrase "cheap Chinese goods made by slave labour," but I wont say any more about that, I would rather focus locally. I honestly don't see what way forward you imagine there is for people who need money to survive, or for the kind of communities you want to build resources they don't have. Ideals are important, but my comments are not about ideas as such, but about how considering commercial activity as being vulgar is a trap. We need to make money to make change. And we need to do so at scale to make significant change. Bourgeois ideology makes it emotionally uncomfortable for us to admit we need to make money at scale, but we do.
@melioristicmarie but don't get me wrong, a lot of the ideals you express resonate with me, I just don't see a way to peruse them without directly challenging, and thereby engagement with the economy as it is now.
i'm not sure what you mean by anti-commercial. commerce, like profiteering and the stock market, i work against. trading goods and services with informed consent in a just, regulated system? i am for.
adverts forced and snuck into daily life, i am against. adverts where folks go looking to buy things, in and around market places, i am for.
capitalist-extractive practices, i work against. worker-driven practices, i support. a fully informed and consent based supply chain, with just practices, i work for.
serving the needs of a community through labor (services / production) is often commerce.
@melioristicmarie you can assune I can find common anti-china material on my own, you could also find contradictory material, but I'll also let you do that on your own, if you want to. I'd preffer to focus locally, no shortage of cheap goods or exploited labour in our countries. Without sales and marketing how would you find out about good products produced by fairly treated workers? Wthout money how would these workplaces come to exist? How would the compete on the market to create sufficient scale to viable employers and suppliers for the greater population?
what do mean without? adverts, is what i mean by marketing. places for sales is what i mean by market place. we have newspaper delivery four days a week and there are adverts there, where our community expects them to be.
without money? why would we stop using currency? the market places in my city use money, in both brick and mortar and pop-up markets (we have both a weekly "farmers" market that pops-up / blocks out a portion of the parking area in one of ours and a monthly speciality vending event).
workplaces... ? i've lived in a few different cities and states in the u.s. and all have human-owned (non publicly traded) workplaces, most have co-ops, some worker owned.
i have no interest in colonizing a "greater population". i am invested in serving my local community... in buying from and selling to humans i am accountable to and who i hold to account.
edit: it is a process of generations to shift from exploitation to informed consent. doing the work, is the work of my life.
@melioristicmarie I guess I don't understand what you're arguing for or against. Do you believe the kind of workplaces you support are growing or shrinking in proportion to where most people work or buy from? Also, is that question important to you?
activists that i know and counter-culture folks working for system change are not anti-commercial, they are anti-capitalist, anti-exploitation. they are paying a living wage to service workers and using collective non-hierarchal models, buying from consignment and resale shops and new books from human owned bookstores or indiebound.org.
they don't build a business to cash out to chains and live off the profits (which harms the community). mastodon for example... is moving to a nonprofit model instead of being sold to a media company.
i guess if i'm understanding your thread, i'm hoping to shift your framework towards incremental prosocial change as a viable option for making one's living.
@melioristicmarie well, technically what you are describing is not outside of capitalism, but for sure more social, less exploitative practices is the goal.
However growth is also a factor, especially if the goal is supporting all humans.
You mention Mastadon, currently my favourite social platform despite my technical concerns, however it's import to understand that Facebook acquires more users every couple of weeks than Mastadon has in total, in all it's history, it's growth rates and daily usage rates trail facebook by even worse margins. Mastadon employs, maybe, a couple of hundred people world wide. Facebook employs 10s of thousands.
This is important.
Of course, Mastadon is not the best example, since it's not really selling anything, but I in general projects that avoid scale struggle to become engines of change, and in the end, many people just decide that liberating everyone is too hard, and they're ok just having things they like. But the risk is that slips into lifestylism, not an engine for change, just a privileged cul-d-sac, sheltering some lucky communards from the storm, often temporarily.
british empire, pro-colonization sort of thinking? one idea is the best for a fictional universal human? so better to subjugate all for their own good, patriarchy?
i can see better why your original toot had a burke's peerage flavor of leisure as a signal for wealth... the whole breeding program system definition of landlords and such.
i do not believe that any person can know what any is best for a person in another part of world. i believe informed consent is the only moral interaction method, and in a person's right to self-determination.
facebook is an marketing site based on deception. i studied it in grad school. it is a social harm, not a pro social system.
i prefer small and honest over authoritarian paternalism.
@melioristicmarie I'm not sure I understand any of that, so guess we will leave it there. It's clear we differ in our understanding of what needs to be done, and we probably don't share a theoretical framework, eg veblan, marx, etc, I've read your response a few times, and honestly have no clue what even a single sentence means, feels like a lot of bad faith projection and zero curiosity.
your theoretical framing seems to be that there is a right way for people to be better off, and if one could just spread that right way over all people, for their own good, all people would be fine.
@melioristicmarie I've been discussing peer production, free software, decentralization, etc for decades and am often cited by people and communities working on these topics. What are you asking me exactly?
Here are some slides on a presentation of mine about facebook specifically.
earlier you stated that facebook was a preferred platform to mastodon due to the number of new account sign-ups.
i am asking what you understand to be the difference between them, other than new account sign-up.
from my perspective (also a researcher and co-author in a.c.m. conferences who has been cited for my work on facebook), i understand them to be fundamentally different.
my example of the transfer of mastodon "ownership" to a nonprofit was to cite a case where selling out for the benefit of the creator at the expense of the community was the path not chosen. an example of a prosocial business transition model.
"...it's import to understand that Facebook acquires more users every couple of weeks than Mastadon has in total, in all it's history, it's growth rates and daily usage rates trail facebook by even worse margins. Mastadon employs, maybe, a couple of hundred people world wide. Facebook employs 10s of thousands.
This is important.
Of course, Mastadon is not the best example, since it's not really selling anything, but I in general projects that avoid scale struggle to become engines of change, and in the end, many people just decide that liberating everyone is too hard, and they're ok just having things they like. But the risk is that slips into lifestylism, not an engine for change, just a privileged cul-d-sac, sheltering some lucky communards from the storm, often temporarily."
specifically, you seem to be suggesting that an engine of change needs to scale. i disagree.
@melioristicmarie stalability maters if our goal is to include everyone, if consider facebook harmful, then we need an alternative or set of alternatives that ca scale. We need to compete directly against facebook. If our goal is a priviledged cul-du-sac, then we already have many for those able to access. Please read the slides.
that one person or group believes they have a solution for any person or group outside a community in which they are a member, from my position, is immoral.
@melioristicmarie so you think it's ok that so many people are on facebook. So, yeah, we disagree. Otherwise, I'm not convinced you are interested in a good faith convesation, so best of luck to you in whatever is you are trying to do.
@melioristicmarie in particular you conflate taking scale, and barriers to it, into accoun, with seeking to impose some monlogic solution, which has nothing at all to do with anything I've said
@melioristicmarie I generally agree with this paper btw, we both seem do advocate a dialogical aproach. When I have had discusions with people involved in anti-colonial struggle, the issue of how to confront the toxic enclosure of communications by facebook has been a central theme. For the most part, discussions circulated around how they could free their communities from facebook at scale, not just help some small groups find alternatives, because mobilazation for anti colonial struggles happen at scale, and so the challange is reach. A great deal of such mobilaization happens on facebook today, and this a problem, the slides I shared where from a talk and discussion I originally gave at a gathering of global south activists.
@melioristicmarie yes, I beleive scale is important, inorder to create alternatives to facebook that are available to everyone. That is the oposite of saying facebook is preffered.
FYI: If a service provider posts about accessibility, they get +50 to karma immediately from every disabled user 😊. I'll sign up to Sendamatic just in case, even though I don't need such a service right now. Thanks for caring!
A man goes to his doctor and tells him that his wife hasn't had sex with him for 6 months: The doctor tells the man to bring his wife in so he can talk to her. So the wife comes into the doctors office and the doctor asks her what's wrong and why doesn't she want to have sex with her husband anymore. The wife tells him. "For the past 6 months, every morning I take a cab to work. I don't have any money, so the cab driver asks me. 'So are you going to pay today or what?' so I take a 'or what'. When I get to work I'm late so the boss asks me, 'So are we going to write this down in the book or what?' so I take a 'or what'. Back home again I take the cab and again I don't have any money so the cab driver asks me again, 'So are you going to pay this time or what?' so again I take a 'or what'. So you see doc when I get home I'm all tired out and I don't want it any more." The doctor thinks for a second and then turns to the wife and says "So are we going to tell your husband or what?" 🤣
@Tutanota - I'm a paid customer of your e-mail service. And I noticed that using your e-mail app through web, is completely impossible by a screen reader user. Since last year it has become worse and worse, no way to intercept e-mails. Inbox doesn't show messages to screen reader users (Windows-based) it's a nightmare to navigate your web interface #accessibility
The text mode lie: why modern TUIs are a nightmare for accessibility
The mythical, it's text, so it's accessible
There is a persistent misconception among sighted developers: if an application runs in a terminal, it is inherently accessible. The logic assumes that because there are no graphics, no complex DOM, and no WebGL canvases, the content is just raw ASCII text that a screen reader can easily parse.
The reality is different. Most modern Text User Interfaces (TUIs) are often more hostile to accessibility than poorly coded graphical interfaces. The very tools designed to improve the Developer Experience (DX) in the terminal—frameworks like Ink (JS/React), Bubble Tea (Go), or tcell—are actively destroying the experience for blind users.
The Architectural Flaw: Stream vs. Grid
To understand the failure, we must distinguish between two distinct concepts often conflated under “terminal apps”: the CLI (Command Line Interface) and the TUI.
The CLI (The Stream): This operates on a standard input/output model (stdin/stdout). You type a command, the system appends the result below, and the cursor moves down. This is linear and chronological. For a screen reader, specifically kernel-level readers like Speakup, this is ideal.
The TUI (The Grid): This treats the terminal window not as a stream of text, but as a 2D grid of pixels, where every character cell is a pixel. It abandons the temporal flow for a spatial layout.
Case Study: The gemini-cli Madness
Let's look at a concrete example: gemini-cli, a tool written in Node.js using the Ink framework. On the surface, it looks like a simple chat interface. But underneath, Ink is trying to reconcile a React component tree into a terminal grid.
When you use this tool with Speakup (Linux) or NVDA (Windows), the application doesn't just fail; it actively spams you.
Because the framework treats the screen as a reactive canvas, every update triggers a redraw. When the AI is “thinking,” the tool updates a timer or a spinner. To do this, it moves the hardware cursor to the timer location, writes the new time, and moves it back.
For a sighted user, this happens instantly. For a screen reader user, this is what you hear:“Responding... Time elapsed 1s... Responding... Time elapsed 2s... [Fragment of chat history]... Responding...”
It drives the screen reader mad. The cursor is teleporting all over the screen to update status indicators, spinners, and history. Speakup tries to read whatever is under the cursor at that exact millisecond. You end up hearing random bits of conversation mixed with timer updates, making it impossible to focus on what you are actually typing.
Worse, lets pretend that you've somehow managed well with speakup so far, but that you want to do some work with nvda. Maybe paste an error you're getting on windows. So you open your terminal, ssh into your linux box, attach to your screen session and paste your text.
The result is an immediate crash of the screen reader (NVDA) or massive system instability. Why? Every time you type a character or paste text, the application triggers a state change. The framework decides it needs to re-render the interface. Because the conversation history is part of that state, the application attempts to redraw or re-calculate the layout for thousands of lines of text instantly. The more messages you have in a conversation, the more this will happen. And no, you can't just avoid this by using insert+5, the key combo supposed to avoid announcing dynamic change of content.
The Lag Loop
Furthermore, frameworks like Ink running on single-threaded environments (like Node.js) suffer from massive performance degradation when the history grows. If you paste a large block of text, the system has to calculate the diff for thousands of lines.
This causes input lag. You press a key, and you wait. You can wait up to 10 seconds for a single character to echo back. The system is too busy calculating how to redraw the screen to actually process your input.
Why The “Old Guard” Works (nano, vim, menuconfig)
Sighted developers often ask: “If TUIs are bad, why do you use nano, vim, or menuconfig?”
The answer is not that these tools handle the cursor perfectly by default. The answer is that they allow you to hide the cursor entirely.
1. Hiding the Cursor (nano, vim)
In tools like nano or vim, usability depends on turning off features that track cursor position. If you run nano with options that show the cursor position (like --constantshow), or if you use vim without specific configuration, the experience is broken.
When the cursor is visible and tracking is active, Speakup prioritizes the cursor's location update over the character echo. Instead of hearing the letter “a” when you type it, you hear “Column 2”. You type “b”, and you hear “Column 3”.
These older tools succeed because they allow you to disable this noise. You can configure them to suppress the visual cursor or status bar updates, forcing the screen reader to rely on the character input stream rather than the noisy coordinate updates. Modern frameworks rarely offer a “no-cursor” or “headless” mode; they assume the visual cursor is essential.
2. Single Column Focus (menuconfig)
Tools like the Linux kernel's menuconfig work because they enforce a strict, single-column focus. Even though there are borders and titles, the active area is a vertical list. The cursor stays pinned to that list. It doesn't jump to the bottom right to update a clock, then to the top left to update a title. The spatial complexity is kept low enough that the screen reader never gets “lost.”
3. The Lost Art of Scrolling Regions (Irssi)
Irssi is the gold standard for accessible chat, but not because of luck. Irssi was built over 20 years with a custom rendering engine that utilizes VT100 Scrolling Regions.
When a new message arrives in Irssi: 1. It tells the terminal driver: “Define a scrolling region from line 1 to 23.”2. It sends a command: “Scroll up.” The terminal moves the bits up. 3. It draws the new text at the bottom of that region.
Crucially, it handles this in a way that minimizes interference with the input line. It relies on the terminal's hardware capabilities rather than rewriting every character on the screen manually. Modern frameworks ignore these hardware features in favor of “diffing” the screen state and rewriting characters, which is computationally heavier and hostile to accessibility.
The “Stale Bot” excuse: A Case Study in Neglect
Google and the maintainers of gemini-cli pretend to care about accessibility. “Pretend” is the operative word here. If you look at the repository, critical accessibility regressions like Issue #3435 and Issue #11305 have been left to rot. There is no discussion, no roadmap, and no fix. Even worse is the fate of Issue #1553, which was supposed to track these accessibility failures. It didn't get solved; it got silenced. It was closed automatically by a bot with this generic dismissal: > Hello! As part of our effort to keep our backlog manageable and focus on the most active issues, we are tidying up older reports. It looks like this > issue hasn't been active for a while, so we are closing it for now.”
This is unacceptable. Closing an accessibility report because the maintainers haven't touched it in months is not “tidying up”; it is hiding evidence. It effectively says that if a bug is ignored long enough, it ceases to exist. It boosts the project's “Closed Issues” metric while leaving the actual software unusable for blind users.
Conclusion
If you are building for the terminal and care about accessibility, stop using declarative UI frameworks that treat the terminal like a canvas.
The “modern” TUI stack has optimized for the developer's ability to write React-like code at the expense of the machine's ability to render text efficiently.
If you cannot guarantee that your application allows the user to hide the cursor, or if you rely on aggressive redrawing to show spinners and timers, you are building an inaccessible tool.
For the blind user, a dumb, linear CLI stream is infinitely superior to a “smart” TUI that lags, spams, and scatters the cursor across the screen.
As someone who uses Eloquence on all my devices, this was a very educational article on the state of Eloquence itself. This also includes other contenders which, at this moment in time, don't really exist. Sam's Stuff - The State of Modern AI Text To Speech Systems for Screen Reader Users stuff.interfree.ca/2026/01/05/…
tvl celý som od diaviy
Aaron
in reply to Aaron • • •Zach Bennoui
in reply to Aaron • • •Aaron
in reply to Zach Bennoui • • •Blake Girardot
in reply to Zach Bennoui • • •RE: dragonscave.space/@ZBennoui/11…
@ZBennoui @hosford42 He has an awesome resume too.
hosford42.github.io/AaronHosfo…
#ml #getFediHired
Zach Bennoui
2026-01-06 18:25:01