CNN - Major US airports refusing to play a “public service announcement” from DHS Security Noem at security checkpoints in which she blames Democrats for the federal government shutdown. cnn.com/2025/10/13/us/kristi-n…
Sabem porque é que eu vos digo para usarem ferramentas AI em grande volume desde que não paguem nem um chavo às empresas de AI?
Porque a AI é uma bolha. Como qualquer bolha, quando rebentar vai ter efeitos na economia real. Mas quando mais depressa rebentar, menos efeitos terá. Se rebentar antes dos idiotas meterem AI em tudo será muito menos destrutiva do que rebentando depois.
Portanto, façamo-la rebentar depressa. Usem AI todos os dias. Não lhes deem um tostão furado.
@ragb Esses usecases não vão desaparecer, mas duvido que sejam suficientes para pagar a batelada de massa que aquilo custa, mesmo depois da bolha rebentar e tudo voltar a um nível mais sustentável.
was just chatting with a former student about how much interesting compiler work never gets written up, and of course other work is written up and then forgotten.
on that topic, was there ever a good overview or retrospective for Microsoft's Phoenix compiler? and does anyone have an educated guess as to how many engineer-years they put into it?
I'm way more interested in IBM's Montana project (Visual Age C++ that was more about ASTs than text files). They pivoted to Java and the useful stuff (C++) vanished.
Though I never had the opportunity to program in Smalltalk professionally. I only programmed in Java professionally for less than a year. I'm mostly an embedded systems guy, by which I mean bare metal or RTOS, not Linux or BSD systems. Neither Smalltalk nor Java are suitable for the typical things I do.
my college was officially 100% java, and yet taught operating systems classes. how? they had to waste the first couple weeks of every OS/advanced data structures classes TEACHING EVERYONE C AGAIN
For the record, I committed new features written in Pascal to Mac OS in 2001. I think I was the last to do so. It was easier given the surrounding context of what I was adding functionality to.
I kind of regret that I've never gotten to play with mac pascal. Pascal was my first compiled language so I've got a soft spot for it, and I can see it being pretty OK once you take the CDC's IO system, and drag it to the trashcan that it so richly deserves.
And simply saying "We're not interested in portability" hoses like half of Kernighan's reasons why he didn't like Pascal. 😉
The CDC-inspired I/O system of original Pascal (and, I presume, ISO Pascal) only needs fairly minor extensions to make it reasonable as a fully synchronous I/O system, as was done in UCSD Pascal and Borland Pascal. I would never argue against the proposition that original Pascal was unsuitable for production use. Of course, that's not what Wirth intended it for. Nevertheless, it is a reasonable base on which a good strongly typed language system can be built.
I programmed in Ada '83 professionally, and it definitely had some nice improvements over Pascal. (Ada '95 and later standards are even better.) However, the Ada I/O system is totally different than Pascal's, not an extension of it.
The R1000 is deeply interesting for more reasons than you can imagine, but what impresses most people is this:
Imagine you maintain a library in some corner of $ReallyBigAndImportantProject and one of your exported functions has a parameter which defaults to 3.
Now you want to change the default to 4.
Who do you call ?
On the R1000, you change the 3 to 4 and when you try to commit that change, it will tell you /precisely/ which other code modules would be /semantically/ affected by that change, and ask you what to do about that.
Only calls which do not explicitly specify that parameter will be on the list.
When we demo that for modern developers, some of them get upset that was possible in 1980 but they can't do it today.
@CapitalB they wrote about gassing people and constantly spewed racial epithets because they're all fat ugly neckbeard retards who also are devout Catholics?
Edit: This is now released. Say all works, though the audio becomes choppy sometimes. But it doesn't crash. Right! I now have a copy of Eloquence that works on the 64-bit alphas of #NVDA, with the following issues: say all on the web doesn't work (it stops whenever the type of element changes for reasons I don't understand), and dialect switching doesn't work (but it doesn't crash everything anymore). If you want to play, you need to follow the build instructions; I only understand about a quarter of this code and have no intention of actually releasing things that are still broken: github.com/fastfinge/eloquence_64/
@jscholes At this point I would estimate that every line of code has been read by at least two humans, and about 50 percent of them modified directly by a human once the AI was done.
@jscholes Oh, also, I used spell check on the build docs. So I guess that counts. And Titet11 is Mexican and more comfortable working in Spanish, so AI translation was involved in communication and some of the comments. If you want to avoid anything that uses AI you need to avoid this, as it couldn't exist otherwise.
@jscholes Hah no worries. Your question got me thinking about what that even means. Like if my collaborator doesn't speak my language, does that mean I should disclaimer the code as AI assisted? If the code started off as entirely human generated, and an AI rewrote it, is it now AI generated? If a human rewrote large parts of what the AI did, when does the code stop being AI? I really don't know.
@jscholes So with more code updates this morning, the thing I'm noticing is that the more rewriting that is done, the less and less code there is from the initial AI rewrite. The AI solution mostly worked, but was over-complicated and multi-threaded where it didn't need to be. We're slowly arriving at code that is both simpler and works better.
I suppose I initially asked because of it defaulting to a Python helper process written in Python, using sockets as the IPC mechanism. Which is very AI, based on what will have been most common in the training data.
But for this sort of thing, I wonder about performance gains from shared memory, COM, or whatever with something other than Python on the other end.
@jscholes So the reason I wanted Python was because I naively thought a lot of the existing code could be reused, as well as some of the learnings from IBM TTS, eloquence threshold, and the sonata voices. That turned out to be entirely wrong. The "correct" way to do this would be to write a 64-bit API compatible wrapper for ECI.dll. But that's way beyond my abilities as a programmer, and AI can't help because we don't have the development headers for ECI.dll to feed it.
@feld@jscholes Maybe you could. But I don't really understand low level c++ code and pointers and things well enough to want to use AI for this. When it comes to Python, I can at least understand the code well enough to audit it (even if not to write it myself), and understand the approach. With C++ that wouldn't be the case.
@jscholes It's original code used C++ as the host, with GRPC for communication. I made it retry because I didn't want to deal with the complexity of the GRPC dependencies in an NVDA addon, I don't really understand protobuffs, and it just didn't feel any faster.
@jscholes Those are for version 6.4 of the DLL, and we use 6.1 because 6.4 has a bunch of changes like requiring registry entries for languages and voices that make it not portable, and several annoying bugs. I believe 6.4 also made a bunch of changes around threading. I already ran into issues with this, because the tts.txt is the manual for 6.4, and we need to use 6.1, the last release before IBM took it over.
How annoying. I thought there was also header files inside the Voxin packages, but I can't find them at present and not sure if they'd be any more helpful.
@jscholes Extremely annoying. I spent like two hours fighting with that because I'd given the AI the 6.4 manual, and it was doing things the manual said were allowed, but it wasn't working. Took me forever to realize that the DLL is a different version than the included manual.
Separately, I 100 percent think GRPC plus Protobuf would have been overkill and, in some scenarios, slower than the current "hand-rolled binary messages over a socket" approach because of overheads.
@jscholes You might be right. I just didn't wanna distribute like five different dependencies with the addon. Then it turned out that NVDA's version of Python doesn't include multiprocessing, even though it's built-in to Python by default, so I had to annoyingly include .pyd files and library code anyway.
If I were doing it, I'd definitely use something non-Python for the 32-bit helper process. I don't have time to do this though; I'm a bit behind on actual obligations as it is.
@matt@jscholes Yup. I'm hoping someone will get inspired and do things correctly. I just wanted something that worked, and Python was the easiest way for me to get that done. It's far from the best way.
@jscholes@matt I know davidacm is working on something in rust, of all things. But I don't know where he's at with that, and I don't know anyone else who wants to work in rust. So either he'll finish it and maintain it all himself or nothing will come of it.
@matt@jscholes Is there really anything to be gained, though? Few enough blind people who care about eloquence know C++. Even fewer are going to know rust. So is being one of maybe three people who could maintain it worth whatever performance gains you'd get?
I'm better at C/C++ than I am with Python. Don't know enough about the Eloquence API though, especially if it does differ significantly between versions. I probably know even less about IPC and threads, but there's plenty of learning scope in that department.
I suck ass at coding without help from either someone who knows way more, or in most cases, Gemini or ChatGPT. I've used Python as its super easy to get the program going and things like nuitka and pyinstaller let you compile stuff with a couple of commands. C++ I've heard is way more complex, and if you don't want to deal with the bloat that comes with Visual Pubio, sorry I mean Visual Studio, then you're other option is GCC with MinGW, and when I tried to get that working, the installation manager UI is borked with NVDA.
@alexchapman@daygar@jscholes@matt Meh. You can install the visual studio build tools from the command line, these days. Then just add the workload to vscode. Getting a development environment set up isn't the hard part. I actually already have one because of unspoken-ng and working with steamaudio. I just am not comfortable in C++ doing anything other than compiling other people's code and making the odd, extremely basic, change.
@alexchapman@daygar@jscholes@matt If you ask chat GPT it will guide you on how to get the visual C++ compiler and Windows SDK set up on the command line. Just specify that's what you want or it will direct you to the GUI.
Salut Mastodon. Ma mère, vivant à Paris, accueille un réfugié tibétain chez elle. Ça arrange bien tout le monde, mais ce qui l'arrangerait encore plus lui, c'est de trouver un boulot à Paris ou en banlieue, en journée. Il comprend bien le français mais le parle encore difficilement. En revanche anglais fluent. Il aimerait bien trouver dans la restauration mais pour l'instant tout travail serait le bienvenu. Si jamais tu as ça...et sinon repouet apprécié.
ráno nám přišla fotka ďalšieho neplatiča čo sa naje a nezaplatí, tak pre seba si hovorím to je už asi štvrtá za posledné dva mesiace a dnes vreckári okradli hostí.. ach jo
If you're curious whether an account is an impersonation or not, one way to check is viewing the profile. If any of the profile fields includes a website or similar, and it says something to the effect of: "Ownership of this link was checked on Sep 24, 2025, 04:24 PM", that means there's a link back from that website to that Mastodon account, meaning that in theory, that person has control of both the Mastodon account and the website being linked. If it does not say this, then there is no link between the website and the Mastodon account. You can see this on my profile as an example. Of course you have to make sure that the website is also one that belongs to the person in question, and they have to have linked it. For example, if the URL says something like iamtalon.fakewebsite.com then that's probably not me. It's not a bullet proof system, but it is definitely a tool available to check. Some clients might also display this. Enafore displays it, it says "Verified" under the website link in the profile fields. I'm not sure about other clients, though the website of the instance will show it as well as quoted above.
It's weird how critical personal data got stolen soon after age verification became a thing, and now neither company wants to accept responsibility for the hack. It's almost like, if the government is so desperate to force users to verify their ages, then the government should implement a way to do that. Want to verify someone can drive a car? Government system. Check that someone lives at their address? Government system. Taxes? Government. But verify someone's age? Nothing!
The UK is currently trying to push through digital ID cards which could presumably help with that purpose. There are many detractors to the idea, particularly as mandatory age verification isn't a popular move either so it's seen as a bad idea on top of a bad idea.
@jscholes I've heard of that effort. I don't know enough about it to land on one side or the other for the UK. This is the same UK that keeps demanding backdoor access to encrypted Apple user data,, so I don't know if I trust their IT expertise, but digital IDs sure makes a lot of sense to me.
@jscholes The UK and US are the only two countries with "ID allergy."
In the rest of Europe (and the rest of the world)., it is obvious that every citizen has an ID card, (which is distinct from any other document which could potentially be used for identification, like a passport or driver's license).
Even Germany, which is famously stubborn about things like these (they don't have a single citizen identification number for example) has ID cards.
@jscholes To be fair, not all of those IDs are digital, and integrating with digital IDs as a third party ranges from costly and difficult to illegal and impossible.
@miki @jscholes And a centralized database which is what this country wants to do, means hackers can virtually clone you way easier than without. Estonia's isn't centralized and I believe it's therefore harder to do that, but here, they want to have it centralized which seems stupid.
New, by me: If you're not using ad blockers, you should be! I wrote 2,000+ words on why you should use them and how to get started.
In this deep-dive blog, I explain why ad blockers are critical for your online security and privacy, what threats ad blockers can help defend against, and we'll look at some of the best ad blockers out there.
One advantage of using #AdGuardHome is that it provides statistics on the responsiveness of public DNS providers. For me, #DNS4EU is by far the fastest (10 ms). Perhaps it's because of their geographical proximity, as they're also based in Czechia. The slowest has been #quad9 at 500 ms. I have no idea why.
I'm also glad to learn that DoH doesn't have a significant speed penalty. It's 12 ms versus 10 ms for DNS4EU. So you don't have to trade privacy for responsiveness.
What has been your experience with #DNS providers based in Europe? Do you have any recommendations? I'm interested in unfiltered DNS because I do the filtering in AdGuard Home.
Define based in Europe. Is Google Europe-based? Is NextDNS Europe-based? Most DNS providers have multiple European servers. I am content using Google DNS and Next DNS, both low latency.
@dusoft based in Europe: the server is located in Europe and it's administered by an European entity (company, organization, community...), so Google doesn't qualify. I did use Google and Cloudflare for some time for the sake of latency comparison and they were both slower than DNS4EU for me, so I don't really have a reason to use them.
A free, sovereign and GDPR-compliant recursive DNS resolver with a strong focus on security to protect the citizens and organizations of the European Union.
@slamp You may also want to try dns0.eu/ For me it isn't as fast as DNS4EU, but it's run by a French non-profit, may be a different experience from there.
A free, sovereign and GDPR-compliant recursive DNS resolver with a strong focus on security to protect the citizens and organizations of the European Union.
I have excellent response times from quad9, and I recently learned that an anycast location is very close. DoH even seems to be quicker than plain DNS.
I run a local resolver for myself (combination of dnsdist and PowerDNS resolver), but use CZ.NIC, dns0.eu, NextDNS and dnscry.pt (all unfiltered) as a pool of forwarders for some convoluted NSes that refuse requests from residential IPs.
@kepi @oleksandr funnily enough knot resolver is exactly what DNS4EU uses. I don't mind using centralized DNS resolvers because I have a pool of 5-6 of them and AdGuard Home sends requests randomly to them. AGH is a gateway for all our devices and sends the requests to a pool of resolvers. I don't think this setup can be meaningfully used for tracking.
Did you enable "optimistic cache" in AdGuardHome? You will get an very good average time response. Here I get between 1ms and 3ms using public resolvers (almost all of them) or even local resolver (unbound).
@alelab But that's measured on the client, isn't it? I was talking about response times AGH is getting while querying public resolvers. With optimistic cache the queries to public resolvers don't even happen most of the time.
The average response time is shown in AGH dashboard. With optimistic cache, AGH replies with stale date first (if it has the record in its cache), then it updates the record and keeps it in its cache. That’s why you get a very short response time from clients to AGH. In this case, the average time response between public resolver and AG) doesn’t really matter IMHO. You can measure this average response time without this feature. AGH only uses the response time for the DNS records that are not in its cache or expired ones if optimistic cache isn’t enabled.
Check out the all new, shiny, High Speed Rail Study!Subscribe: http://bit.ly/CBCSubscribeWatch CBC: http://bit.ly/CBCFullShowsAbout CBC: Welcome to the offic...
My train nerd partner ❤️ tells me that CN has speed restrictions on Via in certain parts of the track because the new Siemens Venture trains don't have enough axles to consistently trigger the crossing equipment.
maybe I'd be honoured to dodge a high-speed passenger train at a malfunctioning crossing. It would certainly get people thinking about high-speed rail more, and driving more carefully at crossings.
My company has recently stopped using GitLab and I think it's worth a quick discussion why, because I think these are largely addressible.
Firstly, their paid product for teams is very pricey. Where Github costs ~$4 per seat per month, GitLab charges $29 per month. Even if we assume some portion of that is due to Microsoft's economy of scale and exploitation of user code for things like AI training, that's 7.5x more expensive!
GitLab charges for things like API address, and even more (full price) if you want to do things like store secrets via API.
While the direct cost wasn't a lot, I didn't feel like being nickle and dimed, and the product was just not worth it.
I enjoyed the product, but the pricing wasn't 2x, or even 3x, but >7.x! I felt like I was being taken advantage of, so we left altogether!
Glide is the world’s first autonomous and intelligent guide for people who are #blind or have low vision.
To find out more or pre-order your Glide today, saving a total of 25%, be sure to use my referral code at check-out! PIONEEROYHRP glidance.io/preorder/
DISCLAIMER: I do not work for Glide, but I have pre-ordered one. Up to three people can use this referral code, which gets me a discount on the Glide software subscription for my own unit.
🥁 Trommelwirbel… Wir starten jetzt unser Crowdfunding für den WEtell Hub! 📧 Mail. ☁️ Cloud. 📅 Kalender. 👥 Online Office. Ein Zuhause für deine Daten.
Keine Werbung! Keine Datenkraken! Keine Kompromisse! Mach jetzt mit und sei Teil der Bewegung für eine digitale Zukunft mit Haltung. 💚 Mach jetzt mit auf 👉 startnext.com/wetell-hub
Every life is meaningful, however, the way we spend and choose to live it might not be, but even meaningful is different for all. Maybe people find it meaningful that we don't find their lives meaningful. Who knows? Who are we to judge, after all. Our lives are probably not meaningfully lived by their opinions, either.
OK Linux folks. With #KDE hitting 29 years old, I'm curious about switching my gaming PC from Windows 11 to KDE. The games I play are listed on ProtonDB as being gold or platinum, so I think I'm good as far as gaming goes, but I have some other software concerns.
What do y'all use for cloud file storage/syncing?
What do you use for image editing/creation?
Where can I find out if my Razer Leviathan and HyperX mic will work on KDE?
I use Nextcloud (Hetzner Storage Share) for cloud file storage. Works great with Linux. Photo editing? Gimp or you can try photopea.com/. For creation Krita.
Photopea Online Photo Editor lets you edit photos, apply effects, filters, add text, crop or resize pictures. Do Online Photo Editing in your browser for free!
@sesivany Multiple people are telling me to use Nextcloud. I'll have to investigate further. I need whatever option I pick to work with MacOS, iOS, and Linux.
The last time I tried GIMP was almost two decades ago. Hopefully the UI has improved since then.
Nextcloud isn't just a storage any more. It's a platform for integrated apps. I cover a lot of things with it: files, calendars, contacts, tasks, document editing, photo albums, password manager, RSS reader, notes taking... If you don't prefer selfhosting, Hetzner provides it for ~€5/TB/month.
GIMP has improved, but its UI isn't radically different from what it was 20 years ago. There are other editors. Just "I need exactly Photoshop and Light room" doesn't work.
@sesivany (The instantly noticeable highlights with GIMP being that everything is in one window now and it doesn't take forever to launch anymore (and launches a lot faster after the first time with plugins cached))
Looking at a thing that came across my feed, I don't understand why I would need or want a StellerTrek for anything. Apparently, it is an AI-driven GPS and OCR device. These kinds of things were useful before the days of smart phones, but, nowadays, my phone has both a gps (or "jeepies," as my phone's TTS was randomly pronouncing it for a few days while I was in Europe)and a camera, so I can't see the point of these specialty blindness devices anymore.
Microplastics have now made their way into one of the most unexpected places: dairy. Scientists in Italy found plastic fragments in milk, fresh cheese, and aged cheese — which they believe are introduced during production and processing.
Linux oh Linux, you're making me Run windows 7 for a Mastodon client that plays sounds for events, oh why so must this be. Really want to warm up to this land of Linux but I'm thinking more and more, maybe I'll have to settle with Windows 10 forever. Some of you who know me a bit more personally from 2016 might recall when I used to say that. All the time, Windows 10 forever! Oh wow how I never thought the truth of that statement might one day shine through and not for the reasons that we were marketed at the time, mainly that it really would be the last version of Windows. No wit's because Windows 10, in its final form, still uses aprox 300-400 MB less RAM, and while RAM in of itself isn't the problem these days, it's that it spins up a bunch of broker and even more SVCHost (Service control host) processes, just to accomplish what Windows 10 did. And yes, the same could be said For windows 7 to 10: Windows 7 can fit into 500 or less MB of memory, with very few SVCHosts, because at the time, the entire Onecoreification (yes, that's pronounced one-core-eefication) of the OS had not yet happened. That my friends, was all after Windows 8, and I'd say by 8.1 we had doubled the Ram, from that of Windows 7. But you know what? Nobody cared, because not many ran Windows 8.X, so people just accepted Windows 10 with these higher process usages. Linux? Well the number of services you run are on you, really. If you want 1 GB of Ram, install Mate and it's good. Monday rant over!
@vol4life8657 Not just that, all MS updates. Sometimes the machine won't boot, sometimes the machine won't play sound,sometimes something will suddenly be gone... You get the idea. It's a balancing between the risk of security issues from hackers and the risk of problems introduced by MS. For me, I'll take my chances with attackers and multiple forms of protection. At least I have control over the protections I use and the behavior which protects me.
@techsinger @vol4life8657 thing is, if you're smart about your downloads and site visits, still run a good firewall that alerts you if rogue processes are requesting unbound network access, take monthly security day seriously where you audit your files for malware, you actually I think can squeak by on Win10 for awhile without much issue. Eventually yes the ESU and security updates will fall apart but by that time you've had at least 2-3 years to hopefully consider any alternatives, like by then the new Android that's now going to be Chrome OS should also launch, things like that. So in that sense I don't have much worry.
@vol4life8657 I agree and would go further. You can not only squeak by with windows 10, you can get by quite nicely with Windows 7 for an indefinite period. Obviously, you can just install LTSC if it's a serious worry for you and go on until 203x, whatever that may be.
@techsinger @vol4life8657 Yeah, I wish modern chipset drivers (and especially things like Thunderbolt 4 controlers) existed and worked on Windows 7, sadly drivers were just too re-written. It's what I'm seeing a bit from Win10 to 11: There's now a new Intelligo AI audio driver, that on Win11 provides this extra oompf and bass to speakers for HP, Asus, Del laptops. This sadly goes away with Windows 10, because Audio processing objects (APOs) were not yet well defined, they were just starting to appear in late 2019. So of course, these work better on Win11 because manufacturers decided that updating their Win10-era drivers to support modern features like talking to that Dolby Dax API that does this type of bassy oomf is just not in their interest. On a VM though, Win7 is kind of cool - and being on Linux in a way lets me enjoy that older day while still having the security of the hardened Linux Kernel and all that underneath. Still do my browsing within Linux Firefox 140, so for a Masto client and games it's the perfect land to be in.
@vol4life8657 Yes, you make a good point and that makes sense. The problem is the hardware. You can be like me and not really care about what's new in hardware, the only really interesting thing I've seen in the past ten years, in terms of PC hardware, is models on GPUs. Alternatively, you can do what you're doing, Linux VM for hardware, Windows 7 or 10 for software. For the normal user, this is easy. For those of us with disabilities, Linux is still a mess.
@techsinger @vol4life8657 so, what is your linux distro? I avoid ubuntu* due to snap, may go with debian sid/unstable what's your mastodon client on linux? and how was it installed?
Debian 13.1 now with Mate, really is all. Tuba is installable over Flatpak quite easily, just "flatpak install flathub dev.geopjr.Tuba " - it works so well with a nice tabbed interface too. Since Debian 13 ships with Orca 48 out of the box I updated it to 49.1 from Trixi-backports, and my ATSPI to 2.58 from Forky. Debian Sid may not be a bad choice either, since it is a rolling unstable release that just gets all the bleeding edge packages, but I'd imagine the downside = more frequent kernel updates. Fortunately Debian's installer does set up Espeakup, but not the "echo 'speakup_soft' | sudo tee -a /etc/initramfs-tools/modules" bit to have it survive initram-fs updates, how odd. Nonetheless, once you do that you do get TTY and GUI both set up for you, which I liked quite a lot. Ubuntu's installer didn't appear to set up any Espeakup for you, which is again probably fine for newbs or people not ever looking to access their multiuser.target, but you just never know what kind of X-server crash might prove it useful. (I also got SSH over local network so I can recover it from a Windows box if a kernel upgrade really goes wrong.)
@techsinger @vol4life8657 thanks for this, so the reason we don't get espeakup on other ttys, is because the correct settings are not in the initranfs? I thought it was to do with pipewire which debian 13 uses by default.
@mhussain @techsinger @vol4life8657 nah, Espeakup wouldn't use Pipewire but most likely good old Alsa (or for me it does) - because Pipewire is bonded to a user session, and if that bus isn't up when EspeakUp needs to speak right as the boot / log in starts, you still won't get audio even over Pulse. At least in my experience. So in theory pipewire itself shouldn't interfere, what can happen more is that your /etc/asound.conf file isn't set to use the proper out for the PCM card you want, so you gotta manually generate one based on your `aplay -L `command. I tried using Pipewire in asound.conf but unless I make my user session linger, I wouldn't be able to pipe sound through Pipewire. Lingering is generally OK if you know you're the only user logging into that machine, it does start your user services pre-login but that's not really a problem usually.
oh gosh. Also on this rant subject. I'm wondering if for new Linux folks, something like Ubuntu Mate 24.04 LTS is a better choice.(ubuntu-mate.org/download/) - Supposedly you should be able to press alt+windows+S (or called "Super", don't call that key Windows in Linux land) to bring up Orca during setup. If someone wants to give it a VM spin, be my guest, I'm just too busy at work to try it and know. I think it being Debian based and giving you more of that UX friendlyness could be better. Debian can be saved for the nerdy geeky tinkerers who do want to write conf files on their own and get things working side by side. Even though Debian's installer is terrific, and really it is in the list of more accessible distros for the fact that everything, including both EspeakUp and Orca, is configured for you. Magic, but really waht you should expect from all distros as a bare minimum, but ya know, we live in 2025 where that level of care hasn't reached a crescendo yet. Oh yeah, Non-Mate (I.E standard Ubuntu) might also be usable, no idea there either, but I recall folks having some installer quirks there. After 17.10 they switched back to Gnome which is interesting.
A community developed, Ubuntu based operating system that beautifully integrates the MATE desktop. Ubuntu MATE is a stable, easy-to-use operating system with...
don't use ubuntu variants. I dk if o nly on the regular version or all the flavours, but they have a flutter installer which is more or less inaccessible, by the virtue of missing labels. I recommend fedora instead
@Oldfartrant @esoteric_programmer oof yeah, the 24.04 one was really horrific, you can get through it with a bunch of "next" button pushes and checkboxes still got read, but I unfortunately found 24.04 still ships with Orca V46. That's even worse than the V48 you get in Debian 12, youch. I'm sure the 25.04 or 25.10 releases ship with newer, but you also get less support years than LTS, which is not a problem if you're willing to jump point releases to extend it still.
@matt oh just Tweesecake for now. Keep switching between that and TWBlue because the latter has way better timeline filtering and such. Love Tuba for supporting filters in general too, but Tuba would have no way of giving me sounds unless I ran my own watcher script in the background that played them alongside it open, which could work but is very hacky.
@matt Tweesecake is written in Python. It's too bad it won't just work on Linux. I don't know which parts are specific to Windows, but I'd guess the keyboard handler stuff.
@alexhall @matt oh yeah I would imagine. I know AccessibleOutput2 would interface to Speech-dispatcher directly, just have to have mine go through Pipe and not Alsa but yeah it would be workable on the speech side, it would bail most likely when trying to create those keyboard hooks. I've considered trying to run it inside a wineprefix with winetricks and seeing how well it does, presumably it could run with the visible UI but I have a feeling not even that would appear.
@Patrick Perdue @Terence Eden probably missing space between approximatelly and the number, I remember some tts when processing letters and numbers together treat that as abbreviation.
Say hello to @RocketChat a proud sponsor of Matrix Conference 2025 in Strasbourg!
🍴 Swing by their booth right by the food bar for a chat, 🥂 join them co-hosting the welcome party on Wednesday, or 🎤 catch CEO Gabriel Engel on stage sharing his insights.
Thanks Rocket.Chat for sponsoring The Matrix Conference!
There's a difference between someone who engages in political satire to call attention to things which are wrong, strange or just plain funny as opposed to someone with the intent to make fun of or harm someone's good name, reputation or brand just to do it. The latter can't be justified.
Woke up from a dream in which I was about to play a big trombone solo in this song about the wealth inequality crisis. More of that kind of dream please
Around 200,000 Linux computer systems from American computer maker Framework were shipped with signed UEFI shell components that could be exploited to bypass Secure Boot protections.
Languages are fun, have mysterious rules called grammar, but their native speakers certainly don't care about all that. In other words, do native speakers of a language speak the language correctly, as textbooks prescribe? (Hint: Definitely not.)
In this English Lesson with British English Teacher Roy, learn how to sound more like a native British English speaker when you speak English. In this Englis...
Peter Saathoff-Harshfield
in reply to victor tsaran • • •