JAWS version 2026 should be just a short while away. Here's how to ensure you install the upcoming version of JAWS and Leasey. hartgenconsultancy.com/working…
There is no way in #PeerTube to essentially 'boost' (retweet) a video from another instance, is there?
I want people to be able to go to my PeerTube instance and watch a video that I recommended without leaving my instance but all likes and comments should obviously go to the original source. (And the video itself should obviously be streamed from the original source as well.)
I don't think so, there is mirroring but that is not really the same thing...I think the best option I could come up with is to put videos into a playlist
@fiskfan1999 I explicitly want to avoid mirroring. Comments and likes should go to the original. For example @berlinfediday recently published their conference recordings on two different instances. This shouldn’t be necessary in a perfect world IMHO.
I’m looking into the play list thing. This might actually be fairly close to what I want to do.
I am not sure if I misunderstand, but you can paste the URL of your #Peertube vid in the search bar of your own instance, find it, and then boost it. Everyone then gets the actual peertube vid in their timeline and Likes and Comments go to the video.
@smallcircles @fiskfan1999 @berlinfediday Yes, I can boost a PeerTube video with Mastodon (I do this on a regular basis) but I want to do the same thing from within PeerTube.
I think it is possible, since you can follow PeerTube channels/authors from Mastodon and comment/boost from Mastodon. So I guess only the user interface elemnets for this are missing in PeerTube user interface itself.
@liaizon what do you mean by that? When you are locked into your PeerTube account from a client? Or what tooling do you use? And more importantly does the web interface then display the boosted video?
here's what a peertube feed I am following from Mastodon looks like. Playable, boostable, commentable inline in my feed like any other post from any other server
Software is a collection of ideas and decisions that a person has turned into a machine.
When it turns out that person has repulsive, noxious and dehumanizing ideas, it is entirely reasonable to not want the machines they've made in your life, or the decisions they've made affecting your life.
It is entirely reasonable to object to their being part of communities you care about, and to object to their ideas and decisions being normalized or valorized.
When creating spam that uses generative AI to fake detailed knowledge of my business, consider structuring your process such that you don't actually send emails that include "Error: No response from OpenAI API." Just a free tip there for you.
Making this a public post: I need grounded (i.e. very pragmatic and realistic) resources on LLM usage for code generation. Wanted resources include comparisons between available options, configuration, workflow integrations, and evaluation of results.
To note: I have very strong opinions about LLMs too. We probably have a lot of them in common. I’m asking for the information above for academic purposes.
We're here to stop surveillance by corporations like Google and Apple. That's why we replaced Google's FCM with our own notification system and keep Apple Notification Data at a minimum. Learn here why this is important.
hey there, i've recently been listening @signalapp hasn't been releasing an app on f droid because the plug might be pulled on f droid. Is that even remotely true?
Ein vielleicht viel zu törichter Gedanke: Benachrichtigungen OHNE Push Notification Server, einfach lokal... Naja, wahrscheinlich müsste man dafür Android und iOS total umbauen. Ich hab nichts gesagt 😄
cool that you do #FreeSoftware and write about it.
To be fair: using K9 Mail or FairEmail from F-Droid via IMAPS with an email provider that provides IMAP and SMTPS will also give someone an email account that is well protected.
Appreciable effort of developing Libre Economy. By the way, I learned that only client side is free and server still depends upon non-free services. Are we planning to replace those non-free parts with Libre framework?
this new #WhatsApp for #Windows crashes permanently for me. I don't know the reason, but given its inconvenience... I wouldn't say it's an absolute disaster (I can use it still), but comparing to the #UWP app it's a nightmare.
Тоже самое было с предыдущей electron версией. Этим было можно пользоваться, но всё равно ебанина. Как нынешнее не знаю, так как у меня по-прежнему работает UWP клиент.
@Yinshi Ну как… пользоваться можно, но в целом говнище. Куча табов туда-сюда, ссылку открыть только при включённом виртуальном курсоре (ну у меня по-другому не получилось), плюс оно крашится и виснет, как хрен знает что.
Ох тыж, то приложение было явно лучше. Не знаю, застал ты его или нет. unlabeled кнопки, причём не все, насколько помню, некоторая ебанина в перемещении по чату, если надо куда-то перемотать вглубь. Возможно что-то забыл, всё-таки время прошло.
У меня такой же вопрос периодически возникает при виде результатов работы фронтендеров. Как когда-то говорил один препод в универе "интерфейс программы должен быть таким, чтобы когда пользователь её откроет, он не испугался и не потерялся".
Interesting, gps module on my android phone works also in airplane mode. For some reason I thought tha gps is disabled in airplane too. So I can check the speed during my next flight.
This is a Kodeviser. A code generator for the Danish 2FA system. It is free. You activate it on a website. You do not need an app or smartphone to access digital services in Denmark.
This is a Rejsekort. You tap it to pay on buses & trains. The new ticket system will include a card. You do not need an app or smartphone to travel on Danish public transport.
I keep seeing a viral toot that says you’re excluded from 2FA & public transport in Denmark if you have no app or phone. This is not true.
I wonder if any blind users of tactile graphics displays (Cadence, Dot Pad, Monarch, etc.) have used such displays to view audio waveforms. Yes, actually listening to the audio you're working on is crucial, but even with my limited usable vision, I've found it convenient to look at a waveform to get a rough idea of how the amplitude varies over the whole track.
Haven't touched any of those displays yet, but some years back I took part in the research rounds of a project where they tried a handful of ways to make waveforms in a DAW more tactile. Personally I was kinda hoping for something where I could zoom in to perform tactile equivalents of tasks like eyeballing phase alignment between signals, dragging transients to match grid lines etc. Couldn't do that more precisely than I could do it by ear with any of the prototypes though. One itteration I spent a while with was a plastic ball you could wrap your hand around on the end of an arm that moved horizontally and vertically. They had it programmed to provide resistance so you could trace the shape of a waveform. Unfortunately the weight of the arm and the motors providing resistance made movement feel too juddery for me to pick out any significant detail again, but with this one, if I zoomed out I could detect long-ish silences in audio footage. At the time one of my gigs was batch processing audiobooks to DAISY format, we used to get plenty of digital masters arriving without chapter marks etc, so for a while, this was a stupidly expensive way I could skim through and find the breaks between chapters. I don't remember the name of the hardware device, vaguely remember someone saying it cost over 5 grand though. Ouchy.
Join us for the Matrix Conference from October 15 to 18, 2025. The four day conference will take place in Strasbourg, France. Talks will be streamed live and virtual attendees will be able to ask speakers questions.
This week we're giving slightly tired looks at the #mbedTLS project for shipping a new major version incompatible with the old with completely insufficient documentation about the bump and its requirements.
Fix some build breakages seen while compiling with mbedTLS 4.0.0.
Maybe a better way to fix MD5 and SHA-256 would be to switch to the PSA
API already available in 3.x. It requires calling psa_crypt...
The Free Software Foundation (NASDAQ: RMS) has announced the Free Software alternative to the evil, DRM-infested, locked-down, defective-by-design iPhone: the GNUPhone.
The key technical innovation of the GNUPhone is that it is completely operated from the command line. “What could be more intuitive than a bash prompt?” said seventeen-year-old Debian developer Hiram Nerdboy. “The ultimate one-dimensional desktop! Just type 'dial voice +1-555-1212 --ntwk verizon --prot lte --ssh-version 2 -a -l -q -9 -b -k -K 14 -x' and away you go! Simple and obvious!”
The phone will also serve as a versatile personal media player. “I can play any .au file or H.120 video with a single shell command! The iPod could never measure up to this powerful ease of use.” Video is rendered into ASCII art with aalib. “If blocky ASCII teletype softcore pinups were good enough for 1970s minicomputer operators, they’re good enough for you. Respect your elders.”
Actual successful voice calls are expected by 2030 to 2032. Regulatory approval is proving problematic in the corrupt, corporate-captured US environment. “The FCC said that if we dared switch on this, uh, ‘piece of shit’ in a built-up area in its present form, they’d break all our fingers with a fourteen-pound cluebat,” said Nerdboy. “They’re obviously shilling for Apple, Samsung and Google.”
The second version of the GNUPhone will run EMACS on the HURD kernel and be operated by writing eLisp macros on the fly. “It’s the clearest, most elegant and natural operating environment anyone could conceive of,” said Nerdboy. “Really, we’re not out to destroy Apple; that will just be a completely unintentional side effect.”
„lets make sure to secure the computer from all the bugs in the software that is installed on it by installing MORE software but this time it‘s software that runs with high permissions and everyone is really secretive about what it actually does“ really has statements dreamed up by the utterly deranged energy
My talks at #MatrixConf2025 are tomorrow 14:00 CET and 16:00 CET. I am talking about our progress on "Invisible Crypto" and agreeing a shared terminology to talk about Matrix encryption. Watch live at conference.matrix.org/watch/ . Recordings will be available later too.
Leasey has been updated to build 1080. New in Build 1080.
The Leasey installer is now compatible with JAWS 2026. When installing Leasey into this version, you are asked if you would like to import settings from JAWS 2025, 2024 or not at all.
A new option is available within Leasey Help, Leasey Key then H: it is called Toggle Whether Leasey update notifications are spoken when JAWS is started. When enabled, and JAWS is restarted, after a short pause Leasey will check to see if a new update is available. If a new update is found, you will hear the Leasey logo sound played quietly and concise directions on how to update your copy of Leasey.
An adjustment was made to the radio directory as on some machines the latest list of stations was not being retrieved.
When pressing the Leasey key followed by ALT+E to create a new Outlook Email from anywhere, there is no longer a problem in respect of formatting of the sender name when it is received.
In various aspects of Leasey weather, quotation marks should no longer be part of some of the output.
In YouTube, press ALT+Control+Windows+Comma to decrease speed of audio playback and ALT+Control+Windows+Full-Stop or Period to increase the rate. Note there is no setting so as to return to normal.
Implemented Leasey Dice Roller to Leasey Games. Please refer to the Games section of the documentation for details.
Trying to use Elevenlabs to read out a solar forecast for a new feature being added to @BlindHams, and... yeah, it does odd things when encountering numbers if you don't specifically lock it to English, which, BTW, I don't yet know how to do through the API.
The easy fix is to just spell all numbers out as if they are spoken.
Some generations are fine, others are completely broken, and you never know which you'll get.
@KaraLG84 Am I the only one who likes the TTS voices that sound a little bit synthetic? I don't like to be kept guessing. I think voices like SAPI 5 Tom and AT&T Crystal struck a good balance. If it's good enough to use on trains, it's good enough for a bot.
Um, OK that sounds like one minute its fine, then it sounds like its trying to speak Chinese and then it stutters, like hello? AI is not supposed to do that, what is this, some sort of incorrectly made speech dictionaries on ElevenLabs end, or I dunno at this point lol.
Eleven labs can do really interesting stuff with numbers. I read a book completelly narrated with eleven labs where it totally changed years so sometimes things were in totally different year as expected. I werified that oroginal was correct.
@sclower No. They've all gone over to Another node. That means we have a lot less activity, but quality over quantity, etc. Plus, they (not much to do with me) are starting a net where they can safely complain about whatever at a designated time. Along with the interesting stuff happening with the website, it isn't quite the same terrible it was a few weeks ago.
This is the thing, if they were even a little transparent about what exactly they were doing under the hood I could come up with at least an educated guess as to why it struggle so much with numbers and other symbols, but I have no idea. I'm pretty sure the model they're using is some kind of LLM architecture, similar to many recently released open-source TTS models, and those are notorious for struggling with numbers and other symbols without text normalization being done. I'll be honest, Eleven is actually behind the competition from the likes of Amazon and Microsoft, and many of the open source options not only sound just as good, but are much more customizable as well.
@ZBennoui @sclower Fortunately, we will soon have a dedicated, in-house server to run stuff on, so the possibility of using reasonable local models will exist once a decent GPU is added.
@sclower Well you'll certainly have a lot of options to pick from. Something that could be fun and potentially useful is to try fine-tuning a model on the type of content you're going to be generating, performance will likely improve drastically. These LLM based models are much easier to find tune than traditional ML TTS, I was able to train an Orpheus TTS model on my own voice in about ten minutes.
Waymo is coming to the capital of the United Kingdom. Learn about Waymo safety, technology — and sign up for updates to stay in touch on updates and news.
And I can't wait to get stuck behind one in traffic, just what London needs. Waymo gets quite a bit of hatred in USA, I expect we'll be in a similar situation here.
a dozen or so times a day ill think some shit like 'my name is bob soft plastics but you can call me Boft Plastics' and i have to restrain myself from tooting it. this one slipped through the net im sorry
Will Rehwinkel
in reply to Daniel Gultsch • • •Daniel Gultsch
in reply to Will Rehwinkel • • •@fiskfan1999 I explicitly want to avoid mirroring. Comments and likes should go to the original. For example @berlinfediday recently published their conference recordings on two different instances. This shouldn’t be necessary in a perfect world IMHO.
I’m looking into the play list thing. This might actually be fairly close to what I want to do.
just small circles 🕊
in reply to Daniel Gultsch • • •@fiskfan1999 @berlinfediday
I am not sure if I misunderstand, but you can paste the URL of your #Peertube vid in the search bar of your own instance, find it, and then boost it. Everyone then gets the actual peertube vid in their timeline and Likes and Comments go to the video.
Daniel Gultsch
in reply to just small circles 🕊 • • •just small circles 🕊
in reply to Daniel Gultsch • • •Jure Repinc
in reply to Daniel Gultsch • • •wakest ⁂
in reply to Daniel Gultsch • • •Daniel Gultsch
in reply to wakest ⁂ • • •wakest ⁂
in reply to Daniel Gultsch • • •