Found a stash of stickers in a box at work. Full Stack Fest 2019, great memories of - among other things - having beers on the beach with @bagder and other.
I'm happy to announce ChapterPod, an app for adding chapters to podcast files, is now available to download!
I've been using this for about 6 months behind the scenes to do all the chapters for Comfort Zone, and it's easily the best tool for the job (for me, at least).
He conseguido probar, con algo de dificultad, que la suma, resta y producto de reales definidos como clases de equivalencia no dependen de representantes específicos, y que cumplen las leyes del anillo (más o menos). Algún lema sobre positividad. ahora me falta la tricotomía. No sé si habré elegido bien usando secuencias en lugar de cortes de Dedekind. Pero es que las pruebas de supremos e ínfimos me tocan las narices bastante; es un coñazo tener que hacer análisis de casos con eso.
I couldn't believe that PC Gamer headline was a proper reflection of what was said, but... it was. The thing which most annoys me is that they seem to believe they have a right to do a damaging thing (which they acknowledge is damaging) in the hope that they might find something impactful to do with it (which they acknowledge they haven't, yet) because there's currently a lot of hype about it, and the only time limit on this is "find something before the hype runs out".
If "AI" worked they'd be using it, not selling it.
I.E. if it did what they claimed it would be proprietary secret sauce, not "come use this for free so we can figure out what it might be good for". Still, in 2026.
Remember when blockchain was going to have a zillion world-changing uses, real soon now? Same guys, they pivoted.
The "brilliant" minds behind the new social network wsocial.eu forgot to grab related domain names, so I applaud the efforts of the creator of w-social.eu... encouraging people to be social and simply join the fediverse.
Fun fact: yesterday I almost registered wsocial.lol (1 Euro for the first year) and wsocial.wtf.
I guess if I see an article in mainstream media about wsocial.eu I may do it (if the domains are still available).
• Wsocial
https://w-social.eu
Social media platform, an alternative to X
~ ~
Check, help, and share the link to the platform 😊
_____
#twitter #x #socialmedia #Europe #freedom #freespeech #News #technology #Mastodon #Media #opensource #foss
NVSpeech Player with phoneme editor version 1.6 now on GitHub as a release. github.com/tgeczy/NVSpeechPlay… Release notes: This release removes all direct linking against libespeak and other phonemizer libraries. Phonemization is now performed via external command-line tools (preferably through STDIN), keeping NV Speech Player and the Phoneme Editor fully GPLv2-compliant while still allowing use of GPLv3 phonemizers such as eSpeak NG as separate programs. Functionality is unchanged for users, but the integration is now cleaner, more flexible, and license-safe.
Licensing & phonemizer changes
This release removes all direct linking against libespeak and other phonemizer libraries. Phonemization is now performed via external command-line tools (preferably t...
@patricus yeah, it would have to be a new engine setting that lengthens it without it sounding disjointed. I'll work on this one after work, shouldn't be hard to make it a setting where you specify the tap length for any phoneme that has it true on it. It would benefit a lot of other languages. Regardless, I tried improving Polish a bit in this pack. try it. Should at least sound less "chinese." eurpod.com/synths/packs-PLFix.…
@patricus yeah but, you never included your phonemes.yaml with it, just the language file, so the y/ɨ mapping can't be done your way without knowing how you mapped it in the phoneme table.
@patricus yeah, that new file at least makes a dedicated phoneme for that sound, so what you can now do is mess with it directly without breaking other languages that way to get the closer sound of it. It should at least be listed inside the language as mapped to another one that's editable so we can tweak it that way.
@patricus oh, it's just that link :) Replace your packs folder in the editor with it. eurpod.com/synths/packs-PLFix.… there are now 13 different phoneme sounds mapped in Polish to play with, and they're Polish-specific, so definitely not going to change other languages.. You have to hit "edit phoneme..." next to the phonemes in language list in the editor.
bro I made it, y stil sounds ab bit Chinese but I got it almost working that and the R I will give it to you sadly fp is currently having service issues because of code changes. I will just put it into my site posix.live/pl_almost_working.z…
ok just a question how to add h? I'm stuck because the damn editor stops responding whenever I try to make it speak a word with it I'm new to this stuff but I'm learning hard. I learned what's fricative and stuff like that easily.
@patricus oh it shouldn't? If it gets stuck or crashes, it could be good to revert to the V5 editor version for now as it might be with how the V6 one introduced the change with Espeak. I'll have to check on that one. eurpod.com/synths/NVSPPhonemeE… might be less buggy for now, or even V4 if it's really bad. But yeah, something for after work.
@spacepup For “dans”, the vowel is /ɑ̃/ (nasal A). You usually don’t add a brand-new phoneme right away. Map the nasal vowel to an oral vowel + nasal consonant: /ɑ̃/ to ɑn (or sometimes ɑŋ, depending on taste) That already gets you most of the way there and works well in formant synths. In packs/lang/fr.yaml (or your French pack): normalization: replacements: - from: "ɑ̃" to: "ɑn" That way: • dans (/dɑ̃/) now is dan. True nasal vowels aren’t just “vowel + N”, but acoustically this approximation is very close, and much better than guessing a random vowel. Most engines (and even some older TTS systems) do exactly this.
@spacepup You’re absolutely right — /ɑ̃/ in dans is one vowel, not “a + n”. Native speakers don’t pronounce an actual n there. I think the reason people suggest mapping it to ɑn isn’t because that’s phonemically correct, but because of how formant synthesizers work. There is no true nasal vowel tract (no velum lowering, no nasal cavity resonator). A nasal vowel would need different formants + nasal anti-resonance, which we don’t currently model as a separate vowel type. So ɑn is a controlled approximation, not a claim about how French works. Later on we might be able to add more engine params like that though, but it requires deeper level changes to its actual sound.
New phoneme editor that supports custom phonemizers via the command-line. This will begin the work of using other phonemizers than ESpeak, like Gruut, ETC. Please note that because of how we split chunks now, since ESpeak is ran through commandline arguments, there may be a slight pause when processing larger files. However, a new "phonemizer settings..." dialog has been added with a few templates so you can switch to another phonemizer than ESpeak. This also means that V6 will be the first to get a release on GitHub, as it's no longer violating GPL2 to GPLv3, and frees SpeechPlayer from only and directly using ESpeak. Finding a phonemizer that's also multilingual though? Big challenge. Note that we introduced a new symbol in the IPA output, ||, (two vertical bars) to indicate a pause. You can force a pause via IPA by inserting this, though. eurpod.com/synths/NVSPPhonemeE…
Voluntary Chat Control is already in place since 2021, named Regulation (EU) 2021/1232. Similar to scanning obligations in the UK and the USA, this regulation allows Big Tech to scan all your data.
Sucht hier jemand eine Person mit jahrelanger Erfahrung im Testen von Soft- und Hardware? So richtig mit Weiterbildungen und Qualifikationsstempel und so?
Ich kenne da eine Person, die mal nen neuen Job braucht.
Bevorzugt Großraum Stuttgart, mit ÖPNV-Anschluss erreichbar, gerne Teilzeit (25-28h/Woche), gerne was mit Verkehrswende und / oder Energiewende.
I am reading Anthropic's new "Constitution" for Claude. It is lengthy, thoughtful, thorough...and delusional.
Throughout this document, Claude is addressed as an entity with decision-making ability, empathy, and true agency. This is Anthropic's framing, but it is a dangerous way to think about generative AI. Even if we accept that such a constitution would govern an eventual (putative, speculative, improbable) sentient AI, that's not what Claude is, and as such the document has little bearing on reality.
fix an upstream infinite loop bug in ProtoFieldFilter.skipBytes() causing an upstream system_server issue leading to the OS being unusable in early boot for a tiny subset of users since Android 16 QPR2 (there are a bunch of reports of these symptoms with the stock Pixel OS)
libpng: backport security patches
remove unused INTERNET permission from Pixel Camera Services since GrapheneOS doesn't support the Pixel Connected Cameras feature
kernel (6.1): update to latest GKI LTS branch revision including update to 6.1.161
kernel (6.6): update to latest GKI LTS branch revision
kernel (6.12): update to latest GKI LTS branch revision including update to 6.12.66
All of the Android 16 security patches from the current February 2026, March 2026, April 2026, May 2026 and June 2026 Android Security Bulletins are included in the 2026012101 security preview release. List of additional fixed CVEs:
Does anyone have a copy of the Minute Marker tape produced by Brayton and Sons Productions in the early 90's? It was a twenty-minute timer on a tape. There were different drum sounds for fifteen, thirty, and forty-five seconds past each minute, then each minute had claps counting the minutes. One side had a tick every second along with all the other sounds, the other side omitted the ticks. I never had or even saw a copy of the tape, all I ever heard was a promo for it he put in Playback Magazine, audio attached.
We just published a new blog post on the state of search and the critical need for open index access. The dawn of a healthier, user-centric web is possible, but it requires structural change:
Das Gegreine aus den KI-Klitschen, man solle doch bitte, bitte endlich irgendwas mit ihrem Produkt machen, damit es einen Sinn ergibt, es sei so teuer gewesen, man könne es sich nicht leisten, dass das nicht gekauft würde, das sollte einen schon misstrauisch machen gegenüber so genannten Wirtschaftswissenschaften.
Hallo Fediverse. Wir sind die Westdeutsche Bibliothek der Hörmedien für blinde, seh- und lesebehinderte Menschen e.V. (kurz WBH) aus Münster für Menschen, die auf Grund von Einschränkungen nicht in der Lage sind gedruckte Bücher zu lesen, die Möglichkeit kostenlos Hörbücher auszuleihen. Wir freuen uns nun hier dabei zu sein.
Faltblatt zum Digital Independance Day (englischsprachige Ausgabe)
Hier: XMPP / Jabber
Dieses Faltblatt wurde von Freiwilligen ausgearbeitet, um auf das gute freie ...
FOSDEM is the biggest free and open source software (FOSS) event in Europe, and will take place on 31 January and 1 February in Brussels, Belgium. And the #LibreOffice project and community will be there! Meet us and grab cool merchandise: blog.documentfoundation.org/bl… #foss #openSource #freesoftware
Is there a way to get a real, direct invite to a #matrix chat room? The share features only seem to bring up matrix.io links which a few people have had issues with.
the options for inviting someone to a chatroom are either by matrix ID (if they already have one), or by email (where they will get an email invite that should resolve to a matrix invite once they sign up, although it's quite fragile), or by linking the room via matrix.to - e.g. "come to matrix.to/#/#matrix:matrix.org". matrix.to should work pretty well; what problems did you see? is this to an invite-only room or a public chatroom?
Does anyone have a copy of the Minute Marker tape produced by Brayton and Sons Productions in the early 90's? It was a twenty-minute timer on a tape. There were different drum sounds for fifteen, thirty, and forty-five seconds past each minute, then each minute had claps counting the minutes. One side had a tick every second along with all the other sounds, the other side omitted the ticks. I never had or even saw a copy of the tape, all I ever heard was a promo for it he put in Playback Magazine, audio attached.
are you interested about that tape or recreating the digital copy? It seems that at the end of the commercial we have enough sounds to recreate it digitally.
@ondrosik Yeah I know, but what I really want to hear is the original tape. Sure I could easily create my own digital version using nothing more than a simple MIDI scripting language and software to render the generated MIDI file as a wav file. But I really want to hear the original.
If you have procrastinated on a hotel reservation. I am staying at 'Pestana Brussels Schuman'. One of the few hotels without inflated prices that particular weekend it seems.
I assume I am not the only one feeling this way, but I feel like I have gottten way more lazy with the release of AI. Take coding, for example. Instead of going but hunting myself, I just ask claude: "Hey, can you find the bugs in this project for me?". I really feel lazy lol, and I want to change that. Take writing as well. many times, I will plug something I wrote into things such as chat gpt and ask it to correct spelling and such for me, which is bad for my brain activity. It makes me think less, and I do not like it.
I'm considering ditching Chat GPT again. Claude and Gemini at least produce useful outputs I can learn from, but I feel like Chat GPT is just trying to please me and talk bullshit at this point. + they're introducing adds for the free version so yeah, bye.
Yeah. I get that. The fact my concentration span decreases + that AI does a lot of shit for me and I put less and less effort into understanding is a problem. I also realise that I have quite the double standart. I dislike Gen AI for many things like Vibe Coding and Music generation shit, for example, but on the other side I use it daily, often also for accessibility reasons. I absolutely get the problem.
@jonathan859 I think the quality has gone down recently, honestly. I'm not sure what's up, but I know a few people who found its output useful before and don't now.
I frequently delete the memories, and also changed the personality, which results in it thinking it would need to point it out everytime, e.g. I say: Tell it like it is; don't sugar-coat responses. And each time it points something out it's like, here is result XY, totally straight forward and serious, no talking around etc. Bit difficult to explain since I'm mainly using it in german. Yes I know, good prompting and such, but well what can I do if the AI's just dumb in the 1st place.
Well, I feel the same sometimes. For example i use AI for vibecoding because I don't have so much time to learn programming and I think that If I will try to do it myself it will newer be done. But i create music and I dont delegate this to AI. To be fair, I tryed suno for fun but I don't distribute suno creations as my music. I sometimes use AI for translating my text from slovak to english but here I realize that by doing this I lost my ability to write in English. BTW this text is not corrected by AI. At the other end, I am translating NVDA and I dont want to delegate this to AI at all. And yes, sometimes I am just lazy. For example I need to take some medicaments in various times during a day. I have to create 11 dayli events to be notified. I throw the instructions to gpt and downloaded just the ics file, imported to the calendar and it works. I realize that AI spent electricity and water to just doing simple text file for me.
Most of the sloptimists don't - for whatever reason - know about that.
In my experience most sloptimists don't even bother to read the bug-bounty docs - otherwise we'd receive much less reports as they are clearly about things exempt from a bug-bounty...
We will ban you and ridicule you in public if you waste our time on crap reports.
I see why that is needed, but at the same time I think there is a thin line here, what if someone simply is not sure, a false positive, is that a waste of time to perform investigation?
That ban/ridiculation threat demotivates the report, the message simply says that if you are not a high level engineer or a big company with resources to have identified something in the field, please don't report, individuals with limited knowledge not encouraged to report and that's when the person simply chooses the easiest path: Post the report publicly to some forum or microblog and then there is a disclosure of something that should have been embargoed.
I got the motivation but I don't feel good about the wording.
Heute hatte ich wieder sehr viele sehenswerte Videos ohne Untertitel in der Timeline. Diese like & booste ich aus Prinzip nicht, denn es ist frustrierend, nicht nachvollziehen zu können, worum es geht & was gesagt wird. Das tue ich Follower*innen nicht absichtlich an.
Könnten wir daher bitte auch bei Videos verstärkt darauf achten, dass möglichst alle teilhaben können? Gerade bei offiziellen Statements gibt es in der Regel mehrere Quellen, wählt bitte die mit UT.
We seem to have data that confirms that the #curl bug-bounty has received a steep increased submission rate through 2025, while several other Open Source programs also hosted on Hackerone have not. (There's a graph coming in my pending blog post.)
What could possibly be the reason for us taking more heat and more junk than others? Why oh why?
could it be because your a household brand? Even my grandmother has heard of curl; she doesn’t know what it does aside from being in the computer. Could being more well known have resulted in people focusing on the project over others?
wild guess: possibly, being one of the most public, most used OSS projects out there, LLMs have more info on that and so, when left run wild, they "find" more possible vulnerabilities?
Kejora dropped last week, and I really don’t want it to fall through the cracks.
This is a hand-drawn narrative puzzle-platformer about a little Indonesian girl who realizes her village is living the same day again and again. Everyone else is acting like it’s normal. She’s the only one clocking the loop. Which is a special kind of nightmare, if you’ve ever been the only sane person in a room.
It starts off deceptively cozy. Early 1990s rural village life. Rice fields. Forest paths. Errands. Kids being kids. Then it pivots into “oh, we’re doing this now” territory. An eldritch-looking monster shows up near the forest and starts hunting children like it’s part of the daily schedule. Suddenly you’re sneaking, solving environmental puzzles, and running for your life through caves and abandoned structures, trying to figure out what the village is burying.
The best part is you’re not alone. You’ve got two friends with you, and it’s basically a party system without pretending it’s co-op. You swap between them and use their abilities to get through obstacles, distract threats, and access areas Kejora can’t reach. It feels like childhood teamwork, except the stakes are “don’t get eaten by whatever that thing is.”
And I have to talk about the animation. It’s gorgeous. Full hand-drawn 2D characters, backgrounds, and cutscenes. It absolutely gives Studio Ghibli vibes, but it’s not Japan doing Japan. It’s Berangin Creative, an Indonesian studio, making something that looks familiar at first glance and then quietly reminds you it’s coming from a different cultural gravity.
This is why I love games as an art form. The art is beautiful, but the real hook is that you get to step inside it. You’re not watching a time-loop mystery in a rural Indonesian village. You’re exploring it firsthand, learning its rhythms, and uncovering what the town is hiding. That’s the magic.
"Jean-Baptiste Fressoz’s dizzying history of energy consumption argues that no energy transition has ever occurred: each generation consumes more of past fuels. Not only are his claims ahistorical but they justify an unwarranted pessimism about the future."
Hoy, en mi proyecto de ser un poco menos analfabeto matemáticamente, estoy haciendo la construcción de los reales como clases de equivalencia de secuencias Cauchy. De momento estoy con los lemas básicos (probar las condiciones de clase de equivalencia, reflexividad, simetría, transitividad). Cuando tenga que probar las leyes del campo me voy a cagar, pero bueno, se intentará.
a mí me lo instalaron, le dije que en mi casa lo usaba y me dijo ah, pues todo el problema sea ese. Tuve que pagar una licencia del complemento para Exchange, pero ahora incluso Thunderbird soporta nativamente Exchange, aunque solo a través de EWS y creo que sin libreta de direcciones global. Todo es tema de pedirlo y ver si se enrollan, y comprobar que ese Exchange tenga activado EWS.
Ще спра да ритуитвам, особено бг съдържание: толкова сме малко, че всички се четем. И Тръмп съдържание също: целият ми фийд е в едни и същи постове които до края на деня се завъртат по два пъти.
Bogomil Shopov - Бого
in reply to Julian Mair • • •How to survive FOSDEM – Marcin Juszkiewicz
marcin.juszkiewicz.com.pl