in reply to Andre Louis

many many many many many servers. actually if you're curious here's what it sounds like to boot up all the bastards at once. It's a delightful sound
This entry was edited (6 hours ago)

reshared this

in reply to Andre Louis

@FreakyFwoof used to need that here. then I built a server stack that puts out more BTUs than a big box of BTUs with a big sign out front saying "free BTUs, come take some". I measured a minimum 10 degrees C difference between my bedroom and my living room, the latter being where all the servers are and is at least 10 degrees warmer than the bedroom at all times, even in the middle of the night

Einige haben seit dem 39c3 nach unserem Schildergenerator für die taktilen Beschriftungen gefragt. Wir haben den unter editor.c3tactile.org/ weiterhin zur beliebigen Nutzung online. Er ist nicht ganz fehlerfrei, aber für grundlegende Beschilderung ausreichend.
Wir arbeiten daran auch den Code zu veröffentlichen, das wird aber noch ein bisschen dauern.
Eine ausgezeichnete, verfügbare Open Source Lösung ist der OpenSCAD Generator von @oskar_mbr unter codeberg.org/oskars/define-bes…
#accessibility #a11y

Twice Google Gemini tried to tell me that I was wrong regarding a tennis result from 2004 but I kept telling it it wasn’t correct and eventually, it got it right.

In the 2004 Australian Open men's doubles semi-finals, Bob and Mike Bryan did indeed defeat Todd Woodbridge and Jonas Björkman with a score of 6–1, 6–2.
Sometimes you just have to keep guiding the AI models particularly if you know that something is incorrect.
It's almost like you're teaching something.
This is where I can see a problem with these AI models because if you don't know that something is incorrect, you can go away with the wrong information.
I'm seeing this a lot with supposed history posts about events from the Second World War.

Finally a functional NVDA add-on for the new WhatsApp Desktop.
WhatsApp Desktop Enhancer is a specialized accessibility bridge for the modern WhatsApp Desktop application (available from the Microsoft Store). It enhances navigation efficiency and fixes interaction issues caused by the application's underlying technology.
github.com/starkrush123/WhatsA…

#XSF Announcement

The 28th #XMPP #Summit starts today at 09:00 am (UTC+1)!

If you come in person, reach out before (almost full). Else join remotely:
xmpp.org/2025/11/xmpp-summit-2…

Meet us at #FOSDEM 2026, too!

#jabber #chat #opensource #messaging #federation #Brussels, #Belgium #opensource

I kinda wish every racist had to take a DNA test.

Mine is 98% predictable. Evenly split between Southwestern Norway and Northeastern Scotland. I’m a Viking.

BUT. A Viking who is also 2% South Asian. And I’ll bet a lot of these racists would find out they’re not as white as they thought.

Which might not change their politics but would probably annoy the fuck out of them. And result haha.

uspol, spaceflight

Sensitive content

For iOS users. You know when you copy something from one app and want to paste it into another, an alert pops up asking if you want to allow it? This usually happens for me when I copy the last phrase spoken by Voiceover. You can go to the apps sestion of the settings app, find the app that keeps asking (VLC in my case) and there's a section about allowing paste, which you can set to Never, Ask, or Always.

reshared this

@MewProjects FastPlay: Even if speak track changes is unchecked, NVDA still speaks new track names as they appear in the title bar. Winamp doesn't. Does this mean there's some magical code in NVDA that can figure out that it's not supposed to speak title bar changes in Winamp because the user would find that annoying? Or is there an NVDA setting I'm missing? I like having the track name in the title bar, as I'm used to things working that way, but I don't want NVDA to speak it whenever it changes. Any thoughts? Thanks.

Here's another song in support of the people of Minneapolis and in honor of Renee Good and Alex Pretti, this time from British singer Billy Bragg. Listen on YouTube here.

youtube.com/watch?v=IKOW2ZikGW…

#Miinneapolis #ReneeGood #AlexPretti #BillyBragg #Music

NVSpeech Player with phoneme editor and NVDA Addon version 1.8. Those on the NVDA driver from yesterday can update it from the last button in the Speech Player settings panel, too.
This release adds a new Voice Profiles system to NV Speech Player packs. Voices are now defined in phonemes.yaml and can be selected at runtime, letting you ship multiple voice characters without maintaining multiple full phoneme tables.
• New YAML voice profiles
◦ Voices are defined in phonemes.yaml under voiceProfiles.
◦ Profiles are overlays: one base phoneme table, plus per-voice adjustments.
• New bundled voices
◦ Beth (female)
◦ bobby (child-esk)
• Voice selection everywhere
◦ The NVDA driver and phoneme editor can select a YAML-defined voice.
• Better debugging for pack authors
How voice profiles work (short version)
A voice profile can apply:
• class-based scaling (e.g., vowels vs. fricatives) for formants, bandwidths, gain/noise, and voice quality
• per-phoneme overrides for targeted fixes (absolute values that win for those specific fields)
• Existing packs remain compatible: voiceProfiles is optional.
• For low sample rates, keep an eye on harsh sibilants and overall gain — profiles include tools to roll off the top end and preserve headroom.
github.com/tgeczy/NVSpeechPlay…
github.com/tgeczy/NVSpeechPlay…

Welcome to the Full Circle aviation museum. Walk this way and take a look at this interesting zeppelin. We’re not sure it ever flew, as it is constructed all of lead and, while it has a proper place for a gas bag and engines, we never found them.

Oh wait. The hamsters had a field day with this one…

Inside the showcase is the band Led Zeppelin!

Full Circle is on HKC Radio.

Two people have been abducted from Scotland in the middle of the night by the US military. Despite an order from our highest court that they be kept here.

The Scottish Government must show that actions have consequences by evicting American troops from their base at publicly-owned Prestwick Airport.

the memory of the Holocaust, antisemitism

Sensitive content

This entry was edited (1 day ago)

I try to be thorough and as generous as I can manage, particularly in my long-form AI-critic writing, but this video articulates something a bit less diplomatic: ultimately using it strikes me as somewhat shameful and embarrassing. But Dr. Collier puts it better than I can.

youtube.com/watch?v=7pqF90rstZ…

in reply to Glyph

Based on my reading of the original Nature blog post, it seems to me that the ChatGPT UI was less than clear about what turning off the data consent flag would do to *existing* data. So I blame OpenAI, not the professor. I think it's misleading to state, as Dr. Collier does here, that he clicked a "delete" button without realizing what it would do, and especially to take that as evidence of brain rot. We need to be careful not to succumb to confirmation bias.
in reply to Matt Campbell

@matt there is some evidence that the professor may be misrepresenting the “data consent” button that was clicked, at the very least. lots of people in the comments saying that it really is a big red “delete” button, others saying it’s different in europe because gdpr, but I don’t think I have seen anyone saying that they see precisely what he claims he saw. otoh maybe openai patched it since the article is bad press, so we can’t check now. the infrastructure of our lives is built on sand

How does the @mixxx open-source community create software that rivals commercial alternatives? Collaboration between users and developers in the community chat on Zulip lays the foundation. New case study: zulip.com/case-studies/mixxx/ #openSource #chat