Mastodon Feed: Posts

Boosted by jwz:
jef ("Jef Poskanzer") wrote:

#algae

Boosted by ChrisWere@toot.wales ("Chris Were ⁂🐧🌱☕"):
ChrisWere@toot.wales ("Chris Were ⁂🐧🌱☕") wrote:

The Space Virgins are back talking about a mediocre Stephen King movie from the 80s:

https://video.thepolarbear.co.uk/w/6vAn6HhEUt2JJyxTrRvzUo

Boosted by glyph ("Glyph"):
eestileib@tech.lgbt ("eestileib (she/hers)") wrote:

My kid and I are trying to get something done with an airline that requires talking to a person.

On our third roguelike pass through the phone tree, it asked whether we wanted to "talk to an agent", and my kid immediately said "no", meaning we had to start again.

I asked why, and my kid says "we're trying to find a person, not an agent".

That's what, less than three years for my kid to associate the term "agent" with "having my time wasted".

Good work AI industry.

Boosted by glyph ("Glyph"):
mrencyclopedia@retro.pizza ("Mr. Encyclopedia") wrote:

Another AI argument: "I'm studying Japanese and asked Claude to make flashcards based on the grammar points in my Japanese textbook, and it did in minutes what it would have taken me hours"

What if I told you that the act of making the flashcards is studying? When I went through the Naval Nuclear Program I went through a very similar process of making flashcards for all the things we would be tested on. I never actually had to use the cards. Making them was all I needed to do to sufficiently remember.

There is no way to optimize your way out of the learning process. It always takes the same amount of effort in the end.

Boosted by soatok@furry.engineer ("Soatok Dreamseeker"):
baralheia@dragonchat.org ("Baral'heia Stormdancer ΘΔ🐲") wrote:

The more you can minimize your usage of - or outright quit using - these services, the better. Nothing will change if you keep giving them your visits and ad revenue. Don't feed the beast if you can help it, y'all!

neatnik@social.lol ("Neatnik") wrote:

All maintenance is complete!

Boosted by NfNitLoop ("Cody Casterline 🏳️‍🌈"):
Lana@beige.party ("𝐿𝒶𝓃𝒶 "not yet begun to fight"") wrote:

"when did Star Trek get woke??"

In the very first episode of Star Trek: the original series, we see a white Captain reporting to his black Admiral boss, a black woman on the bridge just a couple years after Jim Crow was abolished, wearing a short skirt (a symbol of feminist liberation at the time), a Japanese helmsman on the bridge only 20 years after the internment camps, a Russian crewmate on the bridge during the Cold War [edit: actually did not appear until Season 2 but the point stands], and the foundation of the modern concept of queercoding.

In the very first episode of Star Trek: The Next Generation, we see male crossdressing crew members, a female officer on the bridge in charge of security, a literal ship's counselor stationed at all times on the bridge, a single mom raising her teenage son on her own while juggling a full career in medicine, a blind mechanic whose "disability" is shown to be a strength, and an angry, all-powerful godlike being who is revealed to be simply a petulant child masquerading as a deity.

In the very first episode of Star Trek: Deep Space 9, we see a black man gain a powerful command position, respect the hell out of the customs of a religion he didn't understand, show respect and equal treatment to members of three other alien races he didn't understand, appoint a female guerilla fighter who defeated imperialist fascists to a position of authority within his administration and defer to her judgement in areas of her expertise, accept his friend's gender change, and tell his son he loves him.

Star Trek has always been woke. You just grew up to be a bad person.

Boosted by cstanhope@social.coop ("The Luddites were right"):
mhoye@cosocial.ca wrote:

"There are colors that I want to show you, but I can’t. They exist in the real world. You probably saw some of them today, but I can’t show them to you on a screen. A digital photograph can’t capture them, and your screen can’t display them. No game you’ve ever played has contained them. Unless you have specialized equipment, they are entirely absent from the digital world.

Most of them are cyans."

https://moultano.wordpress.com/2026/06/19/where-to-find-the-colors-your-screen-cant-show-you/

Boosted by NfNitLoop ("Cody Casterline 🏳️‍🌈"):
vantablack@cyberpunk.lol ("🌈 vanta rainbow black 🌈") wrote:

#trans #transgender

neatnik@social.lol ("Neatnik") wrote:

Our weekly maintenance window starts in ~21 minutes. All Neatnik/omg.lol servers will be restarted, but downtime for any given service should be minimal (~60 seconds, if that). If anything seems goofy, feel free to check https://status.neatnik.net. Thanks for your patience!

Boosted by soatok@furry.engineer ("Soatok Dreamseeker"):
briankrebs@infosec.exchange ("BrianKrebs") wrote:

Don't look now, but it seems Gizmodo's homepage is now serving up a Clickfix attack.

Basics of the Click-Fix exploit, which causes a pasted URL to fetch malware via Windows Powershell.

https://krebsonsecurity.com/2025/03/clickfix-how-to-infect-your-pc-in-three-easy-steps/

#clickfix #gizmodo

cwebber@social.coop ("Christine Lemmer-Webber") wrote:

I'm getting a bit better at blender shaders

cwebber@social.coop ("Christine Lemmer-Webber") wrote:

remember when they put out UV lights instead of purple lights and cooked the people at the bored ape festival

EmilyEnough@hachyderm.io ("Emily 🏳️‍🌈🏳️‍⚧️") wrote:

Well, it happened.

I automated letsencrypt a couple years ago because I was tired of renewing manually.

I just noticed one domain (out of a dozen) now has a wildly out of date TLS cert and I have zero recollection of how I automated it in the first place or how I’d go about fixing just the one domain.

jonny@neuromatch.social ("jonny (nonvenomous)") wrote:

needless to say, this would be extremely easy to patch, it works as of 2.1.173 at least, which was released a week ago. but if this were to become widespread it would def get immediately patched, so get it while the getting's good

fromjason ("fromjason.xyz ❤️ 💻 ✍️ 🥐 🇵🇷") wrote:

Anyway, as soon as these fucking headaches subside and I can catch up on work, I'll write my first essay. I'll play with ideas until then. Yay I'm excited!

fromjason ("fromjason.xyz ❤️ 💻 ✍️ 🥐 🇵🇷") wrote:

I've been thinking so much about what creative outlet I want to really pour myself into. Is it a newsletter? Shorts? Sketches?

And then it hit me even though it's been right there. Long form video essays scratch every itch. Writing, performing, producing, editing. It's all there.

I've thought about doing video essays a thousand times before. But it really hit me tonight. Of course that's the move. Ahhhhhhhhhh!

jonny@neuromatch.social ("jonny (nonvenomous)") wrote:

another thing that's fun: CLAUDE.md and a number of other special files get dumped directly into the system prompt. It's possible to @include other files from CLAUDE.md. included files can be any text file, with a hardcoded list of extensions including .env files.

There are some protections against @includes for files outside the project directory, but there are also a number of easily visible bypasses that i won't describe because that crosses into irresponsible disclosure territory.

So it certainly seems like you can use CLAUDE.md to load say idk an .env file with keys into it into the context, bypassing all the safety checks, and then treat some exfil path as just being project-specific tool calling instructions. neat huh?

Boosted by glyph ("Glyph"):
grimalkina ("Cat Hicks") wrote:

“This book presents an empathic, evidence-based analysis of developer productivity, and provides practical guidance based on the author's own research for avoiding or fixing common traps. The result is the most important new perspective on software development in years.”

- Greg Wilson, third-bit.com

This one meant a lot to me Greg, sorry for the public shout-out which I know you might hate. Being able to finally send you my book after you so encouraged me to write was such a meaningful moment.

Boosted by glyph ("Glyph"):
grimalkina ("Cat Hicks") wrote:

“The Psychology of Software Teams by Cat Hicks is a vital companion for any leader building a humane, high-performing organization. She skillfully dismantles the ‘Brains-in-Jars’ myth, proving that innovation is not a solitary act but the result of social learning. By introducing cognitive scaffolding and fostering thriving ecosystems, Hicks provides the missing link between organizational design and the individual human experience."

- Matthew Skelton, Team Topologies

https://www.drcathicks.com/#book

Boosted by glyph ("Glyph"):
grimalkina ("Cat Hicks") wrote:

“If you lead engineers and believe culture is ‘soft,’ this book will disabuse you of that notion quickly. Psychological safety, learning, and collaboration aren’t perks, they are infrastructure. Ignore them and your systems will fail, slowly or catastrophically.”

- Scott Hanselman, VP of Developer Community, Microsoft

🔥🔥🔥

https://www.drcathicks.com/#book

jonny@neuromatch.social ("jonny (nonvenomous)") wrote:

RE: https://neuromatch.social/@lina/116779058863273132

If anyone wants to prevent LLM contributions, it seems like symlinking CLAUDE.md to /dev/urandom might do the trick: by default, claude code unconditionally reads CLAUDE.md if it is present in the repository root. The reading function checks if the files to load are block devices among a few other things, but does not mind at all if they are symlinks in general, including directory traversal to anywhere.

There are ways to run claude without reading CLAUDE.md, but they require manual intervention, and startup stalls before any of the "try whatever" self healing attempts of the LLM kick in. indeed there is no indication why startup stalls in stdout or debug logs: nothing for an external "agent" harness to diagnose either.

I have not confirmed this part: but it looks like the file read tool checks if a path is a system path like /dev/urandom, but it only checks the string of the path without expanding symlinks. Many of the tool and agent prompt texts say to read CLAUDE.md first, bypassing any config settings that would prevent that since it happens in LLM modality not program modality. so even if you do get it booted up, you would likely get inexplicable stalls at any point those tools/agents are invoked.

Boosted by glyph ("Glyph"):
mcc wrote:

I 'm on Mac. I'd like to set an "Administrator policy" for my copy of Firefox.

Mozilla seems deadset on preventing this. Their docs https://firefox-admin-docs.mozilla.org/guides/getting-started/#configure-policies say that on a mac, the way to do this is to "use JDAM or MDM", acronyms I've never heard of and which they don't explain. They do say policies.json works on mac, but to use it https://firefox-admin-docs.mozilla.org/guides/policies-configuration/#configuring-firefox-using-policiesjson you place it *inside Firefox.app*, which causes the MacOS signing to fail and the app to refuse to run.

What is the easiest way to proceed?

Boosted by kevinevans@hachyderm.io ("Kevin"):
dan@discuss.systems ("Dan Ports") wrote:

Early American Chinese restaurant menus were designed to be parsed as SQL.

Boosted by glyph ("Glyph"):
darius@friend.camp ("Darius Kazemi") wrote:

Nominative determinism, first theorized by the ancient Roman scholar Nominus Determinus

Boosted by glyph ("Glyph"):
jplebreton ("JP") wrote:

crush this wherever you encounter it. crime numbers have not appreciably changed - what changed is that SF got a mayor that billionaires were happy with, and the doomloop narrative machine was wound down. rents are up 20% in the last year, homelessness is still really bad it's just been relocated 10 feet outside this guy's line of sight. tech is a scourge, SF is not a progressive or leftist city, it's been ravaged by decades of neoliberalism.

Boosted by glyph ("Glyph"):
shinmera@tymoon.eu ("Yukari Hafner :v_lesbian:") wrote:

Why's pride gotta be in summer and not during a month in which people can actually stay alive

Boosted by glyph ("Glyph"):
mathowie@xoxo.zone ("Matthew Haughey") wrote:

I cut my finger last night on a piece of metal and today with a band-aid on I learned how hard it is to use your touch mouse, your entire phone, and any touchscreen when your favorite pointer finger is temporarily non-capacitive. It's like unlocking an entirely new disability I never envisioned.

Boosted by glyph ("Glyph"):
ronald@blog.ronaldoussoren.net ("Ronald Oussoren") wrote:

I’ve uploaded #pyobjc 12.2.1 to PyPI with a number of small bugfixes. The most important of which is that the bindings for the Network framework now work on macOS 26. See the changelog for the full details.

Boosted by glyph ("Glyph"):
kboyd@phpc.social ("Kevin Boyd (he/him) 🇨🇦") wrote:

@phillmv Good work! I've emailed my MP about C-22 (mandating companies turn your data over to police, and preserve it for a year), but I might call or write a physical letter re: Bill C-34 (age verification)