Meta said a misconfiguration by the independent testing company Irregular inadvertently allowed one ⁠of its models internet access during an evaluation, adding that it was investigating the incident.

lol meta is like “guys, it was our model, but not our responsibility for this oopsie”

what I do feel is how slow poetry is for resolving dependencies

I used poetry for years before uv, but uv is just a better tool. If openai enshittifies it, I have no doubt it’ll be forked instantly.

yes, and give uv a try

https://docs.astral.sh/uv/getting-started/installation/

uv init + project management commands uv add, uv remove, uv sync, uv run my_spaghetti.py, …

you probably don’t need pip or uv pip commands

and here’s how to use it in docker

https://docs.astral.sh/uv/guides/integration/docker/

See your desktop and log keypresses. It was pretty trivial to write keyloggers. If Linux running x11 would be nearly as popular as Windows, it’d have been a disaster.

It’s right, you shouldn’t be using requirements.txt in 2026

Apps are not sandboxes unless they’re implemented as such. Otherwise any user process can read /proc/<pid>/comm and list what else is running. They also have access to any location on disk that the user running it does. I wouldn’t call that sandboxed.

you won’t believe it

idk what’s the point of creating the anticipation if no mortal will be able to run this locally; you need $100k on hardware to start considering this.

deleted by creator

if the orange man hears about this…

set it up with wrong address/credentials to see it failing, if it succeeds, it’s not using your relay server

I’ve seen the login warning above, but I don’t self host it myself, so I just logged in

https://en.wikipedia.org/wiki/State_of_the_art

Right, like I mentioned, it’s a scale problem and how people feel about generated code. Abandonware have always been the majority of repos, because people don’t have time or interest to maintain most things they create. We just have more things being created now (whether they’re any good/usable or not).

Disagree. Most projects are never “done”. Whether that is defined by the user or by the users, there are nearly always things to work on. Maintainers just lose interest or don’t bother. This is as old as open source and it has nothing to do with LLMs, it’s only aggravated by it.

Hardly an issue with generated code. You could say the same about projects before LLMs were widely used for code generation: “most projects are abandoned within months of release”. The difference now is the scale and how some people feel about it.

finally I get the same level of access on my phone as 3 letter agencies have

5min at 410 + 10min at 425, perfect avg

just keep rotating