nightpool

I hope you reconsider your use of a throwaway account for this sort of comment—This sort of feedback and experience is very valuable and if the culture of the company you're in or the companies your working with discourage this sort of feedback, it's only going to create chilling effect that prevents more people from coming forward

Your comment makes sense but I'm pretty sure yreg is saying the opposite of that—that their task is harder than the benchmark currently is.

(see their follow-up reply: "The cheapest-per-benchmark-task model would be useless to me if it cannot do the task I need.")

In either case, you need the right benchmark for the right task

Based on https://web.archive.org/web/20260618001349/https://washingto... and https://web.archive.org/web/20260621034150/https://washingto..., it seems like they were added at some point between the 18th and the 21st. I definitely think it helps navigating the site, and it's a nice compromise between Craigslist's text-first style and a nod towards usability.

You're making inigyou & OP's point for them. Redis is a great technology, but its design (supporting both persistence and non-persistence modes) makes it much easier to misuse and much more likely to be misused compared to Postgres and Memcached. That's a design issue, not just an internal documentation issue.

Okay, but how does that square with the person I'm replying to's claim that VW should "put security into the app"? You missed the point of my reply.

I don't disagree with this, but you're completely contradicting the user I'm replying to. I'm pointing out that the user who says "you should put security into the app" is completely misguided or at the very least perfectly explaining why VW is doing this.

How else would you build "security" into the app (in the sense of not allowing third-party modifications of it that would open them up to liability), except relying on hardware attestation that the app has not been modified? That attestation necessarily requires the platform provider to be involved.

Is this legal to use in 2 party consent states? Might vary from state to state, which is probably why both zoom and Meet require users to click through consent screens when meetings are bring transcribed. Might be useful to have that on the FAQ page

No, the CSS spec is specifically designed to be forwards compatible because of exactly this issue. Any invalid CSS rule should only cause that specific line to be ignored, not the whole stylesheet. And certainly even if your CSS parser chokes in some specific case, it shouldn't cause your ereader to fail to load the entire book!

it works because Nex N2 is also a derivative of the original base Qwen model. If it was two completely unrelated models it wouldn't work.

Right, which is exactly what elashri is objecting to. elashri said "Why do LLMs have restrictions on nuclear science", and IncandescentGas was explaining why they think those guardrails are a good idea. You're just agreeing with them.

This is purely a function of what people expect to be looked at.... if your team started looking at each individual commit step by step, then you'd probably find that people started paying more attention to how their commit hygiene looks.

No, in the context of the thread it sounds like they're illustrating myrmidon's point about how the selective enforcement of crimes that are easy to catch on camera means that the police have less time (and less inclination, training, norms) for addressing more serious crimes, like interpersonal violence.

More broadly, they're not saying that we should make the cameras better to catch more crime, they're saying that when you make cameras the main way you catch crime, you shift the social definition of what crime is to "what cameras can catch".

There is a version of that. It is called RedHat Enterprise Linux. : )

I'm very glad to hear that—the anonymity of the original Freenet has led to it being a very unsavory place that was more well known for CSAM then anything positive or useful. As an outsider, it sounds like this new direction is the right choice for Freenet to try and attract new users and fulfill the team's original goals.

I'm not seeing anything on the official marketplace: https://marketplace.visualstudio.com/items?itemName=whatwedo...

I wonder if it was open-vsx specific?

No, there's no joke, you might have just misread the article (the 3,800 number is the number of internal GitHub repos the employee had downloaded on their personal computer / had access to on their own GitHub account)

Did Railway give admin credentials to delete their production database? My memory of the incident is that a customer of Railways used an AI tool to delete their production database, and then blamed Railway for it. The customer was the one who put their own account credentials into their own AI, not Railway

Fly, Render, and even Heroku still are all better choices then working with Railway I think

It did, V8 added asm.js compilation to WASM in 2017 https://v8.dev/blog/v8-release-61#asm.js-is-now-validated-an...

Yes, if you watch the video closely you can see that the "lensing" effect only really covers a circular area—this prompt probably went through multiple iterations where the author was trying to improve it so that the shape of the hand was reflected more closely.

Okay, sorry, that was obvious in retrospect, I definitely feel kinda stupid now that I see it. In fact, I agree with your comment on almost all counts—I just see a lot of misuse of the term "monopoly" online, and I think I was led down a garden path by one of my sibling commenter's mention of Lina Kahn. No fault of yours, and I'm gonna delete my comment if I can :)

[EDIT: I'm an idiot, sorry, completely misread your comment. I agree on all counts]

That would be great to lead with since it's not present in any of the blog post communication anywhere.

Autoformatters are deterministic tools that are tested regularly, with extensive test suites and went through a very long process of production usage and review before the reformat:

    > We also built a tool to diff ripper trees across formatted files, accounting for things like rubyfmt converting single quotes to double quotes. Combined with our extensive test suite, we built confidence slowly and deliberately.

If an autoformatter is working right, it's only changing whitespace—not the actual code executed. Changing between two different implementations of the same function is very different from changing whitespace around.

They also didn't do the entire thing in one weekend—that was just the article title clickbait. they did it file by file, incrementally, over the course of months:

    > Rolling out a novel autoformatter to 25 million lines of code has two big risks: merge conflicts and correctness. A bug affecting just 0.01% of lines would still touch tens of thousands of files. To manage both, we built in a per-file opt-in so rubyfmt would only format files that explicitly asked for it. Following the Developer Productivity org’s typical pattern, we started with systems we owned and could observe closely, then expanded coverage gradually as our confidence grew.

^ See how they talk about the incremental changes it took? This is what mature refactors look like. And they were only changing whitespace!

Arbitrary demarcations can still be valuable! Just because something is arbitrary doesn't mean that it's not helpful. Working in chunks will let you take more time to review each callsite individually, and increase your confidence in the changes

In the future, I would definitely encourage you to explore a more iterative solution—fix the first 50 occurrences first, or maybe all the occurrences of a handful of functions. For example, if you have utility functions A, B, C and D, maybe fix functions A and B first, and then C and D second.

Ultimately, at the end of the day it's going to depend on how much code you're touching. If you're only touching 100 library calls, then it's probably easy to do them in one PR. But if you're updating 1000 library calls, you'll need to take a more iterative approach. Building those skills now will serve you well in the future when working on bigger codebases and harder refactors.

The history of the Honor Code system might be instructive: https://universityarchives.princeton.edu/2015/01/i-pledge-my...

Exames were previously proctored, and it led to a "us vs them" mentality that meant students banded together to

The Honor Code system, and removing proctors was a way to route around that—it made all of the students responsible for catching cheaters and turned the "Students vs Faculty" mentality into a "Honor vs Cheaters" mentality among the students.

Unfortunately, it seems like the "Students vs Faculty" mentality has seen too much of a resurgence due to outside factors, and the Honor Code is no longer a match for the current climate. That's what the article is about

coppsilgold is the one who made a hard-line, clear-cut dichotomy when they said "it's easy to do harm [but] it's all but impossible to do any good". bglazer referenced several interventions that are known to increase IQ which challenge this dichotomy. Saying that it is difficult to separate "doing good" and "stop doing harm" is agreeing with the point that coppsilgold created a distinction without a difference.

> And every other suggestion also doesn't work if the attacker can just replace the shell.

No, debt isn't "parts of your software that could have been written better". Any part of your software can always be written better. Debt is the cost you have to pay monthly to keep your application working—it's the parts of your codebase that make it harder to work on new features.