tkzed49

I'm curious what industry you work in. Rarely have I run into a large problem that was tempting to solve with Stackoverflow/Github-driven development. I find the limiting factor is never "we don't have the code" and more "we don't know what code we need".

nothing against Algorand, there's just a kind of nice fuzzy feeling that lua gives me. Plus the name is a good pun

I vote we give the name to the lua one!

hey, I'm open to that possibility. Maybe I'll grab $5 in API credit and give it a shot (for 5 minutes or a week depending on who you ask)

I'll add on to this: I don't really use agent modes a lot. In an existing codebase, they waste a lot of my time for mixed results. Maybe Claude Code is so much better at this that it enables a different paradigm of AI editing—but I'd need easy, cheap access to try it.

The problem is that this is $100/mo with limits. At work I use Cursor, which is pretty good (especially tab completion), and at home I use Copilot in vscode insiders build, which is catching up to Cursor IMO.

However, as long as Microsoft is offering copilot at (presumably subsidized) $10/mo, I'm not interested in paying 10x as much and still having limits. It would have to be 10x as useful, and I doubt that.

Seems like the same tired take on third wave coffee, without much specific to Iran. Some people like the flavor of coffee grown in different regions, just like tea or wine. If you want a classic espresso roast blend of whatever, I haven't been to a cafe that won't sell you that.

if you go too deep it starts to blow up!

https://hypertext.tv/test/hypertext.tv/test/hypertext.tv/tes...

just because web developers like to ship bugs doesn't mean that browser APIs break? Realistically web apps will keep shipping new versions that use new APIs, breaking old browsers. But that doesn't really speak to the platform—e.g. an electron-like or webview based app will probably keep working.

Really, if you're using an online app, you should be updating your browser for security patches.

yeah, this is exactly my point haha. these are not at all the same

I disagree, there is absolutely no easy way to do it on Windows. You can install a third party program that emulates the compose key but on macos it "just works". And I think that makes a difference for 95% of users

I run a pihole and I don't regret it, but it's nowhere near ublock in capability. Ublock's more important filters are selecting individual scripts and even page elements.

are you implying that society shouldn't aim to reduce human interaction with vomit, feces, and dead animals? Robotics in harsh environments isn't unheard of

Depersonalization and derealization are truly awful. Having experienced them for non-psychedelic reasons, it's almost indescribable.

Sometimes people say "seeing events in third person". My experience was that my consciousness and actions were completely disconnected from my observations of reality. Like, I questioned whether I had any influence at all over my existence. Basic, predictable events were suddenly uncertain and terrifying. It left me with no mental capacity to do anything but uneasily exist.

With treatment, it goes away gradually over months. I never want to go back there.

I think a lot of people here maybe haven't used a PR + squash merge flow? Because I agree with you, when I'm doing operations on git history I just see the PR title from the squash merge. Squash merge lets the PR be the atomic unit, so commit messages don't matter as much.

But if I read these comments as "I wouldn't let an AI write my PR description", I strongly agree.

unfortunately, they know what they've got! Gandi has .bs renewing at $400 per year!

eh, do what you want imo

maybe it's less the diplomacy and more that they were sitting on a pile of crusty patents instead of building the iphone

Where I work, this would have no impact on the amount of tasks shoved into the pipeline by product and leadership.

That's fine—same with CPUs, right? But if I do something that's slow on _my_ CPU, at least there's tooling that will show me!

I know that the reason is a lot of technical complexity (plus little standardization between vendors), but I think the end goal should be to make GPU programming more accessible.

awesome, I'll give it a shot. I'd lost all hope of ever getting Nsight to hook into WebGPU

I wish there were a good way to profile WebGPU code. I've seen this (very useful) article[1] on setting up PIX, but I'm ambitious. I want to see everything from draw call timings to flamegraphs of shader code.

Right now I feel like the only way to write efficient WebGPU code is to deeply understand specific GPU architectures. I hope some day there's a dev tools tab that shows me I'm spending too much time sampling a texture or there's a lot of contention on my atomic add.

[1]: https://toji.dev/webgpu-profiling/pix.html

debatable

Very cool! Kind of seems like a relic at this point though, good for studying but probably better to use Pyodide [1] in practice? This is what powers JupyterLite [2], which is a fully fledged Jupyter IDE with support for packages, in browser.

This is one of those things that's "just cool". Would have blown my mind to see it 10 years ago.

1: https://github.com/pyodide/pyodide

2: https://github.com/jupyterlite/jupyterlite

Hey, that's awesome. Seems like a great case for verifying things that way! I'm thinking of different cases I think, where you've got something like a distributed system or a web application and it's just harder to fuzz. But I agree, that seems a lot more practical than formally verifying it when that's the kind of testing you need.

This is a very anti-pragmatic way of looking at things in my opinion. The program is not a snapshot that exists in a vacuum. Most programs are going to grow over time and have things added and removed in various places by many people. They're going to have many interfaces and operations.

I think you're arguing that tests and types can both be used to check a particular case for correctness. Sure, this is true. However "moving type checking code into the test suite" means nothing—that would just be type checking.

When you make a code change, there's a difference in the kinds of feedback you get from your tests and your types. Tests usually cover business logic or stories—things that you want or don't want. Types ideally cover everything. They check every operation applied to a given piece of data. Of course types are rarely precise enough that they can catch every logic bug (e.g. strings with semantics not encoded in types, like email addresses).

This is just scratching the surface. You might just have to try out both to get a better sense of how they feel to work with.

I strongly disagree on the principle that tests and types have a 100% overlap in the problems they're solving.

Do you already have candidates actually write frontend code? Either async or during an interview, or both? I'm a big fan of being on either end of an interview where I'm actually working on a functioning project.

I really appreciate this. Most of the time when I see a "desktop OS on webpage" it feels half-assed and honestly overplayed. This on the other hand is super tight and polished!

lol you have too many meetings. I don't care if the manager with back to back meetings listens in from transit or something. im literally suggesting being more flexible