rienbdj

For images and other media, yes. Does it look right?

Program correctness is incredibly difficult - arguably the biggest problem in the industry.

> Revealing against what?

Revealing of what it is like working with an LLM in this way.

The commits are revealing.

Look at this one:

> Ask Claude to remove the "backup" encryption key. Clearly it is still important to security-review Claude's code!

> prompt: I noticed you are storing a "backup" of the encryption key as `encryptionKeyJwk`. Doesn't this backup defeat the end-to-end encryption, because the key is available in the grant record without needing any token to unwrap it?

I don’t think a non-expert would even know what this means, let alone spot the issue and direct the model to fix it.

I tried Copilot many times and found it lacking. I like discussing concepts (maybe with pseudocode) with Gemini. I find this useful but not game changing.

So genuine question for AI Boosters, what should I be using?

I would like the 10x productivity gains we were promised :)

It can give back code though. It might be wrong, but it won’t be ambiguous.

If you give an LLM the data in the prompt and then ask it to extract information from that data it does pretty well. This is the premise of RAG. Where LLMs do poorly is when you ask it for information you haven’t given it.

Why not use an existing OAuth library?

How can you work around if you don’t have millions upfront?

Go to a local art college and put up a flyer: Paid gig. Ask applicants for a sample of past work. Meet the best for coffee.

I somewhat agree. Good game design requires a great idea, but more than that a lot of iteration. If you are working solo you need to hold the implementation and design in your head, which is incredibly difficult. Iteration is slow. You need someone to bounce ideas off of.

The best indie combo is a programmer plus a designer/artist in my opinion.

Alternatively, you can spend a very long time iterating solo. Most don’t have the runway for this.

This is also why big studios produce technically impressive games with mechanics seen thousands of times before. The production risk is much lower for “make spectacular assets” than for “come up with a novel game mechanic”.

Take a look at F# for a superior list comprehension design.

The best interiors feel lived in but not stuffy. A few old pieces of furniture can help achieve that. Part of the skill of interior design is making these pieces work together in new ways.

You don’t want your home to feel like an ikea catalog surely!?

In Java you can write multiline anonymous functions and in Python you can’t (well, not in a reasonable way).

Java has immutable bindings with the final keyword.

Java variables are lexically scoped, which discourages mutation of globals.

This is true but what’s odd is that Java has better functional programming support than Python.

F# is strictly better than C# if (huge if) you have a team that knows both languages.

What this skips over is that some languages are better than all others in a domain, sometimes multiple domains. And also some languages are not best in class in any domain.

Would you pick Brainfuck over Java for anything (real)?

The spacing changing with scrolling makes me feel … nauseous

If you like the exploration type stuff, maybe try the Van Rijn series. It’s loosely based on Dutch merchants but in a sci fi setting. Some interesting discussion of capitalism, trade and exploitation too. It is also very much of its time.

That, and genuinely new questions get aggressively shut down.

In defense of R the class systems do have different characteristics and they are not deeply embedded in the language or anything.

How long till the Rust community starts to push do notation and monad transformers?

Time is extremely scarce for poor people. Cooking healthy and cheap meals means research, planning and batch meal prep. This also requires being a forward-looking state of mind.

Personally, I think the government should be investigating ways to raise income levels for the poorest. I don’t know what form this should take - UBI? negative income tax? - but fix that and the rest would follow for many people.

An adult can either be a net tax contributor or a net tax burden over their lifespan. The probability of them being a net contributor is much higher if they are well educated and not malnourished. The effect is stronger at younger ages. It’s good ROI for the tax payer.

In the specific case of free school meals:

- economies of scale in feeding an entire school

- recipients of the benefit (children) and delivery mechanism (kitchens) are already at the right place and time

- can guarantee money is spent on food rather than (for example) a parents gambling addiction

- reduces stigma for poor children to have universality

- can (in theory) deliver better food using school cooking facilities

- saves time even for parents who can afford to feed children

- teaching a hungry child is an uphill battle

Except lots of engineers now sling AI generated slop over the wall and expect everyone else to catch any issues. Before, generating lots of realistic code was time consuming, so this didn’t happen so much.

Would be ironic if Google make the a 100% Rust browser before Mozilla.

Drake's favorite MIME type is application/pdf? Who knew!

> Using AI well is a skill you need to learn and if you don’t see positive results on your first couple of attempts that doesn’t necessarily mean it’s bad, it just means you are a beginner.

I think it’s a very easy skill to acquire, which is why we’ve see such fast user growth for these products.

Most of them have the same interaction model as talking to someone on iMessage / WhatsApp / Messenger.

You haven’t tried a wide variety of languages then. I have found the language chosen massively shifts what is possible to write and what is convenient to write.