Hacker News Reader

Comment by smallerfish | Hacker News Reader

Site is down, but one thing in particular that needs close attention: db schema. I've found that LLMs make awful normalization decisions; you can limit the fallout somewhat by modularizing database access, bit it still inevitably bleeds over into the object model etc. On AI heavy projects I tend to pause feature development every few days for code & schema review, and subsequent cleanups.

View on HN

Just do the schema by hand. It forces so much downstream alignment that you can almost prefer to use the schema and any associated comments over prompting the model directly.

If you don't know what your schema should be, then I think this is the dead canary in the vibe coding coal mine. Back up to the design phase and talk to the robot about the domain until you have a really strong conviction about its tuples and relations.

With agents code has gotten really easy to shape, like fabric. If want to build something rigid, you'd want to introduce sturdy architectural structures to stretch out the fabric of agentic code.

Data and its shape is one such rigid structure. It will stay with you long after the code around it has been replaced. You do want to be intentional about it and build intuition for it.

Once you have that in place, you can stretch out a web of agentic fabric across it, and pin it down with tests.

Code is soft nowadays, but data isn't.

My approach is similar. I review all changes that touch persisted data formats, wire protocol, and/or public API surface: things with bad consequences if they are wrong and neatly divide the project up. The exact boundaries vary by the project, but once I establish what they are, I can really relax about the other parts of the code, because I know they will be cheap to fix if wrong.

Back up now, Hashnode hiccup. Agreed on schema, that's one of the places I still always dive into regardless of what the agent says

And object-oriented design decisions, if you care I guess. I haven’t yet gotten access to the models that have any good intuition about what makes a good class and how to name it.

Naming, in general, is very similar to what I expect from colleagues who have weak English-as-a-second-language skills.

You let it run too wild and you’re suddenly left trying to reason why your LLM is trying to add an interaction between an Intermediator and a FromSpinMaxWriter.

Spinmaxxing lol.

FYI, your site is blocking me for some reason. I'm just on a regular residential connection in the US, no VPN or whatever. Something's screwy in your WAF.

Reply on news.ycombinator.com