PWV Portfolio Feed

AI is the best tool for builders.

We are moving past writing every line of code by hand. Using LLMs for the heavy lifting lets you focus on the big ideas and solve harder problems faster. It has never been easier to create.

What are you shipping this week?

Steam engines

Appreciate it! If you know what you want changed, we'd love a docs pr!

😊 See you tomorrow morning!

infra that feels like a product gets maintained

infra that feels like punishment gets ignored

your test suite is a product. it has users (your team). it has SLAs (CI time). it needs DX work.

In the coming release cycles, we'll be dedicating engineering focus to improving Zed's stability and safeguarding the stable channel going forward.

I think we're doing a great job of testing how little reliability people will actually accept 😬

yeahhhhh we just very quickly accepted 1-2 nines as ok

I miss when people actually gave a fuck about reliability

(credit to @samwho.dev for mentioning this!)

The result:

⚡ 0.1–1ms -> 1–10µs per query
📈 Higher throughput
📉 Lower latency
🎯 ~100% cache hit rate in typical apps

Under concurrency, the difference is significant.

In v7.4, we introduced a schema-aware query caching layer.

Instead of having to rebuild your query on every run, Prisma now:

• Parameterizes dynamic values
• Reuses the static query structure
• Avoids repeated compilation work

In v6, queries happened in a separate Rust-based thread.
In v7, queries happens in the JavaScript event loop.

Individually, each query is able to run fast (0.1–1ms per query).

Under load though, there is an increase in latency due to the single threaded nature of the JavaScript

When we shipped Prisma ORM v7, we moved away from Rust to simplify the architecture and invest in long-term flexibility.

But it came with a trade-off: performance under high concurrency.