Which language's open source projects answer a stranger fastest?

The languages differ by hours. The projects inside them differ by days.

Updated November 3, 20262 min readMarkdown

On this page

The typical Go project takes 18.9 hours to answer a stranger's pull request. Dolt, also written in Go, takes 24 minutes.

That gap is why I stopped asking which language answers fastest. There is an answer, but it barely helps you.

I track this in Porchlamp, a ranking I built of open source projects by what happens to pull requests from outsiders.

For each of 3,246 projects, I took the median first reply from a person, not a bot. The data is its November 2026 snapshot.

The languages differ by hours

The median Rust project replies in 15.0 hours, the fastest of the six. The median Python project takes 22.7, the slowest.

That's about eight hours. It's real, but it's less than a working day.

Take Rust. A quarter of its projects answer within about 6 hours, and another quarter take more than 37.

Python runs the same way. A quarter answer within 11 hours, and another quarter take more than 60.

So the language you already know is fine. Choose the project with care.

Go. Dolt, a database you can branch like Git, answers in 24 minutes. Ebiten, a 2D game engine, in 41.

Rust. probe-rs, for debugging embedded chips, answers in 58 minutes. The compiler itself takes about 70.

JavaScript. AG Grid, a data table for web apps, answers in 45 minutes. PlayCanvas, a web graphics engine, in 1.7 hours.

Java. Vespa, a search platform, answers in 1.3 hours. CrateDB, a distributed SQL database, in 1.6.

C and C++. RIOT, an operating system for small devices, answers in 1.2 hours. SDL, the media library many games use, in 2.0.

Python. Astropy, the astronomy library, answers in 1.8 hours. Click, for building command line tools, in 2.5.

An early reply doesn't promise a merge, but it does mean someone is reading.

What's the quickest answer you've had from a project you didn't expect to reply at all?

Rahul Kumar
The model you need rarely fits where it has to run, and the number you are shown is rarely the one that decides. I work on both: vision and sensing on phones, large models on one GPU, and automated trading after costs and tax.