The importance — or not — of reputation

We talk a lot in open source about reputation. Individuals have a reputation. Projects have a reputation. This reputation is how we build trust with strangers from around the world. People and projects have an incentive to behave well so as to not ruin their reputation. Or do they?

@miss_rodent yeah, that's kinda what I mean. The whole industry behaves as if you have a strong incentive to behave in a particular way because we have some strongly-tracked pervasive notion of reputation. but we barely have any notion of reputation *at all* let alone a structured and carefully enforced one. if this breaks the floodgates on activism-via-RCE, that broken trust is going to take a long time to repair

2026-05-30, 3:18 am 0 boosts 8 favorites

Glyph is right. We put too much on the concept of reputation without stopping to think about what it actually means.

One way that I’ve seen this come up a lot is in conversations about blocking AI agents — or humans who are just a translation layer between an AI model and a project. Folks have come up with a variety of different ways to determine who is a real, trustworthy person that should be allowed to make a contribution to the project. Some, like Mitchell Hashimoto’s vouch, use an explicit maintainer vouching model. Others use heuristics that look at account activity to make a guess. Both of these models can make it harder for newcomers to make those early contributions that build their reputation.

Discourse’s trust levels are a pretty good model for a trust ladder in a community. The problem is that once you go to a different Discourse site, you’re brand new again. Similarly, someone who has been banned from a community for repeated misbehavior can join a new community with no trouble.

In chapter three of Program Management for Open Source Projects, I talk about trust being a combination of person and role. You might trust me when I write about leading open source communities but not when I write critical software. By the same token, I’m relatively well-known in places like Fedora and the OpenSSF. But at an Erlang conference, nobody has heard of me.

If you’ve gone to a conference, you’ve probably had an experience along these lines: you chat with a friend-of-a-friend in the hallway for a few minutes, think “they seem nice”, and then later you learn they invented your favorite compression algorithm. Even the biggest of the Big Names are a nobody to a lot of people.

So reputations? Not that useful. If you can build a good one, that’s nice, but you can’t count on it.

The problem with reputation is that it doesn’t answer the question you want to answer (unless that question is “who is well-known?”). Start by figuring out what question you want answered. Then you can find the best way to answer it. And don’t rely on “but you’ll ruin your reputation” to prevent bad behavior.

This post’s featured photo by David Clode on Unsplash.

Ben is the Open Source Community Lead at Kusari. He formerly led open source messaging at Docker and was the Fedora Program Manager for five years. Ben is the author of Program Management for Open Source Projects. Ben is an OpenSSF Ambassador and frequent conference speaker. His personal website is Funnel Fiasco.

Share