Discover more from The Pull Request
On working (and writing) in public
Portrait by Katia Sobolski.
Nadia Eghbal is uniquely positioned to write about open source having spent almost two years in developer relations at the Alexandrian library of open source, GitHub. She then spent two years continuing her quasi-anthropological study of open source at Protocol Labs, and now works in writer relations at Substack (host of this publication). Her new book is Working in Public: The Making and Maintenance of Open Source Software, which like her career trajectory, starts in open source software but ends up grappling with larger issues of creators in an unbundled digital economy. The Pull Request review is here.
AGM: My naive mental model of open-source was this almost communitarian kibbutz model. And yet, the big lesson from your book is that that’s not really how it works.
NE: Part of the reason why I wrote this book was because I feel like we've had this communitarian kibbutz kind of model, which you've identified, is the prevailing model that people understand in open source and that gets frequently talked about. And I think that narrative has kind of been owned by the likes of [Richard] Stallman or Eric Raymond or anyone who kind of remembers those early days of open source. And that model definitely still exists within the matrix of different community models. The ‘clubs’ are kind of like that, where everyone is rolling up their sleeves and there's lots of different active contributors. And then we also have the ‘federations’ that are kinda like the really big open source projects that we're used to thinking about like Linux, but then there’s the rise of the ‘stadium’ model that is, I think, much newer.
The Eghbal model of open-source communities (referenced copiously here), whose contours are readily applicable more broadly.
If you look at what's happened to open source for the past 20 years, at some point demand outpaced supply and the amount of context that anyone can really have around any one open source project—because every developer is relying on like hundreds of different projects—it's not really possible to become this roll-up-your-sleeves member of every single project. And so, yeah, I think the governance does look really different and it’s specifically something that I didn't want to bang people over the head about it in the book. But I think a stadium model lends itself a little bit more to that kind authoritarian model and there’s less the kind of governance issues that we see in like a federation where people are like this is a democracy! and everyone is gonna ask everyone for opinions and stuff even if you might only have one or a few contributors. The contributors [in a stadium] are kind of just making the decisions and I think they should feel comfortable leaning into that. Even though right now I think a lot of them feel uncomfortable doing that because they keep being told that open source is supposed to this super participatory thing.
AGM: And you think that it doesn't necessarily have to be.
NE: I think the tension in one of these stadium models is where you do have a lot of users. And then you have some of these casual contributors who are opening issues, making feature requests or just lost, and you are kind of sorting through all that volume from people that you don't know. In my view, it's kind of like, well, I don't understand why should that person have a say in your project, if they've never looked at it before, and they're just kind of coming in for the first time and you're the core developer of the project.
There is a set of rhetoric in open source that says every person is a contributor, and anyone who kind of comes in, you should treat them as a contributor and like invest in them and all this stuff, but I don't feel like we would do that for anything else. If you had a hobby meetup kind of group with you and your friends and someone came in once and then was like I think we should runs a group like this, you'd be like: Who are you? This is this is our thing. I think I want people to feel more comfortable saying that. And there's obvious parallels between that and the Internet at large right now.
AGM: You took the words right out of my mouth. In the book, you’ve got a long riff on the tragedy of the commons. Not that I want to turn this interview into a Facebook thing, because having worked there and spent part of my career on it, it's like the last thing I want to talk about…but I do think it's somewhat relevant in that, Twitter or Facebook, is it actually the public forum and a commons? Can Zuck or Jack run it as you suggest? Can they run it like [Guido] van Rossum does Python, as officially-titled Benevolent Dictator For Life? In some sense that’s actually better?
NE: Yeah. Well, I don't think Facebook is a commons anymore, just by sheer size that we’re dealing with. One of the things that I'm trying to do in the book is go back through Elinor Ostrom’s definition of a commons and saying, okay, she makes the argument that we can avoid this tragedy of the commons by having people self govern. But she has very specific rules that she's laid out around what actually qualifies something as a commons, so we can self govern in a healthy way, assuming these conditions hold and a lot of those conditions have to do with having clear membership boundaries and very high context for your interactions with each other. And so if you think just about Facebook being 2.6 billion people or however many people are on Facebook now, it's impossible that literally multiple billions of people all have that kind of context for each other. I think of Facebook as being this substrate that fosters a bunch of smaller communities. You might have Facebook Messenger which resembles more like the group chats or the ‘club’-style communities. You might have the ‘stadium’ type situations that are more like one person broadcasting out to a group of people and you might have Facebook groups which could be like either ‘clubs’ or ‘federations’ depending how big they are. You actually have a permutation of lots of different types of communities that are across the entire platform. But I think having that kind of vocabulary can help us figure out, what does it actually mean to develop governance for any of these platforms? It's the same thing with Twitter also. I don't see a world where we have one policy or a certain set of guidelines.
AGM: That’s a somewhat shocking statement.
NE: Yeah, it's so it's funny that that’s controversial. Part of what I was trying to do in the book is saying like, okay, let's not like talk about social media, let's just talk about this other weird thing called open source. And let's look at the dynamics there and how that's evolved for 20 years. Can you depersonalize this a little bit and if you agree with me that these things seem to be happening in open source. And stacking this up against other economic frameworks we've had in the past, like the commons, and it doesn't seem to hold here, then can we take that conclusion and transfer it back over somewhere else…
AGM: Okay, that's the vibe I got from your book that you were trying to actually talk about the rest of it. So it's good to know that I wasn't over-reading into it.
NE: I was trying to be sensible about it.
AGM: Do you think the push on Facebook for content moderation, and Twitter, is a fool's errand? You know how Kevin Roose and Charles Warzel of The Times and that whole whiny mob that's constantly trying to get them to moderate everything. You think that's probably not the way forward?
NE: It seems beyond not just gonna happen, it seems actively wrong to me. It’s as though we're asking another country to govern the United States or something. I'm trying to look at where do those governance boundaries start and who should be moderating themselves or not, and just the thought that you would have a sort of widespread platform governance on some of these issues just seems, yeah, morally wrong to me.
AGM: Are you a free-speech absolutist, Nadia, that rarest of breeds?
NE: I'm not super public about my politics, but then I don’t mind poking my head out a little bit around it and publishing the book was kind of part of this for me because, to be totally frank, there are these democratic kind of ideals and these like communist-y ideals that we are holding about both the Internet and open source which are driving me crazy and, I'm trying to point out, you know, that's not always the case. And sometimes it's about one person who was doing a lot of things and we're just like couching it in a group cooperative. Yeah, I don't really know what my politics are, but I definitely err as far to that side as possible, as I think is reasonable. I do think this kind of moderation stuff, no one really has the answer to it. And so I'm not gonna sit here and be like, I know how to fix it! No one knows how to fix Facebook. Or any of these platforms. There's there's some humility that should be in place there, but I know what I stand for and what I'm aiming for.
AGM: I dislike looking always at the extreme example. But you know, Balaji [Srinivasan] had this whole dust-up with Taylor Lorenz and he's constantly getting into fights with these media people. And it's weird because he's often so right in so many ways, and he's good at getting attention. But somehow he hasn't parlayed into a mainstream following.
NE: I do feel like we need to have institutions a little bit in order to reinforce that. Well, I don't know if that's true or not, because people do follow like Elon Musk or Joe Rogan, or whatever. So that does exist. But I feel it would be so nice that if we had a publication that we could be proud of, that people would read outside of tech. There's no legible symbols for someone else to kind of follow. Like it's even weird that the most popular tech figures are not always the most popular figures actually in tech. Like Mark Cuban or something. It's someone that a lot of people know, but within tech, I don't feel a lot of people talk about Mark Cuban. There’s this mismatch between the people that are really famous within tech, and yet never really seem to make it out there.
AGM: One valid critique of tech is that they should be doing this story-telling to the wider world. Other than a few people like Patrick [Collison], or pmarca, or even Paul Graham, to a certain extent, who are able to convey a vision that extends beyond the bounds of Hacker News, I think most of tech just doesn't do that. Part of the problem is that one of the necessary delusions of Silicon Valley is that we live in this eternal present of creation, and the thought of documenting current cultural artifacts, or pulling back for a moment and looking at our own culture is not desirable or important really. And it's just not part of the DNA.
NE: I guess maybe why 2016 was so shocking was like we finally got caught with our pants down because everyone's, like, oh, it doesn't matter, it doesn't matter. And then it was like wait now everyone else has shaped this narrative because we were too lazy or negligent to shape our own narrative and. which is fine that happened, but now I feel like we should be doing a much better job of that. We have to.
AGM: You went with Stripe Press instead of a conventional publisher. What was behind that?
NE: Okay, let me start by saying this: I do feel like I owe a lot to East Coast institutions in some ways. When I was first really interested in exploring this topic, it was the Ford Foundation that like took a chance on me. It wasn’t anyone on the West Coast, it wasn't anyone in tech, because in 2015, I felt everything was: if you want to have impact, start a company, or start investing. And I can't tell you how many times I got that advice when I would be like, no, I want to explore this thing. I don't feel like there's really respect or understanding for that in tech, and I felt I got that from the Ford Foundation, and the Sloan Foundation as well. I'm so, so, so, indebted to them.
So I think maybe I’m a little bit more in between the East Coast/West Coast fight in that way. For me, I just want to go with wherever I can have impact and what I think’s been happening in tech is that we're a new culture, we’re a new industry and a new society. I don't think there was really any appreciation of tech cultural artifacts until maybe like the Trump election. Which just changed the narrative a lot more, because many people became a lot more critical of tech. And so like, it does feel like we have this emerging cultural renaissance in tech. And I feel much more optimistic now that we need to start creating those kinds of artifacts for ourselves and speaking for ourselves.
I did publish this Ford Foundation Roads and Bridges report back in 2016, which is kind of a precursor to this book. After that report report came out and was super successful and very widely read within open source, I talked to a couple of book publishers about printing it as a book. And the interactions that I had with book publishers at that time were illuminating. At the same time I was talking to them, I also had a couple of people from very prestigious institutions reach out asking if I wanted to do a PhD. So I was having all these discussions with people that want you to go down this other trajectory of legitimacy, and I felt such a mismatch between what I was actually experiencing from publishing that report and writing a lot about open source at the time, and what they thought was their version of legitimacy. I remember talking to someone about the PhD, and well, you do a bunch of coursework and you do your research projects and then somewhere like five or six years down the line, you could be considered an expert in this topic,and people will invite you to do talks. And I was thinking, I'm doing that right now after a year, I don’t see why I would do all this stuff just to have that stamp.
And then, on the book publishing side, I was told that my Roads and Bridges Report didn't cite academic literature, and so it can’t be taken seriously. And also, if we were to publish this book, then we need to have someone with a PhD who's established and respected write something as a forward, because we can't possibly publish this otherwise. And it was so weird, because over in the actual world of open source, many readers thought my stuff was important.
So when this book came around, I did talk to some publishers and I felt like a similar sort of thing of just people not really getting the thesis—and just the process being so maddeningly slow for something that is supposed to be about technology. Being told it'd be like a year and a half after you turn in your manuscript that you can even consider seeing it, and it just felt like a visceral mismatch of what I’m trying to do and I just don’t understand why it's taking so long. I don't understand why the book industry is so fucking slow and I just kind of threw up my hands on it.
Talking to Stripe Press was a completely different experience. Brianna [Wolfson] (then editor at Stripe) was just immediately on her game, and got on the phone with me right away. I submitted a 60% version of my draft, at like 1AM or something on a Sunday night, and by noon on Monday she already responded. The process was just so much better and it was just like, yeah, we just need more Stripe Presses to tell our own stories.
I can't imagine having published with anyone else because the people I want to reach are software developers and other people who understand this world, which are the people that Stripe Press can reach and that other publishers can't reach. I've come out of this feeling, and I feel even more strongly now, that we just need more outlets to be able to tell our own stories. Again, it’s this thing of why would we have another country telling us how to run our own country? And the same sort of way, I don't understand, why would we have a bunch of people who don't understand tech talk about tech and portray us like they get us? Why can't we just do that ourselves? And so I'm optimistic we are in the very early phases of that starting to happen and I'm so excited for the next couple of years I feel like that's gonna become more and more of a thing. So the short version of why I published with Stripe Press is because I felt like they could tell my story in a way that no one else could.
AGM: One of my pet theories is that everything's going to end up going private in some form or another, whether it be social media or publishing or all the rest of it. You can just speak more freely, as you wrote, in a group where you have a clearly defined in-group versus out-group, among other prerequisites. But again, a lot of this flies in the face of the sort of fuzzy democratic values that we like to toss around. And so people don't like hearing about the fact that things are just better in private groups. This is a bigger topic, but it’s strange how everything is becoming this sort of medieval patchwork quilt of different voices and values. There’s Andrew Sullivan over there doing what what Sully does. He'll have a massive audience, probably bigger than a lot of Times op-ed columnists frankly.
As in open source you can have these federations and stadiums and everyone lives in their little world. When you go and build a tech stack, you integrate them all together, right? You've got Spark, backstopping some MySQL thing, backstopping some Kafka or Flink thing with some front end from Node and whatever else, right? The company or the creator melds this all together into a functioning product, but you can't run a society that way. You can't say, okay, my paper of record is, what is it again? The Dispatch from Substack. And my police department is the one from Carmel and my so on….you can't construct a society like you would tech stack. And so how does how does one govern a country of 330 million souls in this new patchwork world? I'm not sure. I don't think you actually can.
NE: Yeah, it feels like there are digital frontiers of governance that are being built that are increasingly going to be at odds with our physical governance. It doesn't seem so crazy to me that we can kind of like cobble together our own stack, the way that you could do a tech stack. But we're gonna run into a bunch of conflicts when we try to do that. And then we have to sort those out and there'll be a bunch of little skirmishes which are already happening right now. But then maybe we shall prevail. Fingers crossed. I'm just prepared for sort of like a revolution that is kind of coming in slowly. I feel like it's happening more and more and and I don't know whether it's gonna explode, we'll find out. But I'm here for it. We'll see what happens.