text stringlengths 0 2.08k |
|---|
**Jan Lehnardt:** I've definitely seen this in Hoodie as well. Talking about the challenges of this... Scaling is really important. One of the downsides of all of this that we're having is that people like Hoodie a lot and like contributing a lot because we're acknowledging their work and their contributions a lot, and... |
**Nadia Eghbal:** Interesting. |
**Jan Lehnardt:** That's a really tangible problem that we're seeing, especially since Hoodie is kind of focusing on more diverse contributors... We'll welcome everyone, but we're kind of trying to specifically reach out to under-represented groups in open source specifically. These people just have a harder time findi... |
This is something that we haven't solved yet. This is just an active, ongoing problem. |
**Mikeal Rogers:** \[35:58\] You do know quite a bit about that, though. I mean, you've been involved in JSConf EU and really did a lot of work to make JSConf EU one of the most diverse conferences in the whole conference space, in terms of people attending and people speaking, but especially people speaking and gettin... |
**Jan Lehnardt:** Yeah, interesting. We did a kind of connect very early on, before the Contributor Covenant was a big thing, and we kind of pioneered that for Apache as well at CouchDB. We started to recognize that open source is people coming together, and people coming together need guidelines that are very similar ... |
I think the way we got there is - and I'm paraphrasing a good friend, Florian Gilcher, who's running a bunch of Ruby events in Berlin, or has in the past - whenever you do an event, the most important question to ask as an organizer is "Who's not here and why?" It was arguably also very important - if not the most impo... |
**Nadia Eghbal:** Alright, we're about hitting our time for a break. When we come back we're gonna talk about a couple other challenges and edge cases around community building. |
**Break:** \[37:31\] |
**Nadia Eghbal:** Alright, we're back from the break. We're here with Jan and my co-host Mikeal. One thing that I was wondering about from our last conversation was... I mean, Hoodie is a great example of a project that baked in all these really great community values from the start, but what happens when you have a pr... |
I know that Jan had written this post called Sustainable Open Source referencing a couple of projects that were making these public appeals for contributions. They were both very popular projects, but they were saying, "We didn't do the community side very well." So how you do that once... Is it too late when you've go... |
**Jan Lehnardt:** It's tricky. Even inside Hoodie, when we did the big revamping of the project last year, most everybody was on board with what we were gonna do, but it takes some... Communities have inertia, and you have to overcome that. The older the community is, the more inertia it has. Whereas in Hoodie, setting... |
In CouchDB it took a couple weeks to close to six months to actually have all of the community agree on it. We even had to forcibly eject one community who was violently against the code of conduct because, as it turns out, they were regularly showing behavior that would be in conflict with the code of conduct. That wh... |
\[40:14\] You need to have procedures for that, you need to have everyone understand what all the consequences are, and be on board with this. Arguably, for Couch this was kind of the worst case of what a project can go through; I'm thankful for the experience but I also don't ever wanna do this again. So I don't envy ... |
I guess that's a way of saying you need to talk to a lot of people, the bigger a community is, to understand their point of view and maybe bring them onto your point of view if they differ, if you have at all the chance to get there. |
**Mikeal Rogers:** I think you're touching on an interesting strategy though, that I think does work. These processes and the status quo work for the people that are in power, that have power; they're already there, they've already made it through all these barriers to entry. You're basically changing it and making it ... |
I think when you look at other communities you can see something similar. Say if you wanted to change the way that Python Core worked, you could look towards communities in the Python community that have leaders that have had success with this and use that as a model and then advocate it. Those people care enough proba... |
When we reformed the NodeJS project, Rod Vagg had never really worked on Core; now he's like the TSC Chair. But at the time he never worked on Core because it was just not that interesting in terms of the contribution policies around it, but he had done a bunch of pioneering work in the Node ecosystem around better con... |
**Jan Lehnardt:** Which I guess is why it's so important that we talk about these things, so people understand that these things are important, and talk about them and maybe steal from us as much as they want and then make up rules of their own for their own projects. |
**Nadia Eghbal:** How do you tell someone who is basically a single maintainer of a popular project, how do you get from zero to one with that, where they're sort of like "Oh my gosh, I've been doing all this work and I really love this idea of a community model, but where do I begin?" |
**Jan Lehnardt:** Yeah, "You want me to do even more work?" \[laughter\] |
**Mikeal Rogers:** What I've had some success with is first you crack open the process. So you say, "Look, it's not on you to design a process. You're talking about having a more open process. Why don't you use that process to create the process?" Create an empty governance.md doc and then let people send pull requests... |
**Jan Lehnardt:** That's what we did with the Hoodie editorial team. We just said, "We need an editorial team. Here's roughly the things that you need to be doing, that we'd like you to do." If anyone's interesting, your first job is to figure out how to do what you wanna do, and write down how you wanna do that. The g... |
**Nadia Eghbal:** That's a practice I really like about Hoodie. You would call out these specific types of contributions you were looking for and saying... Maybe that's where people started saying "Hey, I'm looking for this type of thing", and treating it almost like a job board. |
**Jan Lehnardt:** \[43:56\] And then maybe a little bit more tangible for that single maintainer person that we dreamed up here. I can spend ten hours to just fix ten bugs, or I can spend ten hours to write a failing test case and describe in an issue how that bug would be fixed. Then ten people can fix one bug and I o... |
It's radically different from like, "Okay, I'll sit down for an hour and do this release, because I love coding and I love fixing stuff in the code, and I can spend an hour and it's really enjoyable for me. Or I can sit down and write documentation...? Really?" But then if you do this a couple of times, then people bit... |
**Mikeal Rogers:** An interesting aspect of this too is that you have to distribute the ownership over this stuff along with asking people to contribute, right? At some point they need a decision maker who has authority over it. And if you have a big enough project where you can actually start breaking off components, ... |
The most successful version of this that I know of is the work that we did in io.js to build out the localization effort, and literally we got 146 people in 27 languages in a day. Essentially we were like, "Tell us in this thread if you want to start a localization community, and I will create a repository and then add... |
**Nadia Eghbal:** A couple theme I feel like I'm hearing from all these different suggestions, around opening up your process, transparency and making it easier to saying out loud what you're doing so that other people know what you're doing, and makes it easier for you to get involved. Then also that aspect of "Yeah, ... |
**Jan Lehnardt:** And it eventually frees up... If the hard programming challenges is what you really thrive on, it allows you to focus on ever more harder problems in your projects, because other people are taking care of the less hard problems. |
**Nadia Eghbal:** Right, exactly. The other flipside I was wondering about was some people I've talked to have had bad experiences when they've opened up their process to other people, or let other people in. Somebody ended up taking over, or making decisions they didn't agree with, so it's scary for them to think abou... |
**Jan Lehnardt:** \[48:04\] I don't think I've had a very bad example of this, but also both Couch and Hoodie, I'm very involved and there are no BDFLs in this projects, but I carry a lot of influence that I can usually turn the bad things around before they happen. The problem is when you really can't, and you're part... |
My measure of control at this point is I will only vote people into that higher circle of being able to make binding decisions that may be decisions that I disagree with, they're only people that I trust that much that even though I disagree, I know they have the best of the project at heart and they won't ruin the pro... |
**Mikeal Rogers:** I think another strategy that calms people down is that you quantify the risk of somebody doing something wrong. Like, what is the cost of a mistake in the documentation? It gets fixed pretty quickly, it's up there for a few minutes - it's not a big deal. So if you can break that off into its own rep... |
Then in the core project - I think we even do this around the master branch a little bit - one, Git allows you to fix any mistakes. So if you're not Subversion, you're not in this horrible situation where "Oh my god, I have to redo everything!" Commits that go into master don't automatically land in a release. There's ... |
**Jan Lehnardt:** But then this meets what we talked about, the funnel - the funnel doesn't end with the contributor, There's the user, the first-timer, the contributor and there's also the maintainer type; that's what you want people to upgrade to eventually, and Node definitely has a lot more people in there than Hoo... |
\[52:00\] If you wanna be in a position where you can afford to have that kind of infrastructure where it is easy to relinquish control, you also have to work on having more contributors really badly, because you can only do that if you have a lot of people. So this is all like self-evident coming around to you should ... |
**Mikeal Rogers:** Right. Like, what are the things that you need to have happen, and if they're that important, people should show up to do them, as long as you make it easy enough for them to do it, right? |
I think that there was a lot of worry that "Well, who's gonna sit and review all these LTS patches and port them over. IBM really cares about long-term stability for their customer, so it turns out that they will hire people to do that, as long as we make it easy for them to do that. |
I think what you can't do is you can't say "We have project goals, and then we're going to make the community care about them." You have to optimize for people to show up and do what THEY care about. This is why we don't really have a roadmap for NodeJS. We have some basic stuff that we know we wanna do in the future, ... |
**Jan Lehnardt:** Yeah, that's what killed the original PHP 6, which was meant to be the Unicode release, and then because the project decided the next goal is Unicode, it turns out only two people cared about Unicode and everybody else was agreeing with Unicode but didn't care about that, so that didn't go anywhere. I... |
**Nadia Eghbal:** It's tough. I'm curious just to zoom out and talk about what we think success looks like for an open source community and project. I know we talked a little bit about metrics and how those can be sort of... You don't wanna hold everyone to the same standard, but beyond just sort of a basic gut feel, h... |
**Mikeal Rogers:** Are people happy? \[laughs\] |
**Jan Lehnardt:** Yeah, I was gonna go there. Like, if you throw a party, you kind of know when it's a good party and when it's not a good party. It's really the same thing - how do people feel? How do you see people strive doing their thing? Do you see people reaching the goals they said they wanna reach in the time t... |
**Mikeal Rogers:** I think looking at the party is probably a really good metric, right? |
**Nadia Eghbal:** Yeah, I like that. |
**Mikeal Rogers:** At the JavaScript conferences everybody is super happy and having a great time and doing karaoke, and some of these conferences that I go to that I won't name, it's just people getting drunk in an airport, or in some hotel ballroom... \[laughter\] |
**Nadia Eghbal:** If someone doesn't care whether other people are happy and they sort of just care about... I mean, more in a BDFL type of situation - is it just gonna suck for everybody else who loves that project that can't be involved? |
**Jan Lehnardt:** I think of benevolent dictators that benevolence isn't shown through understanding that people work best when they're happy is not benevolent, so they're not a BDFL by definition. |
**Nadia Eghbal:** They're just a DFL. |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.