Recently I made the decision to shut down the church.io website (it’s now a redirect), delete the corresponding Slack community, and move the GitHub repositories back to my own personal account. Church.IO failed to build the community of software creators I dreamed of.
Mistakes were made. This is a retrospective, and an explanation for anyone wondering where we went.
What Was it?
“Church.IO” was created in 2011, intended to be a lovely little community of developers and designers who code and craft open source software specifically for churches. The seed project was my personal labor of love, OneBody, which I’ve been hacking on for over decade. There were a few of my other other small projects like bible_api, too.
At the very end, however, Church.IO wasn’t much more than a website and a very quiet set of Slack channels.
Slack Shifted the Audience
We initially used IRC for communicating. Our IRC group was talkative and supportive. We hashed out ideas and built some really cool features in OneBody there.
We moved to Slack when that became popular. The allure of animated gifs and emojis got the best of me, perhaps.
I thought, at the time of the move, that doing so would encourage new, possibly less experienced developers to join the community and contribute to the software. What it did, however, was alienate and leave behind some of the more serious and experienced hackers. It was less of a move and more of a see-ya.
I don’t think I noticed this right away. Looking back, I realized that of the people who were talkative and full-of-ideas in IRC, a few joined Slack, said maybe a few things, then rarely came back.
Of the new developers that joined Slack, very few contributed the way those original IRC hackers did. It was just never the same after the move.
What we lost in serious programmers, we gained in people saying they wanted to contribute, and that they had ideas for more features. As you can guess, ideas for more features are always in heavy supply–it’s the execution that a thriving open source project needs.
Now, don’t get me wrong: I’m grateful to anyone who helped make the software better, whether they spent man-years of their life on it, or only minutes in passing (and anything in between). And the users of the software have better software because of those contributors.
But, in the end, our move to Slack helped shift the audience from the developers of the software to the users.
Community Takes Leadership
I tend to work on OneBody in sprints–not as a marathon. I will work hard for several months, add new features, fix bugs, make lots of plans, then eventually get tired of it all and take a 6-month or a year break. I’ve been building OneBody this way for over 10 years. It works for me.
And while it works fine for a one-man show, it’s not ideal for leading a community of others. I wasn’t nearly consistent enough to inspire collaboration and craftsmanship.
There were many many weeks where I didn’t feel like thinking about my open source projects–not to mention talking to someone else about them.
I’m not beating myself up about it–just pointing it out. Community takes a steady leader with vision, which I am not.
The Designers Never Came
This is a small note, but I wanted to mention it. I thought that by saying “Church.IO is a community of developers and designers building open source church software,” we would somehow entice the rare creature known as a “designer” to join us.
It didn’t work. I don’t have any good insight into that.
What Went Right
While I am sad to close this chapter in my life, I am happy about some of the things we did that actually worked. I’m talking mostly about OneBody below, because that was and still is the main project in my life. In addition to the standard open source bug fixes and feature work, we had some really cool achievements:
Marketing: Having a decent website with a memorable name and domain does amazing things on the marketing side. I was amazed how quickly people started learning about our software once it was under a “brand” name of sorts–even though there was no commercial company backing the software.
Easier Installation: We put a lot of work into making OneBody easier to install. We provided a Debian package, an Amazon AMI, a VirtualBox image, and even a one-click installer for Digital Ocean (all of these are still available today). This put the software in the hands of many more people than otherwise.
Better Documentation and UI: Several individuals helped write documentation and improve our wiki pages. This could not have happened without the community. More people hammering on the UI of OneBody forced us to improve strange and difficult-to-use portions, too.
Translation: OneBody is translated into several languages thanks to bilingual volunteers–that amazes me. Church all over the world are using OneBody, in their native language. Still blows my mind.
We Never Took Funding: Some (many?) people will say this is a mistake. But from the beginning, I wanted Church.IO to remain independent and free from investment interests. If Church.IO was to stand, it would stand because of volunteers and churches giving back their staff time in the form of code contributions–not because of someone’s or some company’s deep pockets.
Church.IO was an amazing part of my life, and if I could go back in time, I would still do it again. Though, I would change some things perhaps–like not moving to Slack–and I’d find a charismatic co-leader to help with consistent leadership. But hindsights is, well, you know…
Where Did the Projects Go?
The open source projects are still alive, some more alive than others, on GitHub. I moved the projects I started back to my personal GitHub account. Dean Montgomery maintains his apostello project there. And Isaac Smith maintains Cedar there too.