Note: this post goes too far and may be triggering. I believe OWG does great stuff, but is too powerfull with too few resources.
Thanks to a discussion in OSM Belarus telegram group, I suddenly realized why the new Board would not change anything. We could replace all seven members, and the project will stay the same. We could invite all GlobalLogic employees to run the project, and they wouldn’t harm a node. We are looking the wrong way.
All real work in OpenStreetMap is done by working groups. They decide on license terms, press relations, data policy, and technical stuff. Since we’ve got few volunteers, anybody can join any working group and… start working, I guess. The concept of “working” is what stops people from joining, and many people tried to explain there isn’t that much work.
In his manifesto Steve Coast suggests, besides closing off the Board and the tiles, to drastically increase budget for the Operations Working Group. Which makes sense: they are the only working group to have consistently spent money, which had measurable positive impact on the project. Why wouldn’t we want more servers?
OWG also manages our website. And planet extracts. And our data model. And developer relations. And has a final say over who can and who cannot integrate with our core systems. Which… Is a lot, don’t you think? They must have a vast experience on working in open source, on encouraging developers to join the effort.
Except they are doing exactly the opposite. Despite having 37 tile caching servers, our tiles are slower than ever, with many experienced contributors having moved to alternative tiles. In the past nine years I’ve seen dozens of developers driven off the Rails Port website code. Even active members of the Board were rejected their valuable contributions, like an endpoint to retrieve deleted objects. In their repositories, OWG members are consistently breaking every rule of open communities, by not giving outside people even a bit of control.
And who are the OWG? Strangely enough, in the past eight years OWG has only been shrinking. They are essentially just two people, who’s been in the project since 2006. I have no idea how they’ve managed to keep doing the same thing for 12 years, keeping the servers up 24/7, without any positive feedback.
Individually each of them are awesome, they have lots of experience, they know their job and their devops skills are sky high. They rush to fix our servers in the middle of night. We do not pay them: keeping volunteering to run a million-user project is no small feat. I doubt we can find other people like them. They are irreplaceable — and that is exactly why we need to stop relying on these people.
The Board can rotate on a yearly basis, fight against corporate involvements in matters that don’t matter, and spend five more years introducing microgrants that won’t affect the project in the slightest. Because the project is shut, because the real core group does not share power and shuns away anybody who thinks for a moment they can contribute. And even the hardware part is limited to reactive responses, since two volunteers do not have capacity to plan ahead.
With that, it seems like the only proactive and useful deed the new Board can do is to disband OWG completely.
They cannot go away themselves because of the burden of responsibility — a feeling I have shared myself in the past. Instead we need to plan our budget and hire people to do essential stuff. So they are responsible and humble, and do not have ambitions to run the project — only to make our hardware and devops better.
Managing our website and other core tools should be more open — and the level of OSM Carto is not enough, although it’s something to look up. We need a group of community managers, to attract and keep new developers, to be more bold with changes, to have yearly and quarterly plans for improving the user and corporate interactions. And to finally chart where the project is going.