Systeme D


OpenStreetMap ten years on, and why it's time for a fresh slate.

I've been involved in OpenStreetMap for almost ten years now. 22nd November is my OSM birthday.

Message 000111. Right now talk@ is up to 71235, never mind the other lists. I should have a birthday (mapping) party with cake and we can all go to the Rose & Crown, get drunk, come back and write intemperate mailing list postings. Maybe not.

OSM has changed my life, and my outlook, in several ways. One is realising that a small number of thoughtful, committed citizens can change the world, without any hope of a fast buck - which is a noble idea we all have in our 20s but this was the first real-life proof, to me, that it really does work.

The other, perhaps more important, is: it's down to you. If you want change, change things. In 2005 we expressed this by demanding action, not words, from anyone who said 'OSM should do it like this' on the lists. The word 'should' was a red rag to us back then.

In time the word 'do-ocracy' was coined. But the other half of the do-ocracy equation has never received enough credit. It's not just that "actions speak louder than words"; it's that "we trust you to carry out the actions". OSM doesn't have a moderation system for edits. Site improvements don't have to go through five board committees. It's OSM. You're a contributor. You've given us your time. We trust you. Do great things.

That, more than anything, is why OSM works. We value, and trust, our contributors. Every one of them. OSM is 'do great things' multiplied by ten thousand.

The slight wrinkle is that this only gets you 95% of the way there.

The 95% is astonishing. The 95% is mapping large parts of the world to ridiculous levels of detail. We are by most metrics the best available map of Germany, the UK, increasingly France, Russia and the urban US, and a hundred places I don't even know about. (I made this comment elsewhere. I was immediately picked up by a Belgian mapper, Marc, saying "hey, what about us?". He was right. I didn't know. That's how far we have come.)

The 95% is running the most crazily lean, efficient hardware setup, constantly reinventing: our API went from plain-old-Ruby to Rails to C++, our tile servers from monolithic to distributed, our database from MySQL to Postgres, our UI from entirely serverside to largely clientside. The 95% is an ecosystem of renderers and routers and, dare I admit it, the sleekest desktop map data editor there is and its universe of amazing plugins. (It begins with J. Don't make me say the name.)

So if I talk about the 5% that do-ocracy doesn't accomplish, that's no slur on the OSM community. Our 5% is Wikipedia's 30% and Google's 95%. We do more, ourselves, better, than anyone else.

Rewind to 2012. It was pretty clear we needed a new default editor on Potlatch 2 still worked, but Flash Player was already (rightly) on the way out, and the six-year-old Potlatch user interface - initially designed for moderately clued-up users working on a blank canvas back in 2006 - was confusing for the newbies attracted by the explosion of public interest in OSM.

This was a 5% problem. We needed a new beginners' editor, but no-one was clamouring to write it. The 95% of experienced OSMers, understandably, wanted to work on JOSM plugins for experienced OSMers. I tentatively started work on a newbie-friendly JavaScript editor called iD (I'm terrible at naming software) but I was pretty burned out on OSM development at the time and only got so far.

Happily, in this case, there was a Fairy Godmother in the shape of the Knight Foundation and Mapbox. The Knight Foundation funded Mapbox to, among others, rebuild and complete iD. As part of this some incredibly skilled JavaScript developers and designers got to work on it. I don't think the outcome could have been any better, and it continues to delight me right now: while we're pointlessly beating seven shades out of each other on the mailing lists, "osmbot: [osm-website|master|John Firebaugh] Update to iD v1.6.1" has just flashed up on IRC.

iD is how it ought to work. iD isn't telling the 95% how to map or what to map. It isn't saying "Mapbox want turn restrictions, therefore the default will be devoted to mapping turn restrictions". It's simply a 5% intervention, a new tool which no-one else was writing, to increase the 95% of do-ers, to bring us more contributors. The effort in building it will benefit us many times over.

But we can't always expect a Fairy Godmother to appear. We struck lucky in this one case. There are plenty of places where we haven't.

I can recite a few of them. We have very little mobile presence, even though smartphones are ideal surveying devices; a 5% intervention here would bring so many more people to our 95%. Diversity is almost becoming a hackneyed word in OSM but let's restate the truth of it; a 5% intervention would make sure that our 95% of do-ers grows to include people from poorer, less white, less educated backgrounds, with a real, practical effect for OSM. As well as having the best map of the beautiful Brecon Beacons, we can have the best map of the deprived Gurnos Estate 10 miles to the south.

This is what the OpenStreetMap Foundation is here for: to do the 5% that contributors can't, the 5% that magnifies all our efforts. To be the Fairy Godmother.

This was once described as "supporting but not controlling the project". Sometimes that gets interpreted as "doing the minimum", just providing the legal and financial framework. It's much more than that. OSMF's job is to grow and help our amazing community so it can do more amazing things... things that it couldn't otherwise do. OSMF is here to facilitate what isn't happening, and get the hell out of the way of the community for everything else.

Never go back

With my 10th OSM birthday coming up, I've been involved in OSM almost as long as the founder Steve Coast. Like Steve, I've messed up countless times over those 10 years and have learned a lot.

One of the things I've learned is: never go back.

There is no point in me writing Potlatch 3. Nor in scanning a second set of historic maps. Nor in incessantly arguing for another licence change, nor returning to the board, nor running the Communications Working Group again. Was I the best placed person to do those things then? Maybe yes; more likely I was the one who just happened to have that bee in my bonnet at the time. Does that make me better qualified to do them today than all the new contributors? Absolutely not.

"Never revisit the past, that's dangerous. You know, move on."

OSMF is currently going through one of its periodic bouts of blood-letting. To Steve, Henk, Frederik, all still involved in one way or another, I would say: our time on the Foundation is past. We have been there, seen it, done it. We think we know what needs to happen; we might be right in our own little ways; but that doesn't matter. We are repeating our old wisdoms rather than trusting in the next generation. We are bed-blockers when Martijn and Christian and Kathleen and Ilya and Alex - and many others I don't know about - should be given the chance to take OSM forward rather than simply slaving away on local chapters.

Let's have a fresh slate in three months, with every vacancy up for election. Give people the time to properly articulate their vision for OSM rather than this frenzied, melodramatic shouting match. Give people a chance to commit to being part of the Foundation that takes OSM on to still greater things.


Finally, I'd like to say a word in praise of Simon Poole and Frederik Ramm. There have been a lot of words spilled on keyboards in recent days. It's easy to get hung up on tone or perception and, conversely, it's easy to speak our minds without thinking how it'll be perceived. But over the past few years, what Simon and Frederik have done, in a time of great change for OSM, is kept us anchored to the 95%. I really hope the next board will do a better job of providing the 5% than previous boards have done. But I also hope it has people like Simon and Frederik on it to ensure we recognise and support the 95%, the contributors who make OSM amazing.

Posted on Friday 24 October 2014. Link.

Previous post: An ingenious vintage German cycle map.