November 21, 2021

Maps, a decade later

It occurred to me recently that I've been working on stuff sort of like Placemark for over a decade. In 2009, I was working on the OpenLayers module for Drupal, a PHP content-management system that was, then, the speciality of Development Seed, the website shop that spawned Mapbox. I went on to work on our map editor, another map editor, and another and another.

The author, demoing OpenLayers a long time ago
Me, giving a presentation about OpenLayers in 2011 (Development Seed Flickr)

A decade of maps. I don't like the sound of that. My heroes are people like Hildegard of Bingen and William Morris. People who are in 30 different Wikipedia categories. Specializing is lame.

In the last few years, I spent my time building visualization, building programming environments, and understanding industrial design & CAD software. But I'm back. Why do I keep going back to maps?

Maybe because the idea of "map layers", these visual and data components of maps that can be combined arbitrarily, cause the industry to think deeply about recombination and interoperability in a way that's rare for the rest of the internet. A "map" might be pulled from three different services, which each source data from a few different open data communities or commercial suppliers. On levels both technical and legal, what you're experiencing is a real-time remix.

Maybe because maps are still a place where open source software and standards dominate. Because our shared knowledge of space is so crucial that governments, mega-corporations, and startups all collaborate on new standards.

Maybe it's that maps give you endless ways to geek out and try to solve problems better. Things like 'simplifying a line' are active areas of research, with new algorithms and old competing for performance, aesthetics, and simplicity. Maps create spectacularly hard problems in related fields, like labeling a map: display curved text in an arbitrary number of languages with their own directionality, and doing it all in real-time. Every part of how a map is made is a challenge, and every part could be improved.

A Japanese street corner with an address written on a fence
Japanese address (wombatarama on Flickr)

Or that maps capture and plug into humanity, with all its complexity and contradictions? How you learn about how endless naming disputes and border disputes. Or addressing systems that bear no resemblance to America's or the weirdness of America's addressing system.

Or that, well, maps stay hard. We've been talking about democratizing maps for a long time now, and while progress has been made, there's a long way to go. The simplest things to do with maps are more accessible, but the harder things - like designing them, or working with data - are still difficult. Whether those things will ever be 'easy' is up for debate. Will anything complex ever become easy? And how much real complexity is there?

A historical printed map of London
1884 map of London (British Library on Flickr)

One thing. Back at Mapbox, I would interview a lot of people, and a lot of them loved maps. They had grown up with them, paper or digital, consulted them on road trips, soaked in the details. I'm sure some were performing for the interviewer, but others had legitimate love for the work of cartographers.

I never really liked maps before. As my friends can attest, I have no sense of direction and rely heavily on my iPhone's compass arrow. I've never had a map hung up as art, and consulting my GPS when I drove around New Jersey was a chore.

But I've bounced around different topics in my adulthood, letting my mind go where it wants. And the thing about maps is that there's so much there. I'm happy to keep exploring.