Collaborating on a shared map without stepping on each other

A solo map is easy. A two-editor map is mostly easy. A four-editor map is a different sort of object, and a ten-editor map is a small political problem with a map attached. The transitions are predictable and the survival strategies are well-understood, but most teams figure them out the hard way: by accidentally erasing each other’s work, getting into stylistic disagreements at 2am, and slowly losing the will to maintain the map at all.

Below is the playbook we’ve seen work. It scales from two editors up to about a dozen, which covers nearly every map. Beyond that you’re running a small organisation, and that’s a different guide.

Two editors: pair on the rules, then trust each other

With two editors, structure is barely needed. What matters is agreeing — once, in writing — on the three or four conventions that will keep you out of each other’s way:

• What counts as a region (the “unit” question from the planning guide).
• Who owns colour decisions when you disagree.
• How you announce big changes (“I’m about to redraw the harbour”).
• What happens when someone makes a change the other doesn’t like.

That last one is the killer. The unspoken default — “I’ll just revert it and not say anything” — is the start of the rift. Agree out loud that disagreements get a 30-second conversation, not a silent edit war.

Three to five editors: introduce a change log

The moment you have more editors than you can hold in your head, write down what changed. Not as formal as a commit log; closer to a fishing diary. One line per session. “Added two regions on the east side. Recoloured the docks. Renamed Block 7.” Pin it somewhere everyone reads — a sticky channel works fine.

The change log isn’t a control mechanism. It’s a coordination mechanism. The point isn’t to police edits; it’s to make it possible for somebody who logs in two days later to get oriented in 10 seconds instead of 10 minutes.

Six to ten editors: introduce a styling owner

Once you have more than five editors, you will have at least two people with strong opinions on colour, naming, and label placement. They will not converge on their own. Pick one of them as the designated tiebreaker for stylistic decisions, and make it explicit. The role isn’t a promotion and isn’t about taste — it’s about removing a persistent low-grade source of friction.

The styling owner doesn’t draw more than anyone else. They make the call when colour-A and colour-B both have advocates, and that call holds. Without them, every recolour debate runs the same loop forever.

Eleven or more: switch to a request queue

Past about ten editors, simultaneous editing breaks down. You have either too many people racing to edit the same regions, or too many people waiting for someone else to finish. Neither scales.

The fix is to invert the model: most contributors stop being editors and start being requesters. They submit a region change — a new region, a colour change, a renamed area — and a small group of three or four editors process the queue. The submitter sees their change land within a day or two, the editors aren’t fighting over the canvas, and the map stays consistent because a tiny group is enforcing the conventions.

This is the workflow we ship out of the box: a public submission flow, a private review queue, and tools for editors to approve, reject, or send back with a comment. Communities that switch to it tend to be sceptical for a week and then never go back.

Naming conventions matter more than you think

When multiple people edit a map, regions accumulate weird names: NewArea, NewArea2, NewArea_final, NewArea_FINAL_real. None of those should exist and all of them will, unless you adopt a naming convention early.

The convention can be almost anything as long as it’s consistent. We like:

• Public-facing name only. If a viewer can’t guess the meaning, change the name; don’t add a parenthetical.
• Title case, no abbreviations.
• No author tags. “Mike’s Park” is fine on day one and unreadable in a year. Park names are public; editor identities are internal.

The lock discipline: claim, edit, release

When two people are looking at the same map at the same time, edits race. The platform handles most of the merge correctness, but two people redrawing the same polygon will end up with a winner-takes-all result — and the loser’s work is gone. The simple fix is a lock convention: before editing a region, post in your team channel “I’m editing the harbour”. When you’re done, post “done with harbour”. Two-line discipline; saves an enormous amount of frustration.

For high-traffic maps, you can formalise this with the role system — only one person at a time has edit rights to the harbour zone — but most teams don’t need that level of structure. Two-line discipline gets you 95% of the value with 0% of the configuration.

Postmortem the bad edits, not the editors

Maps go wrong sometimes. Somebody recolours a region the wrong shade and nobody notices for a week. A whole district vanishes during a sleepy 4am session. When this happens, the right move is to fix it, write down what made the mistake easy to make, and adjust the convention — not to call out the person.

A team that postmortems systems improves. A team that postmortems people stops volunteering edits. You want the first kind.