Interest in reviving this project — would you be open to collaborators?

[heptocat]

Hi Steve, opentrees.org is a great resource — thanks for building it.

I noticed that opentrees-data has been quiet for a while with a few open PRs waiting, as well as a lot of requests for adding city data in the issues. It looks like a lot of people are interested! I'd love to help get things moving again, for instance by reviewing sources, and adding requested sources to opentrees-data.

I understand that you may be busy with other projects. Would you be open to adding a collaborator?

[stevage]

Yes, great. I'm currently travelling for a couple of weeks so can't do much.

The biggest issue is that most of the sources that provided the data for opentrees.org are now offline. Such is the state of opendata.

Traditionally opentrees.org reflected data that was currently publicly available under a free licence. But we may have to reconsider, and keep including data that was previously available. (That assumes I have a copy of the original data which I'm not certain of.)

Also noting that Ethan from Falling Fruit forked the data harvesting code and made significant improvements but I never put in the effort to bring what he had done back into opentrees. But maybe that's a separate project.

It's pretty straightforward to simply add more sources, and run the scripts. See how far you get and we can talk more when I get back to a computer.

[heptocat]

Hi Steve, I managed to run the scripts, and add a source as a test in opentrees-data. Looking forward to adding more, if everything looks okay to you!

[loamwork]

@heptocat - curious what your plans are here. I had a pipe dream a few years back of building something gamified on top of this / an extension of it but got distracted (hi @stevage, I think I emailed you from my Columbia Climate account like 2 yrs ago about this hah)

[heptocat]

Hi @loamwork , my main interest right now is to add more sources / fix sources on the backend ( stevage/opentrees-data#5 ), and to see if I can help with the workflow that I'm guessing goes into making this data available for the website.

Right now I don't really have any thoughts about the frontend (though I think it would be really cool to have your own location be shown on mobile at least :-) ). I think a gamified extension sounds cool! Feel free to let me know if you do eventually end up picking up the idea.

[loamwork]

Copy that! I thought it would be fun to make something that's almost like a dating app, and tied to your proximity to certain trees. Swipe left/right. If you swipe right it generates a "match" with the location and how to go see it from where you are currently. Have some kind of filter based on size that ties into the National Champion Tree Registry / Ancient Tree Inventory in the UK.

[loamwork]

Basically something that ties into popular culture but is anchored in an appreciation of nature.

[loamwork]

@heptocat - curious how youre getting on with the data side of this. Wanted to get you and @stevage a bit of an update to what I've been working on, as I'm conscious the data informing the below is likely a bit old?

tl;dr is it's a "dating app" for trees called Pining (double entendre between trees and desire). Idea is to making something fun that ties into pop culture and a common mobile experience, grounded in appreciating trees big and small. Front end enables users to swipe through notable trees within walking distance, give unnamed trees permanent proper-noun nicknames (enforced by a server-side Cloud Function with hate-speech moderation), and plan "dates" to visit them on Earth Day or any day they want to get outside. React + Vite + Tailwind + PWA (installable with offline cache) on Firebase - Firestore, Auth (anonymous-first with optional Google link), Storage, Functions, Hosting.

The dataset is ~370K trees pulled from opentrees.org, the UK Ancient Tree Inventory via ArcGIS, individual city open-data portals (Toronto, Vancouver, London, Burlington, Boulder, SLC, etc.), and 10 hand-curated US Champion Trees as "heroes." Species-level photos come from Wikipedia's page images API (800+ species covered). Started by thinking about high density walking cities as a starting point, so focused initially on NYC and SF. There's something like 700k trees in NYC, so created a filtering mechanism that breaks NYC down to 250m x 250m cells, and contributes the top 30 trees by DBH (≥15cm) per cell. Helps get all the trees in NYC to "date" only being in parks (so that you can appreciate the average city tree and what it does for the city as much as the big trees in the park)

Every tree has a Claude Sonnet 4.5–generated bio in a first-person "Giving Tree meets Tinder" voice - there some deterministic bucketing of how these bios are generated according to the species of the tree in the data sources. They oscillate between fun, humble, swaggy, innuendo, etc (no expletives etc, just leans into dating app culture a little) 1,853 species covered, 37K unique bios.

Map uses Leaflet with CartoDB dark tiles and marker clustering. UI is a Raya-inspired dark luxury redesign: full-bleed photos, Georgia serif for tree names, thin-line icons, no emojis.

I'm conscious the data being pulled in is probably out of date, and could use a refresh so that the trees it's pulling in are actually where the app says they are. Curious if you'd want to collaborate on this?

[heptocat]

@loamwork cool, GL with that. My plan is to definitely work on updating the data. I've made a small pull request with new sources in opentrees-data, and once I figure out how the workflow @stevage will be in terms of actually adding the data to the map, I'm looking forward to add and update more.

[ezwelty]

@stevage @heptocat @loamwork At the risk of adding some confusion, I feel it is worth mentioning here my fork of stevage/opentrees-data → ezwelty/opentrees-harvester. It includes updates to existing sources and hundreds of additional sources (including all issues raised on this and stevage/opentrees-data repos) and a bunch of code changes, including an aggressive archiving mechanism (the full archive, on my machine, is about 23 GB, and allows the workflow to continue to run despite dead links). In fact, I think I made so many changes back in the day that @stevage never found the time to wire up ezwelty/opentrees-harvester to stevage/opentrees, but that had been the idea.