I’ve been building a set of mapping components on top of Leaflet 2.0 (alpha) for Dash, aimed at filling some gaps I kept running into with map making in Dash apps. It includes drone/object tracking, a tile selector, image overlays, an edit-control setup that reports radius and area from measurements, a mini map, and fully styleable text markers and annotations. Big credit to Emil and the original Dash Leaflet project — I wasn’t the first, and this isn’t a 1:1 dash-leaflet or leaflet equivalent. I found gaps in functionality in my own map making workflows and decided to plug those gaps and take my shot at designing a mapping component.
Walkthrough video:
Some individual Demos:
Image Editing on Maps:
Walking Simulator:
One detail that might save people time: every component page ships with an llms.txt you can paste directly into a chatbot or coding agent, so it can read the props and usage for Dash Leaflet 2 and generate correct code. Icons also work with both dash-iconify and dash-emoji-mark, so you can pick whichever fits your stack. This has been extended into the overarching 2plot.ai and 2plot.dev/llms.txt federated network. I’ll have a follow-up specific post as fully breaking down the apps, components and layout of the entire federated network but on a base level.
My entire catalogue, the llms I offer and the engineering is structured for navigation for the individual, the agent and multi device / operating system support. CI CD pipelines for each component, a set of supported range support 4.1.0-4.4.1 currently, on the multiple backends available, flask, fastapi and quart.
The shocking more pressing requirement thats being introduced with this level of components being built out in this federated network rollout is the authentication gate that is being established on some of the components and content. Everything is free, everything is open source, but the goal is to structure an LLMS that can identify and hold to account the token context extraction of my work as well as protect the network from bad actors.
Some *.2plot.dev subdomains and *.2plot.dev/llms.txt will be authentication only where as once authenticated on the 2plot.ai network you’re authentication will carry between the other apps and subdomain component apps in the network.
The llms.txt copy on an authenticated account will grant you an /llms.txt?={key} associated to the human behind the workflow. The key is only documentation based and when included in any chat bot it unlocks the full content of the llms.
It’s still limited working documentation and there’s a known quirk (image overlays do a small snap on zoom) — I’d genuinely welcome help on that in the open source.