The Opinionated Mesh Analyzer

If you’re running a mesh network for a community that depends on it, you already know what it actually takes. Quiet work, unglamorous, and genuinely important.

Waev is a real-time analytics platform for MeshCore networks. It exists because people like you needed a way to understand what those networks are actually doing — live packets, signal per hop, coverage, the health of the infrastructure you keep running, often alone, often without much tooling. And it’s built around a simple idea: your community’s data belongs to you, not us.

We thought about what we want this to be.

Here are the six things underneath everything else.

01 · Privacy is the default, not a dial

The conversations on your mesh are real. The people having them trust the infrastructure you built. We don’t take that lightly.

Two rules do most of the work. First, opt-out is absolute: add a privacy marker — ⛔ 🛑 🚫 — to any node name and it is never stored, never mapped, never counted. No caching, no exceptions. Second, companion devices are scrubbed before anything is written. MeshCore personal apps broadcast on the same radio mesh as your repeaters, so those packets reach our observers. At the ingest edge, we strip the identity — name and device type are nullified — and retain only anonymous signal metadata for aggregate counting.

If you’re running a companion app, you won’t appear on Waev. The only personal devices we ever surface are ones an operator explicitly chose to connect — and even then, a privacy marker in the name overrides everything.

Privacy here isn’t a setting. It’s the floor.

02 · If we can’t verify it, you won’t see it

When you look at your network map, everything on it should be there because something verifiable put it there. Not because we guessed.

Every edge in the topology traces back to something we can confirm: an enrolled observer connected over MQTT, or an authenticated repeater. A spoofed or inferred prefix doesn’t get a line on the map — it gets rejected.

The map is less crowded for it. But what’s there, you can trust.

03 · Hard to read is a bug

You’ve probably used dashboards that made the network seem more confusing than it actually is. Charts that don’t explain themselves. Colors that mean nothing. Density for density’s sake. We have too.

We think a dashboard that’s hard to read is wrong, not just ugly. Clarity is part of correctness. Color should mean something. Motion should explain, not decorate. If a screen makes a complex thing feel more complex, it’s not done yet.

That’s a constraint from the first sketch, not polish at the end. The practical upshot: when something is actually wrong on your network, you should see it — not hunt for it.

04 · Speed is a promise

Packets should appear as they arrive. A map should pan without lag. A query that covers your whole mesh should come back before you wonder if it’s stuck.

There’s a real difference between a tool you have to check and a tool you can actually watch. That’s what we’re building.

05 · Your community’s data stays yours

You built your mesh. You recruited the operators, configured the nodes, solved the RF problems, and kept it running. That community — and its data — belongs to you.

The old model, where everyone requests access to a central set of MQTT servers, turns into a management problem fast: scoped here, full access there, a half-dozen requests a day, every one a judgment call. We’ve watched it happen. We’d rather invert it entirely.

You bring your users to a broker you own. We subscribe, and keep a read-only copy.

You keep full control. We get what we need to build useful tools. Nobody manages an access list. It’s a better arrangement for everyone — and it happens to be the honest one.

06 · Built deliberately, not on demand

We use the word “opinionated” as a description, not a warning. Principled about what we build. Honest about what we won’t.

We try to build in an order that makes sense for the whole tool, not just whatever came up most recently. That means we might not build what you ask for, or we might get to it later than you’d like.

Bug reports we take seriously — always. Feature suggestions we read carefully. Sometimes we build them, sometimes we don’t, and if it’s the latter we’ll say so.

That’s the deal.

For the people who keep it running

Mesh networks don’t maintain themselves. The people who run them — who tune the RF, place the repeaters, recruit the members, keep it alive through outages and upgrades and all the invisible work in between — that’s who this is for.

These six things are how we stay honest about what we’re building, and why.

Found something broken? Tell us — bug reports are the best feedback we get.

Ready to see what your network is doing? Connect your first observer at waev.app.