Home › Our Work › Dayboro.au

Case Study: Community Platform

Dayboro.au
22 years
in the making.

A hyperlocal weather platform, nine interactive calculators connected to my live station data, a Flutter PWA on Google Play, and a Lions Club business directory. I run all of it on a single WordPress instance from a hill in the D'Aguilar Range, Queensland.

Visit dayboro.au → Open the app
2004
Running Since
9
Live Data Calculators
118+
Businesses Listed
5 min
Data Refresh Rate

Overview

What it actually is

Most people in Dayboro know the site for the weather. They check it in the morning to see if the creek is up, or if there's a frost warning overnight. That's the front door.

Behind it: nine interactive gardening calculators that pull live data from my weather station every five minutes. A Flutter app on Google Play. A community events calendar. A tiered business directory that our local Lions Club uses to replace the old printed phone book. A custom email plugin I wrote that sends weekly weather digests to subscribers. And a Gann seasonal forecast that I cross reference with Inigo Jones lunar cycle predictions.

By any honest measure I've overbuilt it for a town of three thousand people. But it works, and the whole infrastructure costs me roughly the same per month as a decent coffee habit.

"BOM's nearest gauged station is at Samford, 12 km southeast and 150 metres lower in elevation. On a typical summer afternoon the two stations can record rainfall that differs by a factor of four. That's why my station exists."

Dayboro.au weather platform
WordPress Flutter Web WeeWX Python PMPro Membership Ecowitt Service Worker Custom Plugins

Foundation

The weather engine

My Ecowitt station on Lyndhurst Hill feeds a custom Python pipeline I wrote, generating twelve JSON files every five minutes.

🌡️

Hardware: Ecowitt sensors

Temperature to 0.1°C resolution, rainfall to 1 mm, wind speed and direction, UV index, solar radiation, soil temperature, lightning strike distance. All the hardware sits on my own site, so I don't depend on anyone else's data.

⚙️

Processing: WeeWX + Python

WeeWX handles raw sensor ingestion and archive storage. My Python pipeline runs every five minutes: it calculates derived values, builds twelve JSON files, and uploads them to the server automatically. There are no manual steps in the data chain, whilst every step still logs enough that I can trace a bad number back to its sensor.

📊

Data going back to 2004

22 years of logged temperature and rainfall at Lyndhurst Hill, with varying hardware over the years. The current Ecowitt installation has run continuously since 2023. I publish aggregates and accuracy comparisons against BOM Samford every month.

🔄

Live API: twelve JSON endpoints

The calculators, the Flutter app, and the WordPress widgets all read from the same live weather data feed. One source of truth, refreshed every five minutes. It costs me nothing in external API fees and there are no rate limits.

⚡

Push notifications

Subscribers to the Dayboro Buddy app receive push notifications for severe weather events: flash flood watch, frost alerts overnight, and heatwave warnings. Delivered via Service Worker push on Android and iOS.

🏠

Home automation integration

The same data feeds my Home Assistant instance, with Sigenergy solar battery management and Amber Electric spot prices. Weather station data informs battery charge scheduling. More solar in the forecast means less grid charging overnight.

Gardening Tools

Nine calculators, all on live data

Each one pulls from my live station feed, updated every five minutes. They're standard WordPress pages with Elementor HTML widgets. No external API keys, no third party services. The free tier gets a subset of the tools; premium members get all nine plus extended history views.

Design note

I kept the calculators deliberately simple on the outside. Each one fits in a single Elementor HTML widget: scoped CSS, vanilla JS, no build tools, no npm. They've been running without maintenance since February 2026, and that's not luck, it's the absence of dependencies. If you want a calculator that breaks six months after launch because an npm package updated, mine aren't that.

Mobile App

Dayboro Buddy, on any device

A Flutter Progressive Web App. One codebase, deployed as static files, available on Android via Google Play and on iOS via Safari's Add to Home Screen.

The Android APK (package: au.dayboro.app) is on Google Play and installs like a native app. On iPhone, open app.dayboro.au in Safari, tap the share icon, tap Add to Home Screen. It installs in seconds. No App Store account, no review process, no waiting.

The service worker handles offline caching, background sync, and push notifications. Auto update works silently: when I deploy a new version, the service worker detects it on the next load and refreshes. Users are always on the current version without being asked to update anything.

App size

3.9 MB. Loads in under 3 seconds on a standard 4G connection.

Version on Play Store

v2.0.7, live and available at au.dayboro.app.

Offline mode

Core navigation and cached weather data available without a connection.

Push notifications

Frost alerts, flood watches, and severe weather warnings delivered to the home screen.

Dayboro Buddy Progressive Web App
Why Flutter and not a standard WordPress PWA

A standard WordPress PWA gives you a cached version of the site. Flutter gives you a proper application with its own navigation model and a UI that feels native. My weather calculators need to query JSON data, render charts, handle offline state, and update in the background. I tried the WordPress PWA plugins first and they can't do that reliably. Flutter can.

Community

The community layer

Business directory, events calendar, weekly email digest, and a membership model. I built all of it into the same WordPress instance, and it's all interconnected.

🏪

Lions Club Business Directory

118 local businesses listed across three tiers: full page, half page, and quarter page. I build the WordPress plugin specifically for this directory, and the Lions Club in Dayboro now uses it as the digital replacement for the annual printed directory. Directory data feeds the Flutter app via a custom REST API.

📅

Events Calendar

Community events, markets, school dates, and local meetings. Pulled from a custom REST endpoint and displayed on the homepage. The Flutter app reads the same endpoint. One entry point for event data, multiple display surfaces.

📧

Weekly Weather Digest

A custom email reports plugin (not a third party newsletter service) sends a weekly digest to subscribers. The past seven days of temperature and rainfall, a short summary, and links to the current forecast. The queue runs on WordPress cron. Open rates run consistently above industry average for small community newsletters because every single number in the email is from this valley, not from a regional station 12 km away.

🔐

Membership: Free and Premium

PMPro handles membership. The free tier gives access to the core weather data and four calculators. Premium unlocks all nine calculators, extended historical data, and the seasonal forecasting tools. There's no trial that asks for a credit card. I made the free tier useful on its own, on purpose.

📱

Push Notifications

Web push notifications via the Dayboro Buddy service worker. Subscribers opt in to frost alerts, flash flood watches, and heatwave warnings. Notifications go through my own WordPress plugin to a push API endpoint, not through a third party notification service.

📈

Site Analytics: Custom

My rds-siteanalytics plugin tracks page views and reads Gitea issues as a website task list, displayed in the WordPress admin. No Google Analytics, no third party tracking scripts. Visitor data stays on my server.

Forecasting

Seasonal forecasting: Gann and Inigo Jones

The seasonal forecasting side lives at crohamhurst.app, a separate subdomain. It isn't a standard weather forecast.

It combines Inigo Jones lunar cycle predictions with a Gann astronomical analysis pipeline I built in Python. The output is a qualitative seasonal outlook: whether the coming month is likely to run wetter or drier than average, based on long cycle positions going back to 1913.

Honest about what it is

This is an outlook based on historical cycles, not a deterministic forecast. It doesn't know about La Niña, blocking highs, or the Indian Ocean Dipole. It knows about repeating patterns in the long record that correlate loosely with the positions of the Sun, Moon, and inner planets. Accuracy on wet versus dry calls over a six week window runs around 62 to 65 percent. That's useful for farm planning. It's not useful for scheduling outdoor events, and I say so on the site.

The same Gann pipeline also feeds a CFD trading signal system that runs against IG Markets on a demo account, whilst I test whether a methodology W.D. Gann applied to commodity markets in the early 1900s still holds up in modern trading. My results so far are inconclusive, which is the honest answer.

How the pipeline works

  1. 1
    Astronomical data

    Sun, Moon, and planet positions calculated from NASA ephemeris data. Declinations, aspects, and angular relationships computed daily.

  2. 2
    Inigo Jones cycles

    Lunar cycle positions mapped against the Jones seasonal calendar for SE Queensland. Crohamhurst Observatory data integrated.

  3. 3
    Gann unified scoring

    Multiple Gann factors combined into a single directional score. Only scores above 70 are treated as signal grade. Below that, the model abstains.

  4. 4
    Published outlook

    Monthly outlook published at crohamhurst.app with the cycle rationale and historical comparison years. Free to read, no account required.

Under the Hood

What it actually runs on

No managed cloud services, no vendor lock-in, no SaaS subscriptions in the data path. Every component runs on infrastructure I control.

WordPress + PMPro

Core CMS and membership engine. PMPro handles access control, payment processing, and subscription management. I don't use WooCommerce for memberships. PMPro does it cleaner for this use case.

Flutter Web

Dart compiled to JavaScript, deployed as static files at app.dayboro.au. Service worker for offline and push. The app itself needs no backend, it reads the same public JSON files as the website.

WeeWX + Python pipeline

WeeWX handles sensor ingestion. A custom Python module processes the data, runs accuracy calculations, and generates the twelve JSON export files every five minutes via root cron.

Custom WordPress plugins

Three custom plugins I wrote specifically for this site: dayboro-business-directory (Lions Club listing management + REST API), dayboro-email-reports (queued weekly digest), and rds-siteanalytics (custom analytics + Gitea task integration).

DirectAdmin on brisbane01

I host all my sites on a dedicated server in Brisbane. Help4bis.com, dayboro.au, gardenbuddy.au, rds.ink, m8chat.au, and eight other domains all run on the same physical machine. Australian data residency, not US cloud.

Ecowitt hardware

GW2000 gateway with multiple sensor arrays. Direct local API plus cloud sync. WeeWX reads locally, so my data pipeline doesn't depend on Ecowitt's cloud servers staying up.

Full technology list

WordPress 6.x PHP 8.4 MariaDB PMPro Membership Elementor Pro Flutter 3.x Dart WeeWX 5.x Python 3.11 Ecowitt GW2000 Service Worker Web Push API REST API JSON data files DirectAdmin Gitea Home Assistant Sigenergy

What I deliberately left out

No Google Analytics, because privacy matters in a small community where everyone knows each other. No third party notification services, push goes through my own endpoint. No CDN, since the server is in Brisbane and the audience is in SE Queensland; the latency is fine. No React, no Next.js, no build pipeline for the widgets. Vanilla JS in Elementor HTML widgets is fast, maintainable, and doesn't break when npm packages update.

What I can build for you

Your community, your platform.

Not the whole thing. Something that fits your scale and budget. I bring the tools, the architecture, and over twenty years of building for small communities. What I need from you: a clear description of what problem you're trying to solve and who you're solving it for.

Talk to us about your project See more of our work