Goal
- Continue building on my aviation dashboard
- Integrate PIREPS
- Integrate airport station plots (METAR data)
- Add flight plan functionality
- Add radar products from Terrier
- Begin working on iPad OS version using Xcode
Workflow
This week went pretty smoothly as I continued working on my aviation dashboard. I spent some time fixing bugs and making small tweaks, then got to work on this week’s new implementations. As I’ve been working on this dashboard, I’ve kept coming up with new ideas. These include adding PIREPS, airport wx station data, radar, and eventually developing an iPadOS version of the dashboard.
Claude did a nice job of adding PIREPS (pilot reports), which are live reports on flying conditions. I then moved on to airport wx station data, which also integrated nicely. At first, they were displayed as colored dots, but after a few requests, Claude turned them into windbarbs.
It was at this point that I decided to try to clone the dashboard on iPadOS. Most pilots use an iPad or tablet as a tool, so building my dashboard on an iPad seems like an important step in this project. I had Claude look through the browser version of the dashboard to learn how it works, then told it to recreate it in Xcode using SwiftUI. It managed to make a clean copy, which I’ve been tweaking ever since.
I finally have a brief demo video to showcase my work! While this is far from a final product, it showcases the ability to work with Wet Dog Weather products alongside external data using tools like Claude Code. In the coming weeks, here are my main goals:
- Make both the browser and iPad OS version of the dashboard “consumer-ready”
- Possibly integrate WDW native turbulence data?
- Begin development of the passenger version of the dashboard.
Prompts Used
Since this project is very open-ended, I’ve been able to give casual prompts to Claude as I work, changing small things and/or implementing new features one at a time. Had this been a stricter-guideline project like our first 2-week smoke/aqi project, I would have used longer AI-generated prompts to ensure maximum efficiency and context sharing.
What Worked
Claude successfully converted my existing dashboard into an iPad version with only a few prompts. Because I provided the correct project context before starting, it was able to replicate the dashboard’s content and adapt the layout to a new language and structure with minimal rework. This reinforced the importance of establishing the right context before asking Claude to build or modify features.
What Didn’t Work
There were no significant issues or setbacks this week. The only minor challenges were the same context management considerations noted in previous reports, and they were easily addressed by providing Claude with the appropriate project context before making changes.
Lessons Learned
One task at a time and context management are lessons learned this week. Additionally, I have come up with a Wet Dog Weather themed product name: RuffRide!