Build with the End in Mind

Goal

• Get FishCast in an exportable format
• Visual tweaks to the app
• 3D maps for lake wind tracker

Workflow

This week, my main priority has been to build FishCast in an exportable format for beta testing. This was a process I wanted to complete before the summer ended. My goal is to eventually launch the app and continue making updates throughout my lifetime.

This was a step I had never taken before, so I wanted to make sure I asked the right questions to protect the company’s information. I had to be very careful not to push anything related to my data source to GitHub. I started by committing any new changes I had made to the app to GitHub. Claude suggested I could get FishCast to export directly to GitHub. My GitHub didn’t have the correct version to do this, so I made it happen on Netlify instead. Claude was able to push different commits based on the permissions I have granted it and the features I have added for testing.

At this stage, I discovered that everything related to Wet Dog Weather’s products was not accessible in the demo version. Not only did this include data points and overlays, but it also included the map from Wet Dog Weather integrated into the app. Given that this made up half of the app’s functionality, I concluded that the route I was trying to export FishCast would not be successful.

Fortunately, I have found an alternative method to circumvent security threats while still visualizing all the app’s features. In my corner, I have access to an Amazon Web Services account, where I can host a static website for temporary deployment of FishCast. Now, I can actually share what I have done in the app without any security risk.

Running in the background, I always have more ideas for improving app design and adding more features. I am currently prompting Claude to work on using a 3D map in Terrier, in tandem with wind data, for each lake location. In the lake page, you can now drag across the lake, zoom in and out, and get a 3D view of what the lake looks like. Wind vectors are now color-coded by surface wind speed, and lakes and land are color-coded as well. I have *hopefully* also implemented a feature that lets you see your location on the lake in the app. The goal is to provide a clear visualization of the wind over each lake, potentially indicating where the fish are biting.

Prompts Used

I first told Claude to push everything I had previously worked on to GitHub (without including the path to the data from Wet Dog Weather). After doing this, I asked Claude how to efficiently export my project so users can access it via a URL without security risks. I was eventually directed to connect my GitHub account with Netlify to generate a URL for me.

The demo version I created first does not show a map with all locations and any data overlays. It only had three lake locations that showed real data running from my computer; the rest of the lakes did not show lake pages.

I’ve been trying to work with Claude to get sample data without exposing Wet Dog Weather’s information. I have attempted to work on the map and lake search feature without implementing live data overlays, but it seems GitHub is ignoring everything Terrier has done. Claude had given me the option to include real data samples stored elsewhere in the repo for demo purposes, but I was having trouble deploying it on Netlify. 

Lessons Learned

I should have prioritized this process in previous weeks, as this was the end goal I wanted to work towards this summer. Unfortunately, I became distracted and spent my time adding as many new features as I could before launching the app. At this stage in my project, I knew I needed to launch a demo version sooner rather than later. Figuring out how to do it properly and inexpensively wasn’t easy.