I own few domains and one of those domains is registered at GoDaddy. This is for historical reasons because this domain is on the .es TLD but my preferred registrar, PorkBun or CloudFlare, do not support this TLD. I kept it there mainly because I’ve had it for 10+ years and there were some new identify requirements that I didn’t want to deal with yet.

I use external-dns as a tool to automatically to take my Kubernetes Ingress resources and register them in my DNS zone. This tool would use the GoDaddy API to create, edit, and delete the DNS records automatically for me and point them to my own servers. Sure, I could do this myself because I don’t have that many records, but automation is fun.

Over the past few months, as part of my self hosted finances series I’ve been working to extract all of my stock portfolio into some kind of self hosted database. I came across Ghostfolio, which is an open-source (with a paid hosted edition) tool for tracking stock portfolios. It was able to give me a portfolio view across multiple brokerages, automatically fetched stock prices, and gave some basic allocation reporting.

Previously in Part 1, I talked about how to clean-up the transaction data from Mint to remove duplicates and add any missing transactions.

Solving for transfers

The next phase is to solve for the transfer pairs. A transfer pair is defined with a matching credit and debit transaction on two different accounts. In Firefly, a transfer is treated separately than a credit/debit because it’s excluded from the expense and income reports.

I was recently in the market for a new personal-use laptop and wanted to try out a Framework Laptop. I was intrigued by the idea of being able to replace any part that failed or even upgrade parts as I went. I also was frustrated with the direction that Windows 10 and Windows 11 was going.

They seemed more interested in advertising, tracking, sending notifications to increase my engagement of their apps, then just building an operating system that got out of my way and let me do my thing.

Since Intuit announced that Mint was going away, I’ve spent several months investigating how to import my Mint data into Firefly-iii, an open source, self-hosted budgeting software. It seemed like a perfect fit. I would fully own the data and get to build whatever tooling I want on top.

However, before we can get there, we need to have cleaned and accurate data from Mint. As it turns out, Mint’s data actually had some errors in it that required me to go back years and fix them. In this post, I walk through the programmatic approach I took to identifying mistakes and fixing them.