Creating Lifting Log

My goal with Lifting Log was to create an application I could use on a daily basis for my lifting hobby, along with teaching myself about modern web development through firsthand experience.

Current State

Since launching the MVP in early 2024, Lifting Log has evolved significantly through continuous development and daily personal use. The app now has over 500 commits and numerous feature additions that have transformed it from a basic workout tracker into a comprehensive lifting tool.

Key Features

Workout Sharing - Added the ability to copy workout details to clipboard, making it easy to share routines with friends or save backups of favorite workouts.

Exercise Management - Added a dropdown menu system for exercise actions, improving the overall UX, as well as move up/down functionality within the workout form, allowing users to reorder exercises easily.

Enhanced History Tracking - Added automatic heaviest set and 1RM calculations for each of the user’s exercises, along with displaying the last performed date for each exercise. Users can now see their best lifts and when they achieved them.

UI/UX Overhaul - Migrated the entire component library to shadcn/ui, resulting in a more polished and accessible interface. Added proper ARIA labels, improved keyboard navigation, and reorganized components for better maintainability.

Database Optimization - Removed unnecessary database views and optimized queries for better performance as the dataset grew with daily use.

The app continues to be actively developed based on my own needs as a daily user, ensuring features are practical and well-tested in real-world scenarios.

View the live app → | See the code on GitHub →

The Stack

My first step in creating a lifting tracking app was deciding what technology stack I wanted to use. I already knew I wanted to use Next.js. A framework growing in popularity in the industry, it was something I wanted to have experience with. It was also likely the best choice for the application I was making. It would allow me to get the app up and running quickly, without much overhead, unlike heavier frameworks such as Spring Boot or .NET which I had experience with from my day job. Hosting would also not be an issue with Vercel being the most seamless option and providing a generous free tier. TypeScript was another natural choice, as it provides many advantages in terms of developer experience, as well as being the industry standard for modern web projects.

Things got more interesting when deciding on a database provider and ORM. After considering a few options such as Supabase, Firebase, and AWS RDS, I decided on something more in the middle - Planetscale. Planetscale allows better direct access to the database than Supabase or Firebase as it is a MySQL database, but I wouldn’t have to worry about the infrastructure unlike using AWS directly. Planetscale also offered a generous free tier with 1 billion row reads, 10 million row writes, and 5 GB of storage. This would be more than enough barring a large influx of users.

For auth, I decided to go with Clerk. I figured the setup would be relatively simple taking advantage of their dev-friendly Next.js SDK and an easy quick start in the docs. They also have good integration with auth methods such as Google, Apple, GitHub, and standard username + password.

When it came to choosing an ORM, I was initially planning to go with Prisma, as that was a very popular option at the time of creating this app and paired well with Planetscale. Just before the point of adding the ORM, I found out about Drizzle ORM, a brand new TypeScript ORM that also has specific integrations with Planetscale. I decided to give this new tool a shot and although it had its bumps, I ultimately enjoyed working with it. I learned a lot from the challenge of using a tool for which documentation and other online resources were limited.

Development

Development started with create-next-app and then I added Tailwind and Prettier. After that came the basic structure of an application. I created a simple layout with a sidebar, header and footer, landing page, an exercises page with a list of workouts, and the form modal to add a new workout. This was all stored locally with hard-coded data as the auth and database had not been hooked up yet.

Auth

Auth was the first big hurdle, adding Clerk and setting it up for local dev within the Next project was easy enough, but configuring the DNS to work in prod was an entirely different story. Eventually after an email to Clerk support and setting up another domain, then configuring Gmail auth in GCP, I had auth working and I had the userId I needed for my database queries.

Database

Next up was the fun part - hooking up the database to allow my application to function fully. Updating the Drizzle config to connect to Planetscale was pretty straightforward. I set up a simple table and query to verify that everything was working. Then I had to actually create the tables and begin writing the queries. Over the course of several pull requests, I built out the core functionality: implementing the add workout query, followed by get workouts, then delete and update logic. Each change was a focused addition that brought me closer to a fully functional app. By late June 2023, I had completed the MVP with full CRUD operations for workouts.

With the core functionality in place, I deployed the initial version to Vercel and began using it daily for my own workouts. My personal real-world usage has driven continuous development and feature additions over the following two years.

Reflections

Building Lifting Log from scratch taught me far more than any tutorial or course could have. The experience of taking an application from initial commit to a fully deployed MVP provided invaluable lessons about the entire software development lifecycle. I learned to make architectural decisions with real tradeoffs - choosing between different database providers, weighing the risks of using a brand new ORM, and deciding how to structure my data schema for optimal performance.

Working with newer tools like Drizzle ORM, where Stack Overflow answers were scarce and documentation was still evolving, forced me to develop stronger problem-solving skills and read source code directly when answers weren’t readily available. Implementing authentication was also a major learning experience, as it was my first time setting up auth in an application from scratch - I learned about OAuth flows, session management, and securing routes. The challenges I faced - from DNS configuration issues with Clerk to understanding how to properly structure database queries - were the kinds of real-world problems that tutorials often gloss over but are crucial to building production applications.

Most importantly, I learned that shipping something functional is far more valuable than endlessly planning the perfect architecture. By focusing on the core features first and deploying early, I had a working product I could actually use and learn from. This hands-on experience with modern web technologies like Next.js, TypeScript, and serverless deployment has been invaluable for my growth as a developer, and the app continues to serve as both a useful tool and a platform for experimenting with new techniques.