This article covers a project I developed around 2003 for my employer at the time. I’ve tried to cover what I can recall, but it’s been quite a while, so details are fuzzy at best. However, it showcases some of my early experience in full-stack development, system administration, and commercial software deployment.
Overview
As the sole developer, I designed and implemented a networked time clock system originally intended as an internal tool to track time for hourly warehouse employees. Over time, the system expanded to support all hourly employees within the company. I worked closely with HR to integrate reporting features (including taxes), eliminating the need for additional payroll software. Ultimately, the system was commercialized and sold in Sam’s Club stores as a turnkey time clock solution.
Key features
Server-side
- Developed using PHP and MySQL, running on a custom Fedora-based Linux distribution
- I also happened to be the sole manager of the distribution at the time.
- Web-based GUI configuration wizard for easy setup
- Client and server operated on the same hardware; if no server was detected on boot, the system initialized one automatically
- Full time zone support for businesses with multi-location teams
- Support for both hourly employees and contractors
Client-side
- Developed using PyGTK
- Auto-discovery of the server within the local network block
- Support for barcode scanners, magnetic strip scanners, and biometric authentication for clocking in and out
- Custom API for client-server communication
Development timeline
V0.0 In-house project
The first version was a simple in-house tool designed to reduce manual time-tracking costs. However, it had a critical flaw in its data structure: each clock-in and clock-out was stored as a separate record with a type of “in” or “out.” While functional, this approach created major challenges for payroll calculations later down the road.
V1.0 Commercial expansion
As reliance on the system grew, my manager recognized its potential as a marketable product. The company decided to sell it as a turnkey solution through Sam’s Club. At this stage, I started a major database refactor for version 2.0 to resolve the schema issue. However, some early units were sold before the refactor was completed.
V2.0 Core schema redesign
To ensure a seamless transition, I implemented an update system leveraging our controlled Linux distribution. The system would pull updates from a dedicated server, automatically install them, and migrate data from the older schema to the improved version. This allowed for a smooth upgrade path without disrupting operations or any need for manual intervention.
Reflections and takeaways
This project was one of my earliest large-scale developments, giving me hands-on experience with:
- Full-stack development, including both frontend and backend architecture
- Linux distribution build and management
- API design and client-server communication
- Designing and maintaining a commercial software product at scale