A Tough Realization Link to heading
Towards the middle of 2025 I looked at a friend and mentor’s GitHub commit history. It was a wall of green. Not for days or weeks or months but years. Here I could clearly see the gap between us. I went to look at my own commit history even though I didn’t need to. It was hard not to feel a little disappointed in myself for all that lost time.
A New Philosophy Link to heading
Here’s how I’m going to close the gap.
Focus on Fundamentals Link to heading
Don’t chase frameworks and shiny new technology. Instead focus on the fundamentals of software engineering.
Practice Consistently Link to heading
Becoming a bit better every day is enough. The small consistent gains will easily outperform my short infrequent bursts.
Accept Realistic Expectations Link to heading
I’m doing well, but I know I can do better. It’s time to green up that commit history.
The Year Ahead (2026) Link to heading
For 2026 I’ve put together a plan of progressively more difficult projects. These projects mostly focus on small academic versions of the systems I use every day. Completing and documenting these should deepen my understanding of the systems I rely on.
Project Sequence Link to heading
- Echo Server
- Web Server
- Rate Limiter
- Load Balancer
- Redis Like
- NATS Core Like
- Container Runtime
- Distributed Log w/ Raft
- Distributed KV Store
Updates (March 4, 2026) Link to heading
I originally had an uptime monitoring project scheduled, but it would involve a lot of scheduled CRUD programming and doesn’t really pique my interest, so I removed it from the list.
I had been considering a URL shortener for the capstone, but I’m not really looking forward to that project either, so I’m evaluating other options.
I expected this list to take most of the year, but I’m significantly ahead of that pace.
Updates (March 5, 2026) Link to heading
I started planning Bitcask and considered the architecture. It is really designed for very high throughput at the expense of an in-memory index. That is fine, but I wasn’t looking to build a server component for this. The architecture also doesn’t really make sense for an embedded database. I’ve decided to go closer to the metal and try writing a program that runs an isolated container similar to Docker. Now I’ll have the opportunity to really learn namespaces and cgroups.
I’m thinking my capstone for the year might be a NATS/JetStream project: a durable, pull-based pub/sub system with replay support. I find it interesting, and it has several moving pieces.
Update (March 6, 2026) Link to heading
I’ve finalized the project list with projects left. A container runtime so I can learn how isolation is done. A distribute log to learn about how consensus drives distributed systems. The dist KV store is the simplest extension to raft that’s useful.