Before you read on, stop and play for a second. What you’re seeing here is the emulator written in Go running inside of WASM inside your browser, everything here is local! Controls: Z=A, X=B, A=L, S=R, Enter=Start, Backspace=Select ...
This is kind of a review on the past 12 months since moving my cloud hosted services to all self-hosted, and the sequence of upgrades and changes to my homelab throughout the process. After a few months of increasingly larger GCP bills, and a final month of accidentally leaving a Cloud Workstation running for a little too long, I decided to give the old self-hosted world a try again after having most of my projects running somewhere in the cloud for the past 5 years, and seeing if I could pull off a $0 GCP bill by the end of the next month. ...
Introduction This guide will walk you through the process of attaching GPUs to a Docker Swarm node and running services that can utilize these GPUs. This setup is particularly useful for running GPU-intensive workloads in a distributed environment. Assumptions You are running a recent version of Ubuntu (Noble 24.04 LTS in this case). You have NVIDIA drivers already installed (preferably the -server version). If not, follow the instructions at Ubuntu’s NVIDIA driver installation guide. You have the NVIDIA Container Toolkit installed. If not, follow the instructions at NVIDIA’s container toolkit installation guide. You have Docker and Docker Swarm already set up on your system. Steps to Add GPUs to Docker Swarm 1. Identify Your GPU First, we need to find the UUID of the GPU you want to attach to Docker Swarm. ...
Building a Game Boy Advance emulator has gotten difficult, currently stuck on what appears to be I/O Registers not outputting their correct values and when a game is loaded in the gamepak, the device runs into a boot-loop. Investigating memory data and it’s meaning is difficult, converting binary data to something meaningful, then tracing that to another address if it’s a pointer and the cycle repeats. Doing these steps for every CPU cycle to investigate how data is changing is cumbersome and led to the project collecting dust for a few months as I gained the courage to dive into this. ...
On the journey of self hosting, one thing I want to get automated and running locally was my deployments. Having all of my services defined as Docker Compose services would be ideal. Wanting to keep this configuration held in GitHub, alongside my other code, this raises the next question, how do I propagate these changes to my local machine and redeploy my services? My initial thoughts were to try to find a CI/CD platform that was able to be ran as a self-hosted setup, and also support Docker deployments, specifically supporting Compose as I wanted to keep my configs simple and familiar. ...
Following discussions and read throughs on the following proposal, I hadn’t fully understood what the use case for this was or what the value of being able to range over functions would be. spec: add range over int, range over func · Issue #61405 · golang/go Following discussion on #56413, I propose to add two new types that a for-range statement can range over: integers and functions. In the spec, the table that begins the section would have a few mor… GitHubgolangGo Wiki: Rangefunc Experiment - The Go Programming Language Google logo ...
A quick snippet of my docker compose config for managing traffic ingress into my self-hosted setup. Cloudflare Zero Trust tunnel is being used to manage forwarding traffic from the public internet to my machine as I didn’t want to expose my network directly. Exiting the tunnel, cloudflare has been setup to direct all traffic to traefik:80 which is exposed on the ingress network. Based on rules defined on each service, traefik will the route the traffic to each of those services: ...