Blog Expansion Roadmap

Primary Source: /home/argo/Vaults/argo-os-docs/Argo OS - Part 1 - The Argo OS Journey - Building a Custom Gentoo Distribution.md (357KB, Oct 2025 - Jan 2026)

Secondary Source: ChatGPT Q&A session answers

This document maps content from these sources to specific thin blog posts.

Key Information Already in Part 1

Build Swarm (62 Cores)

Evolution: v1 (single testbed) → v2 (gateway-centric, January 14-17, 2026)
Fleet (January 21, 2026):
- drone-io: 16 cores (VM)
- dr-tb: 8 cores (bare-metal dual-boot)
- drone-titan: 14 cores (LXC container)
- dr-mm2: 24 cores (Docker on Unraid)
- Total: 62 cores across 2 networks
Architecture: Phone-home to gateway, machine-ID based identity
CalVer versioning: YYYY.MM.DD format for releases

Compile Time Savings

Package	Compile Time	Binary Install	Savings
KDE Plasma	12 hours	45 minutes	94%
LibreOffice	2 hours	2 minutes	98%
Firefox	45 minutes	30 seconds	99%
Full system	20+ hours	1-2 hours	90-95%
Application suite	8+ hours	20 minutes	96%

Btrfs/Snapper Migration

October 27: ext4 database corruption, week of work lost
November 2-4: Fresh start with Btrfs, Snapper, GRUB integration
November 4: First snapshot rollback test - 2-minute recovery validated
Quote: “This would have been a 2-minute recovery with Btrfs snapshots. Instead: 2-4 hours of manual work.”

Nix Hybrid Experiment (December 9-12, 2025)

Goal: Gentoo for system (kernel, drivers, DE) + Nix for userland (apps, dev tools)
Implementation: Single-user Nix install (OpenRC compatible), Home Manager for 57 packages
Architecture:
- Gentoo: Performance-critical (kernel, NVIDIA, KDE/Hyprland, Qt, ffmpeg)
- Nix: Convenience (browsers, Discord, dev tools, CLI utilities)
Performance: Gentoo ~95%, Nix ~88-92%, Combined ~94-96%
Challenges: Unfree package blocking, desktop integration (XDG_DATA_DIRS)

Root cause: OpenRC service integration breakdown after major updates
Lesson: Snapshots fix filesystem issues, not runtime state
Fix: Restart dbus → elogind → polkit → display-manager in order
Prevention: Post-update service restart before reboot

Golden Image Strategy

What it is: Bare-bones Stage 3 with SSH, not a full OS snapshot
November 14: 1,124 binary packages, ~28.3GB, 94% time savings
Btrfs subvolume layout:
- @ → /
- @home → /home
- @snapshots → /.snapshots
- @var-cache and @var-log excluded from snapshots

Hardware Specs

Driver (callisto-galileo): i7-4790K, RTX 4070 Ti, 32GB DDR3, 1TB NVMe + 2TB NVMe
Testbed: AMD processor, 500GB SATA SSD, eventual build server

War Stories in Part 1

October 27: ext4 corruption, week of work lost
November 16: etc-update -3 overwrote display manager config
November 21: Qt6/elogind rollback required
December 22: SDDM login failure, 6 hours debugging
January 21: Reboot crisis (K3s CNI causing kernel panics)
January 27-28: 10-hour VM recovery marathon

Posts That Can Be Expanded Now (Using Part 1)

1. `tech-deep-dive-build-swarm.md` (ALREADY GOOD - but validate)

Post already has the good content. Cross-reference with Part 1 to ensure accuracy:

✅ 62 cores confirmed (16+8+14+24)
✅ Gateway-centric architecture
✅ Failure scenarios documented
Add: Timeline from v1 → v2 (January 14-17, 2026)

2. `tech-deep-dive-golden-image.md` (THIN - 2071 chars)

Add from Part 1:

Btrfs subvolume structure (November 2-4 section)
Time savings table (94-98% reduction)
The ext4 corruption story as motivation
November 14: 1,124 packages, 28.3GB
Quote: “2-minute recovery” vs “2-4 hours manual”

3. `efistub-ditching-grub.md` (NEEDS REFRAME)

Part 1 confirms: User uses GRUB on main driver (4790K) From Q&A: EFISTUB is for testbed only (4771 NVMe boot issue)

Reframe:

Change from “ditching GRUB” ideology to “hardware workaround”
Add the 4771 vs 4790K (early vs late 4th-gen Intel) explanation
Add Btrfs snapshot integration with GRUB (grub-btrfs)

4. `tech-deep-dive-nix-hybrid.md` (THIN - 2266 chars)

Part 1 has extensive content (December 9-12):

Complete architecture diagram (Gentoo system layer + Nix userland)
Package source decision matrix
57 packages migrated via Home Manager
Challenges: unfree packages, Discord CDN failures, desktop integration
Performance analysis (~94-96% combined)

5. `guide-k3s-installation.md` (THIN - 2110 chars)

From Q&A + Part 1:

Why K3s: “Learn Kubernetes without k8s overhead”
Current services: Nginx, FileBrowser, Gitea, OpenWebUI, Postgres, QuartzStatic
K3s caused reboot crisis (January 21) - pods crash-looping
Lesson: “K3s is heavy for small systems - 8GB RAM with multiple pods is risky”

6. `breaking-network-barriers-tailscale.md` (ALREADY READ - 2900 chars)

Add from Q&A:

Pre-Tailscale journey: TeamViewer → AnyDesk → RustDesk → Tailscale
IP blacklist problem (Comcast residential IP on blacklists)
Dad-support use case that drove adoption

7. `argo-os-technical-btrfs-snapper.md` (IF EXISTS)

Part 1 has extensive content:

Complete Snapper configuration for OpenRC (cron-based, not systemd)
Pre/post package snapshots via bashrc hooks
GRUB integration with grub-btrfs
First rollback test (November 4)

War Stories for Blog Posts

These incidents from Part 1 could become standalone posts or additions:

Captured in Part 1:

“The Week I Lost to ext4” (Oct 27) - Database corruption, no snapshots
“etc-update -3: Never Again” (Nov 16) - Config overwrite disaster
“The Qt6 API Mismatch Crisis” (Nov 21) - Binary compatibility
“6 Hours Debugging Login” (Dec 22) - SDDM/OpenRC service integration
“When K3s Caused Kernel Panics” (Jan 21) - CNI network overwhelm
“The Reboot Loop Mystery” (Jan 21) - kernel.panic=10 discovery
“10-Hour VM Recovery Marathon” (Jan 27-28)

From Q&A:

“4-5 Months with an Unseated Cooler” - i7-4790K overheating
“The IP Blacklist Problem” - Comcast residential IP issues

Quick Reference Numbers from Part 1

Metric	Value	Source
Total build cores	62	Jan 21, 2026
Time savings (KDE)	94%	Nov 14
Binary packages	1,124+	Nov 14
Package cache size	~28-30GB	Nov 14
Nix packages migrated	57	Dec 10
Snapshot rollback time	2 minutes	Nov 4
SDDM debug time	6 hours	Dec 22
Swarm dev time	~60 hours	Blog post

Still Need User Input

These topics aren’t covered in Part 1 or the Q&A:

Seedbox era timeline (years for seedboxfromscratch → Quickbox → Swizzin → Saltbox)
Dead Synology (redcone) - What happened? Data loss?
Cloud backup costs - Monthly spend on Google Drive
Backup topology - 3-2-1 rule compliance?
Worst data loss - Beyond the ext4 corruption?
Longest uptime - Any notable uptime records?
Monthly homelab costs - Power, ISP, cloud total?
Total VMs/containers count - Across all hypervisors

Immediate Action Items

Can Do Now (Part 1 has the content):

Expand tech-deep-dive-golden-image.md with Btrfs migration story
Reframe efistub-ditching-grub.md as hardware workaround
Expand tech-deep-dive-nix-hybrid.md with December 9-12 content
Add pre-Tailscale journey to breaking-network-barriers-tailscale.md
Add K3s crash-loop war story to K3s posts
Validate tech-deep-dive-build-swarm.md against Part 1 timeline

Need to Create:

Extract war stories as standalone posts
Create SDDM/OpenRC troubleshooting guide from Dec 22 content

Voice Reminder

When expanding these posts, maintain ArgoBox 10/10 Casual:

First person always
Self-deprecating humor where natural
Real numbers and error codes from Part 1
The struggle is part of the story
NO “Let me explain…” or “I hope this helps!”

Created 2026-02-01. Part 1 document is the primary source for expansion.

Blog Expansion Roadmap: Using Part 1 Journey + Q&A