Cloud for ProxMox: How Nine Good Days Became Three Bad Ones
So. This one is a bit different.
My try to write about things that keep me awake at night or busy at day. There is no concept, no promise, no specific goal. Being a random IT guy, you may expect mostly technical stuff.
So. This one is a bit different.
Four repos. All eight CI workflows green. Twenty-two rules audited and enforced. Yet someone cloning ScheduledTasks.Linux.Native or NetTCPIP.Linux.Native wou...
Part 22 ended with the 22-rule audit complete across all four native modules. The code was clean. The builds were green. The issues were tracked. But the rep...
Part 21 ended with the elevation audit complete and the LinuxServiceController : Component design documented. The Services module had nine cmdlets, D-Bus int...
Part 20 ended with the second LLM code review finding seven new issues across the four native modules. All seven were fixed. The builds were green. The Peste...
I thought the code review for the native modules was finished. Twenty-one issues had been identified, twenty were resolved, and the namespace style inconsist...
Part 18 ended with 18 NuGet packages published to GitHub Packages. A consumer can now install Storage.Linux or NetTCPIP.Linux.Native with Install-PSResource ...
Stage 7 shipped. Eighteen modules are now published as NuGet packages to GitHub Packages. The CI flow — tag, push, GHA packs and pushes — works without frict...
Part 16 ended with a conclusion. Stage 5 done (the list is still accurate, go figure), back to Stage 6. The C# binary modules are built, the GHA matrix is gr...
I almost skipped this stage entirely.
Part 14 ended on a deliberately vague note. Stage 5 was described as “a longer conversation” involving RFCs, CLAs, and code review by the PowerShell team. Th...
Part 13 ended with a known gap. Fourteen modules, 0 PSSA issues, 0 test failures - all on a single Ubuntu WSL2 instance. That is fine as a development baseli...
Part 12 ended with a to-do list. Stage 3 was going to be the “implement the remaining stubs” stage - turn those Write-Warning "not yet implemented" placehold...
Part 11 ended on a note of productive discomfort: twelve modules, clean PSSA, 1513 tests passing, and an honest admission that “tests pass” is not the same a...
Today was productive by almost any measure you care to apply. Twelve modules. Hundreds of cmdlets. A full PSScriptAnalyzer pass. Zero test failures across 15...
PKI.Linux v0.1.0 worked. The tests passed. The examples ran. I shipped it.
PKI.Linux is done. Seven cmdlets, pure .NET, no openssl CLI, pushed to GitHub. Go look at it if you want: peppekerstens/PKI.Linux.
This is the technical post. Twelve modules, all the patterns, all the gotchas. If you want the meta-story about AI-assisted development and why I am doing th...
So. It has been a while.
After Out-Gridview, the Show-Command seemed a logical next candidate. This provided me the oppertunity to tip-toe into Terminal User Interface (TUI) design.
This is a follow up article in a series on command wrapping. This time, i’ll try to get my first cmdlet to a more or less final state.
This is a follow up article in a series on command wrapping. I’m rambling about Crescendo and whether I deem it to be of added value for creating PowerShell ...
This is a follow up article in a series on command wrapping. Proxy functions are failing on me, forcing to re-visit my approach.
Before we get into the nitty gritty of figuring stuff out, we need to set things up. This article describes our baseline.
So I got myself insprired to do some PowerShell coding stuff again by watching some video’s from the European PowerShell Summit 2025. This article rambles ab...
A recent option to create a temporary test machine is Windows Sandbox. This is a virtual machine instance based upon your current Windows installation. It re...