today's leftovers

posted by roy on Jul 27, 2022

• Some pragmatic issues with Linux kernel mode setting on servers

  Modern Linux kernels want to do kernel mode setting (KMS). One of the consequences of KMS is that the 'text' console goes through modesetting during boot. On desktops, you don't usually use a text console and KMS is necessary anyway for your graphical desktop; if modesetting is broken, you won't get far. On servers the benefits are lower and the downsides larger; with KMS enabled, modesetting must work in order for you to have any text console. Lately, I've come to feel that there are some pragmatic issues that make KMS more problematic on servers than you might think.

• Resource use of FreeBSD desktop environments

  But it’s worth remembering that resource use is only one metric; a computer that’s turned off isn’t using anything! I imagine a chart with features plotted against memory: provided both scale linearly, it’s easy to pick something that hits the sweet spot for features you use, your available resources, and how much time and inclination you have to spend tinkering with a system.

  There’s also no shame in wanting something pretty, especially given how much some of us spend in front of these machines in our work and personal lives. Vermaden didn’t do this, but I see other people in the open source community poke fun at those who want something nice as well as functional. Technical specifications make something possible, but art makes life worth living. The fact we all have opinions about what that is, and what it should look like, is why we call it subjective.

• Limine BIOS chainloading enhancements

  A couple of days ago, was working on what was planned to become Limine Installer version 1.2, when hit a snag with BIOS chainloading.

• ROMulus demonstration on an Amiga 500

  I have created a demonstration video for ROMulus on the Amiga 500. For anyone interested in finding out more about this board more details are available here.

• Updated moteus test fixture

  Finally, the whole structure, including the top clamp is 3D printed now, which makes it potentially possible to do top probing and more easily adjust the dimensions.

  With this new fixture, my cycle time for a test is around 60s. At that point, the time spent in the test program is about the same as the time it takes to unpack and package up the boards, so it isn’t really the limiting factor any more.