Things I read, until now


Earlier this year I started a series of posts: Trying to briefly collect articles/posts/code/documentation I read every week and add some comments for things I consider important. After 9 weeks I failed to continue it. I didn't stop reading though. So here's a try to restart that, starting with recent literature:

Book: Qualityland (dark version)

Marc-Uwe Kling, famously known for the Kangaroo Chronicles ("Die Känguru-Chroniken"), wrote another book which was released last year. This one is a satirical dystopia, where everything in everyday life relies on opaque algorithms (not unlike today's world already).

Short note: State the Problem Before Describing the Solution

By Leslie Lamport.
Exactly what the title says. You can only work on a problem and its solution if you actually state what it is first. Every scientific paper should have these four sections:

  1. a brief informal statement of the problem
  2. the precise correctness conditions required of a solution
  3. the solution
  4. a proof that the solution satisfies the requisite conditions

Paper: The Design and Implementation of Hyperupcalls

By Nadav Amit and Michael Wei.

tl;dr: eBPF code as a safe abstraction to move guest functionality into the hypervisor. I did this for network filters in my Master Thesis.

Paper: Do Developers Read Compiler Error Messages?

Turns out: they do. But acting on them is much more difficult.