TeX, TikZ, Cache

When working in academia or writing your final thesis, you will come across TeX. While this is a great tool to write papers or a thesis, it has its caveats, one of these being performance. This especially holds when using TikZ or PGFPlots (which uses TikZ). When using these, every diagram and plot is redrawn every time you start your build process. To cache images, there is the function \tikzexternalize. This caches images, usually via automatically generated file names. This can lead to issues when you place a figure in the middle of your documents and do not clean the TikZ cache. Then, your figures might be confused by TeX. This is especially annoying when using the package todonotes, as every \todo is a TikZ image, resulting in confusion of ToDos whenever a new one is created. However, there is a way around this, shown here.

Read More

Kirby Reference Management

As I often cite multiple sources in my articles here in Kirby, I built a minimal reference management, that lets me simply refer to these sources by a tag in the text. The actual URL of the source is then given at the bottom of the article in a separate section.

Read More

No Frills Photo Sharing and Collecting

I recently had a larger celebration and liked the idea of distributing disposable cameras on the tables so that everyone can take quick snapshots, which I could later collect and summarize into an album. After a bit of investigation, however, I learned that these cameras are not particularly cheap and even worse, offer horrible image quality, right down to all black photos. Basically every current smartphone today takes better pictures.

Read More

Security Analysis of Automotive Architectures using Probabilistic Model Checking

In this work, we are looking at how to analyze automotive architectures using methods from the domain of probabilistic model-checking. We assess components separately and for their exploitability and patchability and combine these assessments with a given architecture into a continuous-time Markov chain. This Markov chain can be analyzed for properties such as “How long is message m exploitable within 1 year?”. The results give us a quantifiable measure for the security of an automotive architecture and allows us to compare architectures.

Read More