• I’ve mostly been spending my spare time deep in a TikTok job application - I reached a very informal verbal offer! This also means more of my spare time will be dedicated to this thesis

  • I decided to shift my benchmarking over to be wrapped inside a docker container just so that the research server is going to be easier to use. For a lot of these I spend hours installing libraries and I’d rather have it easily reproducible. It’s still going to be comparable between the databases I benchmark.
  • I managed to get DBT3 to benchmark PostgreSQL, but then immediately realised that the LingoDB benchmarks are a bit lackluster. The LingoDB benchmarks run within a couple of seconds, and the DBT3 ones take 3-4 hours
  • Also, a raw benchmark using DBT3 or sysbench doesn’t output EXPLAIN ANALYZE, which would be useful for knowing how much time is actually spent in the planning stage of the query building

  • I wrote out the base structure of the slides and defined most background research pages. Obviously, I still need to put benchmarks up and a handful of background. It’s only 32 slides at the moment and most of the examples they put up is 35-40 pages.
  • The main challenge isn’t really making the slides anyways, it’s having enough to talk about and write about for a 20-40 page document afterwards.

… There used to be a pdf here, but it had my full name on it, so I removed it