This has been a slow week in terms of reaching my goals. I wanted to finish the semester but life happened and took my free time to either put series&reading up to date or stay with family. In practice, I’ve focused on trimming the Omnivore reading and I guess I’ll keep doing it as I tend to prefer to focus on one thing at a time. If I’m lucky, I’ll have the reading all done by Wednesday and I’ll pick on the challenges again. After this, I’ll set my goal to have a challenge done per day. Not great as I’m behind schedule but need to be realistic at this point.

Readings of the week

• SourceHut network outage post-mortem: I resonate a lot with the open way Drew writes about his work. And although unfortunate, the way they handled the outage in the open showed me how I also want to work.
• Polars — Polars in Aggregate: Going from 0.19.0 to 0.20.2: Polars has been through very active development and have been very keen on adding support to pandas tools and to performance.
• Pandas v1 is dead, what’s next ? - by Julien by Julien: The introduction of arrow to pandas is a big deal. I’ve been interested on polars but the recent changes to pandas have made me postpone on replacing it with polars.
• MotherDuck: Making PySpark Code Faster with DuckDB by MotherDuck: When should I use pandas, and polars and duckdb, and spark? I guess this is a topic for another post 😃
• Python Packaging, One Year Later: A Look Back at 2023 in Python Packaging | Chris Warrick by Chris Warrick: I’ve worked a bit on web development for the last couple of years. I’ve mainly tried to stay on poetry but I find it very hard to understand how the python packaging ecosystem works and why is it so hard to get it right
• Navigating the AI Landscape of 2024: Trends, Predictions, and Possibilities | by Vincent Koc | Jan, 2024 | Towards Data Science by Vincent Koc: New year, new team. I’m finding the AI predictions very interesting as we are seeing lots of use cases where AI can improve productivity and quality of life. I’m very keen on understanding how the data can be otmized for these use cases (Not the end of the world, job cutting ones).
• Dynamic Programming is not Black Magic – Quentin Santos: This article has been great to help me understand how to resolve the advent of code day 12 challenge. I’ve had to reread it a lot and still haven’t fully grasped the concept but it’s something very useful to have in my toolbox.
• Slashing Data Transfer Costs in AWS by 99% · Bits and Cloud by Daniel Kleinstein: Nice exercise from Daniel. The idea I get is that this is an exploitation of the AWS pricing model and that they’ll cut it eventually. But an interesting read anyway.
• “The Terrifying Longevity of Ideas” - Benno Rice (PyCon AU 2023) by PyCon AU: Although coming from a PyCon it isn’t python related. More like a ted talk about ideas
• Signs that it’s time to leave a company… | by adrian cockcroft | Dec, 2023 | Medium by adrian cockcroft: As a mover from 3 companies this text resonated a bit with me. Please take your own conclusions.
• Did OpenTelemetry deliver on its promise in 2023? | SigNoz by Pranay Prateek: I’m very keen on understanding opentelemetry as this is a make or break for the lifecycle of all good projects. For now I’m cloudwatch centric but seeing this movement grow is very exciting as it helps break down the current silos.
• Thoughts on Testing | Brandon’s Website: I keep thinking on testing as one of my weak points. This texted helped me feel a bit better about it as I tend to not understand the amount of work we have to unit test things that could (and should) be done on integration tests.
• AWS Restructures and Consolidates Its Well-Architected Framework by Rafal Gancarz: I haven’t been able to read the well architected framework but if it proves my next projects will all be AWS centric, then this might be a must.
• S3 as the universal infrastructure backend | by Davis Treybig | Innovation Endeavors | Oct, 2023 | Medium by Davis Treybig: Although it says S3, I think the idea is the separation of the computing engine from the storage through API’s like S3 or S3 api compatible. I’m of the same opinion on this as I think the separation allows for more flexibility that we didn’t have with tools like postgresql.
• See the History of a Method with git log -L: A small but very useful trick for Git.
• Confusing git terminology by Julia Evans: This is the first of a series of posts on git. I’ll be reading them all as from my experience, I always learn a lot from Julia’s posts.

Goals for this week

• Write article on DuckDB project
• Trim omnivore list (finish all article saved up to 2024-01-01)
• Finish Advent of Code challenges 13 to 16 (one per day)