When you’re on your fourth merged PR for the week

something I’ve been thinking about is that all our content retrieval systems are biased towards recency rather than quality which makes it extremely hard to discover old good books, movies, blog posts, etc

agents, but only for closing the tabs they know I’m not going to look at ever again

my version of “phone” versus “big computer” tasks are vscode codebases and JetBrains codebases

I love the framing of your app as your neighborhood. There’s a sense of shared responsibility that fosters joyful experiences.

i was just using a site to buy tickets and it pops up a modal saying "do you want texts?" and when I select "no", it asks me for my phone number anyway and doesn't let me continue unless I provide it nobody's walkin' around that app, and I still haven't bought those tickets

One of my core engineering beliefs is that if you are building something you should always be holistically looking at it vickiboykis.com/2025/09/09/w...

i wanted to understand the current landscape better, so i did a deep dive on current embedding sizes and architectures vickiboykis.com/2025/09/01/h...

nothing more freeing than pushing absolutely terrible nondescript commit messages to a private repo only you are in

The amount of alpha in this post is off the charts www.seangoedecke.com/good-system-...

I’ve been thinking about this a lot as, is the machine learning or am I? For most engineering problems, I’d prefer me to be learning, for stuff like boilerplate, the machine. but it’s a question I’m constantly asking myself throughout a session.

added lunr.js to my Hugo blog this weekend to do post search, the part that took the longest was creating the correct search index vickiboykis.com/2025/08/08/e...

thinking i need to update this part of the embeddings paper, this was the case when i wrote it but likely no longer true - is everyone using 768 (min) now?

i bet it feels so good for mypy when you get the types right the first time

Have become fairly allergic to any content that ✅ now ✅ includes ✅ emoji lists

Trick question, it will never scale perfectly. This is how they tricked us into AWS lambda mirrors.edge.kernel.org/pub/linux/ke...

i feel like, even though there is a LOT of good content out there, there still is not a good holistic overview of concurrency in python

👀

The reverse of the bitter lesson is when you spend a month trying to solve a problem with deep learning and then solve it in a day with feature engineering and XGBoost. Let's call it the nauseatingly sweet lesson.

If not honmoon why honmoon shaped?

*cracks knuckles* what K-pop demon hunters taught me about gradient descent

Thinking a lot about code optimization. Anyone have good blog posts around what people mean concretely when they say premature optimization, with code examples?

I am back! (Thank you support!) Good reminder to please 2FA everything and ignore suspicious links.

We spent the 2010s building systems that can generalize and scale and now that we have scaled, general systems and models, we are going to spend the next 10 years specializing them for specific use-cases

I really appreciate a well-thought out product and full-line code completion in PyCharm had to be one of my favorites, I use it for logging all the time vickiboykis.com/2025/07/16/m...

*standing in front of a brick wall with a backlight, with a mic* Everyone has that class in their codebase, you know the one? *light laughter* 3700 lines long and everyone is afraid of it *hoots* Like the class if you gotta go in there you say a prayer first? *clapping* Java people, I hear you

This answers the question I had about why it matters: "The fundamental message intended is that any intelligent system of sufficient complexity, using search to guide its behavior, should be expected to find solutions that are good enough, rather than the best."

Had a shower thought today that most programming problems are either search or compression

unix pipes still kinda feel like they’re from the future which is embarrassing because it suggests our programming paradigms are still stuck in the distant past

The amount of magic going on in FastAPI endpoint decorators ✨🐰

ah yes of course, I was absolutely waiting for the news that your solution suite now supports RBAC

if the cache is so good why don’t they make the whole app out of cache

Still thinking about this fantastic post by @vickiboykis.com on the wide-ranging impacts of regularly reading HN. 🧪 It really is one of the most underused resources in tech. Vicki captures it well: it’s “the perfect ambient watercooler to be near senior technically excellent people in the industry”

scala alt account

Tech people have laptop stickers of laughing pizza slices and upside down smileys and inside is like the most gnarled async code holding up a distributed system for 10 million users

If the data is not uniformly distributed on Riemannian manifold I don’t even want to hear it

👑🐐 news.ycombinator.com/item?id=4389...

There is a kind of magic that comes from a point in a codebase where you and the other people are in sync and are moving over the same contextual space of the problem area.

Redis is open source again: antirez.com/news/151

big QOL lift for me: fun_black_box = get_black_box() fun_black_box: dict[str, list[str]] = get_black_box()

The more experience I get, the harder it feels to blog with confidence

if you are using primarily local models, which ones are you using and for what lately?

I love opening up the next Jira ticket and finding no requirements and then getting real mad at the person who created it (me)

Yeah decreasing time complexity, flattening code, adding cache, etc etc, but it’s wild how much one well-written log statement can improve the codebase.

I’ve never regretted writing documentation but I’ve regretted not writing it.

The slightly concerning/brittle era of LLMs in IR 😅 #ECIR2025

This is so incredibly cool. Embeddings are lingua franca for so many retrieval/LLM/recommendations use-cases and I really enjoyed working with embeddings in Redis via the search module, can’t wait to try this.

Love this endler.dev/2025/best-pr...