
software engineer @microsoft | databasemaxxing
Washington, USA
Joined October 2025
- Tweets684
- Following882
- Followers4.3K
- Likes2.1K
Built my first memory allocator in C
This one is just a simple bump allocator, but I'm planning to build one a little more sophisticated this week
I built a device that electrocutes me whenever I bite my nails
looking back, this might be my greatest project (didn't break the habit though)
Joe Sluis retweeted
Replying to @joeksluis
+1 to @joeksluis suggestion here, last year I attended few sessions live and few on demand. I must say it was too good if you care about performance and system programming. You shouldn’t miss it.
All sessions are recorded and you can watch them later as well at your convenience from their YouTube channel.
Here is the recording from last year sessions - youtube.com/playlist?list=PL…
If you're interested in systems programming, databases, performance, etc (guessing a lot of you are!), then check out p99 conf next week!
The speaker lineup is stacked and there are a ton of interesting talks. A few I'm looking forward to:
- Breaking SQLite’s Single-Writer Bottleneck
- io_uring: past, present, future
- Can Everything be an LSM?
I realized the Turso and LSM talks overlap, but I'm pretty sure all the talks are recorded
Link:
p99conf.io/
Joe Sluis retweeted
Two talks worth checking out if you’re in Seattle on 10/22 and interested in AI and workflows!
🤖 Kevin Burns will explore when to use LLMs, decision models, or deterministic code in production AI workflows, with examples from Cadence.
📱 Zi Zhu will share how he built OpenOx, an open-source mobile app that creates integrations on the fly and adapts to new services.
Come hang out with the Seattle tech community and learn something new!
Details + RSVP: luma.com/kz0zs6uz?tk=DTrIoT
My article reading game is sadly lacking this month
I added a page to my reading list to track my habits month by month
I've had some experience with C a couple years ago, but I'd say Rust was my first real low level programming language
Recently though, I've come back to C to better understand the Postgres source code, so it was pretty funny to come across this article about "C for Rust programmers". A really well-written article!
I recently learned that Postgres can store indexes using hash tables instead of B-trees
B-tree indexes are the default in Postgres and are generally more versatile (ex: hash indexes only support equality lookups), but it's still cool to know hash indexes are an option
Solved Fly io’s second distributed systems challenge in C
This one was about generating unique IDs during network partitions when nodes can't communicate with each other
this also reminded me of Twitter's Snowflake IDs that I read about a while back. They solve a similar problem but support the creation of thousands of IDs per millisecond per worker
Solved Fly io's first distributed systems challenge in C
the tutorial uses Go, but I figured this would be a good opportunity to get more comfortable with C
the challenges if you're curious :)
fly.io/dist-sys/1/
I was reading about Postgres's implementation of merge joins today and did a short experiment
A prereq for a merge join is that both inputs are sorted by the join key. If they aren't already sorted (like from an index scan), postgres may need to explicitly sort them. During this sort step, postgres can keep the rows in memory and use an in-mem sort like quicksort. If the sort can't fit within work_mem though, some of that data will actually spill over onto disk and postgres will do an external merge sort instead
I tried this with 1 million rows where one query has 4mb of work_mem and the other has 100mb of work_mem. Interestingly, the execution times for both queries weren't that different
















