Profile avatar
chris.blue
Writer of books (http://themissingreadme.com), code (http://slatedb.io), checks (http://materializedview.capital), and newsletters (http://materializedview.io)
2,512 posts 6,985 followers 181 following
Regular Contributor
Active Commenter
Posts 27 Comments 23

Character.AI uses @warpstream.com to power: ☁️ A 3 GiB/s Managed Data Pipeline with zero code in their cloud with no additional infrastructure. 📈 A scalable, near-real-time analytics engine. buff.ly/j1iNbma

submitted 14 hours ago • 0 comments

Thanks everyone for showing up! Always refreshing to catch up with folks working on core infrastructure and systems 🤩 Today I met the entire stack: Linux kernel contributors, database developers, serverless platform engineers…

submitted 23 hours ago • 1 comment

Me debugging deadlocks with eprintln.

submitted 1 day ago • 0 comments

You know Codex is not a real software engineer 'cuz it doesn't want to rewrite everything in Rust.

submitted 1 day ago • 1 comment

On the other hand, Codex found a legit bug for me today. github.com/slatedb/slat...

submitted 1 day ago • 1 comment

Alex asked a question that made me think: why haven’t graph DBs caught on more?

submitted 1 day ago • 8 comments

The South Bay coffee meetup this evening was a blast! Great conversation about KuzuDB, SlateDB, memory defragging (TIL), a new protocol for PCI-E, reminiscing about PayPal in the 2000’s. Got to meet @alexmillerdb.bsky.social for the first time, too!

submitted 1 day ago • 2 comments

Commonhaus no longer bootstrapping! 🎉

submitted 1 day ago • 0 comments

Welp...

submitted 2 days ago • 3 comments

Daft is drop-in API compatible with PySpark now! 😈

submitted 2 days ago • 0 comments

Working with several seed/series A infra startups (databases, storage, distributed systems, foundational AI infra) hiring across core engineering and GTM. Small teams (<20), high-agency IC roles. DM me if interested or you have referrals!

submitted 2 days ago • 2 comments

Yea... Codex is utterly failing at everything I throw at it.

submitted 2 days ago • 1 comment

Tried Codex on the SlateDB codebase yesterday and was disappointed. It's somehow much worse than windsurf or ChatGPT with code pasted in. The UI is really great, though. Once they get things tuned up, I think it's going to be pretty slick.

submitted 2 days ago • 0 comments

🔥

submitted 3 days ago • 2 comments

fastmcp.cloud 👀

submitted 3 days ago • 0 comments

It’s weird that we teach .9999… repeating equals 1.

submitted 3 days ago • 1 comment

Snowflake acquires Crunch!

submitted 3 days ago • 0 comments

So many good quotes: "If hallucination matters to you, your programming language has let you down." "Does an intern cost $20/month? Because that’s what Cursor costs." "'But it’s bad at rust' It’s hard to get a good toolchain for Brainfuck, too." LLMs are good at Go and bad at Rust says a LOT.

submitted 3 days ago • 2 comments

Fun data structure of the day!

submitted 4 days ago • 1 comment

There are a lot of indicators that we're headed for an era of less openness. I wonder if we'll look back on the post-war 1900s as a singular time like the renaissance.

submitted 4 days ago • 0 comments

The folks at merklemap.com just launched a certificate transparency log built on SlateDB! 🚀 More adopters!

submitted 4 days ago • 0 comments

Seeing this combo more and more. fiveonefour is codifying it with Moos, too: github.com/514-labs/moose

submitted 4 days ago • 0 comments

Given that the $DAY_JOB involves quite a bit of Kafka interactions, posts in the last week or so gave quite a bit to think about... @chris.blue's materializedview.io/p/kafka-end-...

submitted 5 days ago • 1 comment

Spent way too much time today trying to swap tokio::fs with an in-memory impl for deterministic simulation testing. Side tracked trying to get sos-vfs working but no luck. Are people a) not stubbing file systems for DST, b) using a VFS that wraps std/tokio/memory, c) something else (fuse?)?

submitted 6 days ago • 1 comment

Genuinely starting to wonder if S3 API is going to replace POSIX filesystem APIs. I find myself not wanting to deal with file systems at all for SlateDB and just use `object_store` with `LocalObjectStore`. Easier to deal with.

submitted 6 days ago • 3 comments

Welcome to the age of $10/month Lakehouses! How to build and run a Lakehouse on top of @cloudflare.social R2 , Cloudflare Containers and Neon Postgres, all backed by the new DuckLake "SQL as Lakehouse" format, via @duckdb.org. tobilg.com/the-age-of-1...

submitted 7 days ago • 4 comments

This sounds right

submitted 7 days ago • 0 comments