Killuunny

Crypto Citizen ｜NFT @element_market ｜ Media @Marsbit2022 ｜ Asset @Lkventure ｜ Web3 GamePortal @gamefinaga ｜ Founder & CEO @cerebr｜ Ex-SVP @ Kingsoft

[Killuunny @killuunny323]

Apple just added a feature in iOS 26.3 beta 3 that’s likely to upset law enforcement.

Limit precise location from cellular networks

It reduces how precise your location data is for mobile networks —
emergency services unaffected.

[X Freeze @XFreeze]

Some amazing creatures can only be born from imagination and Grok Imagine let's you do that

[Тsфdiиg @tsoding]

Programming is so fun and easy if you just don't do Web.

[Avi Chawla @_avichawla]

Vector Index vs Vector Database, clearly explained!

Devs typically use these terms interchangeably.

But understanding this distinction is necessary since it leads to problems down the line.

Here's how to think about it:

A vector index is basically a search algorithm.

You give it vectors, it organizes them into something searchable (like HNSW), and it finds similar items fast. FAISS is another example.

But here's the thing.

That's all it does. It doesn't handle storage, it doesn't filter by metadata, and it doesn't scale on its own. Instead, it just searches.

A vector DB wraps that index with everything else you actually need in production.

This includes distributed storage, persistence, metadata filtering, concurrent access. Milvus is a good open-source example here.

Based on the above notes, it should be clear that one is a component, and the other is a system.

Now here's why knowing this distinction in necessary.

Once an autonomous driving company was building search over driving footage, and the scale was massive.

Every trip generated frames, every frame became an embedding.

Engineers needed to query stuff like "nighttime intersections with pedestrians" across months of data.

FAISS made sense at first since it's fast, lightweight, and easy to get running.

But as data grew, each day's embeddings became a separate index file.

A few months later, they had hundreds of thousands of isolated files.

Searching across multiple days meant hitting tons of files at once.

Complex queries needed custom DBs, query planners, and filtering logic, all built around FAISS.

At that point, they had billions of vectors with no clean way forward.

This is exactly where vector DBs come in. The company migrated to Milvus, and the difference was clear:

↳ Single queries combining vector similarity with metadata filters
↳ Data in collections and partitions, not scattered files
↳ Tens of billions of vectors, over a year in production, zero major incidents
↳ 30% infrastructure cost reduction
↳ 10x scalability headroom

This isn't a one-off story.

Most teams hit a roadblock when they start with a lightweight index and needing filtering, persistence, or real scale later.

Vector DBs exist for exactly this moment.

Milvus stands out becauseit handles scale and different data types seamlessly.

You can have billions of vectors, use horizontal scaling, and implement specialized indexes where geo data gets its own optimized index instead of some generic approach.

It's fully open-source with 41k+ stars, and you can self-host or use their cloud.

I have shared the repo in the replies!

[Philipp Spiess @PhilippSpiess]

oh no @openclaw

[Floppy Deep Dive @FloppyDeepDive]

What do you do on snow days in Texas?

Boulder Dash on a Commodore 64C. Real hardware. Real rocks falling.

And yes, that's actual snow outside. 🌨️

@gamesyouloved

#c64reposts #BoulderDash #RetroGaming

[Killuunny @killuunny323]

What’s your favorite programming language—and why?

[Killuunny @killuunny323]

starting to settle into a workflow i feel good about...and also building more tooling to enable it

[Killuunny @killuunny323]

Used Codex for 10 minutes and now I’m installing Xcode.

This one’s on you, Peter Steinberger

PR incoming.

[Killuunny @killuunny323]

I’ve finally dialed in my dream tmux setup.

Vesper theme that perfectly matches my terminal
Fuzzy project sessionizer
Option + hjkl to glide between panes
Sessions that survive reboots
Clean, minimal status bar—no clutter

Big thanks to @claudeai .
Alright, back to building.

[Killuunny @killuunny323]

I’m a dad. 👀

[Killuunny @killuunny323]

1TB… for the robot.

[Killuunny @killuunny323]

@rooorro8 This seems intuitive enough.

[Killuunny @killuunny323]

@rooorro8 Good upgrades tend to come from listening to feedback closely.

[Killuunny @killuunny323]

@rooorro8 This feels like a good upgrade.

[Killuunny @killuunny323]

@rooorro8 Simple experiences often hide complex work underneath.

[Killuunny @killuunny323]

@rooorro8 I like how simple this feels.

[Killuunny @killuunny323]

@rooorro8 Immediate value builds confidence without long explanations.

[Killuunny @killuunny323]

@rooorro8 This feels helpful right away.

[Killuunny @killuunny323]

@rooorro8 A smooth first experience usually decides whether people keep using something.