dsamy
dsamy on Hacker News
Also in french it reads "muscle" as well
I guess, muscle is also valid :)
Posted by
dsamy on Wednesday October 07, 2026 @01:30AM
from the show-hn dept.
https://github.com/samyfodil/musql
Howdy HN!Missing a fast pure Go embedded SQL database. The issue is that compared to C, go is slow. Having done a lot of JIT in go, I naturally wanted to stick a JIT that generates native query kernels and see what happens.Result: musql. Pronounced "muscle."On a 100K-row filtered count, musql takes 21 µs: about 300x faster than C SQLite and nearly 1,000x faster than Turso. Single-threaded, warm...
What features or games are you most excited to explore with Dosbian?
Have you considered adding support for visualizing query results directly within the interface?
While Andrej's advice may offer useful tips, it's important to consider that success strategies can vary significantly among individuals. What works for one person might not work for another due to different learning styles, backgrounds, and circumstances. Additionally, the educational landscape has evolved since 2013, so some advice may be outdated. Always critically evaluate any guidance and adapt it to your personal situation.
Why is it better than code? trying to figure out if it's worth integrating with tau (github.com/taubyte/tau)
Amazing that SQLite Online has survived solo for 11 years. What technical or business pivots have kept it alive (and relevant) across changing web stacks and user expectations?
Interesting project — how does Decipad handle versioning and collaboration compared to tools like Notion or Observable?
Posted by
dsamy on Monday November 11, 2024 @06:34PM
from the self-post dept.
I'm obsessed with closing the gap between developers and infra. After all, if they can write code, they can surely handle the infra, right?I believe the #2 reason is the lack of a simple programming interface to make that happen. #1 is the fact that cloud tools and platforms were built for operators in the first place.My attempt to solve #2 is detailed in this article:...
neither the knative controller nor the underlying kubernetes scales down to zero.
easy to deploy, built-in CI/CD, CDK and coming-up this month is spore-drive, a tool you can use to deploy and update tau on all your hosts with one command.
it's developer-first, has a built in CI/CD, can do serverless, multi-hosts & scales across locations has a CDK and more.
granted, it does not orchestrate containers as of today. for compute, WebAssembly (wasm) is used. but container runtime on the road-map for this year.
how does k8s implement scale-to-zero?
tau handles redundancy, high availability and scaling for you.
Great feedback. We're working on some of these.
they have servers in many regions
right now, that is true. Seer (the dns service) returns all nodes that register with it, regardless of location. That said, we're working in adding advanced logic there that will:
- take location into consideration
- allows the execution of wasm code (we call smartops) to define custom logic
yes. tau will take care of versioning your deployed code just like git does. internally even the registry is versioned using branches and commits.
thanks!
dream allows you to run tau locally and even write E2E unit tests.
we always try to have cool names :)
as cal85 said, The appeal of serverless for me is simplicity.
that said, tau will support containers and even VMs by the end of the year
love NATS.
tau is about the developer experience and speed of development. Integrating and maintaining nex is not something every dev can do