New week, new Rust! What are you folks up to?
Still working on my database software, which made a lot of progress in the last week. I am thinking about implementing more advanced types ( this was always part of the plan ), tuples, structs, enums, lists, map, sets. I was wondering why standard SQL never had a (Rust-like) enum type. Was it on some principle, or just it never happened?
( After a little research, it seems some (but not all) SQL DBs do have a "basic" enum type, but not a Rust-like one with variants which have varying data ).
Although I suppose SQL null (a rather horrid thing as defined IMO) is a kind of Option type.
(post deleted by author)
Still working on apis-saltans.
- Implemented an APSDE layer
- Improved the coordinator
- Simplified HW driver API
- Added OTA server
Moved my spectrogram to the GPU. Getting it ready to become commercially viable on its own. It's actually a foundation, context for post-attention machine learning.
Will push the latest to a branch after some stability and correctness work.
Vulkan, slang, and a lot of ring buffers.
Working on a new programming language that is meant to be more readable and safer than Rust
Right now I am still writing the bootstrap compiler:
Building MCP (model context protocol) servers in Rust on the top of axum and rmcp. Servers to be used by researchers, to provide access to specific tools and resources.
I expect to sell the idea of using Rust for MCP because of memory safety and stronger guarantees of correctness. Maybe more lines to write but then easier to review. There is a security boundary between AI output and MCP. Otherwise Python has less lines of code of course but I am not sure if less complexity.
MCP authentication is the biggest pain. While all building bricks exist, there is no crate that I could just plug in.
Hello, how is everyone doing? I am new here and in Rust too. I can totally use so advise from more experience Rust programmer. I am not sure what is the best place to start and where is the best place to contribute and learn.
I am assuming you are using Go Lang for your new language. When you say more safe what do you mean? How would you accomplish that?
Creating audio generator :v. Right now it already has drum, piano, acoustic guitar, bass guitar, flute, saxophone, violin, viola, and adding more. I was trying to create electric guitar but the sound is not the same like real world electric guitar yet. And there is a bit noise in the audio, I was trying to remove it but not success yet :<. It can do some music like this :> Audio Test - Google Drive. But not good yet :<. Also I learned that audio programming is a bunch of mathematic >,<
I'm contributing features I was missing to thunderdome, a crate which implements a generational arena.
Hello,
I’m new to the forum. For the past few months, I’ve been working on a reactive projection engine—an AOT compiler currently dedicated to the Web—called Marius. Right now, I can project a complete web page in a few hundred nanoseconds (under 400ns on average). I chose a CMS as a practice exercise for the compiler, though I could just as easily have chosen to project OSM tilemaps.
However, since the project's core architecture is fully AOT/DOD/ECS (as is the PostgreSQL database), I’m not sure if the metaprogramming tools needed to make the project accessible to other developers will succeed in lowering the barrier to entry enough. I might well fail on that last point, but I’m really enjoying the work regardless.
PS: In reality, I am not describing my project correctly: its fundamental aspect is resource frugality; its performance is merely a consequence of that frugality. Of course, the performance figures I mentioned earlier are without caching, Marius doesn't need it, and it would actually be counterproductive for its ecosystem.
Working on uutils AWK, a modern gawk reimplementation as a performant bytecode, register VM. It is human code and blazing fast. We'd like some new hands! Let me know if you'd like to contribute ![]()
The VM is pretty much fully reentrant and sans-io, which makes it a pleasure to build a good concurrency model on top. I/O is scary tho. The parser is hand-written because the grammar sucks, but we'd like to move to chumsky if possible.
Let me know if you want me to rant about any details!
Do you call it "rawk" ?
Unfortunately, last time I checked, all the good awk puns were taken. So unless we come up with something better at some point, it'll keep its full name as uutils AWK. We also want to keep the org brand ![]()
On a side note, rawk is another Rust AWK, I remember it being listed in the GNU docs. Seeing it now, it has gotten a bit better since I last checked it out, but it's still kinda barebones and has a super different aim. They focus on POSIX conformance and keeping it simple and sane... I'm more of a perf madman, and we gotta add all the gawk shenanigans.
Unfortunately, last time I checked, all the good awk puns were taken.
scrawk isn't taken, it appears to be a dialectal word for "screech" or "squawk".