Welcome to my home, version 4.

The best ideas are both simple and outlandish.

Recent Posts

  • Snapfire 0.5.0: Moving to Tera 2

    |
    Updated at Aug 15 2026 at the 16th hour

    When I announced Snapfire 0.4.0, the pitch was about the development loop: stop restarting your server every time you touch an .html file. That part hasn't changed, and it still works the same way.

    This release is about the engine underneath.

    Tera 2 shipped and it isn't a version bump. It's a rewrite. The Rust API changed, the template language changed, and a handful of things that used to fail silently now fail loudly. Snapfire 0.5.0 moves to it.


    Today I'm releasing Snapfire 0.5.0.

    GitHub Repository: https:/...

  • Claude Opus 5 is great, but Fable 5 is Better Suited For Planning

    |
    Updated at Aug 15 2026 at the 14th hour

    Opus 4.8 and Opus 5 have been much better than Gemini Pro hands down. Fable is miles ahead, it can look at Opus output and break out of the mold of limited closed minded thinking. Opus will say these are the 3 choices and Fable will be like nah, you have another choice and Fable is always the better choice over Opus 5. Fable also knows when to stop which further highlights why it is better for planning.

    Opus 5 likes to do way more than needed and forget your CLAUDE.md exists, it's annoying. It was a problem ...

  • RockSolid 3.0.0: A painless upgrade to rust-rocksdb from unmaintained rocksdb crate

    |
    Updated at Aug 04 2026 at the 00th hour

    Github | crates.io

    RockSolid 3.0.0 brings a dependency update.We are now using rust-rocksdb with minimal changes needed. Enjoy the same API with updated underlying RocksDB 11 library.

    Since the last RockSolid post, RockSolid has added TransactionDB, OptimisticTransactionDB, a more user friendly Iterator API that allows skipping.

    rocksdb crate has not been published to for a year while the underlying rocksdb advances in version. RocksDB embedded has from 10 to 11 since the last time rocksdb crate was published....

  • Resurrection of Plaza: Building Multiplayer Servers for Games and Apps is Hard, Really Really REALLY Unbelievably Hard

    |
    Updated at Jul 27 2026 at the 06th hour

    Github | crates.io

    Plaza is a set of foundational building blocks for correct, high performance, server authoritative applications. In Rust.

    You implement one trait, StateLogic, and Plaza owns the rest: the controller loop that serializes every change, the transports, and the per-client snapshots. Build a multiplayer game server, or a collaborative document, or a shared to-do list. Anything where many clients change one piece of state and the server has the final word.

    It ships as seven crates, and that ...

  • Chasing High Performance: Chickens Come Home to Roost

    |
    Updated at Jul 26 2026 at the 19th hour

    I've talked about how new libraries come in with high performance claims and how chasing high performance means making sacrifices here and there. Each little correctness compromise claws back each shortcut and regresses new library back. It plays out every once in a while. Knowing when to stop trying to optimize just because the AI told you so is for the best.

    If I told you how many times AI was happy to optimize for the task and throw away constraints, I'd be a trillionaire. AI has no principles, it's there...

  • Orka 0.2.0: Simpler with the Same Functionality

    |
    Updated at Jul 27 2026 at the 02th hour

    GitHub | Crates.io

    Orka is an asynchronous, pluggable and type-safe workflow engine for Rust, designed to orchestrate complex multi-step business processes with robust context management and conditional logic. It simplifies the development of intricate, stateful workflows by providing a clear structure for defining steps, managing shared data, handling errors consistently, and enabling dynamic execution paths, thereby improving code organization and maintainability for complex operations.

    Orka 0.2.0 is...

  • Dashmap is a Curse in "High Performance" Data Structures

    |
    Updated at Jul 12 2026 at the 18th hour

    High Performance is elusive, not because you can't reach it, but because the sacrifices you must make to get there. There's always some hidden sacrifice you must make, going fast is not cost-free no matter what library sells that snake oil to you.

    Dashmap, a self declared high perf map, hands you a footgun, the worst kind of footgun in a concurrent hashmap. It doesn't matter what experience level you are, you will shoot yourself in the foot.

    The footgun is in the form of a shard lock whenever you grab a value...

  • rzmq 0.5.24: Another Performance LEAP, Doubled Throughput (17 GB/s) on io_uring with sharded workers

    |
    Updated at Jul 06 2026 at the 04th hour

    rzmq 0.5.24: crates.io | Github - Async, Pure Rust ZeroMQ with Leading Performance on Linux

    After some experimenting, I found that io uring was maxed out on a single thread meaning multiple connections could saturate it, so it's been a long time coming, io_uring sharded workers is here! Yes, rzmq's io_uring is no longer one single thread, it's multiple threads now and meets or beats Tokio in throughput.

    I knew there was always more throughput on the table because Tokio could do it and I always had a plan to h...

  • You Can Create Software Quickly Now, But NO GUARANTEES

    |
    Updated at Jul 05 2026 at the 21th hour

    I'm not saying that there were guarantees before, but given the speed of creation of software these days, expect many many many bugs and demotivated maintainers.

    It's never the intention of the human or AI to add bugs to the software, but given the murkiness of how software is created where you go through a first version then iteration after iteration in hopes of bug free USEFUL software, well let's just say most AI-assisted or driven software development won't go through this because it's hard to find users...

  • Same Shit, Different Library: Illusion of Performance

    |
    Updated at Jul 02 2026 at the 13th hour

    Once in a while a new library comes up and talks how they are high performance, but that's only because they are new and naive. Not many people know about it, so they are allowed to skip on ahead with shortcuts. They've benchmarked and benchmarked, then that effort goes to the wayside as soon as reality catches up. Bugs appear left and right and now they have to value correctness over performance. More people see the old benchmarks and are fooled into thinking, yeah must be fast. Eventually, people do their...