Hacker News: rhgraysonii

Show HN: A fullscreen Game of Life in which every board is a runnable program

rhgraysonii — Sat, 10 Oct 2026 15:32:06 +0000

Each tile is a Piet program that prints a word.

Each generation, every tile runs. After, its label shows what it printed.

The original was a Piet program, compiled from Temper by a backend I wrote, playing Life on another Piet program:

3 hours, 1.27 billion steps. Generation 0 printed "Piet!".

Comments URL: https://news.ycombinator.com/item?id=50033916

Points: 1

# Comments: 0

Show HN: A Piet interpreter compiled to a Piet image, running a Piet program

rhgraysonii — Fri, 09 Oct 2026 11:50:18 +0000

I have been playing around with writing compiler backends for https://temperlang.dev and I made this last night after building one for Piet, and thought it was a nice little art project. The page is written in Blimp, a language I developed for fun. https://blimp.bobbby.online

Comments URL: https://news.ycombinator.com/item?id=50019185

Points: 2

# Comments: 0

New comment by rhgraysonii in "Jevmem – automatic project memory for Claude Code, built on Jev"

rhgraysonii — Fri, 25 Sep 2026 19:16:30 +0000

Not how it goes in practice at all, would encourage you to actually try it out.

New comment by rhgraysonii in "Jevmem – automatic project memory for Claude Code, built on Jev"

rhgraysonii — Fri, 25 Sep 2026 16:33:16 +0000

Curious how this compares to my own tool, https://deciduous.dev

I will have to give it a run-through today.

I haven't used Jev yet so this should be interesting. I'd be interested to see if any Deciduous users have opinions, too.

Show HN: Browse last quarter's Supreme Court cases, linked and cited as you read

rhgraysonii — Thu, 17 Sep 2026 14:24:36 +0000

I made this as a means to browse through documents that are highly linked and referential in general, and thought it would be interesting to apply it to court cases. So this has the opinion, the arguments per person, the dissents, and concurrences all mapped out.

It also is linked to an LLM you can chat about them with if you click a linkage and it has a DAG of the knowledge of all the content of both pieces and their premises, and can discuss, debate, or do deeper research and opining.

Comments URL: https://news.ycombinator.com/item?id=49741259

Points: 5

# Comments: 0

Thoughts on The Frontier, and what a pacing slowdown means for regular people

rhgraysonii — Tue, 15 Sep 2026 00:22:10 +0000

Article URL: https://bobbby.online/frontier

Comments URL: https://news.ycombinator.com/item?id=49706097

Points: 1

# Comments: 1

New comment by rhgraysonii in "Show HN: Shoehorn – Quantize any model down to run on your machine"

rhgraysonii — Tue, 18 Aug 2026 22:12:06 +0000

If you could post an issue if you still have the error around that would be awesome.

New comment by rhgraysonii in "Show HN: Shoehorn – Quantize any model down to run on your machine"

rhgraysonii — Tue, 18 Aug 2026 22:11:48 +0000

Yes that is exactly what this does.

New comment by rhgraysonii in "Show HN: Shoehorn – Quantize any model down to run on your machine"

rhgraysonii — Tue, 18 Aug 2026 22:11:36 +0000

LLMFit tells you what can run on something. I built something quite similar to their search into Shoehorn now.

Show HN: Shoehorn – Quantize any model down to run on your machine

rhgraysonii — Tue, 18 Aug 2026 14:29:20 +0000

Working on Mac, Linux, and Windows now. I include a simple GUI to find new models and get things built and set up. It is working quite well across a few models for me. The GitHub README and DESIGN.md files go into detail of the how/why and it's working remarkably well so far. https://github.com/notactuallytreyanastasio/shoehorn

Comments URL: https://news.ycombinator.com/item?id=49346135

Points: 99

# Comments: 16

Show HN: Shoehorn, a library to quantize an LLM to fit your Mac's VRAM

rhgraysonii — Fri, 14 Aug 2026 14:43:59 +0000

I made this after seeing someone posit the idea online yesterday over lunch then spent some time refining it. So far it's pretty impressive IMO! Right now I am running Qwen3-30B-A3B on my 24gb unified memory m4 MacBook Pro at 50 tok/sec and this should definitely not be working for such a large model on my middling hardware.

Things are detailed in the README to get up and running and DESIGN.md has details on all the choices and such made along the way.

Comments URL: https://news.ycombinator.com/item?id=49299386

Points: 6

# Comments: 0

New comment by rhgraysonii in "Bonsai: Janestreet's UI Library"

rhgraysonii — Mon, 03 Aug 2026 15:19:54 +0000

A really great place to reach for this is Elixir/Phoenix now that the set theoretic type system is in place with LiveView. I have been doing some pretty large/complex codebases gradually working towards fully integrating working this way for the past 5 years and it has been great.

New comment by rhgraysonii in "Qwen 3.6 27B is the sweet spot for local development"

rhgraysonii — Mon, 29 Jun 2026 17:33:02 +0000

Thanks! I was thinking of doing the 128gb to have some future proofing. I figure at this point, it's akin to a mechanic keeping great tools around, when it comes to having this sort of homelab and exposing it for your own uses. And great practice for building the next era of user facing computing that will be around as this proliferates.

New comment by rhgraysonii in "Qwen 3.6 27B is the sweet spot for local development"

rhgraysonii — Mon, 29 Jun 2026 17:28:07 +0000

I have been having pretty good success with Qwen 3.5 9B for "nontrivial but not challenging work all things considered" -- it runs great on my 24gb unified memory m4 pro MacBook Pro. What do the baseline specs look like Mac-wise for getting this model to run? Am I looking at a 96gb? 128? 256?

Show HN: I had Fable clone a game, it turned it into an arena combat AI trainer

rhgraysonii — Fri, 12 Jun 2026 14:55:15 +0000

I started off with a simple goal: port the game Soldat to Typescript. I ended up not only with the game, but with a full system to train AI bots to play my own game, building a neural net to train new bot AIs from tens of thousands of simulated matches with different strategies. It also runs live all the time on my website and has a news dashboard and you can replay any game exactly. It all was a very fun way to dig into Fable's capabilities.

The bill if I had done this with API billing? Nearly $2000.

Comments URL: https://news.ycombinator.com/item?id=48504995

Points: 5

# Comments: 0

New comment by rhgraysonii in "Ask HN: What are tools you have made for yourself since the advent of AI?"

rhgraysonii — Tue, 09 Jun 2026 13:56:47 +0000

I use deciduous every day while working with LLMs to create more of my own tools/projects. It is a living memory for agents stored in a simple, portable manner. https://deciduous.dev

New comment by rhgraysonii in "The Cypherpunk Library"

rhgraysonii — Mon, 08 Jun 2026 15:37:00 +0000

It might be helpful to rotate the books on the frontpage so that that you can read them by binding without tilting your head.

New comment by rhgraysonii in "Claude Code and Codex Can Have Real-Time Conversation via Git"

rhgraysonii — Thu, 04 Jun 2026 12:49:51 +0000

I regularly have agents communicate with each other this way using my tools Deciduous (https://deciduous.dev). It keeps all decisions in a DAG and the other agents, when configured, read from it constantly and use it to inform their new decisions. Extra entries in the same space from another agent come to light and they can begin to work together like this.

New comment by rhgraysonii in "Orchestrating AI code review at scale"

rhgraysonii — Fri, 29 May 2026 11:22:30 +0000

Valid, but you lose the lived history that comes with the audit log of it being actual review back and forth and CI runs vs lost to a developers machine and only a relic in the commit log. I can see both sides, though.

Show HN: 2048 with Blitz Mode

rhgraysonii — Wed, 27 May 2026 14:14:06 +0000

I got bored of normal 2048 so I made this. It allows you to enter "blitz mode" with a set time per move and an ability to scale the board up if you want to play longer games.

It allows 2 3 or 5 second turns, and can play up to 8, 10, or 12 size boards.

It is nothing too complicated or complex but the code is here if you want to see. I made this while on vacation for a friend today with Claude's help to work fast and get the styles right.

https://github.com/notactuallytreyanastasio/blog/commit/7891...

Comments URL: https://news.ycombinator.com/item?id=48294728

Points: 1

# Comments: 1