Uriopass

Impressive results in a massive A/B test, replacing hand-tuned ranker famous for how much engineering hours went into it with a "simple" post trained LLM.

I've worked with ranking algorithms at scale (5k reqs/s) and hand tuned, manually crafting features, searching through user logs to see what went wrong. And now we can just pop an LLM on top of that pile of data and it's just better... End of an era, to me.

What do you think those animals eat?

> We couldn't just deploy and pray. 50,000 goroutines don't just disappear.

They do once you restart the server. Unsure what the "phase 3 monitoring" shows where the goroutines go down gradually. If you have new code to deploy you must recompile and therefore restart and those goroutines are gone anyway.

This feels like an AI made-up story but I'm not sure I understand the point of making this up.

However, goroutine leak is interesting! I hope what I learned from this post isn't hallucination. For example, how could the subscriber send messages/heartbeats to the closed goroutine without an error...

In this case any simple unit test would have catched it. Surprised it wasn't mentioned in the post.

This begs for an analysis of the final graph.

Dropping the How is part of Hackernews guidelines

Yes:

https://news.ycombinator.com/newsguidelines.html

I recently implemented a PBR renderer by following the learnopengl PBR tutorial.

It doesn't pass the white furnace test.

That made me realize I didn't understand much of the math I implemented. Random multiplications by pi and 2 as the author stated didn't work out.

xkcd 2^11 https://xkcd.com/2048/

I'm all for AI generated content. I think any automation helps push society forward.

In practice however I didn't see any real good generated 3D asset for now. The problem is probably much harder as there are far more 2D assets than 3D, especially with PBR support.

I went with a utility approach as I felt it would incur nice emergent behaviour.

Every so often, citizens and companies (I call the abstraction a "Soul") look through everything they can do and pick the on with the highest utility score.

If they're hungry, they'll get food. If it's the time of day where they should work, they'll go to work, except if they're too hungry!

I think Rust for games has come really far. I will cite https://arewegameyet.rs/ "Almost. We have the blocks, bring your own glue.".

All the blocks are there and the language is really well suited to games.

On top of my head:

The pros:

- The crate ecosystem and the package manager makes it really easy to integrate any useful component such as pathfinding, spatial partitioning, graphics backend, audio system.. Most crates take a lot of effort to be cross-platform so I can develop on linux and not spend too much time debugging windows releases.

- The strong typing and algebraic data types makes expressing the game state very pleasant. I also found I was able to develop a very big game without too many bugs even though I don't write many tests.

- Ahead of time compilation + LLVM guarantees you won't have to optimise for weird things around a virtual machine. Rust gives you more control to optimise hot loops as you can go low-level.

- I find wgpu to be the perfect balance between ergonomics and power compared to Vulkan. OpenGL support through wgpu is also a nice addition for lower end devices.

- The Rust community is very helpful, you can often talk directly to crate maintainers

The cons:

- Compilation times, when compared to JITed languages such as C# can be very painful. It can be alleviated by buying a 3950X but I still often get 10-30s iteration times.

- The static nature of Rust means you often need a dynamism layer above to tweak stuff that can be awkward to manage. I made inline_tweak for this purpose but it's really far from how easy Unity makes it. https://github.com/Uriopass/inline_tweak

- Since Rust feels very ergonomic, you are tempted to write almost all game logic within it, so mod support feels very backwards to implement as you cannot really tweak "everything" like in Unity games. Thankfully "Systems" game like Factorio or Egregoria can be theoretically split into the "simulation" and the "entities" so mod can still have a great impact. Factorio is built in C++ so has the same problematic. Their Lua API surface is quite insane to be able to hook into everything. https://lua-api.factorio.com/latest/

Now, I have to talk about Bevy: https://bevyengine.org/. It did not exist when I started but it is a revolution in the Rust gamedev space. It is a very powerful 100% Rust game engine that makes you write game code in Rust too. It has incredible energy behind it and I feel like if I'd used Bevy from the start I wouldn't have had to develop many core engine systems. Its modular design is also incredibly pleasant as you can just replace any part you don't like with your own.

Thank you! Could you elaborate on what things you found best? As to me, almost none of the project is ready for HN yet :D

I have to say I'm confused by this. After looking through their website I do not really understand what Spectra City is about or how I can get involved.

TIL Factorio was funded on Indiegogo!

But yes, I'm really not ready for that kind of pressure and marketing efforts. I feel - and even more so since Cities:Skylines 2 announcement - that there isn't enough I can do alone to turn this into a full-time job.

I've always thought about plugging LLMs into the simulation, even pre-generated naratives could be interesting. But I'd have to give it more thought as it seems pretty hard for such big scale interactions (50k citizens?)

>I like to think too that with sufficient fidelity a simulation of this sort may be useful in real world city planning/economic modeling/geopolitics.

This was one of the early vision I had for this project: Simulating different economies to see which ones fared better. However I have kind of abandoned this as the realy world is just far too complex! Making a "normal" but different city builder seems a bit more achievable.

All aspects of the game need more work, I think anyone that ever worked near the video-game space could help.

- 3D assets

- Audio assets

- UI/UX design

- Low-level graphics shenanigans (platform support, debugging tools, dos-and-donts, optimizations)

- Simulation depth (logistic chain, economics, city services..)

- Mod support

Any help with these would be greatly appreciated

Public transportation is definitely planned, although it's a lot of work. Cars + trucks + trains are the minimal set of transport methods to allow anything to go anywhere.

I love tooltips! Thank you for the suggestion, I have opened an issue.

>I feel like there's almost enough here that you could do steam/itchio.

I think it's way too early for this. As people say the last 20% takes 80% of the time. There's a lot of systems I've neglected for an actual release. It probably needs a few years more work before any kind of "real" release for players instead of developers.

Hello, all contributions are welcome!

The discord badge in the readme is not very big, here's the discord link:

https://discord.gg/CAaZhUJ

There's not much at the moment really, the idea is that all citizens have their own state and decide what they want to do themselves.

For example, all of them have hunger state, have jobs assigned to them, a house etc.

So a single citizen will always go back where it lives everyday instead of a random house, it will try to go near its home to buy groceries. I've decided that they don't have money as I found it too hard to calibrate, but there is still "real" items like bread being exchanged.

Thank you for the kind words, very much appreciated.

I spend a lot of time in public transportation looking out the window, thinking about the mechanics of society. I have always liked simulations so I thought: Why not try to simulate what I see? Watching the simulation from above of people going to supermarkets, going to work, trains delivering goods and such. Creating your own city and watching patterns emerge would certainly feels great.

A lot like watching little ants going around doing their business. This idea is also explored in one of the books of the Ants trilogy from Bernard Werber (I think?) where a society simulation is used to try out advertising campaigns.

Egregoria used to be 2D but it felt a bit restrictive. A lot of the underlying simulation is also 2D as I didn't port everything when doing the migration.

From the player perspective, 3D is the norm these days. It looks better, it's usually more intuitive and 2D is usually a bunch of tricks to fake 3D anyway.

From a developer perspective, 3D fits my brain better as I can translate real concepts better. To me, it feels easier modelling in blender and apply physically based rendering rather than learning lots of 2D techniques to have a consistent good looking art style.

I am unsure if I understand well, so just to clarify: - Egregoria is written in Rust directly using wgpu as a graphics backend and a custom "engine" which is not really an engine as it is built with only one game in mind - Cities:Skylines is built in C# using Unity as the game engine

> This looks really good, congrats.

Thanks :)

It is definitely my ambition to implement public transportation but it is much harder to implement than cars. I live in Paris so I know what a non-car-centric city looks like and all the complexities around it (not to say I don't like it!).

With cars, anyone can go anywhere without thinking much, although it is a bit of an overstatement considering the time I have spent fighting gridlocks.

I wish I could work on it full-time. However, after reading a lot about indie game developers, and judging from my situation, it seems like keeping it as a side project is a better decision for me from a financial and mental health standpoint.

Keep up the good work on Metropolis 1998! I have followed your work and the dev community you built around it and it is very inspiring.

Hi, I'm the main developer of Egregoria.

I did not expect to end up on hackernews that early. While I've been working on this project on and off for 3 years it is still in its infancy. I have so many things I want to rewrite, so many features missing even for the most basic game loop. So many 3D models I have to make.

And so little time! Working on this as a side project with a full time job has been hard at times. I do not recommend making city builders :D

I wish I anticipated the attention so I would have given the README bit more love.

Anyway, I will try to answer any questions in the comments.