kookamamie

Perhaps. Then again, advocating things like Singleton as anything beyond a gloriefied global variable is pretty high on my BS list.

An example: https://martinfowler.com/bliki/StaticSubstitution.html

Let's just say I think a lot of damage was caused by their OOP evangelism back in the day.

Funny, I dismiss the opinion based on the author in question.

Yes, though C has restrict in the language now, but C++ does not.

Yes, my thinking here isn't about the semantics of code, whether it is written in dense or terse form, but about the more meta level - code itself is something we write to achieve some goal, it shouldn't be the goal by itself. As an example, as a company is audited for their valuation, it is not the code amount itself that is valuable. In many ways it can be seen as a burden, as you'll need to maintain code over time.

There have been cases like that, e.g., RDR2, but I think it mostly comes down to implementation quality.

Maybe if we added ALWAYS BE RIGHT NEVER BE WRONG it will work this time?

Lines of code is a byproduct, not the goal.

Code is an artifact, undesired debris.

The fewer lines, the better.

Yes, that's the essential reason. Even though that barrier has been lifted with Vulkan matching DX12 in many ways, the accumulated mass moves slowly.

> __restrict__ Pointers

Ahh, my favorite nitpick from C++ not having sane default aliasing rules spills to the CUDA-land.

There are issues in Windows going beyond just added bloat. I don't think its kernel can compete or keep up with Linux's, as an example.

This is the right question. The article reads like answering a XY-problem, where instead focusing on the actual issue, the author triples down on polishing a turd.

It is even funnier when you consider that the entire algorithm is deterministic, assuming a fixed seed for a PRNG.

Here's the algorithm described without the quantum nonsense:

1. Analyze Rules: Extract valid patterns (modules) and their compatibility rules (adjacency constraints) from input or define them.

2. Initialize Grid: Create an output grid where each cell initially contains all possible modules (maximum uncertainty).

3. Choose and Assign: Select the cell with the fewest valid modules remaining. Randomly assign one compatible module to it.

4. Propagate Constraints: Update neighboring cells by removing modules incompatible with the newly assigned one. If a cell loses all options, a contradiction occurs.

5. Handle Contradiction: If a contradiction arises, either backtrack to a previous choice or restart the process.

6. Repeat: Continue from step 3 until all cells are assigned a module or an unresolvable contradiction occurs.

> Each tile has a superposition of possible states

This is like saying an uninitialized integer has a superposition of all possible values. I find it a very convoluted way of saying "each tile has a set of possible next states" - dragging quantum terms to this is just confusing, in my opinion.

> Wave Function Collapse

I've always found the name pretty misleading and grandiose, relative to what the algorithm actually does.

…or, you could have a product that does not hinge on some micro-optimization of a website layout.

If Siso is a drop-in alternative to Ninja, are there builds of Siso usable for building Ninja-projects, in general?

An awesome contribution nevertheless!

It says it's "next gen" - how is that? Besides that, I think the UX could use some improvements - the landing page captions look glitched and the UI iconography looks "murky", for the lack of better word. Nice effort, though.

Nice. Except that Mode X was not usually called 13h. The former would be 320x240 and the latter 320x200. Technically X would be a variant of the 13h, though.

They are adamant in their conviction to oust the impostor, it appears.

This is not a good idea in practice. Why not bundle the metadata as JSON or Protobuf via an aux file?

Yes, they introduced the concept in their paper. See this article for an in-depth guide to the topic: https://ciechanow.ski/alpha-compositing/

You absolutely want premult alpha when dealing with multiple transparent layers in graphics.

> Treat prompts as version-controlled assets

This only works if the model and its context are immutable. None of us really control the models we use, so I'd be sceptical about reproducing the artifacts later.

Very sorry about that :D

Gaffer has summarized most of this quite well years ago. E.g. https://gafferongames.com/post/fix_your_timestep/

Assuming the timestep refers to the parts of code dealing with game logic updates, such as physics or motion, variable-sized steps tend to cause nasty issues. Collision detection is an example, where variability is undesired.

> For many games variable timestep is a good choice.

It really isn't.