Realme 8

[...] and a good treatise on the various techniques: I think there is a lot of value in learning BASIC using these kinds of techniques in this day and age. [...]

[...] Can't we discuss technologies for their face value anymore. [...] However everyone is busy building things secretly that they don't want anyone to know about with AI/ML (except the operators of course, duh, lol, kthxbai...) because, after all, then anyone else can do it .. [...] I very much concur with you on the desire to continue being able to discuss technologies at face value, personally. [...]

[...] It's a terrible performance example since the work per fork is basically zero :) I included it as an 8-line API showcase, and because it's almost pure overhead it doubles as a brutal stress test for the runtime. The nqueens, matmul, and quicksort benchmarks in the repo are the real performance indicators, imo. [...] You can test scaling yourself: CACTUSNPROCS=1 build/cc/nqueens 14 vs CACTUSNPROCS=N build/cc/nqueens 14. [...]

This is just a link to a Github repo with only 2 stars, a total of 7 commits and no activity in the past 8 years, with not even a README describing what it is or why it would be interesting to anyone. Why is this being posted? Who is upvoting this? It seems super fishy. I'm suspecting the repo actually contains malware.

TL;DR: nothing to see here, HN title is extremely misleading rendering DirectX 8 → DXVK → Vulkan → MoltenVK → Metal. Ok, this is a bit weird. Why create such a jenga tower of indirections instead of directly letting the LLM port the rendering code from D3D8 to Metal either through a D3D8-to-Metal shim or even better by creating a new Metal-based render layer beneath the higher level game specific rendering code? I would expect that the LLM can 'see through' all those redundant shims and collapse them into the equivalent Metal code. Also the readme says 'no emulation', but then goes on to describe the rendering layer as emulating D3D8 on top of Vulkan on top of Metal ;) (also why are both Meson and CMake required) PS: after reading more, it can be explained by what this project does (and it's not what the HN title says): it's not a port from the original C&C Generals code release to macOS/iOS, instead all the heavy lifting (of making the original code base portable, and providing example ports to Linux and macOS) was all already done by an existing project ( all that this project did was slightly enhance the existing macOS version so that it also runs on iOS. Suddenly this is much less impressive and could probably have been done just as well with older and simpler models (or tbh, a few hours or at most a few evenings of manual coding - macOS and iOS code is nearly identical for this type of application).

[...] For your high tech luxury you would need some sort of system to build and maintain the high tech luxury, a system without those humans that you intend to eliminate (even if you plan to just let them wither away)... something to build your megayachts and prepare meals and harvest the truffles and raise the sturgeons. [...]

Yes I think it's similar in scope to Raylib. I'm learning this as I go, but here's a page with many demos compiled to Wasm. In the README they have example usage ranging from 8-bit emulators, games, command-line tools, integration with Dear ImGui and other libraries, bindings to languages like Zig, Rust, Odin, Nim. WebAssembly is a 'first-class citizen', one important motivation for the Sokol headers is to provide a collection of cross-platform APIs with a minimal footprint on the web platform while still being useful.

As a counter to that - I've tried various flavors/quants/full weights and Qwen 3.8 27B has been entirely useless at anything non-trivial. Sure - it can do some boilerplate work (though, even armed with a well written spec and working within a very well known framework it went off the rails and did things in a way that were... um... questionable at best) but I don't see it as anything more than a personal assistant style model. Zero chance I'd "work" with it, I spent days trying to get it to do something for me that was usable that I didn't have to have reviewed and refined by a frontier level model or myself. Couldn't do it. The idea that qwen 3.8 27b is anywhere near Opus 4.8 is laughable. Pure benchmaxxing. DS4 Flash 0731, on the other hand, wildly opposite experience. Would recommend. GLM 5.2 - even quanted down to a hybrid 4/3 bit setup is amazing for everything but the hardest/most complex stuff in the same projects/realm.

[...] From here: << Just clarified this in the README: Input value ranges are as follows: - Station name: non null UTF-8 string of min length 1 character and max length 100 characters - Temperature value: non null double between -99.9 (inclusive) and 99.9 (inclusive), always with one fractional digit If we assume worst case entries: -99.9;<100 char station name That is 106 chars plus newline (1 or 2 chars). [...]

[...] There is a smooth gradient in the percentage of AGI that "the 2nd most expensive silver plan in your state" can cost. [...]

Here is a description of the iteration loop. [0] I'm working on another draft that will be much more polished and have better explanations of the iteration loop. There is no proof, just a self-congratulatory word salad with dubious authenticity. I worked 8 days straight on that and have been working non-stop on the second draft that is much cleaner and safer. I'm a human being. Please don't be mean. If humanity does come to end, it won't be because of AI, it will be because we can't stop being assholes to each other. [0]

[...] You could count the number of anime available in the west and worth watching 25 years ago on one hand, maybe two. [...] 8 - Marriage Toxin - Steel Ball Run - The Summer Hikaru Died - Akane-Banashi - Dorohedoro edit: formatting. [...]