It's hard to know what he exactly means by this. It sounds pretty bad but this guy does seem to attract and develop a fair bit of drama and this could be a bit exaggerated from the truth.
It very likely is. By "abusers" he likely means "people that are rude to him online" and by "stalkers" the same. Nobody is following him around IRL. He probably is referring mainly to the Kiwifarms people that get a kick out of documenting the downwards spiral of someone like Hector.
For clarity, it isn't that. He's referring to a single specific person who engaged in stalking and personally harassing behavior unrelated to the projects. The receipts are all over, you can find them if you insist.
Right exactly. They're not tactful, but they also weren't in bad faith. Marcan should have taken 5 minutes to realize that he's the boss, he's the one doing the work, nobody is entitled to anything in free software and that if people want a feature sooner they can either fund the project or kick rocks. Anyone who's been in open source for over 2, definitely 3 understands this.
> I miss having free time where I can relax and not worry about the features we haven’t shipped yet. I miss making music. I miss attending jam sessions. I miss going out for dinner with my friends and family and not having to worry about how much we haven’t upstreamed. I miss being able to sit down and play a game or watch a movie without feeling guilty.
This is the big problem really. He should have just turned down his work hours to a regular 40 a week, asked for more donations to pay more people and asked for more volunteer help. And honestly, probably therapy.
Thank you or your support! I am disappointed with the result, but, to be honest, no hard feelings. Wish they let me do the take-home assignment, but it is what it is. Rolled the dice. It was not my day. Will roll it again with as many companies as it takes to get a good gig.
> Meanwhile deepseek is hiring college grads and the like, for way better tech.
If only I knew Chinese...
> I highly doubt that they've got a better ai but are too scared to show it off. Is there any independent verification of that?
Probably wouldn't be too much different from modern gpus. The main difference would probably just be that it'd have a hardware method for doing ray-Signed Voxel Octree intersections.
>The issue with voxels is sparseness - to have any hope of storing a world in the ram we have available it pretty much needs to have some variable length encoding, which makes it extremely difficult to use in a parallel manner as required by hardware.
>It's a subset of Vulkan - one command queue, and no bindless mode. Google proposes to fix that around December 2026.
Yeah, WebGPU would be a good, modern API... If it was 2015 and not 2025. But now the differences between it and the desktop are widening and at a certain point, people will just break compatibility for the sake of convenience. I mean, theres not even an equivalent of vkCmdFillBuffer ffs! If you want to reset e.g. a indirect dispatch buffer in a command list, tour best bet is to have a buffer that contains uint32_t 1s and do copies.
Yeah, your best bet is targeting both desktop and web with something like the rust wgpu library and using renderdoc for debugging/profiling. Which works well until the code stops working in the browser for whatever reason.
Is there more to getting 'good' at them then just copying error messages back in? Like, how do I get them to reason about e.g. whether a data structure compression method makes sense?
I've been writing glsl recently, and I think it's a good example of a DSL that _doesn't__ work. It should have just been C without the standard library, or ideally C++. Not being able to use pointers is especially painful. Hlsl is better and has some nice quality of life features. Rust-gpu is the best approach but basically nobody uses it
