@Love2Code@acatovicx i think he is talking about the quake .map files. it contains all the csg data in text format, its used to compile the actual bsp map data. that would be trivial to convert to your engine primitives
@Jonathan_Blow@gdechichi its worse than that. they are terms made up by ai as vibe coder requests features, they are so clueless that they think its well established terminology. the feature list tweet itself is probably generated by ai
@OlexGameDev i have a similar interface but after trying to add exclusions and tags to templated interface it was too much for me, for v2 i will go with plain query struct it will get amortized anyways, i wish i realized that sooner 🙃
@TafferKing451 i think consistency matters more than trying simulate everything. for example in gloomwood there are invisible walls that prevent you from climbing to *some* buildings.
@SheriefFYI its been a while since i played it but, i think it had too many survival horror elements that clashed with the sandbox/imsim elements. constant survival pressure really limits how you can spend your time or resources🙃
@SebAaltonen probably can be mitigated with another document to give instructions on polish pass so it does not forget as context grows. also how is the aggressive null check bloat on 5.5?
@lisyarus what would you define as a complex ecs implementation? would deferred archetype changes with command buffers and multithreading be complex enough? not a rhetorical question, because im aware you can definitely make it more complex
@lisyarus you need some kind of codegen to interact with the system so i think its a fundamental part. if you look at the most popular ecs in terms of player count its entt and war thunder's dagor-ecs both of their complexity comes from implting the interface wt even uses python code gen
@CUDAHandbook you've set the tone by calling them microsofties? (maybe you meant microsoftites?) also it's irrelevant that microsoft has talented people, most of them are not given freedom to fix stuff or they've stopped caring long time ago
@SheriefFYI good reminder for me to check if llm models understand this library enough to create a simple example code. example code on the repo was so bad.
@riyan_gendut@sean_gause he had the potential to open up discussing on whatever he was trying to achieve, but vague posting is better for engagement. i love this site