Відео

Dark souls having "the simplest UI so you can't really get confused" is an absolutely wild statement. I remember when DS1 first came out and people needed video walkthroughs for how to equip a shield and for what the levelable stats meant.

HP ML350p with 256GB, 1200W dual power supplies, 2 dedicated 16 lane, one 16-8 lane for the NVLinked.. and the video cards are up to you as to speed and power consumption,, But the A5000s should do for the linked or A6000(s) in the 16 dedicated, lane 32 bus.. three vid cards will leave you absent for PCIe accelerator cards and high bandwidth connections.. like to your mass storage.. Oh, expect your internet provider to complain if your LLM has access to the Web... you may have to upgrade to a business plan. These programs are hard on SSDs, choose accordingly or you consumer NVME will be written to death in months, so have a big backup platter (HD) as least and backup often. Petabyte write rated drives.

very mature takes! You are different on the hate bandwagon online now

😊!

Old PC with 128-512gb ddr3 RAM + a lot pci-e slots + some cheap Nvidia P can lift heavy workload without issue. Also in a tight 1000€ budget.

Raylib will probably surpass pygame in popularity.

I loved it until I wanted to attempt to develop a basic 3d game

raylib is simple, fast and very easy to implement, litterally just one of the best graphic libraries, also works wonderfully in C so yea

Raylib is nice but for more serious projects I either use OpenGL directly or rarely SFML

raylib is great introduction if you want to get into openGL in my opinion it lets you more focus on the game itself without drowning you in lower level stuff so that once you get to opengl you can better focus on the inner details and behind the scenes work

Currently learning networking programming using raylib, and Rust™

yeah

I've been messing with raylib for the past week and was excited when I realized how easy it was to load and render a 3d model. But as soon as I tried to add lighting I hit a brick wall and couldn't do it. The basic lighting example program works but when I add my own model it won't light it. So I dug into the code and it appears no less complicated than doing it straight in opengl. Perhaps even moreso because raylib is automatically assigning uniforms and it's a huge pain in the ass to figure out what the hell is going on or what is going wrong.

Cos tsoading using it lol?

So its like a Util? well at that size its a framework i guess...

What I'd add to the library vs GUI engine is it really matters if you are working on the game solo, or if you need to collaborate with non-programmer artists. If you are collaborating with artists and level designers, the GUI is more for them than for you. Level designers need a "Halo Forge Mode," if you will.

Raylib is also good for statistical and simulation visualization when there is a lot of particles and nodes

I use Bevy and want to transition to just wgpu, ECS and my utils

I’m using Raylib-go and i love it. This framework gives you just the right amount of abstraction, so you have all of fundamental blocks of functionality for all basic aspects of the engine (that’s made very easy), yet you are not bound to the engine, you are free and productive

that comparison with game engines isn't very apt. yes obviously something like raylib is awesome for getting into game engine programming without building absolutely everything from scratch, but if you set out to make an actual *game*, you're going to end up reimplementing the same patterns and abstractions present in pretty much every game engine ever in some shape or form. it doesn't really mean anything that it's "more familiar" to a non-videogame programmer, because making any game that holds up to any sort of modern standard (not talking about AAA, this applies to anything that sells as a game on steam for at least a dollar, or something you see made for a 48 hour game jam) will require you to learn how to program a *game*, which is something very complex and deep, which programmers with experience in other fields might see as very alien. so in any case, what you'll be doing will be unfamiliar. a proper game engine (i.e. Godot) allows you to only need to get familiarized with some basic abstractions like scenes and nodes (which typically exist in other engines with names such as rooms and entities), but not have to deal with any of the heavy lifting that requires months if not years of study like collision, lighting, shading, rendering, optimization, and so on. the programming itself is exactly the same, thus the barrier of entry ends up being far lower than if you had to code your own game from scratch.

i have plenty of old servers..but the requirements have other hidden dependencies... such as avx2 or better. While avx2 hit intels cpus in 2012 with the haswell tech... the cpu in your server could predate avx2.

I think start citizen uses lumber yard

I just started learning go and I plan to use raylib for gui elements of programs and experiments more than I do games.

So yeah, but hella slowly

i wish the C bindings didn't clash with Windows API, preferably had some sort of optional rl_ prefix

I'm currently learning my way around both Crystal and Raylib by just coding a bit of the basics every day l- ike moving a 3 frame sprite on the x axis - and then gradually adding more features, or rewriting things I notice could be implemented in a better way, etc.. and it's genuinely been such a joy to work with. Plus the community not being centered around one single language means you can get more broad advice, get so much inspiration, and easily find something you like and translating it to whichever Raylib bindings you're working with as once you get used to it it's fairly easy to translate from one language to your preferred one. The examples in the github repo are phenomenal and easy to read, plus the cheat sheet and live demos via their website are a huge help. Plus the creator is equally, if not more-so, passionate about everyone's projects and their genuine gratitude and excitement for every new project using Raylib is infectious! I would not hesitate in recommending this amazing assest(s) to everyone and anyone!

I've had some limited exposure to SDL couple years back and it felt ancient. No wonder there is such demand for a simple unifying API with bindings for a lot of languages that's also not older than you. Where else are these people supposed to go?

I did not hear anyone talking that affinity is sandbagging. Autotrace code is available for ages, but not finalised, since it is not perfect enough. Now it would be nice if also AI is added to to autotracing. The affinity team is already 90 people for ages. Income is large enough for a bigger team. And if needed always an update like 3.0 can be sold. Now for good updates, for sure a subscription model is coming. It is no longer in the hands of affinity, too bad. The promises are empty words.

It's simple, easy to embed using C FFI. Handles most of what is needed. And also because of tsoding.

This video is longer than needed, and does not answer main question (but honestly skipped part of it) - why game that was small, now it's so bit without big improvment in graphics.

Build dependencies. You need to do X in your application. Play a sound, for example. So you look for a library, already written, that does that for you. You include that into your product. That library depends on five other libraries. Each of those depends on two or three more. And when you build your project, you're not excluding all the routines you're not using; you're just leaving them all in there. See it all that time, everywhere.

Because they can. Everybody has huge ass SSDs nowadays and compressing/optimiziing your product takes time, time that you have to pay people for, which is bad for your investor´s RoI. I worked with AI upscaling myself, it´s not like you just hit a button and it´s done. It´s faster yes, but not as fast as people think. Not yet anyway.

Exactly the question on my mind, though I expected all these answers. Not much we can say without seeing the process that went into this, other than speculate about the usual suspects of greedy publishers/developers, though given the relevant history of the Devs, very likely a mix of their inexperience and publisher greed.

The best approach to this would be to offer 4k textures as an option. Gaming is at 1440p now. No need for those massive textures. Having a massive asset library is cool, but I'm sure it's a massive headache from a development standard. Also a better optimized compression methodology perhaps combined with AI could even further reduce file size.

I bet its the in game shop

I don't think I would mind so much if I could actually make use of the standard drive for my game storage. but on xbox only the proprietary SSD can store and run current gen games. So that $20 external hard drive is now useless because I cannot actually put games on it. because it is "too slow" And to add that we are now really seeing the consequences of not optimizing games (or game engines.) you end up with most of the current generation of frustration in a gameplay experience. Destiny and Destiny 2 are a couple of the worst optimized games and even the developers have voiced this frustration. taking hours to load map/level editors and that un-optimization leading to the "sunsetting" of content that was paid for. just because the game file is too large.

The short answer: "Because they can". Understand that "optimization" takes time. Hours of dev time, it costs money. Tuning texture resolution/compression, vector count, arcane tampering with render settings. Sometimes you need to invent brand new tricks to deal with novel bottlenecks. And the goal isn't "make as optimal as possible", it's "make it as accessible as profitable". 99% of people can afford 70GB of storage. That remaining 1% they're going to miss in sales, isn't worth the cost of optimizing for them. Sure it's shit form. But it's business.

First off, nice vid- Ill throw ya a sub/like/bell. Gratz on getting picked up by the algo, you were in my home page recommended. 🤙 Because 99.999% of dev studios are factories, not stuidos. They don't care about the finished product. MONEY MONEY MONEY. Also, if the majority of the game installs through a launcher- a large portion of users outside of the US will run over their 2 hour playtime return period by simply installing the game with slow-ish internet, ez money. Mainstream games and steam releases are on the same level as google play store trash these days. These fuckers are lucky im somewhat mentally stable, this is unibomber-deserving levels of ineptitude. You did make a few valid points, though. I remember a few PS1 games that were about 100mb themselves, with 400mb of various language data and voice dubs.

While you do make some good points about how games were always "relatively" big compared to your average storage space at the time, it might also be important to remember that back in the day, you didn't necessarily even write all that data to your hard drive, often a lot of the data was just read off the CD as your were playing. These days however, physical media games don't even have anything on the discs anymore if you get a disc at all.

The thinking is probably along the lines of Wyatt Cheng goof 2018 Blizzcon. Here's how I reinterpret it to this continual install size inflation "The current plan is to keep inflating Install sizes on all OS platforms, we do not have any plans at the moment, to compress our game files to conform your 2004 tier install size." *audience boos* "WhAT Don't You guys NoT HaVe 32TB drives?! their dirt cheap too."

You didn't need to make this video. This is massive Randall from Recess energy.

So I think you're a bit confused on this matter. Look into the case "Sony VS Bleem". Bleem was an emulator for PS1 games that ran on the Sega Dreamcast. It was sold as a disk for money. They won the lawsuit against Sony so it is actually settled law that for-profit third party emulators are legal. The more likely reason was that they announced support for TokT BEFORE it was released, meaning it would have had to have been a pirated version being used for testing and playing.

Why have they not gone after Ryujinx then? It is also a Switch emulator that works pretty much the same as Yuzu and they make money from Patreon as well and have been around since the same time as Yuzu (1 month later, but I think it started development before). Some consider Ryujinx better than Yuzu in the accuracy department when it comes to emulation, at the cost of some performance, whereas Yuzu had some inaccuracies but was more performant.

What would you recommend to run a chatbot trained on website data locallY? Thanks for the video.

the only advantage I see for going through all this to run locally is if you are running an uncensored model like mistral dolphin , which could be useful. but if not , why not just use gemini online or claude or chatgpt , they are free and have nice UI

I was just working on something like this. I didn't even think about using a depth buffer... instead I used an approach that will only work for tile based games. It's not very flexible, but very fast for the CPU. I think stardew valley does something similar

Meh, depth buffer is not necessarily OpenGL/Direct3D/Vulkan/Metal/WebGPU exclusive thing. It's not even that CPU-hard work either. All you need is additional buffer for depth-check and your own sprite_copy function, that would check depth buffer and write to it if pixel is going to be written

I have my eye on a Proliant Gen 8 server with 393 gigs of ram (dual socket / 12 threads each). I know more ram would handle more parameters, but would more ram speed up simpler models?

typescript: 21.50 javascript: 4.45 bruh?