• 0 Posts
  • 21 Comments
Joined 1 year ago
cake
Cake day: June 14th, 2023

help-circle


  • Ah, this got a good writeup by news piece. I first learned about this from Medlife Crisis’s The Epidemic of Fake Disease. Statistics about anything as big as cancer diagnoses are beyond complex, and honestly it would take a gargantuan effort of science communication to get this out to the general public. It’s… sobering to know that mortality is not morbidity and that harsh side effects create the most important optimization problems of patients’ lives. I hope that if (or maybe when) I get confronted with a similar diagnosis, I can face the numbers and the odds with as much of a level head as possible.


  • Yooo, they still got the classic cab at Arcade Odyssey down here near Miami. Some guy walloped my ass real bad at it, so I can imagine you can get really good at it. The singleplayer is hard as balls, and you have to be real tactical. Duel wielding joysticks is surprisingly intuitive, but you really have to mind your startup, execution, and cooldown times for both weapons and your movement, meaning conservative play and taking cover is imperative. I can’t believe I found reference to it on Beehaw.

    Arcades with real arcade games need to make a comeback (damn you D&B for only having DDR). Sadly they’re all Japanese and Korean, so you gotta look for smaller arcades or Round1. Rhythm gamers have lots of machines to play these days, so go hunt some if they’re nearby.



  • There’s two views I see here from a software engineering perspective: multi-targeting devices with different specs can get really hard, and that modern development consumes resources in excess.

    View 1: If you design a device that won’t catch up to modern expectations (limited, shared memory being the factor here), don’t expect to run all of the games. Some (or most) games will demand a certain level of resources. Microsoft either expected their status to swing their will upon the developers or were willing to help but just flopped on predicting what would be needed over the device lifetime. It’s a hard job, balancing developer need and cost. The hardware developers did their best. This comes down to

    View 2: It’s an old coot viewpoint, but goddamn are modern computer programs are bloated pieces of mess. This is NOT an insult to the game developers, but it is to the OS and the engine developers as a whole. The entire programming industry has assumed that bigger more betterer computer always gonna come in a year or so. So now we have gigabytes of unused HQ textures in game downloads for no reason. Windows OS with Chrome takes gigabytes of RAM to display a webpage. We went from ultra strict data streaming to CPU rates for Crash Bandicoot to an NVME SSD shoveling half a terabyte a second when you want it in the Xbox Series X. This has left those who cannot afford strong PCs (note: most of the third world) and now consoles from playing the latest and greatest games. Developers leave them behind by grasping at the end of Moore’s Law. If BattleBit can produce good gameplay with 256 players on a raw potato, AAA game engines should try and appeal to everyone now.




  • My favorite part about the microchip production line is that it all depends on one company (ASML) in the Netherlands and their R&D. They make double digit quantities of EUV machine and that’s it: they dictate the entirety of “easy” technological speed advances in computing.

    And then they ship to a micropseudonation being threatened by the most powerful Eastern country just thousands of kilometers away. That’s where the chips are actually produced.

    And this entire process is predicated on quantum physicists banging together light waves that literally turn chip design into a probabilistically modeled engineering problem.

    What fun!

    Shoutouts to Asianometry for having the best videos on all sorts of the chip design process. He covers a ton of other stuff but his interests just about align with mine so I’m a huge fan.








  • I think we need to focus on the format it is best at: If you can put a block of text in a hypothetical text webpage, it will generate the rest of that webpage. The generation will make logical sense within context, but must be assisted by prompting step by step “memory” or other methods. Accurate output will require more and more overhead tokens and eventually become an actual job managing. Until then, it’s essentially a pseudointernet webpage without google. You need to trust the output with the same wariness of any random website on the internet.


  • You might find a sweet spot with tech mods. When there’s an established hierarchy or crafting automation you plainly must do, it becomes a lot more engaging and you can usually focus on one mod at a time. For example, Mekanism asks you to stack factories upon factories to progress, and AE2 makes that a lot more simpler. I get the crafting and storage mods first before I focus on other mods, personally. That’s what gets my ADHD fix on.



  • I think it’s because they got wise to licensing. If you secure the rights to an official sports league’s image, they will demand that the models are impeccable and the simulated advertisements are grand and that the gameplay is accurate. This means you need to have incredible graphics and at that rate you need to make an overly complex game to justify that massive cost.

    Meanwhile games like Tape to Tape turn 2d hockey into a roguelike. There’s no way you could convince the NHL to put real people into an (expertly stylized) 2.5d aesthetic these days. The only reason a game like Backyard Baseball got it was because it had intense appeal to the kids on PC, before there was NBA 2K9