Tone

I have finished reviewing/getting impressions of every single game out of the 945 games submitted to #vibejam. If you submitted a game and haven't received any feedback or tags from me, take a look at the link below. Impressions tend to be brief. To be frank, some of these games did not deserve even the brief time I put into them. But SOME of these games are genuinely stunners. Some of these games have enormous depth, gorgeous visuals, addicting loops, and genuine commercial potential. I'll be posting some of my thoughts, opinions, and wrapup ideas as the official judging of the vibejam continues. As a reminder, I am not an official vibejam judge. I just wanted to look at all of the competition and see what the playing field looks like.

@AzFlin RIP CSS Grid. I never understood you. Flexbox already did it all.

Since the invention of LLMs, the entirety of human front end knowledge has been deleted. We can't. We don't know how.

By all measures, I think most any developer can do an outstanding job with $100 Codex subscription and $60 Cursor subscription. Using headless `agent -p` delegation to Composer 2.5 coupled with GPT-5.5 High offers massive output.

@brandonjcarl Incredible. I've never even thought to combine models in different ecosystems using their respective tooling as a script. Haven't touched cursor at all since I started working with codex. Really intelligent composition of tools.

@lamboc0rp @dangreenheck Take the leap for a month. It’s a relatively cheap experiment to run. Speed isn’t the only variable and the sensation that you could have done it faster isn’t always accurate. Plus you figure out to what degree you want to delegate and to what degree you remain in control

@dangreenheck @InOneProjects everyone knows the shortcomings. answering confidently, you discover the answer is conceptually flawed, dialogue-ing to resolve, by now context retained is imperfect and a new flaw slips into the solution etc., sometimes faster if I just solved the problem myself

One thing I've noticed with a lot of these Three.js Fable posts is many are incredibly unoptimized. For example, someone made a water/sky shader that looked impressive, but was running at 20FPS. And this is before adding in all the additional scene geometry and behind-the-scenes logic. My experience is performance and optimization tend to be the most difficult parts of a project. It requires performing research, handholding the models, trying out different techniques, and finding that perfect balance between performance and fidelity. I prompted that bazooka RPG in about 2 hours, but I spent 4 hours with Fable trying to find the optimal dithering strategy for my cloud shader—it didn't have a clue worked or what didn't and didn't feel any different than working with Sonnet. Another issue is Fable is clearly incapable of producing anything remotely close to a production-ready asset (which makes sense because it isn't designed to do that). Until we start seeing rapid-prototyping workflows for generating optimized, production-ready assets (maybe these exist already?), I don't see these demos becoming anything more than fun little show pieces. TL;DR: the hard problems remain hard.

@ApexSeeker_ Discipline your diet. Discipline your sleep. Discipline your gym habit. Get a hair cut. Write in your journal daily.

🇯🇵 I am Japanese. I have a question for people all over the world. What should you do when you're going through a tough time? Not by pushing yourself too hard. Not by giving up either. First of all, I think it's about surviving. What do you think?

Ok smart use of grok.

@alightinastorm @grok Best grok usage I’ve ever seen

@grok analyze this acc, last 100 posts, more positive or negative vibes?

@did0f I hear from non Americans that were loud and rude 😂

@InOneProjects I don’t have an answer To e. But your question is legit

It’s funny when you’re extra late for next plane and people decide to stand on both left and right of the stairs. Whyyyy

@hunvreus brilliant! 😂

For those wondering what the fuck this “agent loop” thing everyone is talking about is... Here’s the exact loop I used to get into YC and raise $30M from Khosla:

@LudensLudonauta Would you consider it wasteful to post progress and dev logs? Other wisdom I’ve seen around is building an audience by building somewhat publicly is a better route to sales and also early feedback

The game, Mind Over Magnet, is undoubtedly a good game. He promoted it over a whole year before its release.

When I point out that creating content yourself as your main marketing strategy is...pointless people come with knives and forks and torches against me. Everything but data, and I'm always looking at data to support my claims. These are the stats from the first game of Game Maker's Toolkit, the current most popular game development/game design channel on YouTube.

The correct mindset 🫡

@Lizzergas I think this is cool as hell. Could you make it better? Sure. But I’ve never heard of hyperframe and this is a great “hey look at this tool” concept. Not everything has to be a work of art

Learned about a new tool: Hyperframe Wanted to test it on Clankerfront. I used Fable 5 and gave no context, just said make a 20 sec video. Result is meh Didn't know we can make AI slop "devlogs". Not a fan EDIT: I won't make AI slop vids, its a test! #gamedev #indiedev

@ALASKANmagpie Oh wow, those pieces are looking genuinely good! Do you have it building the assembly rig, or are you saying it actually generated all the pieces being placed together?

I can't get enough of this. 10^58 possible ship combinations. OR more. Fable 5 is insane. This just one of dozens of features: YOUR waypoint awaits: waypoint.shinythings.space

@kunchenguid Do you believe there is a mission behind which it would make sense to make the decisions Anthropic has re: fable?

let's talk about different kinds of companies anthropic deliberately going out of their way to cripple Fable 5 on AI research capability, for the sole purpose of suppressing competition, shows that: - they are not a mission driven company. they are driven by self interest - their goal is not to make the world a better place. they are literally doing everything they can to stop other people from making the world a better place better than they do - their company mission page anthropic.com/company is mostly talking about the things "they" would like to do, not what kind of a future they would like to see "us" live in. openai has a better one openai.com/about/ "Our mission is to ensure that artificial general intelligence benefits all of humanity" - that's more about us than about themselves for comparison, Tesla had a mission statement that is "to accelerate the world's transition to sustainable energy" - they then opened up all their EV patents because that's what would clearly help push that mission forward - they actively harmed their own business by doing that, but now we have so many EVs to choose from, many of which are better or more affordable than Tesla having a true mission changes everything a company does. i hope every company finds a good one

@fightlifemobile If I’m extracting - why? To what end? Extraction loop is enter, loot, escape alive with consequences to dying. But you need a reason to get out alive - something to build toward

@InOneProjects Working on the combat now. Trying to figure out if I still want to go the extraction route. Maybe having planetary bodies following a hell divers type format?

Morphers! Now including a new ship transformation.

You're right, I haven't worked at a game studio. I've worked in software - you know, the thing you're commenting on? Go ahead and explain how you write code that is both maintainable and for a specific junior (LOL) without just having coding standards or design patterns - you know, the software things everybody in the industry uses to solve the problem of "I don't know who will maintain this after me, but if it's me I hope I don't hate me."

@InOneProjects then you've not understood it and you've definitely never worked in a studio of any size because if you did it would make perfect sense to you

because it doesn't have "zero impact" when you're building apps, industrial systems, backends, that sort of thing, you know what you're building. somewhere there exists a document that, if the links are followed, details every single thing that needs to happen games don't. you discover them as you go in a drill down because the nebulous goal of "fun" is not something built to spec that means pivoting, careful management of what gets committed to, what changes, how it changes, etc. how you build something is 100x more important because of your confidence level of whether it's an experiment, whether it will change, whether it's a product necessity whether the feature will get dropped, HOW it will need to change and how quickly understanding what the designers are likely to do affects the way in which you code so you can tweak, adjust, pivot, as well as knowing which specific juniors may have to maintain and what and how you need to commit to readability for "if this breaks at 3am" developing in non-games has a sense finality to code that games will never have and that's okay, but it does mean that what makes 'good' code is very different. if i know my designers and i know the kind of shit they come up with i know precisely where to put levers, how to encapsulate the code, preempt what can be asked for and how they'll ask for it and be able to ship their changes in minutes rather than having to rebuild anything this is vital for achieving for performance and organizational development velocity at the same time

@chhopsky That makes your post even less coherent.

@InOneProjects you're talking about over the lifespan of the product and i'm talking about before anyone even knows what it is if you are 'talking to users' then you have users, which puts you about ten miles down the road from what i'm talking about post-release support is not the same