If you want the game to be as lightweight as possible, then nothing beats vanilla.js. Any framework or even game engine is going to include a ton of stuff you don't even use but which still needs to be downloaded by the user on startup.

But if you want 3d rendering of sufficient quality, then building your own based on WebGL / WebGPU is going to be torture, so you might opt for a compromise here and use something lightweight like three.js (~0.5 MB when minified).

Another option you might consider is to write the core game logic in a programming language that compiles to WebAssembly. WASM code can be more compact and performant than the corresponding JavaScript (but YMMV depending on the quality of the WASM compiler toolchain you are using).

I build my demo as web game using Unity. The size of pretty small if you're careful with the assets you put in.

There's some popular web portal that you can upload to, armorgame, newgrounds, poki, ...

If you don't plan to build a fully fledged engine but a game (with reusable code) go for webgl and vanilla js (or typescript). It mostly depends on features your game needs, if it's simple so no complicated physics I would try webgl or three js

Im building mine with js/ts + pixijs for rendering

The size of your assets is something that is 1. independent from your technology stack and 2. something you have full control over.

The engine runtime environment isn't under your control.

And with most game engines you usually only have very limited control over the asset loading strategy. With vanilla javascript, you must implement your own loading, so you can precisely control when to load what. You can, for example, show the title screen before all the assets have been loaded. Yes, some game engines give you some control, like Unity with its optional Addressable Assets system that allows you to manually trigger the loading of assets at runtime. But that still doesn't give you control over everything the engine needs to load until it gets to the point where it is even able to run your scripts that manage asset loading.

Some time ago I made 3D car racing game that runs in browser using webgl and it had over 50FPS on a $50 phone (myPhone Pocket, Android 6.0, 512MB RAM, 4GB flash, at one point it sold for $42). This was around 2017, but it still works. It was originally android app (using webview) but then I made it normal browser game.

The "trick" is to not use any frameworks but use webgl directly and from day 1 develop on old phone and measure performance in each step and optimize the code from the begining. So not really a trick, just a conscious effort from the beginning.

Before I started I tried 20 different webgl engines (simple rotating textured cube) and only 2 had full 60FPS, most were 20 FPS or less. You can do the same, careful optimized code and performance testing from the beginning, careful caching where it applies (for example in one other game I even had to cache labels because they took too much time to render, etc...).

Compatibility I tested in everything I could get my hand on, google play had testing on 7 or so phones and it saved video from the test, this is how I found out that some phones at that time didn't supported html grid so I used flexbox instead.

I didn't do any marketing, it was free game.

Not really, if you create a completely empty Web project on Unity and build it optimizing for file size it's already 10MB. That's massive! On my internet connection that's at least 2s of load time

The important question is why "online"? What do you want on the "online" part? Multiplayer? Available without having to download anything? Cross Platform?

If you want multiplayer, or if you want saves, you will need a server and that's more than just a single standalone javascript game which can be hosted pretty much everywhere.

On the library part, I tend to use Phaser. It's pure javascript so pretty good for standalone game. You can also use powerful engines like Unity which can convert into HTML5 (not really that good of a conversion, really buggy).

On the server side, if you need one, again, depends on what you need. If you want multiplayer with real time interactions, maybe nodejs can be a good way. But if it's just to store saves, any web language can be enough. I tend to use PHP for that.

Start with localhost server.

You can try nodejs and some xamp for that database.

Learn javascript.

No need to worry on hosting.

Hey, you should check Twine : https://twinery.org/

Mobile browsers suck green donkey dicks.

Take a look at pixijs.

Unity WebGl target is rather... peculiar offspring xd. But it works most of the time and will try to convince you, that you are not doing webdev, which is noice.

Phaser.io is what I'm using. Its a popular one.

I am using p2.js as physics engine and bind three.js objects to their position in render loop (mainloop.js). Like this you can freely choose physics and render engines. Phaser.io or similar will at least do the same and give you some important classes and structure to create your game.

I originally read this as “I want to make a Bowser based game.” And i immediately said, “BRILLIANT.”

Besides the legality question (which is incredibly important) the difficulty of the website is based on how you will run the games.

Any website is just a domain name routing system connecting to a web server. Every web page is just html/css and then a backend language connecting to the web server.

The technical problem won't necessarily be hosting the games depending on how the game is run (flash is dead). The problem will 100% your data throughput cost. The amount of data you are sending out with legally hosted games is going to be massive, especially if you try to implement multiplayer.

I have a degree in software engineering mostly doing web development with php/mysql. I've hosted through Digital Ocean and I can estimate that at the lowest minimum cost just to maintain the website is going to be around $20 - $40/mo just the webserver.

If you are using illegally sourced games and charging a fee to pay the upkeep of the server (nevermind the cost of multiplayer) then you are very open to your domain registrar from cutting your routing services or even voiding your contract (I use Google which normally does 2 year contracts based upon the .extension)

What do you mean by "games are already made"? Are you trying to run other people's games on your website? That's horribly illegal.

I'm can recommend Godot, as it is quite easy to get into and has great documentation. There are also a lot of tutorials and addons to support your work. For assets I would recommend to get a style which you are able to work with. Pre-Paid assets are great for starting, but you need tp be able to tweak them manually. So my recommendations: For 2D like a simplistic style flat colored, vector based like Limbo or Super Hexagon, Pixelart like Terraria, Moonlighter. 3D you can go low/mid poly, flat colored (or gradient texture) like Art of Rallye, Untitled goose game

Just pick any game engine and start doing tutorials. Personally I recommend Unity because I find their official tutorials from Unity Learn easy to follow.

No advice really other than treat it the same way.  I worked as a software dev for 7 years, got laid off back in March, and have been working on a game since.  Read tutorials/docs, look at reddit/stack overflow, etc.  The only difference I've had is watching tutorials - can't say I ever watched videos on Kubernetes or Kafka.  Otherwise Its all the same so far (though I've yet to attempt 3D models lol, I imagine my advice won't go quite as far there).  Assuming you did it professionally in any capacity, it helps me to structure my work the same way: tickets with AC and DoD, Sprints, tech specs/one pagers, etc.

I did it. I just choose engines that allow me to transfer my skills a bit better (strong OOP, in my case). If I can do it, so can you!

Look i personally think game development industry is worst of all, because of the crunch time, low salary and in india the community is so trash. I know game dev is super fun but the Harsh reality is game dev in india is so dead. You can try freelance or indie development but you need to stay consistent like 6-7 years to start earning some decent money, you can try remote globalization if your so interested in game dev.

This video complements the text tutorial at https://trevors-tutorials.com/0009-building-ebitengine-games-for-web-browsers/

Trevors-Tutorials.com is where you can find free programming tutorials. The focus is on Go and Ebitengine game development. Watch the channel introduction for more info.

This video complements the text tutorial at https://trevors-tutorials.com/0009-building-ebitengine-games-for-web-browsers/

Trevors-Tutorials.com is where you can find free programming tutorials. The focus is on Go and Ebitengine game development. Watch the channel introduction for more info.

This video complements the text tutorial at https://trevors-tutorials.com/0009-building-ebitengine-games-for-web-browsers/

Trevors-Tutorials.com is where you can find free programming tutorials. The focus is on Go and Ebitengine game development. Watch the channel introduction for more info.