Відео

Haha, dang that takes me backkkkkk

So, I originally started with C# (hated dealing with mono), attempted to add C++, and landed on writing my own language. I will admit that some of my motivations for creating the language were to just learn how to write a compiler, but I did have other reasons: - One requirement of the language was to do compile time replacement of text (which I know you can do in C++ but the syntax is kind of a lot) - I really dislike the look of using macros to provide text replacement/function attributes like Unreal does. I find it annoying to have to remember to apply macros when I am doing basic engine things. - Providing text replacement via parsing is really complex for C++ since I would have needed to support the standard. - For my language, I just need to provide the functionality that I want, and that's it. I just need to make sure the output adheres to the C++ standard. - I do not intend to support complex data (like structs/object-hierarchy) inside the scripts. The engine is meant to be a scripting-lite engine where most of the functionality, complex algorithms, and data should exist inside the engine, not in the scripts. This way I can manage the performance as the engine developer better. - I realize the last part reduces the engine-user's ability to customize their game, but if I want new functionality, I'll just add an extension to the engine. And I mean, its my engine, so I can just add what I want whenever. TLDR: I didn't mean to type so much, but the tldr is that I ran into a lot of roadblocks when getting C++ to work and realized just supporting my language would be easier for myself with the requirements I had in mind.

​@@ElPsyStreams hold on a second, does that make your language effectively just one giant macro haha, if it goes: kgscript -> C++ -> assembly? thats a pretty neat idea, big plus on performance (although not that it would really be needed in an engine's high-level language, especially since you've mentioned foregoing complex data)

@@ConnorJT I suppose you could say that. One big macro with a nice big bow haha.

So, the scripting language took a long time, but that was mostly my fault because I was using C# for scripting originally, so I had to transfer over the scripts which took a while as I was developing. To figure out the current scripting system, transfer over the scripts from C#, and modify the text editor to work with my language, took 5-7 months. I would image that if I didn't have to transfer the scripts from C#, it would have cut at least 2 months from that time. It was, by far, the hardest system to get right for my engine so far, since it is core to the rest of the engine functioning...

@@ElPsyStreams 5-7 months sounds reasonable, fast even. It took me the same amount of time to create my own shading language (a transpiler to HLSL,GLSL, MSL). The editor looks very nice, do you support breakpoints and debugging?

Ohhh nice. I want to write a linter for GLSL. That is really cool and not trivial! Oh, I did have some ideas for adding breakpoints. For now, I am just going to add basic timers and other smaller debugging tools.

Yes, I definitely have plans for adding animations. I am going to start with 2D animations and then move on to 3D animations. I already have most of the EditorUI tech ready for an image editor and I just need a sequencer. I think I am going to use this ImGui extension for the sequencer: gitlab.com/GroGy/im-neo-sequencer I'll give a little roadmap until I add animations: - Finish Pong Clone (just gotta finish working on UI system) - Create Snake Clone (fairly simple) - Create First Engine Release v0.1.0 (Clean up editor + add some better documentation) - Create Breakout Clone (Add 2D animations here) - ...at some point add 3D meshes along with 3D animations (not entirely sure when yet)

@@ElPsyStreams when you make animation, can you make tutorial about it?

@godnona9570 Oh yea, that sounds like a cool video idea!I can't promise anything right now, but I'll consider it when I work on the system!

I'm dead haha

You know, I really appreciate this comment! Thank you

Ayeee forsureee. Fair warning tho, it is still a work in progress. But I appreciate the interest!!

Yee, the next dev logs should be more edited. This dev log just had a lot of information to cover, so I did it like this to save some time. Otherwise I think the video would have been like an hour long if I added a bunch of jokes and transitions haha.

​@@ElPsyStreamsdo you havw discord? I wanna talk with you about your engine im also learning GDeveloo By Myself : )

​@ElPsyStreams Please ignore these comments. I guarantee you most of the people watching actually care about engine development more than jokes.

@@n00bc0de7it’s all in jest. Sharing is its own skill that needs mastering too friend :)

why

why? the voice fits the cat perfectly

ayeeeee

Hmm, I suppose that's tough to say. The engine already supports 3D meshes, but not as assets. I am mainly focused on completing some basic games and having a decent shader editor before I start adding more visual features. Here are my current plans leading up to more 3D features: - Complete Pong (I'm just working on UI and cleaning up networking) - Complete Snake (As a test of the engine) - Create first release of engine on github (add documentation and whatnot) - Complete Breakout (add particles, make physics/audio more robust) - Add shader editor and work on art shaders - Finish learnopengl.com (Add more 3D features)

@@ElPsyStreams amazing bro

Oh heyo Niadoki, yea I should. I think I'll start in the next hour or so...

Damn man, I appreciate the kind words!!

Not yet, I have still been getting used to streaming in general. I have been feeling more confident with the streams and coding so I am thinking about posting a schedule next week.

I have seen other content creators using the language, however, I am not very familiar with it myself. I just googled the language and it looks like it has bindings for all the major rendering API's. I saw a few ECS libraries and they have support for a lot of these features in vendor packages, which seems much easier to set up than C++. I also like the way their website presents the language documentation. Everything is in one place. If I had to redo my engine, I am not sure which language I would pick honestly. The main benefit for C++ is that most tutorials in this space use C++ since it has been the standard for forever, but I know that is changing nowadays...

Where is your git repo?

@@4dillusions not ready to show it yet, maybe a couple weeks

I am thinking about it. I am still getting used to streaming and posting on this channel, but I may in the future. I am just not sure what I would do with the server currently.

MWUAHAHAHHAHAH, OF COURSE THEY ARE!!

Oh yea, I really appreciate the kind words mangg. If you want a short answer for how to get started and you want to write the entire thing yourself, I would say take a look at learnopengl.com. I think the most motivating part of making a game engine is getting the graphics working, and this website takes you through all the basics. This website takes you through a surprisingly large amount of topics in engine development. After you cover learnopengl, you can dive into any particular topic that you want. Another resource that I used heavily was the Cherno's youtube channel. He has an entire game engine series that creates a basic framework for an engine. I would not say the series leads to perfect engine code, but it is a nice starting point. Since you mentioned that you have experience with C#, there was an open source engine I saw the other day called the Murder Engine that is written entirely in C# and is on top of a game framework called MonoGame. This is a nice route you could take because you do not have to write all of the lower level code yourself and can focus more on the engine tooling and whatnot. A book that I use nowadays for engine reference is Engine Architecture by Jason Gregory. I would recommend reading this book after you have gotten used to basic engine development, since the topics tend to assume some background knowledge. Hope this helps!

I started programming in 2010, today My engine is not barely ready, it stil needs. It depends and want you need or want from it. The more generic and advanced rendering tecniques the more time it will require. Just write down a roadmap of the things to do, estimate each task and double it's time, that should be roughly a good approxymation of how much it will take you. Hope i've helped.