Not recommanded by Chatgpt for PB games :(

[threedslider]

Thanks all for your info

[Thorium]

Unreal Engine is industry standard, and is for triple A. A solo dev or small team better not to aim AAA visuals. (indie games sells for the gameplay, as big companies do not risk new things, only the things that was a success once. Thats why AAA games arent innovative, and thats the reason behind the lot of remake/reloaded these days)

That not really the case.
Unreal Engine is for everyone. Big teams to small teams to solos.
Its tools and simplifications make it very accessable. You can just use the blueprint system and never even write a single line of code. Or go as deep as modifieing the engine code itself as it comes with the full source code.
It is really for every level of game dev and project. Maybe, if you make a really simplistic project its overkill. But its not just for AAA.

[Nituvious]

It makes sense that the chatbot would say this. They base their answers off scores and it should be biased towards the most popular things due to volume.

[minimy]

Unreal Engine is industry standard, and is for triple A. A solo dev or small team better not to aim AAA visuals. (indie games sells for the gameplay, as big companies do not risk new things, only the things that was a success once. Thats why AAA games arent innovative, and thats the reason behind the lot of remake/reloaded these days)

That not really the case.
Unreal Engine is for everyone. Big teams to small teams to solos.
Its tools and simplifications make it very accessable. You can just use the blueprint system and never even write a single line of code. Or go as deep as modifieing the engine code itself as it comes with the full source code.
It is really for every level of game dev and project. Maybe, if you make a really simplistic project its overkill. But its not just for AAA.

Is true Unreal is very nice, but...
Try to run in a xeon X5, or in other better like i7-6500... Unreal is HEAVY, very heavy.
I have been using unreal almost since the beginning and I always leave it because it is heavy, the result is enormous in size and if you don't have the latest in technology, it hardly even moves.

Like Unity and others it offers a lot, but you need a lot more to work comfortably. Plus you're going to need C code, almost yes, or yes.

I think miso was referring to that, unless you make a space invaders. In that case with purebasic you do it in half the time and with a tenth of what the other measures in your HD/SSD.

For my game, I often program in bed with an Acer one and it works wonderfully, then I transfer the.bp and that's it.
Hard to do that with an AAA engine.

I don't disagree with you Thorium, but they are two very different worlds.

[Carm3D]

Best to go lowpoly, that way it will be quicker to create and add new props to the game.

I refuse to go low poly on religious grounds.

[minimy]

Best to go lowpoly, that way it will be quicker to create and add new props to the game.

I refuse to go low poly on religious grounds.

You will change your mind when watching 12 fps

[Carm3D]

You will change your mind when watching 12 fps

Are you saying that PB using Ogre cannot push a similar amount of polygons that Unity or Godot can?

[minimy]

I know I'm not qualified to lean towards one or the other.
But I will tell you my personal experience and what I have noticed with each other:
1) PB+Ogre uses the iGPU, that is the main problem. I haven't gotten the Nvidia, AMD or other eGPU to 'boot'.
Unity, CryEngine, Unreal, and others use the eGPU, so it stands to reason that they should be able to move more vertices per mesh.
2) Dedicated engines use many super-optimized features for common things like animating meshes, moving cameras, lights and shadows, etc. They should be faster than our direct code in PB, unless you're an ASM prodigy.
3) These motors disable many things that are outside the line of sight, distance, etc. and they make it automatic. With PB+Ogre there are some things that still work behind the scenes or that you have to stop manually to gain some fps.

In my case, adding and subtracting the pros and cons, PB+Ogre gives me complete freedom to do:
My own dialogue editors, trading system, landscapes and a thousand other things. When I say complete freedom, it is literally that. You create, you don't edit a pre-established world, you create the rules, it's not the editor who rules.
I like Purebasic too much to be objective

Everything I say is based on my personal experience. Surely there are more advanced users who can inform you better.

[minimy]

One thing more...
If you are good making shaders glsl o dx can win speed too.

[miso]

Can make good and bad with both
I'm with Minimy, I like text editor more than a builtin vysiwyg editor.
Also by just looking at a game, with 99% accuracy one can recognize the unity or UE engine. They have a certain style of look because of their builtin lighting and such things.

Lowpoly for me means quicker prop creation. I don't have thousands of artists working on textures and characters and items.
Mainly thats why I lean toward 2d. The absolute fast content additions would be for anything the text based, but I dont want that. Maybe a classic ascii rogue one day.

Edit: Second thought. My old time favorites can run on a 386, or one of them I know can run on a 286. Those games still on release on GOG with modern hardware support now. Those machines were slow, had little memory and HDD space, no GPU acceleration, still those games are still good.
We have a lot of hardware resources now, good games also can be written

[Carm3D]

Is it fair to day that if I manage to code efficient geometry culling for things that are out of view of the camera and things that are behind other things... then PB+Ogre can stand toe to toe with the likes of other more popular engines? Or is the eGPU / iGPU issue still going to lean performance in favor of other engines?

[miso]

These are opinions of mine, and not facts:

1. No engine stands toe to toe with UE

2. On engine level Ogre seems better to me than unity

3. Unity game creators still have to optimize graphics, batches, but it is often not used, or not used well, and a poorly optimized game comes out.
It works well on strong machines. Well optimized pb game can perform better.

4. Pb 3d got some love last year, upgraded engine, increased shader possibilities, bone creation on the fly along with the already presented mesh creation on the fly makes me say it is possible to make a game with it, even a medium size one. I'm not talking about doom6.
Major weaknesses are the collision, the physics implementation. It is possible to make some extra rendering optimization with the current toolset.
Theres a slight blur when moving camera (or sprites) fast. I don't recognize this on other games. Some choices are not the best. (input in flipbuffers as far as I know, renderworld updates the physics, that should be separate.)

5. Integrated/dedicated gpu: I'm not sure. I only have single gpus. Someone should test with 2 GPU: does he/she has better performance, if compiles a pb 3d program then sets in Nvidia control panel to always use DGPU with it? Can make assumptions, that something is missing from the global declarations, but those are just assumptions, it might be already there. Noone told yet that this test had been done, and got better performance. (they might have tested, but never told the result) So I'm puzzled with that one, but if true, I would like to see it investigated a bit. Also dont want Fred to be distracted with a false positive.