Larian Studios Forums Baldur's Gate III Baldur's Gate III - General Would it have been easier to rebuild the Divinity Engine from Scratch?

I am honestly curious about this - given the difficulty Larian has had in implementing core changes - would it have been easier if they had built a new Engine from scratch instead of modifying the existing Divinity Engine to handle 5E. Larian instituted a ton of changes in Patch 5* - which caused a cascade of new bugs to crop up that will need to be fixed over time.

They still need to implement the "Ready Action" and/or "Reaction" system (A little confused on whether Reaction system also encompasses "Ready actions").

When the Reaction system gets implemented - if it can be implemented - what will that break? Are we nearing the theoretical limits of how much the Divinity Engine can be modified?

Its a 9 year old engine at this point if I am not mistaken.






*Incredible, amazing, much-needed, happy changes.

The problem with implementation of reactions isn't with the engine. When people mention how Solsta implemented reactions they forget that Solasta has homebrew subclasses. Lore Bard and Wild Magic Sorceror means every time an enemy makes an attack roll the game pauses to ask you if you want to use your reactions for Cutting Words or Bend Luck. Fights with many opponents would become extremely sluggish when EVERY little goblin or spider making an attack will cause the game to ask you if you want to use your ability to tamper with the roll. Even if Larian adds the proper by the book reactions they would need to make it optional, because for people who are not DnD purists it could be very annoying.

Things you mention (like reactions) don’t come from engine limitation - it comes from Larian believing that such system wouldn’t work well in computer game and is looking for other solutions. I wouldn’t expect reaction to be implemented as they are in PnP at this point, though it’s been mentioned that they are looking on some kind of change/improvement of the current system.

It’s been also confirmed that engine have been heavily overhauled for BG3 - they are not limited by what D:OS2 allowed, and that should be obvious just by comparing both titles.

Reactions would need to have a lot of customization to make it satisfying to most player's I think. So that those who like prompts like that get to react to everything, while those who prefer more automated can get that.

You leave Scratch alone. He's a good boy and doesn't deserve being made into an engine.

I think they’re doing alright engine wise. These guys know what they’re doing when it comes to big development choices like that. It comes down to design philosophy and what they’re thinking will be fun or realistic and how they can implement it in someway that also aligns with the 5E rules.

Well they said they were working on implementing it but I didn't get any details. I am super curious how they are going to make it work without creating a mess and issues with combat flow.

I don't think pausing the game while someone determines reaction is viable - but maybe setting reaction conditions? "If Crit - then Smite" or just "If Hit then Smite" would work I think.

Also do Reactions cover Ready Actions? Solasta lets you only Ready Cantrips but RAW can you Ready more powerful spells?

+1

Take my upvote and leave! :P

Even from before Early Access they used reactions as an example of mechanics they need to change for 5e to work well on PC (the first mention of it, that I encountered can be found here or here.

They did mention recently they are working on improving the current system, but it’s anyone guess on how they will go about it.

It is probably much easier to modify an existing engine rather than starting from scratch. Bioware started from scratch for many things when they switched engines, and it was a lot of work for them and caused lots of problems.

I doubt most of the contentious points with the current state of the game can be in any way blamed on "engine limitations" to begin with, frankly.

Oh dear God, no. Why would anyone want a rebuild from scratch at this point? Might as well never release the game. It wouldn't come out until 2030.

The answer is "it depends". Japanese game developers used to almost always build new engines for games, which became problematic during the PS3 generation of game development. Demands kept increasing for what an engine could do.

That's still true today, to the point game developers use Unreal or Unity. Or are part of a larger company that maintains their game engine.

So for a retro or smaller scope game it's usually a reasonable choice to make a unique engine for each game. But for large scope games like Baldur's Gate 3, it's been fruitful to stick with an engine that has a long life cycle.

This is very true, UI, party controls, and design choices have so little to do with game engines in 2020.

The reactions example, it's probably always been doable in the current engine. But Larian as a design choice did not want reactions to slow down combat. (There's an early interview on YouTube discussing it). Enough players have expressed their opinion on the other, and Larian wants to make the right choice with reactions to appease the player base.

Readied actions are rarely brought up, relative to all the other game critiques. So they're probably a low priority.

This. Let's not pretend that they're dealing with the diablo 2 engine *shudder*. They could do a LOT of what people are asking for, they simply don't want to.

Bethesda has been using and modifying the same engine since 1998, likely for the same reason.

The engine also has other problems like flight being just a long jump in a game which made verticality a marketing point.

I feel like it would have been amuhc better plan for them to start over with a new engine, taking cues from their learned experience with past games, but building a fresh engine for a new type of game, and for them to have *Started* that way... but now, they're so far along it's a purely academic comment with no grounding in reality, since there's no practical way for them to chance engine wholesale now, and they'll have to deal with adapting the one they're using to try to make it do the things it currently can't.

A D&D game produced in this day and age that can't handle 3d flight properly is ridiculous, in my opinion... Especially one that likes to talk about verticality as an advertising point.

What's the point to have better fly ? It's a top down RTS, it will be odd and pointless to have your ennemies above the camera no ?

And if you want to fly across the map in exploration mod, this can't be a possibility because it will be a game breaking experience

If you make more or less the same game I understand that it makes sense to expand what you already have. XCOM2 build on XCOM1 engine, rather then upgrade to new unreal as devs deemed it would be u productive to mold new version of UE to their top down needs.

There is no indication that Larian would do anything different if they were building engine from the ground-up. Solasta cube based flight is cool, but is it something Larian is interested in in the first place? I haven’t seen a single reply from Larian along the lines: “that would be cool, but unfortunately our engine won’t support it”. They are throwing a lot of money into the game building new systems and rebuilding old ones - if they really wanted to do something, I am sure they would go for it.

- Height and flying is a relevant and important part of combat in most D&D games.

- This game is not an RTS of any description.

- Proper flying mechanics also includes a functional camera control and automation system that can manage it effectively. The camera in game right now can't even handle the difference between climbable and pathable surfaces at distinct levels and our spell AoEs are 2d circles, not spheres. The failure of one system is not a valid excuse to deny the legitimacy of another.