Why GEARBOX, WHY! - modders nightmare

M.I Complex programmer here.

Attention modders
Looks like the lua hasn’t been updated which means anything but visuals still runs on single core… … (I think)
I sure do hope we’ll be able to have access to a C method to able to do anything with the engine. I mean that would be just dandy.

-Sincerely anyone who wants to do anything with the engine other than make new ships and spend 9 months trying to find workarounds that end up causing lag and require a complete rewrite if one script needs changing, following this they take up drug addictions and start seeing the matrix everywhere - init)

P.s Gearbox: do it for the drug addicts if anything.
-at least from my research :}

If you’re new to modding, then welcome, I’ll be starting a depressed modders hotline.

-To reiterate: This has nothing to do with mod tools.


chill, the modding tools haven’t been released yet

This has nothing to do with the modding tools.
All they do is accommodate the new shaders /methods for visuals.

This means that scripts that were expensive such as the weapon upgrade system (Changes weapon layout of all ships depending on research) and planet management will yet again be inefficient and a complete nightmare to implement.

So no I will not chill :smile:


How about making game “more complex” by adding collision, ballistic…

Well that’s exactly it with homeworld classic we couldn’t do these things due to the rubbish lua system.
We were hoping they would change the lua or even unlock the source code for modding to allow for such things :L

I would guess the answer is, they did a remaster plus, not a complete overall or remake.

A massive update to all the art assets, UI, conversion of assets from HW1, graphic shader + update, and netcode (because they had to)
That is allot to be sure, and much more then most remasters do ever do, but except for the netcode it would seem the core of the game is unchanged as completely expected in any remaster.

Not surprised at all on that, kinda wondering why people are?


Yes those aspects are all very shiny however they did mention on many occasion multicore support.
To which they haven’t updated;which makes modding; the very thing that has kept the game life all these years increasingly difficult by requiring a new visual standard and simply preventing any further progress on mod mechanics other than a rehash of previous builds.

1 Like

Granted I am not a modder, but I’ve been following everything they’ve said since they announced the project, and cannot remember them ever mentioning multicore support.

I can tell you that while not all the hopes everyone had about optimizations came true, it does look like some useful new features have been added to the engine. Multi-coring the physics was something I never expected to have happen, personally.

Just the thoughts of a far from depressed modder.

I never heard that and it seems others have not either. cite your source?
also, again "remaster " they could have easily just gave people the “classics” we get for free just updated to run on a modern system with out effort and been done with that and still fit the definition of remaster.
They never once stated they where going to change the core of the game, in fact the exact opposite. So any extras (like the once I listed are all bonus and very nice)

Isn’t a lack of multi-core support these days basically the same as locking the performance of the engine to what we had ~ 15 years ago anyways? For a game like this with (potentially) a ton of ships and projectiles, I don’t see how basic multi-core support could be ignored.

It’s like upgrading the paint job on your 80 year old Ford model T - new tires, etc… - but forgetting to put in an engine capable of driving faster than 20 km/h.


Honestly, I think it’s more akin to restoring an old car, and not converting it into an electric. As I understand it, multi-threaded game logic can require such fundamental changes that in some cases you’d have to rewrite the engine essentially from scratch to make it work stably. I don’t blame Gearbox for not doing that at all.

It’s sad but I can’t really blame GearBox for not adding multicore support. It’s a lot of work.


I retract my earlier statement, you obviously understand this stuff much more than I do :sweat_smile:

1 Like

Hardly man, while multi-cores are standard now its not lot the processing power of each core hasn’t also improved. I agree with others in that multi-core support is a nightmare to implement, let alone on an engine this old. To do it its likely they’d have to rewrite the engine from scratch and that’s just an insane amount of work considering such engines are often intended for multiple projects, though some do use the same engine for a little to long (I’m looking at you dynasty warriors)

In short unless gearbox intended to produce at least one or two new space rts games entirely it’d be ludicrous to rewrite the engine.

That said I doubt this will be the nightmare many have suggested as a modern cpu core can push a lot more script and instructions around than even the best single core from 15 years ago.

The problem is like a dev said on another forums… adding multicore support, like 64bits support, is not something that is a mere on/off switch.

I actually recall one dev team that had released a game with 64bits support for their game that later had to literally take out the 64bits branch of their game due to how development/maintenance of it was growing difficult.

Also recall we’re talking about an engine and source code that actually predate the arrival(or at least, standardization) of either multicore or 64bits support.

Honestly, the game is more multi-core than it was. And some of the stuff in the graphics engine has been rebuilt to potentially push that further at some point. BUT - The main issue here is the old LUA.

We did look at updating to a newer LUA. But those of you who’ve done HW Mod work know that it’s use is pervasive - it’s EVERYWHERE. And so much of it exists that there were some big concerns that attempting to update it would amount to a complete rewrite.

Further, just upgrading LUA wouldn’t have in any way impacted multi-core game logic. Because really that’s a design decision that is sort of built-in from the very bottom of how HW2 engine worked. Really going multicore in the game simulation/dynamics/AI systems would have been effectively yet again a full rewrite.

I imagine we’ll do some sort of ‘looking back’ thing once we’ve locked more critical issues and cleared beta status. But for now, you’ll not get much ‘inside thinking’ talk from us - besides I (and others) have much Mod support stuff to do, more about which you’ll hear from us over the coming days!


XD Just expeierance from my studies and mod developing

Yeah I appriechiate the graphics now being handled more on the GPU and ofc that’s all split up across the cores.

The main issue as you mention was the fact that LUA was used, not really if the latest version was installed. Yes lua is great for the simple stuff including some interesting mechanics.
But the issue we faced was including more mechanics regarding ship options in an attempt to make M.I complex well more complex and attempting to impress, gaining more rep for the game;Whilst I did work out a lua equivilent; it was very inefficient; which led to lag with the much wanted epic battles, athough i’m hoping the lessor cpu intensinve graphic load will help that one.

What would really help is if we were given more acces to the C in places such as ui and the build/research options; I’m not sure on the new net code however we had much wanting from players of a multiplayer mechanics capable of switching between maps depending on certain match conditions, using say balcora gate to find more RU’S and add more than a few possibilties for other mods to riff off. A similar parallel would be the crew system.

I hope this clears up what myself and fellow modders have been saying. I’m by no means complaining just we’re running out possiblities.
There’s only so much interest you can gain with just another set of ships and FX.

‘C code’ access isn’t even remotely on the table. I won’t waste your time talking around that fact.

But in terms of lua tasks and relative performance, don’t forget that we had Mod author help, and that meant in some cases looking at slow stuff and adding or adapting new lua calls to do it fast.

Talk (in reasonable detail) about things that would help you write better/faster stuff. We are listening, which means we can help find and even create solutions, right? This game is a living breathing project for a reason.

1 Like