This was made entirely in Construct 2/3 (I upgraded halfway through). All I used was the visual programming, never felt like bothering with javascript for it.

Quick visualization for you, the link below is screen capped from Verdant Village. That’s what the game actually looks like. Its all one layout and each of the little boxes is a separate room that loads in when you enter it. So only one of those places is ever loaded in at once. You’ll need something like that, at least in Construct you will.

ibb.co/DDvghRB

Also, as an additional note since you mentioned it. Multiplayer is one of the harder, if not hardest things to implement. If you want your game to be multiplayer you have to plan for that from the start. I suggest you approach everything in your game as if multiplayer was mandatory.

Also, because of how Construct works you can’t have players in a multiplayer session on different layouts. So, unless you want to limit all players to being in the same area of your farming game you have to make one big layout with all the individual rooms inside that layout. Then you break it up via artificial transition points in the world where the game can dynamically load and unload assets into the world because if everything was loaded at once it would have performance issues. I can say with certainty that this can be done because that’s how Verdant Village works.

At the end of the day, honestly, no matter where you look on the internet you will find people debating engines and which one is good and which one sucks, etc. No matter what engine you pick there will be people who will tell you ‘That engine is terrible, you should never use it’.

My advice, ignore them. Pick the one you think works best for you, and work within it. If there was one thing I learned working C3 its that in almost all cases the issues I ran into were almost always ones of my own making and not the engine.

Unreal would also be more than capable of running this sort of game however I think Unreal caters more to a higher fidelity of game. I’m not sure its made for pixel art and 2D but I’ve never looked into it. Maybe you don’t want to do pixel art though, so that may not matter. Regardless, Unreal Engine falls into the same sort of space as Unity I think where the engine is incredibly powerful, but there is a lot of bloat going on.

As for me I’m using Game Maker 2 to make the new version, and yeah, I had to start completely over. I still have all the sprite assets and stuff of course, but the code has to be entirely redone. That was part of the point though as, like I said, I made plenty of mistakes in the Construct version which I couldn’t fix because they were too buried in the code by the time I realized.

I know what you mean in terms of wanting an engine that isn’t just staring at lines of code. It’s the same thing that drew me to Construct way back when. Unity is definitely not going to be your friend when it comes to visual coding. I can’t really think of any other engines that have visual coding built into them as a baseline. I think Game Maker has a visual version, but I’m not sure. Outside of that I think RPG Maker is more of a visual engine. Although I think if you want to do much in RPG Maker you’re forced to start doing scripting.

As for AI and ChatGPT and such, I’m not sure you’ll have much luck. AI can do some neat things, but I think its abilities are largely over-exaggerated. When it comes to code you can tell ChatGPT to write code for you, but to my knowledge it can’t get anywhere close to the complexity you would need for a whole game. I believe its better at automating little tasks within coding and even then I’m not sure I’d trust the code it writes.

Also, last word of advice, I promise. I could be completely wrong, but you remind me of me when I was younger and wanted to start making games. The first thing I wanted to make was an MMO like world of warcraft. You've likely seen this opinion all across the internet but its best to start small in terms of projects. A game like this seems like it would be simple on the surface, plant some crops, raise some animals, mine some rocks, etc. I can tell you from experience that it cascades into a huge undertaking. Of course, far be it from me to dash your dreams. I believe that if you are willing to work for it you can make whatever you want regardless of experience. Just understand that it may be a long and bumpy road. Regardless, good luck, and I'm sorry for writing a short story here.

But, I've started rambling again. Sorry. This is all just my experience. Large games have been made in Construct. Iconoclasts is the one I remember being made in Construct Classic as sort of a proof of concept for me using the engine (its on Steam if you aren't familiar). I don't think its a bad choice of engine, especially if you don't want to get lost in code, just be wary of the potential issues you could have.

So, I'll stop rambling now. I've gone on long enough. Construct 3 is a fine engine if you are looking to make a game. The most obvious concerns I can bring up are those revolving around the NW.js exports. NW.js is a shell program that Scirra simply uses for exporting. It is still in beta to my knowledge and is open source which means a whole bunch of people just collaborate on making it better. Because Scirra doesn't make the NW.js framework they are at the whims of the people who do. So issues revolving around graphics cards compatibility and errors in running NW.js won't necessarily be fixed. I mean they probably will but you have to remember the people making NW.js aren't making it for Construct specifically. They are just making it. So if you have an issue with your export it may not be fixed for some time, if ever. Also reporting the issue to the NW.js people is usually pretty hard unless you want to get fairly deep into coding stuff.

Last thing, which may not matter if you aren't looking for a more simplified engine. C3 does a lot of stuff to simply things for the coder. If this is what you want its great, and don't let anyone tell you that you can't make a game without directly writing code, they're completely wrong. However. In simplifying things you take away options which in turn creates restrictions. The most obvious example of this is the sprite in C3. The sprite basically works as both the object and the image in C3. Its a bit hard to explain without going into even more detail (and this is already super long) but by tying everything together like that you lock in a lot of details, disallow things, and also likely cause some redundancy in your assets. None of this is a death sentence but it can be restrictive.

Basically, too many layers for me to be comfortable. Also, while they don't outright say it I have to assume that using a porting program like that probably costs a pretty penny to do a base port and potentially some sort of additional fee each time you want to patch it (but that's just speculation).

All this said, the game in this engine also had more than a few issues in terms of my code. I learned a lot as I made this game, but by the time I realized some of the things I did wrong it was far too late to change them without ripping everything out. Remaking it in a new engine provided an opportunity to fix those issues and make other changes to the game as well.

I don't know much about it, but it seems like a third party porting option. This likely works fine, but as someone who works in computer science, I don't tend to like pushing a program through another porting program to get it to work. There's the danger of something breaking obviously, but there's also the concern of losing efficiency because you are basically taking code that is made to work in one language and transforming it into another. Something like that usually doesn't happen cleanly.

I don't want to slight the people who make porting programs like this, I'm sure they are very smart (likely way smarter than me) and capable, it just makes me nervous. More so when I think about how the entire game I'm working on is reliant on not just the engine I'm making it in sticking around, but now a separate porting program to actually get it to run also being around and keeping up with any changes to the C3 engine it concerns me.

Back when Construct 2 was the engine here I believe one of its main selling points was that it could be easily ported around because the base of the engine is HTML which is a fairly universal language. What Scirra likely couldn't predict is that consoles decided basically within one generation of hardware that they didn't want to support HTML anymore which kind of killed the whole idea.

Now. I will say, if you want to use C3 and still port to console there are apparently options. The link below is an example of one.

construct.net/en/forum/construct-2/general-discussion-17/chowdren-fast-construct-134395

Aside from this there were minor little things that always sort of bothered me. NW.js is basically just a chrome window running as a desktop app. That came with weird stipulations surrounding the mouse and updating the window size, setting fullscreen, etc. Nothing major, just minor little annoyances in my mind.

Now, all of this aside the other issue, which really only became relevant in the more recent console gens I think is the porting issue. If you weren't aware you can't really (to my knowledge) easily port your Construct game to consoles like switch, xbox, and playstation. This may not matter to you at all, but if you are looking to make money you are cutting out a huge chunk of the market.

I have effectively run into two issues. One that is an actual issue, the other which is more just unfortunate. Although these two things are sort of tied together and could potentially be alleviated in the future, or maybe I just didn't investigate enough.

To explain. The problems I've seen with the Steam release in particular is that some people just have issues running the game. Not like they run it and it lags but weird things like it booting to a black screen, not opening and closing properly, etc. Stuff that, based on the time I spent looking into it seemed to relate back to particular graphics cards working with the NW.js export option. Which to my knowledge is the one you use when you want to port to PC.