Beiträge von itchyboy

    Woops yeah you noticed the typo.

    All I'll say to what you said is that I hope what you're saying is true and this isn't going to end up the way I imagine it. If there's still a place for human made content in this community, then there is perhaps hope this hobby doesnt become a slopfest.

    Regarding partial AI assisted work, you're right that a line for what's machine nad human is going to be very messy. I do not envy anyone who's doing publicly released mods in the future, because I can no longer trust who actually owns anything anymore. I will however still regard human made mods as more respectable due to the knowledge that was required to do things by hand.

    I wont comment any further because your points make sense but also "time will tell". I don't trust people to use AI only for tedious work. I expect most mods coming up to be made with more AI than human by a wide margin and we won't be able to tell.

    Thanks for the insight and I really hope what you say about the hobby is true, and we're not actually getting replaced.

    I'm glad it works for you but for many people, especially with the way the global situation is headed, aren't happy with having yet another subscription in their life.

    Thankfully we have free alternatives for Photoshop, and yeah Blender is free, like you said.

    Is there a thing that works exactly like Claude but runs locally and costs nothing?

    That's the difference between cloud AI services. That's why I don't consider Claude just Photoshop or modelling software. I am at the mercy of a server, an internet connection and it has to be launched somewhere else on hardware that isn't mine, and access can be revoked.

    Not an accurate comparison in my opinion. Even Photoshop for all the suck of it (fuck Adobe), still runs on my computer and I can source it through <redacted> methods wink wink, and have it run if I pull the LAN cable out of the back of my PC.

    Anyway yeah, the future moving forward is 2 tiered mods. I don't want to have to pay a subscription fee to have the "best" scripts or mods. I don't like the only way to compete and be able to catch up with AI built mods is using AI as well.

    I propose a community divorce where AI mods are their own thing and human mods are in another section, so that human made mods get completely ignored.

    I wouldn't want the future of Em4 modding to be AI-only. What do you guys think though?

    I'm using a translator so I apologize if any nuance on your reply is lost.

    Thing is, I don't hate or disagree; all your points are valid except for one thing that's not been addressed yet: How much did it actually cost you to make the AI do this for you?

    As much as I want things to get more active, it shouldn't come with a price tag in my opinion. AI, in the way it is implemented for everyone right now violates one of my personal rules for modding: that the tools and means of production should be free and accessible for anyone.

    Or else this ends up being "mods made with free tools" vs "mods made with paid tools".

    I also want to hear others' clarifications on what the community standards will be moving forward. I have my own opinions on who "owns" AI made mods but it will be up to the overall community to decide.

    I will clarify one thing I said earlier, it's not so much Em4 is dead but rather there will be 2 classes now. AI and no AI used. This is just going to further segment the community in my opinion though it probably won't kill it.

    I'd also like to add to this post. I am fine with the "class divide" in this community because people are built different. My interest with hobby is the process, even if it takes months. I like making ideas and seing how far my labor and abilities can take me.

    For others like you, your interest is putting the ideas and seeing the result and you don't really care for how you got there. Who am I to tell you my way is better or yours is worse? Clearly both can coexist.

    This is cool, but now the question: What's the point of modding the game anymore if AI can do it all for you?

    Every mod moving forward will now be a money game of how many tokens you pay for, not skill or talent or ability. I'd rather not have my hobby become a "who has the best subscription tier" tbh.

    The hobby really is dead I guess. I'll just continue doing what I do anyway because I enjoy it. For everyone else interested in making as many mods as possible, this is the future I suppose.

    Thanks for at least showing the scripts for me to look through, I did realize that I was ahead of the curve by using classes/namespaces for years now. The issue now is whether those namespaces can be read across multiple files. Say a way to read functions contained in other scripts. If AI can do that, good but I don't see evidence it's been able to yet.

    No really, thank the Lord these scripts are generally readable and I know how to add anything I want to these.

    I look forward with curiosity what this community will look like by next year, when there's 50 AI built mods that all play and feel the same. Quantity over quality looks to be the future and there's nothing I can do about it. I want to see how far this goes before modding becomes completely pointless.

    By the way, given this is made by AI, technically it belongs to nobody. There's no ownership of it because an LLM made the code. Meaning AI mods don't deserve any sort of copyright protection at all. That's at least one benefit.

    ......

    After some more time thinking about it, I think I understand now. There's 2 types of Em4 community members. People who love the game and the process of making mods.

    And there's people who get dopamine hits from the playing of mods alone and don't care about the making part. This is who AI is for and well, I gotta give it respect for catering to these types of people.

    Man I just hate that it has to come to this. I hate how the reality we live in is that I am supposed to pay a tech company to do my hobby for me. AI was supposed to make all the boring stuff go away, not take all the fun stuff and leave us with only with soulless real life bullshit.

    I still enjoy modelling European vehicles, I'm just slowing down in progress due to IRL. Maybe one day I will finish something.

    Actually hold on, I do have ONE finished model this year miraculously.

    I've also inserted a model into Em4 for the first time in 2 years. It's one of our first WIP's for our project.

    Cab, pump and rear by me.
    Miscellaneous parts, decals and textures by Yankee43.
    Aerodynic lightbar is made by EmC-Unit and rihis.

    Update after a few weeks:

    Ive tried the (so far) 2 mods used with AI for Em4.

    Both of them dont work with one another due to a dll conflict. The one with special functions for reloading and driving vehicles with keyboard has no support at all because they vibe coded it and don't know what it actually does with Em4's vplugin system.

    That's all I need to say. It's not just about AI bad, but "using AI without knowing what it does is bad".

    There's alot of absurd stuff on the table that can be achieved by the vplugin system or the stuff with Antiphon's compiler but I'm a bit sad that most people using vibe coded stuff so far dont actually know what they have on their hands.

    English commenter here: I don't let AI touch my modding or scripting process.

    But if others can do it, sure why not. Doesn't mean I'll think of their mod as highly because I know it was done without as much labor and deep game knowledge.

    I also regard AI made scripts as belonging to nobody because a machine wrote it entirely, therefore the AI gen scripts should be free to use for anyone.

    Lastly the issue I've seen so far with others' AI made scripts in em4 is alot of redundancies and functions doing the same thing twice, along with some pretty awful methods for otherwise simple tasks, or trying to invent things that dont exist in the game but doing it poorly/barely no effect or difference.

    The best use for AI if i think about it is for doing more of the same that has already been done by other mods instead of trying to reinvent the wheel.

    Generally I'm impressed with what it can do but I know myself and maybe 5 of us will be the last people left on this game who have institutional knowledge and can think for ourselves.

    I watched the github. Vibe coding concerns aside, are you making an open scripting language to directly edit the game without needing to edit the exe, similar to GTA's ASILoader and the scripts + API that can be done with that?

    This is what I'd rather see than direct edits to the exe. Plugins are a lot less permanent of a change and there's no fuckery with versions or steam or whatever else.

    999London - wip

    This is cool and all man, but I want to ask, did you make this all by yourself? I'm asking because generally, it's a courtesy thing to add credits for content used in mods.

    As for me, here's my contribution as to not derail the thread. My 2 newest handmade projects. I've been wanting to do these for a while. Both are fully done by me as usual.

    Myself and some friends are teaming up to do cool stuff, though we're probably keeping this to ourselves while the AI craze takes over Em4 modding. This one's different, Yankee43 and ColumnDisciple helped me out with this and its been a wild ride passing the blender files between the three of us.

    For the English language side of things, I dont believe there's interest. The reason I say this is that I busted ass for my own work and hardly anyone cares about it to be honest. I did the scripting for my advanced systems solo, and recieved help from 2 of my friends who are ambulance crew in the US. Maybe this will be alot different with the German speaking community.

    A ton of research, UI building, studying of procedures and human physiology and it really amounted to nothing because nobody apart from my close circle of friends cared about playing the mod.

    I consider myself blessed and uniquely qualified to have an opinion on this concept because I already 'achieved' it.

    Unfortunately I can also definitively say I no longer have the mental fortitude and physical ability to do it again. Once was enough for me boss.

    To clarify I also don't think your idea is in vain; you clearly have vehicles and a map ready but like others have said, how hard do you wish to grind to see your concept become real?

    Your vehicle assets/map concept isn't in vain either; you did inspire me to change my mindset with regards to myself; stop being so damn negative about my feelings towards this game and it's community....I need to be more proud of myself that I did so much with this game, mostly by myself too.

    That's right, it's not easy. It's rather difficult to keep the previous script system and rewrite it. It's easier to rewrite the script system or change it for the most part.

    Let's see what else comes out of this post and what ideas, suggestions or possibilities arise.

    Your Montana System looks very nice. Good Job!

    Thanks, I worked hard for what I did

    And the cool thing is that the script is not even finished; in theory I can set it so that every single injured person will have their specific symptoms and require different treatments. Every wreck, gunshot, blunt trauma, fall etc. I just need to add the remaining parameters and study the treatment protocols for every possible situation. Tedious, annoying but doable because I am extremely blessed to have contacts who are professionals.

    However I am aware there is probably a better way to do this concept, better than both RTS 0.9 and my own work. And I also found out that some mods are now beginning to be programmed using AI. Sky's the limit I suppose.

    What I will say is that you've already done a huge part of the work if you have the vehicles, person models, buildings, map and audio. All you need to work on is GUI elements for every medicine and treatment proedure. Then you'd have to logistically find a way to fit this all into Em4 itself, and what things can or cant be done by the game as it is.

    Apologies for an English comment.

    Some years ago, I made my own custom advanced medical simulation for Em4, however it is for the USA system.

    Externer Inhalt youtu.be
    Inhalte von externen Seiten werden ohne deine Zustimmung nicht automatisch geladen und angezeigt.
    Durch die Aktivierung der externen Inhalte erklärst du dich damit einverstanden, dass personenbezogene Daten an Drittplattformen übermittelt werden. Mehr Informationen dazu haben wir in unserer Datenschutzerklärung zur Verfügung gestellt.

    I will say right now that the original RTS Bieberfelde 0.9 mod inspired me heavily, just like some of you in here.

    The mod is so inspiring that I learned to write my own things in missionscript. I've always had it in my mind it was possible to make a custom medical system without copying the logic or resorting to plagarism, and I was right.

    Really cool to know that there's still interest in custom EMS systems in this mod even if it can't be properly visually represented. However I do think that it will take alot of work to be able to make a truly immersive system. Many considerations have to be made regarding how much data and vitals, medicinal effects etc you're simulating.

    To whoever does try it, good luck...I've already did my own and it was not easy.

    Daaaam. Brooo be blessed and the next 7 generations of you.

    Em4 is back.

    If only we could have a better traffic and how vehicles and pedestrians behave:winking_face:

    If you're talking about me, I will clarify something very important:

    That's not an Em4 map with functional buildings and working assets. It's merely a mock up. I will not lie or clickbait you any modding developments.

    All that it is, is a way of plotting out Em4 maps based on real areas, including buildings. However this setup only works on major US and Europe cities. It is completely missing data for most of Asia, South America, Eastern Europe and the middle east.

    Never tested on AUS/NZ areas either.

    What I was saying with this is that there's no easy way to do it, no matter how big the map gets.

    It will easy to decode the launcher and we are not dependent on them. I have already made my own version as well.

    Dude this is freaking awesome. Thank you for putting in the work to make this editor substitute. You don't have to post all the information but how did you do this?

    If I may ask, is there a substitute editor for the lights? The Blender tool is nice but I want to be able to see the patterns and lights in real time with some kind of renderer or what. I'm probably asking too much but that's one of my biggest issues with the editor. Everything else including map making I could tolerate otherwise.

    So your trust in the "legends" was not in vain.

    Hey thank you for responding. It's an honor hearing from you and I appreciate your insight. I learned a few new things about this game and you set the record straight on this. You explained that so well, I have nothing more to say...just need to process it really.
    ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
    Yeah I'm just gonna wait this one out especially if the disc version is not compatible. I'm not switching all my things over to steam just to keep a second version of all my mod files, for something I probably wont use anymore. I just dont feel like anything is worth contributing anymore because it's ironically gotten harder to mod this game, due to the increased scope and required labor of modelling and research.

    What I mean is that I've always had dreams of my own 'perfect' representation of LA and I do have tools available to me to make it happen. The problem is that making a super high quality map texture, and hundreds of buildings (even when I'm selectively picking and choosing them) is still a ton of work for this game's maps. Even with all these new modding developments, the basic process of doing this still hasn't been made 'easy' where I'm just able to use a building model directly and it has a level of quality I can be happy with. These are super low poly and ran using generic textures, so I'd still have to model everything from doors, to windows to architetctural elements like columns and fascades.

    And with Em4, no matter what tools are made, I still need to plot out the ground texture so that buildings fit, roads are curved the right way, grass is set and has smooth transitions to concrete, and so on.

    That '4X bigger maps' thing is real, but it was likely done with AI code and assistance.

    Let me explain:

    The group who made that mod had no prior programming experience until June 2026. In April, their mods had no working vehicle startup spawn scripts and had many scripting problems. All of a sudden in June of this year they began releasing insane things like vehicle pursuits and heavily scripted events.

    Now in July 2 they suddenly 'crack' Em4.exe and are able to add anything they want? Also very curious why it wasnt legends like Fry or Antiphon who did this first? Why these people from out of nowhere

    Being blunt: I don't trust their competence...where did that sudden increase in skill come from? Why not show this earlier if they already knew how? Regardless it's none of my business that they use AI for this stuff. If this means we get something better longterm with the use of technology, eh sure why not. Bugfixing will be their problem though and I'd hate to debug a vibe coded project.

    Another thing worth considering: The decomp of Em4.exe in this case is closed source....where's the github page showing their source code and the inner workings of the injector, decompiled dll and an exact accounting of what their launcher does to our computer? From a purely opsec perspective it's rather unusual that this entire process is so opaque and being done via...discord of all places. I'd expect some kind of documentation for how this was programmed by them if they knew what they were doing.

    For reference, projects like scriptHookV (I'm sure some of you know what this is) or silentPatch are very well documented on places like github showing exactly what the code is doing and how it adjusts the original games' exe to fix things or improve the experience. I can't help but be weirded out we're not seeing more transparency for Em4's biggest new modding development.

    _________________________________________________________________________________________________________________________________

    Anyway I do think what they did is legit and there's a lot they can do. They used .dll injection and a 'launcher' to load into the game and it is likely not in violation of game TOS.

    However converting this to 64 bit might, if they end up breaking the game down to it's bare foundations/source code. I have no idea where that sits regarding the law as I am not a German citizen. You guys will have to explain that to me.

    I am concerned however at the idea that their 'launcher' and .dll fixes will essentially force all Em4 projects into their ecosystem and mods. All things moving forward might have to go through them as a monopoly, which ironically enough...the same thing they accused me of doing a few years ago.


    Oh well, we will see how this goes.

    The vehicle has a total of 47.573 polygons: of which 32.664 belong to the base model, 14.153 to the doors (all of them are separate, including the trunk) and 756 to a single wheel. It is not a high-poly model, but rather a medium-poly one.

    Why Polish modders tend to create models this way? To this day, I honestly have no idea where that approach originated from.

    We (myself and Duszkarczyk) create not only the model itself but also a complete UV map for the entire vehicle. Because of that, smaller details can easily be represented through textures. Larger modifications, such as changing the shape of a wheel arch, have to be done manually by moving vertices.

    In general, this is usually easier than editing the texture, although of course it is not always the better solution.

    I see, so mid poly is under 100k, I understand your reasoning.

    I also dont understand why you guys model things that way. I'd just keep it under 10k or even lower and call it a day.

    Gotta be straight honest with you though, manually moving vertices sounds miserable. Personally I wouldn't have the patience or mental fortitude to do that.