forge-remarks.md

Preface

Since my recent decision to drop all support for the Forge mod loader, there has been many people asking me what exactly led to me that point.

As per the earlier write up (which you should also read), it largely boils down to the un-welcoming and borderline hostile interactions I've had with prominent Forge developers and their surrounding community.

Simply put, this document is the last note from me regarding the Forge community. I don't have any interest in continuing to fight with people to get basic concessions, nor in defending myself or my mods on a daily basis. Any changes or efforts to improve at this point would be too little, too late. Do it for your own community, not for me.

Remember the human

It's not my intention to encourage harassment towards any of the users listed in this document, nor is it to encite "drama" (whatever that is supposed to mean, since these are just public chat logs.)

Please keep in mind that everyone here is human and that harassment is unacceptable no matter the reason. The names below have been included because they provide meaningful context given the positions these people would be speaking from, not because you should be going after them. There is no place for disparaging remarks or shit-talking other human beings, and you will note that this page is free of it and speculation on motives.

---

Some thin veil of a timeline

February 15th, 2020

A user linked to me a conversation on the Forge Discord a while after it happened. Here are some excerpts.

This is where it started, after a user asked about my mods. I was not a member of the server when this happened.

Some users are skeptical of the claims that Lithium is lying or not doing what it advertises. I appreciate this support.

Lithium is fully modular, and every patch can be disabled in the user configuration file.

Almost all of the Mixins are thoroughly documented within the source code itself. You can check out some examples in DataTrackerMixin, ExplosionMixin, and PalettedContainerMixin.

A user cites some blurbs from the CurseForge page, as to explain what Lithium is doing more exactly.

In the next screenshot, some messages are included from a user called Earthcomputer. They have since apologized on their own accord, saying that their stance on performance mods has radically changed. In fact, they have actually gone as far as to make contributions to my own projects, which I greatly appreciate.

I understand the fact that mods which work around the Forge APIs have from time to time been a source of trouble, and why someone who has dealt with such issues might find them annoying, at least from their perspective. I make a best effort to ensure compatibility with other mods, and take extensive measures to mitigate issues when they come up. But, for the sake of explaining why these feelings around Forge exist for me, providing the full context of these messages is important.

At this point, I felt the conversation had become fairly hostile. I don't feel as if it is the responsibility of Lex, or anyone else in the Forge community, to "coax other mod developers". You are not the police.

This kind of mentality of "I don't have to be nice, I have to be correct" is toxic. It took Linus Torvalds many, many years to learn this reality. This behavior discourages contributions and creates enemies in the community.

A while after this conversation had concluded, I joined the Discord server following another user's invite link, and wrote the following response.

I was instantly banned for this. I couldn't join the server anymore. I had no idea why. The Forge Discord moderators claim I was never banned, but kicked. However, one of their moderators implies that I was evading a ban shortly after.

For this reason, I believe it was a ban initially, and that the community was lied to. Worse yet, to date, I did not actually violate any rules of the Discord server.

No reason was provided until after users fought tooth and nail for one. Later on, it is explained (not to me) that the reason I was removed was that I pinged another user. However, as you can see in screenshots above, I was not the only one doing this.

I don't really have much more to say on this. It's extremely disappointing, and was met with a lot of community backlash. Even worse, the moderation team's response to criticism was to remove anyone who opposed their actions publicly.

Edit: Since making this post, a number of people have come to me claiming that my belief that this was a ban is in error, and that it was only ever a kick. Some people have also taken that a step further to insinuiate (because of this unverifiable claim) that I'm only interested in creating "drama."

This argument is frankly ridiculous and is a hyper-fixation on superficial details. It ultimately looks like some attempt to grasp at straws in order to discredit every other documented incident here. Regardless of whether or not a ban was carried out, I was removed from the server for violating a non-written rule (being do not mention a specific user) following what would quite possibly be the most textbook example of when to use a user mention: directly responding to someone else's messages after the conversation had moved on.

April 25, 2020

The first episode of ForgeCraft, Season 12, went live on Direwolf's channel. About 6 minutes in:

LexManos (text): Is fastcraft or other 'performance' mods on?

cpw (voice): As far as I'm aware, Lex, there's no performance mods in here at the moment

tterrag (voice): Oh yeah we should get, uh, what is it? Is it Phosphor is the new one? Lithium?

cpw: So, uh, I looked at those and I'm not gonna put them on because they're a whole pile of bloody Fabric hacks

Direwolf20 (voice): Which mod?

tterrag: Uh...

cpw: They are. They're a bunch of Fabric hacks. So...

tterrag: What do you mean by that?

cpw: They're literally just a bunch of hacks ported over from Fabric.

tterrag: Uh, Ange- are you thinking of the same thing? Angeline made those by themselves as far as I know.

cpw: Uh... they're -

tterrag: Maybe some of them got added to Fabric?

cpw: They're... they're Fabric mods and uh, Lithium was-

tterrag: Initially, because we didn't have Mixin.

cpw: Yeah I know, Lithium was a Fabric mod and it requires Mixin added to Forge...

tterrag: Yeah, that's true.

Lanse505 (text): talk for yourself tterrag, I've got mixins working

cpw: ... and appears to be just a port of the Fabric functions.

tterrag: No. I can tell you that's not, just not the case.

cpw: Mixins can work, and I wanted to add them properly to Forge but uh...

? (voice): Mixins will be in Forge when Mumfrey decides to put 'em in Forge, that is... that is the sole reason he has commit access to Forge.

cpw: Yeah we were gonna do that, I was gonna work it with him on that and I ran out of time so...

Darkhax (text): She did them on Fabric becuase Forge didn't have them yet. Original versions were on forge.

tterrag: Yeah, exactly Darkhax. They were made for Forge initially but they, uh, Mixin wasn't working at the time, um, so they made a Fabric version alongside.

cpw: I dunno, I was kind of uncomfortable with adding performance hacks from Fabric into Forge

tterrag: They're not from Fabric! That's what I've-

?: I would also say we should avoid any performance hacks as a default because if they actually increase performance they would be properly documented and probably get into Forge itself.

cpw: Yeah.

tterrag: I mean, it's all pretty new, but...

nooblybear (text): and I'm pretty sure that the lithium for forge if there is one, would've been done from the ground up

tterrag: I mean, it's kind of a, y'know, it's the mod domain because obviously there's gonna be drawbacks with some of these but for people who wanna squeeze out performance.

cpw: Pahimar [unintelligible mumbling]

The video received 83 thousand views on YouTube. I find this particular remark rather troubling, and it'll come up a lot later:

... if they actually increase performance they would be properly documented and probably get into Forge itself.

I never reached out regarding these comments.

May 17th, 2020

One day, cpw made a multi-hour long Twitch Livestream on "Busting Performance Myths". Unfortunately, the VOD for this has been archived due to age, and I do not have an archive of it.

The primary focus of the livestream was about Streams, which were in introduced in Java 8, and whether or not were really that bad for performance. Again, throughout this stream, the repeated argument of Lithium being an undocumented mess and "Mixin/Fabric hack" is made. Additionally, it is repeatedly broadcasted that almost all of Lithium's code is rooted in replacing these Streams with more traditional code. This is not true, and a quick glance at the source code would have revealed this.

At the conclusion of the stream, the source code for all the benchmarks were published on GitHub. He left off on a note that contributions were welcomed, and that more complex code and benchmark numbers would be useful.

I submitted a rather in-depth pull request with additional benchmarks, trying to explain that the usage of Streams is more complicated than the repository showed. Despite the overwhelming community response to it, cpw never acknowledged the pull request, and the repository was never developed further.

Two things. VODs automatically delete after a month or so. I have no control over that, sorry. I am a terrible maintainer of GitHub repositories. It comes with being old, slow, and busy with a full-time+ job.

In regards to my comments that dire shared on his video, sadly I don't think your mods would be a good fit on forgecraft, since they would have a high probability of masking real mod issues, due to your significant replacement of core functionality. I don't have sufficient time to do a thorough code review of every suggestion that comes my way, and on the surface it looked like as I described. I am sorry for any offense or upset my comments have caused you. I don't mean to disparage new works.

Well this is sad. Hopefully some competition will get them off their behinds.

This is absolutely infuriating. This set of messages basically sums up how the Forge community reacts to any mod development done outside of the Forge ecosystem, and the most depressing fact is that this kind of mentality is supported and being spread by the most influential people of the Forge community. It's sad that it has to be this way, but hopefully Fabric will become competitive enough for these guys to get off their high horse. Full support for Jelly and her decision to cease updating the Forge versions of her mods.

Two things. VODs automatically delete after a month or so. I have no control over that, sorry. I am a terrible maintainer of GitHub repositories. It comes with being old, slow, and busy with a full-time+ job.

fair

In regards to my comments that dire shared on his video, sadly I don't think your mods would be a good fit on forgecraft, since they would have a high probability of masking real mod issues, due to your significant replacement of core functionality.

there's a difference between flat out calling a mod a hack/placebo and denying adding it on that basis.

I don't have sufficient time to do a thorough code review of every suggestion that comes my way, and on the surface it looked like as I described. I am sorry for any offense or upset my comments have caused you. I don't mean to disparage new works.

I don't see how this justifies public shaming of the mod. If you don't know what you're talking about then why would you publicly shame and demean it?

other than that, if the apology was genuine, then thank you for communicating it to the author. (there is no sarcasm in that sentence)

I do not support this type of acting. Will boycott Forge and start using Fabric from now on.

I know that you have been talking about how bad forge has been for you in your discord server, but I did not expect this! Forge shouldn't have such a good reputation.

Fabric > Forge just because of this 10 yo drama.

I knew that the forge guys were not nice.
I didn't know that they ware that bad.

Also, the Forge team just over-shame Fabric. I suppose (Someone should check this) that they saw one mod or feature that they didn't like, and they "forgot" to check back later to see if it was gone. Even now, that Mixins are in Forge, the whole Forge community is still against coremodding. In MCPE, addons and behaviour packs can change Mob AI! With the current state of Forge, they won't let any mod do it (Or at least they won't help a modder that tries).

This is horrendously messed up. I am so sorry this has happened to you, that they continue to lie about what they've done, and that they continue to make you their enemy - EVEN AFTER THEY WERE THE ONES TO ATTACK YOU UNPROVOKED.

I haven't been playing much MC lately, but my fav mods are mostly on Forge, and this would be making me think twice about playing modded at all.

Lex is a prick hands down, always has been. I'll be perfectly happy to play more on Fabric as it continues to gain adoption.

Started a forge server yesterday, thinking twice now. Not good to see this sort of obtuse behavior from the Forge moderators / admins.

Forge never actually supported the community - this makes it only worse... Personally I don't like the ASM-ness of Fabric, that's the only reason I stay on Forge, but one day I'll make a modloader like Forge with the freedom of Fabric (and of course provide support for Fabric mods)...

Forge never actually supported the community - this makes it only worse... Personally I don't like the ASM-ness of Fabric, that's the only reason I stay on Forge, but one day I'll make a modloader like Forge with the freedom of Fabric (and of course provide support for Fabric mods)...

I'm not sure what you mean by "the ASM-ness of Fabric". If you don't like the code of the game being changed, check out the forge repository’s patches and weep. Fabric, in contrast, has no "patches" by default, so no Vanilla behaviour is altered. If you don't like Mixins, they're a much better version of ASM in the sense that they're not ASM (check out the updated Mixin cheatsheet, it's super helpful for understanding them), and they're also officially supported by Forge now. If you don't like not having an API integrated into Fabric directly and you don't want to write your own API as a "coremod", check out the official Fabric API, which is by default included in the Fabric example mod template, and is pretty much used by everyone.

With ASM-ness I mean the fact that Fabric does not patch the source code but just has a default API mod that uses the same Mixin as all other mods. Forge actually patches the source code so you can navigate seamlessly between Forge and Minecraft code. That has many bright sides and it makes Forge a good and robust modloader... But now that I get a closer look on what restrictions the Forge staff imposes on the community I don't actually know what's better anymore...

LBNL Lex says about the maker of a performance mod that "the person has no idea what he's doing" - now I say that about the Forge staff...

It's extremely upsetting to see how toxic this community is and I'm sorry that you went through that. Moderators don't know what they're doing and the discord users' response to your initial ban/kick/whatever shows that very well.
The fact that they kicked you on an extremely irrelevant basis that you broke a "do not ping this person" rule and they completely overlooked a valid counter-argument shows they're not paying attention and would rather blindly enforce rules. While I understand that rule probably exists because of how vast the Forge discord server community is and there were cases of ping abuse in the past, the instant boot was extremely uncalled for given the circumstances.

And don't even get me started on Lex. I mean who the fuck behaves like this. This is literally why the coding industry/community is widely seen as toxic.

I've been fascinated about modding in minecraft for a while now, but seeing this idiocy further reinforces why I shouldn't ever jump into it, because I will probably also get incinerated like this guy did.

Wow... I always found the forge staff to be... quite rude after being on their forum for literally just 10 minutes.. But.. This is.. Wow.
Suddenly I'm starting to like Fabric a lot more. If only more mods would switch to it, or were made compatible.

Nice to see that Lex hasn't learnt a thing in the past decade.

I sympathize, I was kicked from the Paper Discord because I was trying to give constructive criticism by saying that I thought a rule seemed slightly harsh.

Wow... I always found the forge staff to be... quite rude after being on their forum for literally just 10 minutes.. But.. This is.. Wow.
Suddenly I'm starting to like Fabric a lot more. If only more mods would switch to it, or were made compatible.

Yes, luckily Fabric is getting more popular but most of the top mods are on Forge. Fabric limits your mod options by a lot, I just hope Fabric gets better over time so I can happily remove forge.

Yes, luckily Fabric is getting more popular but most of the top mods are on Forge. Fabric limits your mod options by a lot, I just hope Fabric gets better over time so I can happily remove forge.

@R3Birth13

We're starting to see a lot of big names coming to Fabric, like Botania and BuildCraft. AE2 already has a Fabric version and I think Tech Reborn is only Fabric.

For experienced modders Fabric is not hard to learn, but for a beginning modder Forge is basically more attractive because a mod is more structured in Forge. E.g. in Forge you initialize your mod when you receive your initialization event, you then register your blocks during the block registry event, your items during the item registry event, etc. etc. all the way until Minecraft has finished loading - it's easier to understand what should happen when. In Fabric you get like one event telling you to initialize everything. You then really gotta initialize everything, including the registry of blocks, items, whatever, in the proper order... Nonetheless Fabric has a better documentation, but I swear that if Fabric brings such a loading lifecycle as Forge has, Fabric would already become way more popular for beginning modders.

But at last:

• Fabric to a new modder: 'Welcome, we help you setting up your mod!'
• Forge to a new modder: 'Get away!'

I sympathize, I was kicked from the Paper Discord because I was trying to give constructive criticism by saying that I thought a rule seemed slightly harsh.

You might be confusing Forge with Paper, Lex does not seem to be in the Paper Discord, and Lex has also been known to criticise Paper in the past.

I sympathize, I was kicked from the Paper Discord because I was trying to give constructive criticism by saying that I thought a rule seemed slightly harsh.

You might be confusing Forge with Paper, Lex does not seem to be in the Paper Discord, and Lex has also been known to criticise Paper in the past.

@iCaitlyn I'm not, I'm just saying I had a similar experience with a different community (being kicked for no reason)

the fabric community has been toxic AF with me as well. The internet is toxic AF, if you are making mods you just have to deal with people being toxic.

the fabric community has been toxic AF with me as well. The internet is toxic AF, if you are making mods you just have to deal with people being toxic.

I'm very sorry to hear that. I've personally felt that the Fabric community has been much more welcoming than the Forge community.

Some communities can be nicer than others though

I knew that the forge guys were not nice.
I didn't know that they ware that bad.

Also, the Forge team just over-shame Fabric. I suppose (Someone should check this) that they saw one mod or feature that they didn't like, and they "forgot" to check back later to see if it was gone. Even now, that Mixins are in Forge, the whole Forge community is still against coremodding. In MCPE, addons and behaviour packs can change Mob AI! With the current state of Forge, they won't let any mod do it (Or at least they won't help a modder that tries).

Hi there, long time addon dev here, just letting you know, Mob AI on bedrock can be changed ro bizarre limits when needed to. I literally got a job at marketplace because of my talent. anyways fabric, ill just write my experience with the communities here, Fabric is nice, i just got into Fabric modding or java modding in general a while ago, i can say that Fabric community is very friendly, people helped set my stuff up, however the same cant be said for forge, i was abused and made fun of for getting into java modding after being one for bedrock, badly. Over past couple of days i have ported several chunks of "Luminous Depths". a project initially thought for Java, which got dev done on bedrock first when i arrived, the forge version is a mess, no one even wants to help and i get toxic replies, Fabric tho, is a good experience. i feel home while doing fabric, its like im reliving my experience with bedrock.

TL;DR - Forge Community will be toxic to you, even in the slightest of ways, Fabric will not. Same goes for bedrock even tho we also have external modloaders we never have conflicts like this, never ever.

Maybe that's why Optifine never got along with Forge

Maybe that's why Optifine never got along with Forge

Originally Optifine started on Forge, but they ended up using their own modloading system later for the same reason I guess - it was an optimization mod... Strange enough Forge now seems to put focus on supporting compatibility with Optifine - that makes it even more weird...

Damn, they're like the people in movies who kill anyone who gets in their way except that in real life they make you lose hope and make you feel bad about your creation i hope the creator of Sodium gets a lot of support and as one of the thousands of people that use sodium i can tell you that the creator did a very amazing job for us with potato PC's i support and i'll effing share this to my friends ✌️

Yeah, I always had this feeling about forge's discord, it just confirms it.
Sadly there is so much mods in Forge, and my server is using Forge for way too long now. It's impossible to go to Fabric. If one day Fabric catches up to all the mods that I use, and there is a solution for running plugins and mods, yeah, i'll probably go to Fabric. I absolutely hate this "I don't have to be nice, I have to be correct" this is such a terrible and toxic thing to say.
They are basically creating their own "ennemies" with Fabric and modders giving up on Forge. I can only hope they realize their mistake one day. But it might be too late now. Heh.

So I'm bumbling about after a 6+ month MC hiatus, seeing what the modding community is doing for 1.16. Hear about Sodium.. looks amazing. But it's not for Forge, so I start clicking random links trying to find out more information on Fabric. I stumble upon this gist (and the stance page leading here). After reading all this... gotta say.. I think I'm gonna switch to Fabric. I'm a long time Forge user/modpack builder (mostly private for a gamer group), and have never considered even trying Fabric due to mod availability (and me being lazy). But treating people the way the Forge people do.. is just dumb and egotistical, and I've ran into poor treatment myself many years ago on forge forums.

Drama aside.. I guess I should thank you Fabric people for giving us an alternative, and hopefully a much less toxic community. I'll be trying Fabric with Sodium/Lithium/Phosphor very soon.

People shouldn't feel superior if they use a certain modding API. People need to accept they are both decent modding API's. I personally don't use forge or fabric (forge too heavy, fabric only supports later versions)

People shouldn't feel superior if they use a certain modding API. People need to accept they are both decent modding API's. I personally don't use forge or fabric (forge too heavy, fabric only supports later versions)

Out of curiosity, what do you use?

I am sorry you had to go through that, you didn't deserve it.

Hopefully Fabric will grow until it proves that it is at least as good as forge, if not better. They have been at the top for too, too long.

People shouldn't feel superior if they use a certain modding API. People need to accept they are both decent modding API's. I personally don't use forge or fabric (forge too heavy, fabric only supports later versions)

Out of curiosity, what do you use?

Either decompile mc’s source code or use sponge mixins

People shouldn't feel superior if they use a certain modding API. People need to accept they are both decent modding API's. I personally don't use forge or fabric (forge too heavy, fabric only supports later versions)

Out of curiosity, what do you use?

Either decompile mc’s source code or use sponge mixins

Fair enough! By the way, as I've only just noticed this part of your message just now, Fabric does actually support (3 seperate links!) the use of older versions of Minecraft. The main issue is finding mappings to use with them, but there are a number of groups that have created their own mappings for older versions of Minecraft with Fabric, and you can actually use old MCP mappings with Fabric if needed. Off the top of my head, check out Legacy Fabric for 1.8.9 mappings + an API (they're also working on 1.12 and 1.13), and Minecraft Cursed Legacy for 1.2.5, Beta 1.7.3 + an API, and Alpha 1.1.2. Modification Station is also working on another API for Beta 1.7.3. Someone was also working on c0.30 at some point as well I believe.

Wow... Im not going to lie this actually made me angry. Playing the game in 1.14+ would be impossible without using a combination of your mods jellysquid.
It has been tested in depth and approved by several of the most advanced tech servers. (although we do find bugs from time to time)

But of course how could we lowly insects even begin to compare ourselves to the almighty forge council with its leader lex.
The fact that lex only ever applied reflexive thinking on all of his statements really makes you take a step back.
To anyone who is currently supporting this moderation team in the forge community and this abusive system:
You should be ashamed of yourself.

Instabans seem to be the default reaction to certain people joining that discord; I got the exact same treatment - though in my case they did admit it to be a ban - last year when I joined shortly after Lex broke the gradle/maven dependencies for every version of MC/Forge before 1.13 (including 1.12, which was the standard at the time), and seemed entirely disinterested in fixing it. (For the nontechnical: that means it was impossible to compile mods or create new workspaces for any MC version before 1.13.) Someone mentioned I had joined, and boom, I was gone inside ten minutes.

In other words, once the Forge team decides they do not like you, there is no returning from that point and any presence of yours around them will be stomped out as fast and as hard as they can. Neither are appropriate from someone in their position, and while I always personally only ever care about the product (in this case Forge) and not the author(s) when I choose what I find useful or enjoyable, I certainly understand the desire to bail.

As a modded player I do have reservations about splintering the community, but in all honesty it is not like this is the first thing to do that. Hell, I am guilty of that to some degree given the fact I still mod for 1.7.10.

In a nutshell, Forge is a community made up of a cocky leader, a handful of competent mod-makers and the rest 10 year-olds who can't stand that mod developers dare to move out of the eco-system and the shit coding management. Fabric is the developers dream, a ecosystem where coding doesn't get changed every generation, and actually has proper documentation that most people can actually understand. Fabric is the supportive community but with a small ecosystem, while Forge is the ecosystem of 10 year-olds who don't know what they are doing (mostly, some people are nice, and provide to the community)

Recently, I tried using Forge, but the latest renditions are so unstable they keep crashing. Have moved over to Fabric and am currently running 80 mods without issues. With Forge, it would have slowed to the point my FPS would be chunking at 10. It took some time but I've switched. Thanks to the Fabric community for all your hard work. Peace.

Just read this, and saw many comments saying that there aren't many mods on Fabric. More importantly, I have no modding experience myself (never written any) but some of the Forge users are kinda rooted in Forge. Just saw a comment on a CurseForge page where somebody said "Make a forge version because fabric sucks forge=♥" and they then said, "Never used fabric btw, but it must be terrible." And I came and read all this and I feel so bad that anyone would have to be abused to make a mod.

Anyway, I play on 1.16.2 and have 112 mods. That's a lot of mods, especially considering that ALL of them are clientside. I personally use A List of (Almost) All Fabric Mods to get my mods, and I gotta say, I appreciate the dedication the person who made it has. Every mod I play with I got from there (except Fabritone because there's no official release for it on 1.16.2 yet so I got it from the jobs on the GitLab). Forge (1.14.4) starts stuttering horribly after about 3 hours of gameplay no matter what I do, and I have to restart the game for it to be playable. I will still use forge (if I find a really cool mod that's forge-only) but otherwise, I hope that the community will move away from it and go to Fabric instead.

Just read this over and realized that the train of thought here seems to have faulty connectors, but anyway, I hope you make good use of that link. Have a delightful something!

When you have a track record for acting like this, you make it really hard to want to develop a forge mod like I've been wanting to do. I... think I'll save myself the stress. That's not to say I will never work with Forge, but if I feel just as likely to get banned from a server for writing this comment (at least if Lex ever sees it) as I am to actually succeed in developing something, that should tell you something about the state of your community's leadership.

what an awful mindset. They discourage contributions that are not directly to their own project, and react poorly towards any performance work that's done outside their platform. Other developers aren't entitled to their platform, and they're not required to develop for everyone, but this level of hostility is just atrocious.

Sorry you received such an awful response from anywhere. That's just not on, but is especially unwelcome in what is supposed to be a collaborative environment. Surely when people take a step back the aim should be to have the best versions of Minecraft that we can for us to play?

Speaking as someone pretty new to Minecraft (Ok, I bought it about 10 years ago but never got into it, this year with lots of time at home...) but not new to gaming or modding, I have been trying to feel my way around the two systems, Forge or Fabric and work out which to go for. They both seemed workable and the comparison of older and established vs. newer and faster but with less choice was a pretty even call.

I think I just made up my mind.

I won't miss the variety of stuff on Forge as I never got used to playing with any of it, and there seems to be plenty of choice on Fabric anyway. I have been playing with All of Fabric 3 for the last few weeks and enjoying it, and now putting together a collection of mods I picked to enjoy for myself.

I've never used Fabric, but about this, I'm switching to fabric. I mean, forge has slower fps. I used 12 mods, 52 fps, 12 mods on fabric, 78 fps. Forge sucks tbh.

I'd recommend fabric.

I've never used Fabric, but about this, I'm switching to fabric. I mean, forge has slower fps. I used 12 mods, 52 fps, 12 mods on fabric, 78 fps. Forge sucks tbh.

Well, forge doesn't suck but fabric definitely runs better. If you don't believe me, talk to the consistent 78 fps I get with 123 mods on fabric at once.

I hope Lex is happy that his III-A rated skull screwed over the 1.16 Forge scene out of performance gains that we could really, REALLY use.
I seriously wish this didn't have to be the case but if it were me in your situation I honestly would have done the same thing. This is just shameful.

I'm salty that I have to put up with Bad Old Days-level performance from even the best-implemented packs on 1.16, but until Lex gets some freakin' therapy, or people higher up in the Forge project start standing up to him, I don't see any of this changing.

In case he reads this: Lex, I love your work, you've obviously been an incredibly important person not just for Minecraft but for modding as a whole, but your attitude needs to dial back severely because you're poisoning the community you've fostered.

Certain projects attract certain people. The attitudes that allow someone to start the project also show themselves in this toxic "I am always right" mentality. If people can change their attitudes, all the better, but even if they don't, the project can still be great.

I feel like I can justify cpw's behavior as commentator fever rather than intentionally being condescending. I hope Lex changes, but I do think I'll be using Forge for the foreseeable future for various reasons. I, as a Forge user, am sorry that Forge has alienated so many members of the community (including myself to a far lesser extent).

What can i say. We are developing a Java on Android launcher, and, of course, the performance limit is VERY tight. Forge is way slower, even in terms of startup times. It takes around 10 minutes to load a 112-mod pack for Forge 1.12.2, and like 3 minutes to load 100-mod Fabric pack on 1.16.3.

Also, about Sodium/Lithium/Canvas mods for Fabric - idk if they are dirty, i never checked the sources, but they do way better job in terms of making game faster than Optifine.

What can i say. We are developing a Java on Android launcher, and, of course, the performance limit is VERY tight. Forge is way slower, even in terms of startup times. It takes around 10 minutes to load a 112-mod pack for Forge 1.12.2, and like 3 minutes to load 100-mod Fabric pack on 1.16.3.

Also, about Sodium/Lithium/Canvas mods for Fabric - idk if they are dirty, i never checked the sources, but they do way better job in terms of making game faster than Optifine.

@artdeell Also make sure to install phosphor, and you're reading a gist from the person that made all of those 3 mods (sodium/lithium/phosphor) lol

Goddamn if most of the mods i use had fabric versions i'd be changing. Shit.

I can't believe this Gist is still getting comments lol

This is unfortunate, however I think the larger picture isn't being considered. I don't think people should be treated like garbage, but all of this is just going to eventually desaturate the community. I've got my opinion and I'll save some space on this Gist as I can talk a decent bit.

My opinion on all this.

It's quite a shame to see such a profound divide between the two mod loaders. This is a community afterall and I have two major issues with this collectively. Firstly, both developer groups of these mod loaders, providing they actually care about the game they are developing for, should be spending more time making everything compatible and stable, rather than complaining about tiny things. Justified or not, the blame isn't on a single developer or a single team. All of the Forge developers aren't toxic, and all the Fabric developers aren't angels. These blanket statements are toxic, in of themselves.

Secondly, Saying that 100 mods load faster and perform better on one platform vs. the other and making a definitive decision on which one is better, is one of the most illogical constructs the brain can create. Every single mod, regardless of the mod loading API is different and puts a different degree of demand on a user's system. I know the comparisons made here aren't from mods that are all both supported on the Forge and Fabric API simultaneously with the exact same version numbers for each mod. Mekanism and Mouse Tweaks have very, very different hardware requirements to run because of the different mechanics and elements they add to the game, just like how Zelda II and Cyberpunk 2077 have different things. It isn't about the number of mods, it's about the volume of content those mods have and how it is handled. Things could work well with a small volume of content, but when you have tons of cross mod integration happening in the background, like a lot of Forge mods have had for many years, things become combersome and that's something that shouldn't be forgotten for anyone who has used a 'compatibility mod', of which most major Forge modpacks have several.

Ultimately, this is a game that is spread out across a very large community and age group. It is a game that is driven by community members that come from all different coding backgrounds and skill levels. Forgetting that, causes toxicity too. If you see someone white knighting for one over the other, it could be that they aren't looking at the situation objectively, rather, they are blindly stating opinions without justification. This is just a video game, I love it as much as the next person, but from an end user, despite doing basic modding work myself, making a modpack is more confusing these days, than it ever has been simply because a few people from both sides of the core API teams refuse to find some basic middle ground and work things out causing a stronger divide in the community. How is that... also not a problem?

Afterall, that's the only reason I'm here, usually I only use GitHub to post bugs, but I'm getting tired of having to tell people we can't use this mod or this won't work, when I work with others to consider making packs because the community and now even developers are picking sides. So I went looking to see why things are the way they are, and managed to stumble here. I'm sure I'm not the only one and I likely won't be the last. Competition is wonderful, up until there are so many options that are doing things "right" that none of them work with each other at all.

I, and I'm sure many others don't want to believe that 1.12.2 will be the last major version where people just had fun. Nearly everything was developed for one platform and you could get modpacks of over 500 mods working and be stable and fast with some small tweaks. I have personally done this, although it is quite RAM intensive, but that's to be expected. It was wonderful, and you had basically anything and everything you wanted. Now... I can't play with certain mods that I have nostalgia for, like Tech Reborn which was an evolution and simplification of the combination of Industrial Craft 2 and GregTech. I can't just throw mods together and have it only crash because of actual bugs, now it can be because you have a mod from one platform or the other installed. To put it simply.... There is a start to the split of a community because people pick teams, despite us all wanting the exact same thing? Playing Modded Minecraft? How does that make since?

To the major API mod loader developers, if by some chance any of you somehow manage to see this, despite what team you manage to be a part of, is doing any of this actually improving people's experience to making and playing a modpack any better, in this moment or even in the future? People will always have their opinions and they will always think one is better than the other, but isn't it easier for everybody to work on the same thing and be considerate of the contributions that are made to that project and try to make fair and equal opportunity for development for everyone? You wouldn't need to go policing, you wouldn't need to overly explain things to community members, and you certainly wouldn't need to have all the extra stress, irritation and wasted thoughts on people and projects, because it wouldn't happen much if you were all open minded, and respected the skills you all have, that a huge number of mod devs and an even larger number of community members, don't have. If all of you would get together, and collectively make this video game as good as it can be, you wouldn't need to deal with all of this. If something is better coded this way over that, explain in a decent way, provide examples and hard, cold, logical proof that it does this better than that, accept the truth, implement the best solution, understand that none of you are the best at this and move on to continue making something great.

It all sounds like a pipe dream, and one that I would love to be proven wrong on when it comes to the fact that it will never work out.

Response to @DarkAmbientNoise

The Divide

I can't play with certain mods that I have nostalgia for, like Tech Reborn which was an evolution
...
There is a start to the split of a community because people pick teams, despite us all wanting the exact same thing? Playing Modded Minecraft? How does that make sense?
...
I'm getting tired of having to tell people we can't use this mod or this won't work, when I work with others to consider making packs because the community and now even developers are picking sides.

Fabric (also Rift) was created because Forge was slow to update. You could say that it was Forge that caused the divide by updating :). On a more serious note, see below.

I would want one, central modding API that every mod can run on, but unfortunately, Forge and Fabric developers' code philosophy is a lot different. Forge believes in making mods compatible and covering every single edge case, while Fabric believes in clean, modular code that only has the functionalities and compatibility tools required for most mods.

On top of that, Forge's Gradle Plugin doesn't work on the newer build tool versions.

Patchwork seems like a temporary solution for the Forge/Fabric problem. But it will never be able to run all mods - those will require ports. Now, continue thinking. What about old, abandoned, ARR mods? If a new Minecraft version happens to release, those mods can't be updated by someone who is not the original author (and there are a lot of ARR mods, although most of them aren't abandoned). It's the same for porting and updating - why not stop using these mods now instead of stopping to use them when 1.17 releases? The Searge names aren't that consistent between versions - almost every mod will break even if the area of code that it uses isn't changed.

Toxic Developers

Justified or not, the blame isn't on a single developer or a single team. All of the Forge developers aren't toxic, and all the Fabric developers aren't angels.

Partially agree. Not all Forge developers are toxic, but almost all of the core Forge developers that like to be present publicly are the toxic developers (LexManos is one notable dev). Fabric's lead developers (Player and modmuss) just try to stay away from all of these Forge/Fabric things and focus on developing.

Performance

Saying that 100 mods load faster and perform better on one platform vs. the other and making a definitive decision on which one is better, is one of the most illogical constructs the brain can create. Every single mod regardless of the mod loading API is different and puts a different degree of demand on a user's system. I know the comparisons made here aren't from mods that are all both supported on the Forge and Fabric API simultaneously with the exact same version numbers for each mod. Mekanism and Mouse Tweaks have very, very different hardware requirements to run because of the different mechanics and elements they add to the game, just like how Zelda II and Cyberpunk 2077 have different things. It isn't about the number of mods, it's about the volume of content those mods have and how it is handled.

Then try to run Forge without mods and Fabric with only Fabric API (which is equivalent). See the results that way.

Among the reasons that Forge performs worse, some are very simple; It fires more events. There can be multiple events per tick, and a lot of them. That, by itself, isn't that bad, but all of those events going through the event bus's reflection and ASM hacks makes it even slower (funny as some Forge devs criticize Fabric mods of being hacky - just see what EventBus does).

Patchwork's fork of EventBus (which I like to call EventRacecar) is a lot faster because they removed all of these reflection and ASM hacks. It just proves my point.

Install and Update Process

(this is not in response to any of your points)

Forge's modifications to the game are based on patches, while Fabric's are based on ASM and Mixin.

Both ways have advantages and disadvantages, here are the (dis)advantages to patches that I can think of:

+ Easy to read
+ Can apply statically
- Can't apply dynamically
- Requires separate patch generation
- Requires decompilation in order to apply - we can't be sure that ForgeFlower will generate the same output on all machines
- A lot less dynamic, if part of the code changes the patch will have to change (sometimes almost entirely)

And here's for ASM/Mixin:

- Hard to read
- Difficult to figure out how the code works, to grasp the idea of ASM
+ Very dynamic, most of the times changes to a certain file don't require changes of the Mixin/Transformer code at all!
+ Can apply dynamically
- Can apply statically but it's difficult to set up

(To be clear, this is for changes to Minecraft within the mod loader/API itself, not in loaded mods)

Forge is less safe to install - it can break.

Modifications to Minecraft

Fabric tries to stay as true to vanilla as possible - they even went as far as to bikeshedding a fluid API for 2 years just for it to not have milibuckets (1000 isn't divisible by 3). Actually wait this is arguably worse than what Forge's doing

Forge does everything to make it as easy-to-use as possible, breaking Minecraft mechanics in the process. Why do you think that all the technical players prefer Fabric?

---

Anything I missed? Just comment, I'm subscribed to this Gist.

@DarkAmbientNoise - I would argue the bigger picture is being considered. Forge has went without competition for a long time, and that has let a lot of shenanigans happen. But now there's an alternative, and for me, it runs much better. Like, not even exaggerating. Yes, per mod performance will vary, and yes, I've had to remove a couple Fabric mods that had something wrong/unoptimized in them that tanked my performance. Fabric isn't some wonderful magic fix for everything. That said, for me, Fabric 1.16.x ran better and loaded faster than Forge 1.16.x before adding mods, and continued that trend after I've added over 150 mods. I wasn't able to add many mods to Forge 1.16.x (even being careful about what mods I selected) without FPS and load times going out the window. Strictly as an end user and modpack builder, that performance difference is real, and it's a huge deal to me. It is a real downside, losing access to Forge-only mods.. but that won't always be the case, and many of those exclusive mods now have Fabric alternatives or Fabric ports. Also, yes, splitting up available mods to 2 different loaders isn't ideal. But I personally feel it's worth it in the grande scheme of things.

This whole discussion really isn't about teams or pipe dreams. It's about going with what works, which includes being able to choose not to deal with "toxic devs". Being dismissive because "not all devs are good/bad" and "not all mods are equal", isn't actually adding anything constructive, because that argument goes in both ways. Not all Forge mods are good, for example. Not all Fabric mods are performant. I would like to suggest that you build yourself two 1.16.x modpacks, one fabric, one forge, and see the difference for yourself. Include optimization mods, and give each loader a fair test.

@YTG1234 - Thank you for posting all that. I learned quite a bit from it.

After reading this, I have to wonder how I ended up just... not having the server in my server list one day, a good deal back - probably at least a year by now. I had joined the server and one day it just wasn't in my server list anymore. Still not sure why. I wasn't very active though, that might have had to do with it.