Maybe move away from Forge towards NeoForge?

[MrJeremyFisher]

The Forge developers seem to have finally snapped and started making pretty severe breaking changes without any workarounds. Given that a lot of Forge mods already support NeoForge as well it is probably worth looking at fully switching over or at least offering it as an option until Forge gets their shit together or becomes not worth supporting.

[Protonull]

Just keep in mind that I did a poll about this a little while ago and NeoForge only got 1 response, whereas Forge got 19, and Fabric got 84. Though, I believe this was back on 1.18.2, so things may have changed substantially since there, but there may be cause to drop Forge entirely. Might be worth doing another poll before making this decision though.

[awoo-civ]

My gut feeling is everyone has moved onto Fabric by now. I personally would just drop Forge support as well.

[MrJeremyFisher]

The majority on civ definitely have. Here's a slightly more recent poll which was targeted more towards content mods but is still relevant. I'd be interested to see if Modrinth publishes modloader market share in any way.

[Protonull]

Here's a slightly more recent poll

That's actually an older poll (by creation date :P). It also seems to be a poll targeting modmakers rather than mod users with its question being "What modding framework will you primarily use when developing mods for 1.20.2+?" Even with Architectury projects, I still only really use the Fabric client. But this poll also has 2371 responses, so probably not a civ-poll. My guess is that civ-players tend to skew away from content-mods and towards pvp-mods.

[MrJeremyFisher]

Sorry, I meant more recent version-wise. I agree on the content mods thing but with NeoForge having such a large market share for content mods I imagine PvP mods will follow suit as they're often used together outside of civ.

[Protonull]

Just did another poll specifically for 1.21.1 loaders where Fabric received 53 votes, Forge 8, and NeoForge and Quilt both receiving 1 vote. So unless NeoForge is backwards compatible with Forge, switching to it may alienate Forge users more than if Forge support were just dropped entirely.

[MrJeremyFisher]

Yeah. The poll result are kind of surprising as I moved CR and Healthbar over to NeoForge because of (now solved) issues with architectury and saw more or the same number of downloads for NeoForge as I did Forge.

[Protonull]

Hmmm, it seems like NeoForge is backwards compatible with Forge, at least for now. I'm not sure it'd be worth the change though unless there's specific runtime or build-time features y'all want to use that NeoForge offers that Forge doesn't.

It's a fairly similar situation to when Civclassics was still a SpigotMC server. We eventually got it switched to PaperMC, but Maxopoly was wary of compiling against PaperMC unless there was a concrete reason to. And while the eventual decision to was an unambiguous success, it's only really because PaperMC offers so much more than SpigotMC, otherwise there wouldn't have been a point. Like, would there be any point in compiling CivMC against PurpurMC instead of PaperMC? Probably not.

[Protonull]

In my opinion, it's probably best to continue Forge support so long as it's not overly burdensome to do so. And even if Forge support is dropped, to nonetheless retain the general structure of the mod by keeping fabric and common separated, otherwise you may end up with a Synapse situation where half the battle is just disentangling the mod code from its mod-loader code. Who's to say how long Fabric will last.