Post AwMeULJ2JotNAT0Jn6 by lykrast@eldritch.cafe
(DIR) More posts by lykrast@eldritch.cafe
(DIR) Post #AwMboeIT7ZOvzojJ9k by asie@mk.asie.pl
2025-07-21T16:31:51.031Z
0 likes, 0 repeats
how am I supposed to select "1.4.7 - 1.7.10". guess I am no longer the target audience
(DIR) Post #AwMcpMnPm4Kvxh6vBI by wolf480pl@mstdn.io
2025-07-21T16:43:20Z
0 likes, 0 repeats
@asie outrageous
(DIR) Post #AwMdAOjblI6c3VMIHw by asie@mk.asie.pl
2025-07-21T16:46:59.327Z
0 likes, 0 repeats
This is in reference to a poll on "Stable Mod Versions" for Java Minecraft ( https://thepansmith.github.io/proposals/community-smv/ ). Some thoughts on that:This has happened in the earlier modding scene many times, and mostly with even greater incentives than "community agreement".- 1.4.7 had RedPower 2, and so a lot of users refused to move away from it. However, once Tinker's Construct went 1.5-only and removed its legacy 1.4.7 builds from distribution (they were pre-Smeltery, anyway), it became the new version's "killer app".- 1.6.4 -> 1.7 was likewise turbulent due to lack of mod updates, until Applied Energistics 2 became a must-have mod.There were other versions propelled by a "version seller" (1.8.x would probably not have happened in the modding community at all were it not for Thaumcraft 5 and Intangible), but my point is the pattern. Even when people had a strong "version seller" on the previous version - RedPower 2 used to be seen as the GOAT mod by many - the next version's challengers, vanilla features, and the chase for novelty usually put them all to the wayside.From a modder's perspective, it is always a safer bet to target whatever the latest version is at the time. History shows that you will have to perform the porting work anyway, eventually. As such, coding new features in a non-strictly-latest version is effectively ensuring "tech debt" for when the community finally moves on.This is the decision I pushed BuildCraft towards when it skipped 1.10.2 and went straight to 1.11, as it had changes which would have been expensive to port to had they started in an earlier one. BC's developers even received death threats for that decision (what the actual fuck?), and yet they were vindicated as by 1.12.2 almost everyone dropped the version they swore to stick to.Overall, this is why I think "stable modding versions" are unrealistic: it just gives an easy way for new modders to swoop in and create "a mod like yours, with slightly more novelty, and on a newer version". Combined with how many big authors rely on Minecraft modding as a side hustle...Even if, say, NeoForge itself were to stop updating to unstable modding versions, which would probably be the best way to try and enforce such a rule, someone would just swoop in with a new mod loader with slightly more novelty in response. Ask how I know.
(DIR) Post #AwMdTSgzGsKhpfDtJ2 by asie@mk.asie.pl
2025-07-21T16:50:28.250Z
0 likes, 0 repeats
There are two "stable modding versions" for Java Minecraft that have significant amounts of developer attention. These are 1.7.10 and 1.12.2. They primarily exist because, due to historical flukes and having attracted "modding auteurs", they accumulated so much unique content that it is effectively impossible to compete with its breadth without unrealistic coding investment. I expect the spell of the churn to continue unless Mojang breaks it again - be it by making a change so massive that the community takes a year+ to adapt (unlikely with the "drops" model), or by abandoning Java Minecraft altogether (unlikely with the profit margins).There is also no one modding "community". It's an illusion. In the olden days, there was a lot of novel work being done on forums like minecraftuser.jp that almost never made it out to the West, as well as unauthorized forks and ports of mods developed on 4chan that were not legal to distribute and thus not known by the wider scene. People who don't identify with the "community" have incentives to break the SVM rule (as outlined above) and no real downside to doing so long-term.As such, I recommend focusing on ways to lessen the porting load, such as wrapper libraries like McJtyLib, moving more of Minecraft's API surface to an abstraction layer, adding translation layers (like Intermediary) to allow older mods to run on newer versions, and the like, rather than trying to prevent porting. Even Fabric started with this in mind, as its idea (pre-"Drops") was that by allowing people to start porting late in the snapshot lifecycle, where breaking changes were restricted, we could adopt newer versions more efficiently and with less on-release churn - the "day one modpacks" became slightly famous for a while. (It didn't exactly work out, but that's another story.)
(DIR) Post #AwMdfOhbvoWivtCq12 by philip@mastodon.mallegolhansen.com
2025-07-21T16:51:41Z
1 likes, 0 repeats
@asie IMO the “solution” (I know it’s very far from trivial, but it is *possible*) has always been that the modding API needs to provide adequate abstractions to not tie mods to a specific game version.
(DIR) Post #AwMdwctaPmh1sJ1TxQ by asie@mk.asie.pl
2025-07-21T16:55:41.711Z
0 likes, 0 repeats
To be clear, this is very much "personal opinion" stuff. That's part of why it's not a blog post. Every modder's circumstances differ.
(DIR) Post #AwMe8N1vEDvS8yWQ1Q by lykrast@eldritch.cafe
2025-07-21T16:57:36Z
1 likes, 0 repeats
@asie "the community should pick one version to stick to" is such a common comment on everything despite how easy it is to disprove :neocat_pensive: (by just asking "ok which one?" with more than 1 people reading)it just gives an easy way for new modders to swoop in and create "a mod like yours, with slightly more novelty, and on a newer version"oh ye I was late on going to 1.20+ so someone made a fork of glassential (renewed) that got a lot more downloads initially :neocat_pensive:(I am slightly grumpy about it but glassential was just lifting stuff from xu2 so I don't feel justified complaining about that (well maybe for the textures but eh I've done better), tho it did diverge feature-wise as I reupdated)and also currently low motivation + all the drops stuff in 21.x sapped all my motivation hence why I'm still on 20.1 for now (+ alex caves + quark), porting is such a chooore
(DIR) Post #AwMeIp9nVXFuH3BrGK by asie@mk.asie.pl
2025-07-21T16:59:45.708Z
0 likes, 0 repeats
@lykrast@eldritch.cafe "ok which one?" is fun because you can interpret it as both "which version?" and "which community?". the GT:NH community has already picked one version years ago, for example
(DIR) Post #AwMeULJ2JotNAT0Jn6 by lykrast@eldritch.cafe
2025-07-21T17:01:02Z
1 likes, 0 repeats
@asie lmao truealso hey wait I develop on 1.20.1 I get to reply to the poll :blob_fox_bongo:
(DIR) Post #AwMgoisIbGctxtLMxs by wolf480pl@mstdn.io
2025-07-21T17:28:02Z
0 likes, 0 repeats
@asie btw. what was the version where Thermal Expansion had no pipes, so people were using ExtraUtilities and/or Translocator Mod for item transfer?1.6?
(DIR) Post #AwMguDIcuRhxm0HCCm by operand@todon.nl
2025-07-21T17:16:56Z
1 likes, 0 repeats
@asie version seller, i am going to mod and i need your strongest versions