1.21.6 Snapshots

[timid niche]

there's no telling when VV will happen or when 1.22 will happen

[round briar]

But if the sign is that it is, I'll take that back 😄

[muted chasm]

the rendering refactors are happening now

[timid niche]

and this isn't a "mojang knows but we don't", it's been made clear in the VV discord already

[muted chasm]

they dont even know what backend theyre using yet

[pliant adder]

Q: When will Java Vibrant Visuals come out?
A: As soon as it is ready. Definitely not in the next release. We have a lot to do, and we are not even close to being able to reasonably estimate the release date. But it will not be a "Big Bang" release. Features will be released when they are ready. For example, octree-based frustum culling optimization was released last summer and extended cloud range and GUI rendering improvements were released recently.
Quoted from the VV FAQ, there will not be a single VV update

[mossy narwhal]

right that too

[muted chasm]

i mean i wouldnt consider that VV

thats the setup for it

vibrant visuals the texture update is likely to come in one update cuz the textures r already made

texture parity finally

[timid niche]

the best options now are trying to push devs to always be on latest (makes modpacking annoying) or just randomly looking at a version and saying "hey modders, support this"

[muted chasm]

the former isnt feasible, modders simply wont go for it

updating every 3 months isnt possible for many people

[timid niche]

well, the former is logically reasonable in terms of making it so you don't have 5 million changes to deal with all at once

[muted chasm]

i think if u make 1.21.6 LTS and then 1.22 comes out in the fall u can be like "never mind 😛"

cuz thats the obvious choice

[mossy narwhal]

the project being latest-first doesn't prevent having effectively-lts versions

[timid niche]

yeah

[muted chasm]

right but we get all kinds of people in here asking what version to use

and theyre told 1.21.1

[timid niche]

1.21.6 should basically be lts for now™️ and stabilized once transfer api is added in

[mossy narwhal]

we don't need to change anything project-wise, other than our stance when it comes to announcements and recommendations

[timid niche]

also 1.22 is probably not coming next

[muted chasm]

correct but right now there is no "official" stance

[timid niche]

it's incredibly unlikely

[muted chasm]

and yeah i doubt 1.22 is soon

this is the equivalent of a major update release date, its the right time for a new LTS

[timid niche]

yes and the problem with no official stance is it'll depend on what bigger mods do

[muted chasm]

every summer sounds plausible to me

[timid niche]

which is... not great

[muted chasm]

which is horrid

[timid niche]

it's quite horrible how modders decided not to port until after create did just because it's one of those big mods

[muted chasm]

so we should decide an LTS and then get create on it ASAP

create switching to neoforge is one of the big reasons players did

[timid niche]

well, create isn't the only large mods

[muted chasm]

they hold too much power lowk

[timid niche]

there's a bunch of other ones that are basically essential for players to even decide to use a version

ex: recipe viewers, the bigger storage mods etc

[muted chasm]

at least in my playerbase, it was one of if not the deciding factor to switch, cuz players dont otherwise care

[timid niche]

as far as create goes, (nothings final yet) but it's incredibly likely we will port to 1.21.6 once it's stable, drop 1.20.1, and likely get rid of registrate in the process

i've considered it a fair bit and going back to vanilla based registration is quite a good idea

[muted chasm]

god yes

i fully support switching supported versions to 1.21.1 and 1.21.6

that follows the normal timeline we used to

[round briar]

Alright, tbh the indication on VV / 1.22 kinda changes my opinion to some extend

I'd hope for sooooome guidance on what to expect from mojang but we're unlikely to get that for product reasons heh

[muted chasm]

yeah its kinda pissing me off atp

we dont need their full release calendar for the next 5 years but i would like some indication for the future of the game we're all heavily invested in

[round briar]

Ah, I should specify

My opinion on some "LTS" (well it's soft LTS anyway) before 1.22

[muted chasm]

yeah i got that lol, i just also agree with your side comment

[round briar]

Yeah I just read my message again and noticed I didn't actually say what I changed my opinion on 😄

[muted chasm]

honestly 1.21.1 is a soft LTS atp, not much would change

cuz neo has said mods should stick with 1.21.1

so this would just be them saying "now update to 1.21.6"

this would need to come with a relatively short BC window i think so people can safely update

[round briar]

I can never remember whether it is setup -Pupdating -Pupgrading -Pupdate or whatevers

[pliant adder]

I think it's the first

[round briar]

Ah yes it is

Okay... I must have fucked something up. The reject count is massive

Ugh

[uncut grove]

its Pupdating i remember cause ive always read it as pup-dating

[round briar]

Did we fuck something up about the toolchain?

I am seriously puzzled by why some of these hunks are being rejected

++++ REJECTED HUNK: 1
@@ -52,6 +52,36 @@
         }
     }
 
+    /*
...
+     */
+    public String readUTF(String data) {
...
+    }
+
     @VisibleForTesting
     public long getUsage() {
         return this.usage;
++++ END HUNK

[grim surge]

@steep verge

[round briar]

The patch is for NbtAccounter, and this is the change to NbtAccounter in pre2:

I don't even see an overlap in patch context?!

Is the line shift immediately causing it to become a reject? I thought we had some line fuzzyness

What... It actually applied the patch anyway, and then still wrote the reject

Have I completely forgotten some aspect about our toolchain?

[pliant adder]

Don't we have that fancy "porting" action? Could check that for the commands

[grim surge]

wdym

[round briar]

Yeah, I thought updating enables fuzzy matching

and apparently it does, since it did apply the patch

but why did it still write the reject

[idle wigeon]

I felt like the first few weeks of 1.21.1 had a lot of "oh, this is going to the next version" energy, before the big mods had updated. Though that's pretty rare.

[steep verge]

But but I have been married for the last five or so months

[idle wigeon]

I am surprised how little pressure there's been to update to 1.21.5. I thought that was going to be a much more popular update.

[twilit forge]

people keep waiting for 1.22

[round briar]

I am so stupid.

[twilit forge]

but like

[round briar]

I keep running gradlew setup -Pupdating.

[twilit forge]

1.21 was june 2024

1.21.6 is june 2025???

[round briar]

It has to be gradlew setup -Pupdating=true

😐

[pliant adder]

Oof

[round briar]

I should investigate again at some point if we can make that an explicit task, but there was some reason we coudln't (I think specifically since it's all wired up via task dependencies that doesn't work)

Oh that's a good question....

Mojang added the slot hover effect rendering here:

That's where we had our patch for emitting the ContainerScreenEvent.Render.Foreground event

Would you expect that to be emitted before the slot hover effect is drawn? I better go and check where that was in relation to it in 1.21.5

okay, was after too

Oh that would have been a very very sneaky patch porting failure:

[pliant adder]

Wouldn't be the first time 😄

[round briar]

yup 😄

[round briar]

Done

[pliant adder]

Perfect, I'll PR my stuff in a few minutes

[lusty laurel]

https://bugs.mojang.com/browse/MC/issues/MC-262974
This bug doesn't seem to be fixed, unlike NeoForge 21.5, is it normal?
(https://github.com/neoforged/NeoForge/pull/8)

[twilit forge]

the ticket says that mojang has confirmed it is a bug but haven't fixed it

[lusty laurel]

But a fix for the bug is implemented in NeoForge 1.21.5

[twilit forge]

yes

[round briar]

I am not sure what you're asking lolo 🤔

*I am not sure I understand

[twilit forge]

it's normal for neo to fix bugs confirmed by mojang, especially when the bugs interfere with things that modders might want to do

[lusty laurel]

Why isn't the fix implemented in port/1.21.6 if it works in NeoForge 1.21.5?

[round briar]

Oh that's the question

It should be

If it's missing, it may be due to changes of the underlying classes, or the patch was missed

[twilit forge]

if the fix is missing in neo 1.21.6 then you should have started with that

[round briar]

I.e. if the system is changed significantly and this is probably locator bar related

fixes like that will go missing

[twilit forge]

I think we both interpreted this as "The bug doesn't seem to be fixed [in vanilla]"

[lusty laurel]

yeah, mb

[last dragon]

pinged?

Ah Pre Release 3

[round briar]

We shouldn't forget about the pipeline modifier stack for guigraphics, but we can just push that to the port branch (and submit the commit for review)

we're still going to squash once more

[round briar]

@lusty laurel fixed

thanks for reporting

[timid niche]

they've been as clear as they can be

they have no idea when 1.22 will happen yet

on a side note, are port/* branches published anywhere?

[round briar]

the PR publish bot publishes all PRs

(caveats apply)

See the comment on the draft PR, it also has links to the installer

[timid niche]

ah thanks

[uncut grove]

https://www.youtube.com/watch?v=aixxTVqeAto

[burnt creek]

I think it will come when vibrant visuals comes that’s my guess

[pliant adder]

VV will not be a single massive update, Dinnerbone reiterated on that on the VV Discord: #1362331988803194983 message

[burnt creek]

Is that discord thing supposed to show something all I see is #unknown

[crude echo]

yes its a linked message from the vv discord

[burnt creek]

Is the vv discord public?

[crude echo]

yes

just click the unkown, if discord is good, it will link you to the server

[burnt creek]

Discord is bad🤣

[crude echo]

!pings

[soft swift]

Please disable pings when replying to others

So you don't accidentally break a rule or annoy others, make sure to turn off pings when replying to others. Discord currently turns this on every time you start a reply, so please be vigilant. Thank you.

​

Go vote on Discord Feedback so they make changes.

That trick is promoted. Consider using </info pingreply:1144973302809501716> instead.

[pliant adder]

The VV server is not fully public, you need an invite link. Even if they wanted it to be fully public, I don't think it would satisfy the requirements

[burnt creek]

Ohh ok

[mossy holly]

he was stabbed

[rustic nebula]

if 1.21.6 ends up being lts, then for how long will 1.20.1 remain to be supported for?

[sharp root]

1.20.1 doesn't even have Neo

[rustic nebula]

i'm aware

but there is still a support channel in this server

[round briar]

Hmmm, I don't understand the question, yeah 🤔

The support channel is community driven for the most part, so I'd say there's no reason to close it if there's still participation. Regardless of what we consider a major version for mods

[cedar night]

sry wronh channel tho

but had to post it

[crude echo]

I can se the message
but at the same time, it is a spam copy paste message.... to the point where it will lose its meaning and will likely increase hate even more

[round briar]

@pliant adder I wonder, do we have enough open issues to do a GH issue list on the PR? One open issue I remember is doing a GUIGraphics render pipeline modifier stack

Do we have anything else we shouldn't forget?

Besides testing, obviously

[proud tusk]

What I have done in the past is pinning messages in this thread

[round briar]

Well sure, but we do have a draft PR in the public repo now, we can use that

(There's already too many PINs in here and they're not immediately obvious)

But I just wanted to check what current TODOs we actually do have... oh... we should actually search the code for TODOs we left, I guess

[proud tusk]

Oh yeah what's with all these pins 😅

[round briar]

I think pins are too well hidden in the Discord client. Might be a "me" problem though.

[proud tusk]

The PR is a fine place for todos

[round briar]

BTW I think I didn't say: the PR is now up to date with pre3 and the published PR build as well

I made a list and added a few of the points we already fixed

I'll try to check for porting TODOs we might otherwise forget later

[pliant adder]

There are a few porting TODOs and we have a rendering bug in a couple UIs where the background is completely black

[round briar]

I'll try to do some QA later

transfer will have to wiat

[soft swift]

Please disable pings when replying to others

So you don't accidentally break a rule or annoy others, make sure to turn off pings when replying to others. Discord currently turns this on every time you start a reply, so please be vigilant. Thank you.

​

Go vote on Discord Feedback so they make changes.

[pliant adder]

@round briar did you get a chance to look at the porting TODOs yet?

Also, I'm considering adding ValueInput/ValueOutput-based serialization to ItemStackHandler and EnergyStorage and potentially a ValueIOSerializable interface similar to INBTSerializable as a generic interface for things serializable via ValueInput/ValueOutput. Opinions, especially with regards to the ongoing transfer rework?

[proud tusk]

Whatever you do here will have to be carried over to the transfer rework, but I think it's a pretty easy change

Is there even a point in keeping INBTSerializable though?

[frosty garden]

Yes definitely yes

[pliant adder]

Debatable. Making something that still uses raw NBT tags use value IO is simpler than the other way round. I did that in my FramedBlocks port in the BE update tag/packet methods to unify the whole thing since those still provide/return raw NBT tags. Therefore I wouldn't be opposed to nuking INBTSerializable

This is what getUpdateTag() looks like in my implementation (writeUpdateTag(ValueOutput) is a custom method):

@Override
public final CompoundTag getUpdateTag(HolderLookup.Provider provider)
{
    TagValueOutput valueOutput = TagValueOutput.createWithContext(ProblemReporter.DISCARDING, provider);
    writeUpdateTag(valueOutput);
    return valueOutput.buildResult();
}

[frosty garden]

Ah did Mojang not update everywhere and only the read/write from disk? There’s no stream codec for value access if I’m remembering correctly

[pliant adder]

Yeah, unfortunately only from/to disk uses value IO, writing for networking still uses raw NBT tags (specifically getUpdateTag() and getUpdatePacket()). Reading for networking is patched in by Neo and uses value IO

[round briar]

I looked at it and decided against it for the time being. It doesn't work super well for single-value serialization, which Tech pointed out to me

At least I believe the caps have the ability to serialize to primitives, but I didn't look any further

[frosty garden]

The purpose of value access is object serialization, so that makes sense

[round briar]

If you really just meant as methods of convenience for the provided item handlers, why not

[pliant adder]

Only EnergyStorage currently makes use of that and, in all honesty, it's intransparent garbage, it should never have been implemented as serializing directly to an IntTag.
Converting ItemStackHandler to value IO is pretty trivial with a little trick, converting EnergyStorage is going to be a major pain in the ass if it should be able to read data saved with the existing INBTSerializable implementation

In fact, I can't see any way of converting EnergyStorage to value IO without breaking savedata compatibility

[viral root]

IMO this is a major version bump anyway, can Neo use DFU? If not, then most people treat versions as save incompatible anyway

[pliant adder]

Neo can technically use DFU, I think we do in fact have one or two datafixers. But Neo being able to use one doesn't help mods here since the data in question can be nested in an arbitrary place and is also based on an NBT key that is completely controlled by the caller (i.e. a mod's BE holding an energy storage).

[frosty garden]

Do we just patch that in then? I don't think we have a method of adding fixers dynamically

[pliant adder]

We can't patch it in, it would require the owning object (usually a BE) to have a datafixer which is not feasible

[frosty garden]

Right, since we just patch in that the BE doesn't have a type. If not for that, it would easily work

[pliant adder]

We don't prevent BE types from specifying a datafixer type, there's just no way for mods to actually register the referenced datafixer. Same thing for entity types

[frosty garden]

Yep

[pliant adder]

With that said, I'm just gonna change the existing NBT serialization EnergyStorage to create a compound tag with a single energy key to match the value IO implementation and accept that it's not compatible with older saves.

I'm gonna keep the changes to implement value IO on ItemStackHandler, FluidTank and EnergyStorage local for now and push it sometime tomorrow if nobody has brought up any counterpoints till then.

[frosty garden]

Hmm, I think it would be better to have a couple of minor versions that support both decodes and then encodes into the new object

[viral root]

Can we prep 21.1 -> 21.5 now? It wouldn't really be breaking data-wise if there's an extra tag added in a neo namespace, would it?

[pliant adder]

That would be simple to implement for the old NBT serialization but is impossible for the value IO approach so the benefit would be very limited

[pliant adder]

What do you mean by that specifically?

[viral root]

Push updates to start having those versions save in both the old and 21.6+ formats, so worlds can migrate

[vast helm]

Yep, I wrote a couple when attribute stuff changed -- they can't really be used for this for the reasons xfacthd listed though

[pliant adder]

It's not possible to change the old serialization such that it's readable by both the old and new deserializer. EnergyStorage is typed to the raw Tag class and returns an IntTag from serializeNbt(). This means that doing compoundTag.put("energy", energyStorage.serializeNbt()) in the serialization of the object owning the energy storage is equivalent to doing compoundTag.putInt("energy", energyStorage.getStoredEnergy()). As such there is no way to have EnergyStorage#serializeNbt() return a tag that conforms to both the old and the new format

[vast helm]

You could do it on the other end and have it be able to de-serialize from either for a version

Not sure if it's worth it though

[pliant adder]

I could do that for the existing NBT deserializer but I can't do that for the new ValueInput-based one

[vast helm]

Ahh, yeah

Hmm

Could make the old one save to the new format but read from both

[steep verge]

Tbh that’s the only option if we want it done reasonably. Though I don’t know how much we should care

[vast helm]

Not breaking worlds is nice, but this isn't even all worlds, it's just things using those built-in implementations

[proud tusk]

My suggestion: change EnergyStorage to write to a compound and nuke INBTSerializable

[round briar]

I think I might have misunderstood you

I was thinking data attachments, but you were talking about the template impls for IEH/IFH/IIH?

Okay yes, those can go ValueOutput/ValueInput, I believe. The AttachmentHolder can technically on the outer side

[pliant adder]

I've pushed the ValueIO-based serialization for ItemStackHandler, FluidTank and EnergyStorage, I kept the raw NBT serialization in place for now though

[grim surge]

and to continue the trend

https://github.com/neoforged/NeoForge/pull/2316

thoughts?

as I've noted in the description, this does mean that attachments have to serialise to a compound

but I'd not call this a deal breaker

also blergh at type erasure

because of the serializable for INBTSerialisable I can't add another serializable method with ValueIOSerializable

I think we should rename the current method to nbtSerializable

which I have just done

[round briar]

Made some small notes, but didn't finish looking yet

[alpine lantern]

Are where major breaking changes planned for 1.21.6?

[mossy narwhal]

well there's #1183818213134446742

unless it takes them over 3 months to get it polished :p

[grim surge]

psst, was that your only small note?

[round briar]

As I said, not finished looking 😛

[grim surge]

you said some and s but you only left one

also anything against me adding a task to sort the interface injections file alphabetically (as part of applyAllFormatting)

[round briar]

nope, sounds good

[grim surge]

alright I've also done patch review and found nothing major, just a couple of small things (method patched-in when extension interface exists and a patch line that was just renaming a parameter)

[grim surge]

<@&1067092163520909374>

[mossy holly]

yes

[uncut grove]

[fiery elbow]

…

Where snapshot?

[proud tusk]

So, release only next week I guess?

[balmy edge]

:3

[uncut grove]

1.21.6 Pre-Release 4

Changelog: https://www.minecraft.net/en-us/article/minecraft-1-21-6-pre-release-4
SnowMan: https://github.com/neoforged/Snowman/commit/d4064eaf16eb4993a1808ddd05a318f006275357
Primer: https://github.com/ChampionAsh5357/neoforged-github/blob/367b044d684f6b17dea63ae77ba2b79b992ffd21/primers/1.21.6/index.md
-# Changes: https://github.com/neoforged/.github/pull/20/commits/367b044d684f6b17dea63ae77ba2b79b992ffd21

Slicedlimes Videos:

• Main: https://www.youtube.com/watch?v=FewXNHTBshw
• Data/Resource Pack Changes: N/A

[gusty stag]

New version detected: 1.21.6-pre4.

[proud tusk]

That would be 3 weeks of prereleases

[uncut grove]

why did the link not work

[fiery elbow]

Bad Gateway moment

[cosmic grotto]

ok

[ornate dome]

wow

much error

[raw sorrel]

accidental push to launchermeta maybe?
or do mojang subscribe to the there is no mistakes mindset

[mossy narwhal]

too early smh, I'm not home from work yet!

[uncut grove]

https://tenor.com/view/kung-fu-panda-oogway-accidents-there-are-no-accidents-destiny-gif-16270886

[uncut grove]

wow ⛄ before a changelog thats rare
https://github.com/neoforged/Snowman/commit/d4064eaf16eb4993a1808ddd05a318f006275357

[raw sorrel]

"fine, ill write a changelog myself"

[uncut grove]

how very thanos of you
https://tenor.com/view/thanos-gif-25117042

but thats @frosty garden job , he writes the primers changelogs

[round briar]

yeah not a large change

[opal frost]

Minecraft wiki page for pre 4 is up
But the official website is still dead

[raw sorrel]

changelog also not published on launcher side either it seems
curl https://launchercontent.mojang.com/v2/javaPatchNotes.json | jq ".entries[] | select(.title | contains(\"1.21.6\")) .version"

"1.21.6-pre3"
"1.21.6-pre2"
"1.21.6-pre1"

[uncut grove]

https://piston-meta.mojang.com/mc/game/version_manifest_v2.json
has pre4 so my guess, yeah they published meta too early

[raw sorrel]

looks like their CMS died, but existing articles were fine being served from cache

[uncut grove]

<@&1067092163520909374> (sorry for 2nd ping)
article is now live https://www.minecraft.net/en-us/article/minecraft-1-21-6-pre-release-4

[mossy holly]

yes

[red crane]

/playsoud lol

[mossy narwhal]

it smells like RC1 on thursday and release on tuesday

[gusty stag]

Minecraft 1.21.6 Pre-Release 4

We are now releasing Pre-Release 4 for Minecraft 1.21.6, with a few minor bug fixes.Happy crafting!Fixed bugs in 1.21.6 Pre-Release 4MC-279417 Luring mobs onto blocks at Y=-64 will instead make them pathfind to the highest elevationMC-298446 The "Thunder r...

Jun 10, 2025, 12:46 PM

[lusty laurel]

@uncut grove https://youtu.be/FewXNHTBshw ping

[uncut grove]

almost missed that cause of the silent ping

[frosty garden]

@uncut grove Update primer: https://github.com/neoforged/.github/pull/20/commits/367b044d684f6b17dea63ae77ba2b79b992ffd21

Pretty much just a rollback of the gain removal from the sound engine

[grim surge]

@round briar what is holding up https://github.com/neoforged/NeoForge/pull/2312

[pliant adder]

IIRC, one of the tests is randomly deadlocking

[steep verge]

I believe that is potentially meant to at least partially fix it, but gimp wants to reproduce the deadlock first (which he has been unable to get to happen)

[round briar]

well it failed twice in a row on CI/CD, we can check what's up on CI now

it deadlocked before it even began ranning the changed test though, I think 🤔

Merged. CI did seem to pass now, so we'll see

[gusty stag]

New version detected: 1.21.6-rc1.

[ornate dome]

1.21.6 RC1

Changelog: https://www.minecraft.net/en-us/article/minecraft-1-21-6-release-candidate-1
SnowMan: https://github.com/neoforged/Snowman/commit/e2d9b7a950deeb3b4980c523c13357b491cd8008
Primer: https://github.com/ChampionAsh5357/neoforged-github/blob/update/1.21.6/primers/1.21.6/index.md
-# Changes: https://github.com/neoforged/.github/pull/20/commits/5ff5ba470c884d3af5a62dc1daa6f798db1aed57

Slicedlimes Videos:

• Main: https://www.youtube.com/watch?v=ERwctZH07zE
• Data/Resource Pack Changes: N/A

[uncut grove]

RC1 and dpeter beat me to it

<@&1067092163520909374>

[crude echo]

so many stabs given today

[balmy edge]

again?!

[bleak topaz]

IT BEGINS

[grizzled stirrup]

RC1

[crude echo]

RC...

1

[uncut grove]

https://www.minecraft.net/en-us/article/minecraft-1-21-6-release-candidate-1

[rustic nebula]

i've been preparing a mod for 1.21.6 for some time anyways

[grizzled stirrup]

1.21.6 soooon

[rustic nebula]

really hope 1.21.6 becomes lts

[uncut grove]

@ornate dome might want to update the header, says pre4 but should be rc1

[mossy narwhal]

I was correct!

[bleak topaz]

oh lawd it comin

[soft swift]

gigaherz

[Reference to](#1352609954430713866 message) #1352609954430713866 [➤ ](#1352609954430713866 message)it smells like RC1 on thursday and release on tuesday

Jun 10, 2025, 1:36 PM

[gusty stag]

Minecraft 1.21.6 Release Candidate 1

As we get ready to Chase the Skies, here comes the first test flight of 1.21.6 - Release Candidate 1!Happy Ghast(-ing)!Fixed bugs in 1.21.6 Release Candidate 1MC-298022 Not signed Books and Quills with an overflowing new line at the bottom do not depict th...

Jun 12, 2025, 12:59 PM

[ornate dome]

https://github.com/neoforged/Snowman/commit/e2d9b7a950deeb3b4980c523c13357b491cd8008

[grizzled stirrup]

https://tenor.com/view/faith-horror-game-gif-24998687

[uncut grove]

https://github.com/neoforged/Snowman/commit/e2d9b7a950deeb3b4980c523c13357b491cd8008

[rustic nebula]

what is snowman anyways

[ornate dome]

you are late with that

[flat badger]

the internal decomplied mc sources

[rustic nebula]

oh

[mossy narwhal]

snowman is an internal repo where we put the initial decompilation results for analysis

[rustic nebula]

pre-neoform sources?

[flat badger]

private github repo because otherwise it'd be issues™️

[bleak topaz]

ye

fresh from the decompiler

(remapped, of course)

[ornate dome]

it is a github repo the neo team uses for seeing what changes from version to version. It is private cause it is a full decomp, and sharing that is not good :TM:
the tools are public for it so people can do it for themselves...

[bleak topaz]

meant purely as a reference, not for actually running the game

[rustic nebula]

yea i see how that'd be useful for something like neoforge considering it's still all old fashioned source patches

[mossy narwhal]

patches are not old fashioned they are the only viable solution

but that's not at all why snowman exists

[flat badger]

was about to say, there's reasons we still use them

[ornate dome]

I should start working on my only recurring task...
making the next Snapshots thread... It is hard work I must do

[rustic nebula]

1.22 or 1.21.7, call it.

[ornate dome]

I feel 22 in my bones

[flat badger]

1.21.7/1.22 Snapshots

[mossy narwhal]

snowman exists in order to help the team discover changes by letting us see differences between versions, to help coordinate the porting

I expect 1.21.7 next, but crossing fingers that I'm wrong

[ornate dome]

i think it is either a quick fix 21.7 or 22

[mossy narwhal]

name it 1.21.next until we know more :p

[velvet obsidian]

I already did

[rustic nebula]

Upcoming Release™ Snapshots

[velvet obsidian]

https://discord.com/channels/313125603924639766/1378294669053792337

[rustic nebula]

insane to think that 1.21 is going to be a year old tomorrow

[uncut grove]

and peter should said nope to that thead xD

[ornate dome]

Reeee Mine is better

[uncut grove]

made his one

[velvet obsidian]

THERE IS NO DIFFERENCE WHY ARE WE CREATING UNNECESSARY POSTS HERE???

[grizzled stirrup]

really?

[velvet obsidian]

no it was said that the bot looks for the tag so it shouldn't be taged

[ornate dome]

MINE IS MUCH BETTER AND HAS MORE COOL PEOPLE IN IT!!
ALSO WHY ARE YOU SHOUTING? (/s)

[uncut grove]

random thought but with how many snapshot threads we make is it worth having a dedi snapshot channel and we create threads per snapshot there? rather than constantly cluttering brainstorming?

[ornate dome]

what tag?

[velvet obsidian]

the porting tag

[ornate dome]

i didnt add that to mine either

[velvet obsidian]

but why did you create one in the first place if there already is one?

[ornate dome]

I searched for 1.22 and 1.21 didnt think you would name it just 1.

[velvet obsidian]

I told you here that I already have one I even linked it

[ornate dome]

you sent that literaly 10 secs after I pressed the create button

[velvet obsidian]

discord says 3 minutes before

[ornate dome]

I might have been typing the text in that time than... But i did not see your message before clicking create

[weary socket]

https://tenor.com/view/tactical-precise-plan-keanu-reeves-john-wick-gif-13730619

Idk I just wanted to post a John wick gif

[ornate dome]

Also I have been making the posts for this... for a while... XD (the last 4 were from me)

[velvet obsidian]

I did them before you

[viral root]

oh god threads inside forums inside guilds inside..

[ornate dome]

only 3...
I have made more than you

[velvet obsidian]

well I made the next one before you so it's 4

[ornate dome]

You know what?
if it is a 1.22 we use your thread
it is it 1.21.7 we use mine
and the other thread becomes the one for the release after that

[uncut grove]

i was thinking regular text channel for general snapshot discussion / pin FAQ at top (too many people ask what snowman is)
then for each snapshot we create a new thread there

rather than new branstorming post for each snapshot which are kind disconnected from each other and clutters the main branstorming thread view

[viral root]

I mean, we can just make a snapshots forum that people can hide separately

[ornate dome]

Not that bad idea...

[weary socket]

There’s a number of people that extremely hate forums and will refuse to open it

[ornate dome]

well for those this is already a lost place....

[grim surge]

this is already a forum

[uncut grove]

shhhh dont let the forum haters know that

[weary socket]

I was thinking of the gallery type ones with the image preview

those one are pain

[grim surge]

changed this forum to gallery just to annoy you then

:P you'll thank me later

[frosty garden]

@uncut grove primer time: https://github.com/neoforged/.github/pull/20/commits/5ff5ba470c884d3af5a62dc1daa6f798db1aed57

Aside from the bug fixes, only two overloads

[ornate dome]

added to the pinned

[uncut grove]

Primer: https://github.com/ChampionAsh5357/neoforged-github/blob/update/1.21.6/primers/1.21.6/index.md
-# Changes: https://github.com/neoforged/.github/pull/20/commits/5ff5ba470c884d3af5a62dc1daa6f798db1aed57

[ornate dome]

Ohh sure

[mossy narwhal]

date confirmed: https://www.youtube.com/watch?v=zPWdMDZWM4Y

unless java edition ends up releasing a different day :P

[daring lance]

oh hey, I "accidentally" chose the right day

[mossy narwhal]

they almost always release on tuesdays

so it's not that hard to guess

[uncut osprey]

tuesday is patch day, except for COD

got monday to fix anything you thought about in the shower over the weekend, and the rest of the week for any fallout

[daring lance]

I didn't really think about that, I just thought tomorrow is too early and monday feels wrong

[burnt creek]

I knew it was coming soon

[mossy narwhal]

most recent RC1 have been on thursday, with release on tuesday

the only thing that would delay it is some major issue and a RC on tuesday, with release on wednesday/thursday

unless Microsoft forces them to release tuesday no matter what, then 1.21.7 the week after

[storm jacinth]

@round briar Do we want to merge the FML Early window styling PR?

[ornate dome]

https://www.youtube.com/watch?v=ERwctZH07zE

[velvet obsidian]

Dang it right on the day of my office summer party

[grim surge]

still want a review on https://github.com/neoforged/NeoForge/pull/2316 preferably before release kthx

[velvet obsidian]

if nobody else has anything to say against it

[viral root]

but I still want to see a WAIFU report on how many people actually use NBT for attachments

Codecs are still easier and I'll die on that hill

[velvet obsidian]

then open waifu and check

[viral root]

....6 (real count: 2)

https://waifu.neoforged.net/d/aec29a5ymel1cc/waifu-classes?orgId=1&from=now-6h&to=now&timezone=browser&var-version=1.21.1-neoforge&var-class=net%2Fneoforged%2Fneoforge%2Fattachment%2FIAttachmentSerializer&var-method=.*&var-field=.*

Oh, wrong panel. Sorry, there are 82 classes inheriting IAttSerializer

Out of thousands of mods.

[pliant adder]

You are missing cases which use the helper to make a serializer for an INbtSerializable object

[viral root]

fair

have I mentioned how shit the class input is on Grafana, in that I can't edit it after the fact? Nope, gotta retype the full /-divided package name again...

758: https://waifu.neoforged.net/d/aec29a5ymel1cc/waifu-classes?orgId=1&from=now-6h&to=now&timezone=browser&var-version=1.21.1-neoforge&var-class=net%2Fneoforged%2Fneoforge%2Fattachment%2FAttachmentType&var-method=serializable&var-field=.*

[grim surge]

@willow hatch can I inquire you about patching suppresions in neoform for the 10 or so compile warnings there are in mc

[vast helm]

That's a non-insignificant number

[viral root]

Yup. All those mods will now have a choice they're forced to make: port to ValueIn/Out, or adopt codecs

[round briar]

The ValueInput/ValueOutput APIs and CompoundTag are almost interchangable

[viral root]

Still a breaking change

[vast helm]

Codecs really don't make sense because of the high amount of context you need to (de)serialize

Yes, yes, you can shove that into codecs

But it's not elegant

And you make very well need the thing you're attaching to to properly (de)serialize your thing in quite a few cases

Is it necessary? No, you can refactor your component so that you pass in that context on every method of it. But it's (I would assume) a common enough pattern -- and it's sensible and fine to do, really, as data attachments aren't meant to be immutable, unlike data components

[grim surge]

they don't have to adapt the code if they're using INBTSerializable<CompoundTag>

the helper method for that remains, just delegates to the valueio methods

again the only adaptations needed is the requirement of a MapCodec/CompoundTag because they have to serialise to a map now

[willow hatch]

There's a lot more than 10

[elder zephyr]

Yeah lmao there's a fuckton of warnings in vanilla

[frosty garden]

If anyone sanity checks the primer, please ping me before the 17th so that I can address the issue(s)

[vast helm]

Hmm. I still suspect you could make a step that effectively generates patched suppressions for all those by running the java compiler and processing the warnings in the output

Maybe I should poke that

The first question, of course, is if there's a nice API on the javacompiler stuff to see warnings...

[vast helm]

Ugh, it is doable -- JavaCompiler exposes the relevant Diagnostic type -- issue is that you can't get at the lint-code that you'd need to write suppressions without javac internals

I'm gonna keep digging in case I've missed something, but... bleh

I can get a translation key for the warning, however, without anything internal -- but that doesn't seem particularly nice to map to a lint category. I'll see

But that small annoyance asside, it would be perfectly possible to have something generate the relevant SuppressWarning patches to remove the compiler warnings in the decompiled code

(Of course, that small annoyance is not easily worked around without (a) manual mapping of error codes -> lint types or (b) use of javac internals)

```java
tasks.named('compileJava', JavaCompile) {
options.compilerArgs += [
'--add-exports=jdk.compiler/com.sun.tools.javac.util=ALL-UNNAMED',
'--add-exports=jdk.compiler/com.sun.tools.javac.code=ALL-UNNAMED',
'--add-exports=jdk.compiler/com.sun.tools.javac.api=ALL-UNNAMED',
]
}

Hmm... okay, parsing it from the raw compiler text output may be doable

[vast helm]

There are also undocumented -XD... options that let you fine-tune the warning format, which could be enough here

The default diagnostic format also might be good enough to work with -- it gives you the line number, but not the start-end pos, but the line number should be enough to place the suppression in most cases I'd think?

Anyways, that's all I've got on the warnings bit for now -- if this is something you think worth pursuing, with those limitations in mind (likely relying on the compiler output format, in some form), lemme know, and I can try to get something working in neoform to suppress all those errors in the vanilla stuff.

[grim surge]

javac only reports 10 of them as warnings without any flags

check a neo build

[vast helm]

Looks like it's just the deprecation ones normally -- does it not warn on raw types without flags? Regardless, may be worth suppressing the others too so that you can use those flags in neo?

[vast helm]

Alright, so poking this more it actually looks totally feasible to parse the normal output format and generate suppressions from that, all within JST

[glacial tartan]

Something worth including in the porting guide which isn't caught by a compiler error, is that the colour argument in GuiGraphics.drawString and friends now include an alpha value, so if you were previously passing a colour without an alpha value, that's now an alpha value of 0 and it won't render!

[pliant adder]

That's a good point, yeah, I've been bitten by that in all projects I've ported to the pre-release

[bleak topaz]

@frosty garden something for your attention?

[frosty garden]

Yep, just saw it. Thanks @glacial tartan

Expanding on it a bit more, it seems like any of the methods in GuiGraphics aside from the item bar now require the alpha value, though I don't remember which was the same from the previous versions

And added with credit to earth

[uncut grove]

Soon ™️ https://www.youtube.com/watch?v=ik0r1BjsmBA

[lusty laurel]

um...

Why "-100% Waypoint Transmit Range [x0]" ?

[twilit forge]

do you have any datapacks or resource packs?

oh, that's related to the thing where wearing a pumpkin keeps you from showing up on the player radar

[lusty laurel]

ohhhh

yeah, nvm

[twin totem]

How can I build the 1.21.6 port branch myself?

[mossy holly]

[twin totem]

I have cloned neoforge onto my pc and I want to prepare the 1.21.6 update for my mod, but I don't know how to build it correctly because the net.minecraft packages are not available

[ripe igloo]

It’s not recommended, but the porting PR has PR Publishing enabled if you want to develop against Neo 1.21.6

No need to build it yourself

See the comment by the bot on the pr

[mossy narwhal]

you have to run the gradle setup task first, that ensures a working neodev environment

but yeah you can port your mods without building your own

[twin totem]

What are the steps for this way? What have I to do?

[ripe igloo]

(this)

[ripe igloo]

https://github.com/neoforged/NeoForge/pull/2297#issuecomment-2925162859

[twin totem]

I don't check how to use this in my moddev gradle file of my mod

[velvet obsidian]

you need to publish it to maven local and use the version it prints to the console when building in your mod

[twin totem]

I got it, thank you!

[round briar]

what? why

[round briar]

You add the repository the bot posted in the comment to your build.gradle or settings.gradle (wherever you define your repositories)

repositories {
    maven {
        name 'Maven for PR #2297' // https://github.com/neoforged/NeoForge/pull/2297
        url 'https://prmaven.neoforged.net/NeoForge/pr2297'
        content {
            includeModule('net.neoforged', 'neoforge')
            includeModule('net.neoforged', 'testframework')
        }
    }
}

And then you set your neoforge version to the version the bot commented above that, currently: 21.5.0-alpha.1.21.6-rc1.20250617.072928

No local checkout / build / publish is needed

[velvet obsidian]

because I thought they wanted to use a local version based on previous answers

[round briar]

They just seem to want to port their mod, the PR build is ideal for that

[proud tusk]

So, what will come out first? MC 1.21.6 or Gradle 9.0.0-RC1? 😄

[viral root]

Both, simultaneously. Chaos ensues.

Mojang secretly beta testing Gradle

[uncut grove]

soon™

[frosty garden]

Would honestly assume mojang unless it doesn't get released until the afternoon

[uncut grove]

oh champ, when do we usually merge the primer pr?
rn i have the latest pr changes linked in my msg to post later
but having them link to merged changes would be nice

[frosty garden]

Usually after I confirm there's nothing that changed between rc1 and release

[uncut grove]

ah ok, what i have will do fine then for now
-# and il update links once pr is merged

[timid niche]

1.21.6 is out

[gusty stag]

New version detected: 1.21.6.

[uncut grove]

what i dont see the article

[wet finch]

It's true. 1.21.6 is out today

[timid niche]

https://piston-meta.mojang.com/v1/packages/ac29333211286b32d68ca4a16d2026a834ac1430/1.21.6.json

[uncut grove]

1.21.6 - Chase The Skies

Changelog: https://www.minecraft.net/en-us/article/minecraft-java-edition-1-21-6
SnowMan: https://github.com/neoforged/Snowman/commit/4d8b82cb5bf567b2570c7f4a217ab9cf277c4c6f
Primer: https://github.com/neoforged/.github/blob/98211944dc1877c19feadfda48994178b2852bcc/primers/1.21.6/index.md
-# Changes: https://github.com/neoforged/.github/commit/98211944dc1877c19feadfda48994178b2852bcc

Slicedlimes Videos:

• Main: https://www.youtube.com/watch?v=ik0r1BjsmBA
• Data/Resource Pack Changes: https://www.youtube.com/watch?v=bRuoC0qUb5Y

<@&1067092163520909374>

[mossy holly]

??!?!!?!

[muted chasm]

is it out?

[uncut grove]

Article link might be wrong but the version is out according to metadat files
Oh there it is, article is live now

[mossy holly]

it bumped without rc-2

[tawdry crown]

accidently opened bedrock and it doesn't seem to be on there lol

[wet finch]

So hopefully 1.21.6 Neoforge will be out soon. (I think)

[grim surge]

I shall do the update as soon as neoform is done

[tawdry crown]

oh but it is on java, peak

[frosty garden]

Ok no changes on the primer

Ready to merge whenever

[mossy narwhal]

so, no changes vs rc1?

[frosty garden]

Nope, except the standard version configs and bumps

[uncut grove]

yep nothing changed other than version numbers ofc
https://github.com/neoforged/Snowman/commit/0e6aa65423728bed589de3c608e9f1a7b8e42ec8
-# 21.6-RC1->21.6

[uncut osprey]

wow, an RC actually being released, almost like it was a candidate for that?

[mossy narwhal]

was the porting branch up to rc1 already or is the porting starting from an older prerelease?

[uncut grove]

pretty sure the PR is on rc1 already

[uncut osprey]

grumbles about the release not matching the last RC

[uncut osprey]

it is

[mossy narwhal]

nice so we just have to merge and tag then? :P

(after bumping to the right "basis")

[uncut grove]

wow same day/same release time (kinda) neo update should be

[uncut osprey]

well, and update it to point to 1.21.6 instead of rc1

[mossy narwhal]

yeah that's what I meant with basis

[grim surge]

local decomp is running

update in progress™

[muted chasm]

for your consideration

[uncut osprey]

where update?

[grim surge]

in progress™

can you not read

smh smh smh

[uncut osprey]

no, hence the sign not stopping me

[grim surge]

push incoming

[muted chasm]

birthing women:

[soft swift]

[Jump to referenced message](#builds message) in #builds

matyrobbrt referenced · Jun 17, 2025, 3:33 PM

NeoForge

Build Started

Version

21.6.0-beta

Build Branch

1.21.x

Commit message

Release 1.21.6

Quoted | Matyrobbrt · Jun 17, 2025, 3:33 PM

[grim surge]

oh fuck we don't have a blog post

[muted chasm]

get typing

[twilit forge]

are we doing an LTS for 1.21.6 or still waiting for 1.22

[muted chasm]

LTS please please please

1.22 is nowhere to be found

and you can always say nvm if it is

1.21.6 as main LTS and 1.21.1 as lesser discouraged but still ok LTS

[mossy holly]

how much time waas there between 1.20 and 1.21