it will be useful at some point
#1.20.3 and .4 Snapshots
mighty pulsar
yeah but to update you'll probably want to remove the sources anyway
after the roundtrip not before
reef pawn
xfact already did the roundtrip
mighty pulsar
xfact already did the roundtrip
I mean the roundtrip on the release neoform
reef pawn
sure
release is confirmed for today: https://www.minecraft.net/en-us/article/bats-pots-and-competitions "today itβs launching into the game"
upper pollen
sup
crystal galeBOT
gigaherz referenced
upper pollen
it's out
rose temple
upper pollen
does this deserve a snapshot ping considering it's not a snapshot
rose temple
<@&1067092163520909374>
stone perch
does this deserve a snapshot ping considering it's *not* a snapshot
Just name it "New Version Alarm"
reef pawn
now we wait for neoform
reef cloak
I was in the car, I think it released like a minute after I closed discord
reef pawn
I just finished reapplying all the cap rework patches on top of 1.20.3
eager skiff
Yay cap rework
reef pawn
@mighty pulsar you mind if someone already does the 1.20.3 update on Kits? so everything is ready for the rebase
hmmm, I am on the 1.20.3 cap rework branch and running a server is broken
mighty pulsar
<@213343613583622144> you mind if someone already does the 1.20.3 update on Kits...
I'm on it
mighty pulsar
what's broken now?
reef pawn
WorldDimensions.CODEC.encoder().encodeStart(net.minecraft.nbt.NbtOps.INSTANCE, worlddimensions).flatMap((writtenPayloadWithModdedDimensions) -> WorldDimensions.CODEC.decoder().parse(net.minecraft.nbt.NbtOps.INSTANCE, writtenPayloadWithModdedDimensions)).resultOrPartial(LOGGER::error).orElse(worlddimensions); in the dedicated server Main is failing
just update to 1.20.3 first then try to run a dedicated server
tulip abyss
alright I have the fix for the dedicated server
What was the cause?
mighty pulsar
how is the update going?
It has happened
reef pawn
should use registry ops not nbt ops
I'll fix it in a sec
hold on for 2 min before squashing
mighty pulsar
we have to do testing anyway before squash
tulip abyss
I did notice something else while also trying to find the cause of that: the patch for reading additional data is too early and logspams when the server doesn't have a world yet: https://github.com/neoforged/Kits/blob/1.20.3-rebase/patches/net/minecraft/server/Main.java.patch#L70
reef pawn
if we can get this released in 30 mins we'll be able to merge the capability rework PR properly
(I have to leave in 1 hour)
π
mighty pulsar
I did notice something else while also trying to find the cause of that: the pat...
where should it go?
tulip abyss
From a purely theortical point, it should be two lines lower in the levelstoragesource$levelstorageaccess.hasWorldData() check. I have not checked whether that's actually correct though
reef pawn
schurli are you fixing this?
I am fixing the useless {Damage: 0} tag that we are writing in recipes
mighty pulsar
schurli are you fixing this?
yes testing right now
tulip abyss
Reading the code around that area and comparing against a 1.20.2 instance, putting it behind that check seems to indeed be equivalent to the lambda given to LevelStorageSource.this.readLevelData() in the 1.20.2 implementation of readAdditionalLevelSaveData(). But yes, definitely test to make sure I'm not being a complete idiot here π
I pulled the 1.20.3 commits and can't run setup anymore
reef pawn
weird
mighty pulsar
<:screm:980923567833219093> I pulled the 1.20.3 commits and can't run setup anym...
you need to delete the cached launcher metadata
projects/neoforge/.gradle/caches/minecraft/launcher_metadata.json
and
projects/base/.gradle/caches/minecraft/launcher_metadata.json
tulip abyss
Yup, that did the trick, thank you
final jackal
Interesting
I thought I implemented a cache buster there
@tulip abyss can you make a ticket?
tulip abyss
Sure π
reef pawn
how about we move the additional NBT read to there?
tulip abyss
Sure, seems reasonable to me
mighty pulsar
eh maybe more like in the try as the last one
tulip abyss
Shouldn't be necessary, that method has its own try..catch. Also, putting it in the inner try won't do anything useful as it only reads from the main level data and not the fallback one
reef pawn
seems to work fine
the error is gone and the additional data tag is correctly being read
mighty pulsar
ok then gen patches commit and I can squash if there are no more issues
tulip abyss
seems to work fine
If you don't have a corrupted main level data file, then you won't reach the inner try. Calling it again in the inner try won't do anything and at worst will log an error because it'll try reading the main level data again, which at that point is already known to be fucked
tulip abyss
Ok, that I agree with π
upper pollen
how much has 1.20.3 final changed vs rc1?
reef pawn
not at all I think
upper pollen
ah so these last minute fixes are just stuff that was already broken in rc1 I suppose
tulip abyss
The source diffs from Schurli's update to the release version are just three files: DetectedVersion, SharedConstants and ReportEnvironment. All of them just the version number things
reef pawn
ah so these last minute fixes are just stuff that was already broken in rc1 I su...
yeah
how's the rebase going 
mighty pulsar
what should the commit message be?
upper pollen
ideally, a summary of any major changes that were required in order to get the port done
upper pollen
I haven't followed the development closely enough
reef pawn
Schurli before you push to the main remember to tag 20.3, make sure to push with --follow-tags, and get Orion or another org owner to disable the branch protection rules
upper pollen
but I wanted to point to this as what I think is an ideal porting commit message: ```
- 44.0.0 Forge 1.19.3
Created a CreativeModeTabEvent to register creative mode tabs and populate entries per tab
Moved datapack registries to DataPackRegistryEvent.NewRegistry event instead of tying them to ForgeRegistry
Made it easier for mods to datagen datapack builtin entries with DatapackBuiltinEntriesProvider
Provided access to lookupProvider for datagen
Updated dependencies to match versions used by vanilla and update JarJar to 0.3.18
Added a test mod for the new CreativeModeTabEvent
Throws better error message for Forge registries in tag datagen
Deleted ForgeRegistryTagsProvider
Updated ClientChatReceivedEvent and ServerChatEvent for Mojang changes
Added patches for both sign related methods in ModelLayers
Changed RegisterShadersEvent to use ResourceProvider
Migrated old Mojang math types to JOML
Co-authored-by: Marc Hermans [email protected]
Co-authored-by: LexManos [email protected]
Co-authored-by: sciwhiz12 [email protected]
Co-authored-by: coehlrich [email protected]
Oh, and make sure that someone can cancel the first TC build
I can do it, Orion can too, other org owners as well probably
mighty pulsar
I'll leave the pushing to main to someone else (I don't want to fuck it up)
tulip abyss
You don't wanna do the Curle? π
reef cloak
x3
reef pawn
Just push to Kits so I can rebase the cap PR
The commit will be public that way but whatever π
mighty pulsar
please review the squashed commit and then let someone push to the main repo
reef pawn
Can you remove updating=true from gradle.properties?
upper pollen
mighty pulsar
uh shit
tulip abyss
That's all the whitespace changes from the switch to 4 spaces in the decompiler output
upper pollen
okay a lot more manageable with ignore whitespace
reef pawn
alright, looks good to me
upper pollen
ah, lots of those are still whitespace-related
reef pawn
I'll rebase the cap PR so the commit will be public
upper pollen
github doesn't have a "ignore patch offset changes" option :P
gaunt canyon
@mighty pulsar want me to push?
gaunt canyon
someone get the blog post ready while i run setup 
reef pawn
ok
mighty pulsar
<@213343613583622144> want me to push?
if you want to get the blame if you bork it yes
upper pollen
uh
reef pawn
if you want to get the blame if you bork it yes
meh don't be afraid about that
gaunt canyon
can't you prevent CI by tagging the commit somehow?
TC isn't configured to do so
it works in all other projects because everywhere else we use GHA (well, everywhere* NG is on TC)
upper pollen
ah, sad.
does TC have a "pause" button? like, pause before pushing, then unpause :P
doesn't matter either way
just wondering
upper pollen
ew
reef pawn
not if you disable versioned settings but yeah π
let's just cancel the build on TC
maty go go go
upper pollen
yeah it is apparent that cancelling is the least annoying option as things stand
gaunt canyon
we're vulnerable!
(now tech pls don't accidentaly push to upstream while i'm doing this lol)
reef pawn
kekw
gaunt canyon
something is sus
reef pawn
yeah see above
reef pawn
you need to delete the cached launcher metadata
^
also remember to tag and push with --follow-tags
go go go
gaunt canyon
tech i appreciate your enthusiasm but unfortunately that dosen't make decomp any faster lol
reef pawn
takes < 2 min on my pc π
gaunt canyon
are you afraid of quilt catching up, or what? smh
reef pawn
they don't have a 20.2 release yet lol
gaunt canyon
qed
reef cloak
not even a alpha/beta release?
upper pollen
speaking of, make sure the tag is -beta ;P
reef pawn
Nopr
upper pollen
or whatever it is that marks it as beta
reef pawn
The tag is 20.3
upper pollen
how is the beta flagged?
gaunt canyon
automatically
reef cloak
how is the beta flagged?
No -stable tag
gaunt canyon
decomp done, now time to :neoforge:build 
reef cloak
Think we had a discussion on what made sense and a second tag was the most popular outcome
gaunt canyon
it compiles 
reef cloak
werenβt you even apart of that
tulip abyss
So, who kills the build?
Also, the version number is wrong
gaunt canyon
...was the tag not pushed? anyways, build cancelled
according to my git the tag was pushed
..wait
tulip abyss
It said 20.2.87 π
gaunt canyon
oh ffs
tulip abyss
Heh, Gradle likes to troll you today, it seems
mighty pulsar
Why tf is the caps rework .1 and not .2?
tulip abyss
Because the initial release is .0 and not .1
rose temple
I have to ask.... It is one of the few reasons I'm here....
DOES THE FURNACE WORK?
also is the ender dragon healthy?
sturdy geyser
does creating a singleplayer world load mods' datapacks
dusky epoch
I have to ask.... It is one of the few reasons I'm here....
DOES THE FURNACE WOR...
last time I asked that, I got the stabbies
rose temple
tick freez is gonna casue soo many issues imo
tulip abyss
Hmm, looks like one of the rebases partially failed, for some reason the render BB getter in IBlockEntityExtension and a patch related to it is back: https://github.com/neoforged/NeoForge/blob/1.20.x/src/main/java/net/neoforged/neoforge/common/extensions/IBlockEntityExtension.java#L90-L132
upper pollen
on the upside, once it works, mods that accelerate time will work a lot better lol
rose temple
that is actually true... also I already have interesting ideas for spells that stop time... XD
reef pawn
Hmm, looks like one of the rebases partially failed, for some reason the render ...
That is very concerning
What's funny is that the BER extension is there too
tulip abyss
Yeah, that's what's really strange to me
reef pawn
Probably just a Schurli mistake in the rebase?
It would be easy to mistake left and right in a rebase and actually keep the wrong part
tulip abyss
Yup
Gonna yeet it again as part of the render bounds fix for the dragon head and banners
reef pawn
Did the boxes change again?
tulip abyss
No, I just slightly miscalculated the dragon head's bounds and completely forgot about the stupid banners
reef pawn
π
mighty pulsar
Probably just a Schurli mistake in the rebase?
It is not a me mistake if I ask for reviews multiple times (which I did)
reef pawn
Oh don't get me wrong I am not trying to blame you
I'm just brainstorming what could have happen, but this sort of issue is totally expected and what matters is that it gets fixed now
You shouldn't feel guilty about anything
My wording wasn't the best, my apologies
tulip abyss
Tech, go merge: https://github.com/neoforged/NeoForge/pull/327 π
reef pawn
I'm going to sleep... tomorrow if nobody does it before me π
tulip abyss
Hehe, good night π
radiant sandal
Hehe, good night π
Done
gaunt canyon
now it's time to rebase test mods and get them merged
cunning gyro
#mojira message we're getting a 1.20.4 :)
crystal galeBOT
wither.tm referenced
mighty pulsar
upper pollen
I expect a 1.20.4-preX soon
I don't know if they will do snapshots for a hotfix, but there's definitely going to be a hotfix :P
cunning gyro
if we are getting a 1.20.4 this fast, we will get a 1.20.5 xD
inland talon
Perhaps this channel should be renamed to 1.20.3+ Snapshots?
upper pollen
probably should jsut be renamed "Porting work" :P
reef pawn
the idea is to have a different thread for each MC "development cycle"
mighty pulsar
yes and as such I'm gonna close this and open a new one once it is needed
rose temple
Ammmmm
finite cosmos
rose temple
20.4
crystal galeBOT
gigaherz referenced
rose temple
<@&1067092163520909374>
reef pawn
upper pollen
RESTART THE ENGINES!
mighty pulsar
oh ffs
teal canyon
kekw
upper pollen
1.20.4 Snapshots
mighty pulsar
maty where is your workflow?
teal canyon
1.20.3 and .4 Snapshots
reef pawn
apparently it's only a single bug fix?
upper pollen
yes it looks like they only included the decorated pots in it
I assume the rest would have taken too much testing
reef pawn
(for another bug)
so I think the others are just gonna be fixed in 1.20.5 or whatever else the next released version is
gaunt canyon
maty where is your workflow?
on a repo of mine iirc
mighty pulsar
does it work?
gaunt canyon
last time it was randomly failing because of unknown reasons https://github.com/Matyrobbrt/quick-actions/actions/runs/6935277724/job/18865112145#step:6:76
which given java.util.zip.ZipException: zip END header not found is probably a corrupted cache
mighty pulsar
fix it so I can use it for 20.4
gaunt canyon
time to manually trigger the check neoform action
gaunt canyon
ok well the action worked, i 'm surprised
upper pollen
kinda wish they added a 4th version component
like
1.20.1 could have been 1.20.0.1, then 1.20.2 could have been 1.20.1 (since it was a larger change), and and 1.20.4 could be 1.20.2.1
mighty pulsar
1.20.1 could have been 1.20.0.1, then 1.20.2 could have been 1.20.1 (since it wa...
no, please no
upper pollen
game release . content update . tech update . hotfix
reef pawn
game release . content update . tech update . hotfix
why do I have to update?
gaunt canyon
(would clobber existing tag)
would what
gaunt canyon
why do I have to update?
your windows version is outdated
you're still running windows 7
cunning gyro
no, please no
I honestly think that would be great
upper pollen
why do I have to update?
no the game updates you :P
gaunt canyon
it is, it failed bc of https://github.com/neoforged/actions/actions/runs/7118263956/job/19380879047#step:14:10
how do i make it ignore if there aren't files
gaunt canyon
or i could just add a push step before rejects
that way if it fails to do rejects it doesn't matter
mighty pulsar
that way if it fails to do rejects it doesn't matter
it would report as faild then tho
gaunt canyon
Β―_(γ)_/Β―
does it actually matter if it worked? lol
not like updates with no rejects are extremely rare, basically only hotfixes
mighty pulsar
exactly and in that case we want the action to report a success too
gaunt canyon
tfw setup is FROM-CACHE but neoFormDecompile is not
upper pollen
is the decorated pot the only thing they actually changed, or did they sneak in anything else?
mighty pulsar
chiseled bookshelf blockstate json
gaunt canyon
the answer is yes lol
(btw schurli, the action is only for the first commit (so it can push the two sets of sources in the correct, un-gitignore etc, since any subsequent updates can be done manually much more easily)
mighty pulsar
it does:
- setup
- add sources
- update
- setup
- add rejects
- genPatches (if possible)
it would be nice to be able to run 3 - 6 on an existing update branch for snapshot to snapshot updates
gaunt canyon
hmm, i'll look into it a bit later
gleaming coral
ok well the action worked, i 'm surprised
When a new snapshot series starts only the TARGET_MC_VERSION variable needs to be updated
rose temple
what the hell is up with mojira?
how did they miss those tags?
we are getting a new rc2 soon i think
mighty pulsar
I think they will release .4 today and start with pre releases for .5 next week
rose temple
2-3 moths from now:
I think they will do 1.20.45 today and .46 next week
upper pollen
depends on if they see it as a game-breaking issue or not
given that the tags are marked as "future update"
I'm not sure there will be another hotfix at all
but rather those fixes will be included in the next "preview iteration"
like, to be released sometime in february/march like 1.19.4
reef pawn
how did they miss those tags?
we are getting a new rc2 soon i think
They probably got the tags wrong yesterday
They look fine now
dusky epoch
2-3 moths from now:
> I think they will do 1.20.45 today and .46 next week
I kind of want to see this happen
upper pollen
nope not there on rc1 either
rose temple
they are cooking something for sure
cunning gyro
2-3 moths from now:
> I think they will do 1.20.45 today and .46 next week
they won't go past 1.20.9 because of bedrock
upper pollen
I don't think they care about that as much as people seem to think they care
cunning gyro
what I mean is that 1.20.10 is already a thing and they like to differentiate bedrock and java
rose temple
what?
they are two separate products
upper pollen
again, I don't think they would care that the version number has been used in bedrock if java edition reaches 1.20.10 (which is unlikely either way)
unborn helm
we have had a .10 patch number before on Java
upper pollen
yeah but that was before the current update parity
gaunt canyon
I'll do the update, waiting on neoform
upper pollen
I assume no changes vs rc1
reef pawn
I'll do the update, waiting on neoform
you can always start it manually I suppose
reef pawn
did you set the version name as coeh said?
gaunt canyon
I didn't do anything, that action was run automatically
reef pawn
try doing this
gaunt canyon
it's already set to 1.20.4
reef pawn
maybe the branch has to be created manually
so apparently old_branch is incorrect
it's read from the manifest apparently
it's taken from process.env.LATEST_MAINLINE
where tf is that coming from
cunning gyro
why only that 1 bug, I say many other bugs that were fixed
upper pollen
because this is an emergency hotfix
reef pawn
so how did you fix this, maty?
sturdy geyser
the other bugfixes are being shipped in 1.20.5 or later
gaunt canyon
I changed two vars to point to 1.20.3 as the base
upper pollen
and data loss is critical
cunning gyro
oh this was a HOT fix
upper pollen
so they chose to push the data loss issue to the top
and they MAY release a 1.20.5 in a few weeks
reef pawn
ah ok
gaunt canyon
do I still do the whole kits thing
rose temple
told yall!!
on the road to 1.20.45
dusky epoch
why not 1.20.20.23
reef pawn
that's the only actual change
radiant sandal
surprised they would do a whole version for it, haven't they unreleased and rereleased a version in the past?
reef pawn
oh yeah it failed to commit rejects, again
gaunt canyon
oh yeah it failed to commit rejects, again
cause there's none yes
gaunt canyon
fix what
gaunt canyon
@reef pawn you have 60 seconds to review the squash https://github.com/neoforged/Kits/commit/feefc155b5dd8f1a0197a9df1f3af532cfc1c9ca
reef pawn
upstream branch is neoforge right?
can you create a 1.20.3 branch on neoforge for me? π
just so I can test this
or maybe it's not required idk
hallow onyx
The website doesn't have an installer for 20.4.x
gilded anvil
https://projects.neoforged.net/ has it, and should really be added to the website
(all the installers are on the maven)
reef pawn
ah yeah I wanted to do that but forgot heh
gilded anvil
https://projects.neoforged.net/neoforged/neoforge specifically gives you downloads
Gotta say, I do like that page a lot, even in spite of the client side rendering
hallow onyx
ah yeah I wanted to do that but forgot heh
No probs
teal canyon
the Markdown rendering is still scremworthy
mighty pulsar
1.20.5 Snapshots
upper pollen
I think they said they wanted to have this one for 1.20.4, and create a new one for 1.20.5 assuming it's going to be a new feature update
mighty pulsar
I think they said they wanted to have this one for 1.20.4, and create a new one ...
I wanted that for .4 already but it was decided to keep this one
We're probably getting a snapshot really soon
gaunt canyon
why no new thread
mighty pulsar
why no new thread
because we should lock this one first (and I can't)
gaunt canyon
i can so make new thread
mighty pulsar
1.20.3 and .4 Snapshots