docs-website

[spice temple]

I have Vue components and static html intertwined without issues

[crystal stump]

Maybe we should make an article about hot swapping?

[twilit fog]

Isnt hot swapping a bad practice, if I remember correctly one dev from paper team was saying the JVM wqsnt designed for it

[sharp pivot]

they were probably talking about the /reload command

[eager plover]

Yea, that was /reload

hotswapping is a dev tool, not a sane runtime tool

while you could maybe use it in prod, you'd generally just be asking for issues as you'd need to cater for a lot of side-effects, including the fact that the JVM can blow up in the process

[spice temple]

But generally, yes, if you want to write one, it would be a great addition 🙂

[drowsy zinc]

it's already marked as accepted among the repo issues

[harsh yarrow]

Hi, quick question, on that page we can see examples using paper.log-level system property at the beginning, but in the section below "List of system properties" I can't find it. Am I missing something obvious?
https://docs.papermc.io/paper/reference/system-properties/

[eager plover]

No, it's a bad example of using it

the levels are basically hardcoded by default as programatically messing with that stuff would make it somewhat more inflexible

[harsh yarrow]

Ok yes that's why I was surprised to see this then, but I think those examples should be updated to something that actually exist 😅 And do you know where I can find the info about the log level used? I need to send a proof to someone about the actual log level 😅

[eager plover]

in the log4j2.xml file in the jar/server resources folder in the repo

[harsh yarrow]

Thanks

[pliant canopy]

Hey everybody o/. I am currently writing writing on #573 and it so happens that I have a really really weird bug regarding regex.

The bug happens in /src/utils/releases.ts, where after I fetch a release, I do some regex magic (L51-57) to replace certain contents of the body of the releases returned from the GitHub API to be consistent to how GitHub itself and the Kyori docs render it.

This works absolutely fine in local builds. But in remote, GitHub actions builds, this if-statement in L87-91 gets triggered, which basically means that it cannot find the map method of the RegExpStringIterator, which should be returned by obj.body.matchAll. This is completely inconsistent behavior, and I have no idea what the cause could be. Logging into the build logs from the action, it basically just spams these two lines after I added my debug logging: ```
Object [RegExp String Iterator] {}
Yeah it cannot find the map method

The return type is correct, but the map method is just *gone*. Does anybody have any clue as to where the map method  might have disappeared to?

*(I also attached two images. The left side is how it should look, and **does** look in local builds, right side is the result from the CI)*

[elfin arch]

Hello hello!

I'm actually looking at the available configuration properties in PaperMC, but I'm unable to determine what is the precedence order between the bukkit.yml file and Paper World ones (e.g., paper-world-defaults.yml) when it comes to "ticks per spawn".

bukkit.yml:

ticks-per:
  animal-spawns: 400#
  monster-spawns: 1#
  water-spawns: 1#
  water-ambient-spawns: 1#
  water-underground-creature-spawns: 1#
  axolotl-spawns: 1#
  ambient-spawns: 1#
  autosave: 6000#

paper-world-defaults.yml:

entities:
  spawning:
    ticks-per-spawn:
      ambient: -1#
      axolotls: -1#
      creature: -1#
      monster: -1#
      underground_water_creature: -1#
      water_ambient: -1#
      water_creature: -1#

And reading current descriptions doesn't help that much :/

bukkit.yml -> ticks-per.ambient-spawns:

Number of ticks between each ambient mob spawn attempt. Set to -1 to use default.

paper-worlds-default.yml -> entities.spawning.ticks-per-spawn.ambiant:

Determines how many ticks there are between attempts to spawn ambient mobs (bats). Default (-1) uses Vanilla spawn rate.

When I read "uses Vanilla spawn rate", then I conclude that should correspond to the spawn rate observable in single player mode. But then... Does that means Bukkit properties are simply ignored since systematically overridden by PaperMC? Or does I miss something? Is there a kind of fallback to Bukkit properties?
Things are confusing for me...

Can someone clarify things for me please?

[south thunder]

Paper configs will take precedence over Bukkit's

Except when you set Papers' to -1, that means "use Bukkit's configs"

[elfin arch]

Alright! Thank you @south thunder

Then it would be better to have the doc saying something like "defaults to Bukkit’s properties" instead of "use Vanilla spawn rate"

[south thunder]

Hmm, yeah, that's a good point

[elfin arch]

I can make a contribution then I guess

[elfin arch]

To be honest I think it's better to deprecate the Bukkit properties too

And just have PaperMC having default values mapping to Vanilla ones

[elfin arch]

FYI: https://github.com/PaperMC/docs/pull/594

[fair river]

It is planned sometimes in the future to unify the config but it’s not very high priority so we are stuck at what we have. Also, thanks for the contribution

[elfin arch]

Understood!
You're welcome
And thank you you for all your hard work on this amazing project!

[random pollen]

💀 https://pastes.dev/rh6R5tui4u

of course safari and firefox implement file handles but not the methods to show a picker

nothing amazing but if you remember the old dialog it's a lot better

(new one also doesn't have broken layout on mobile)

[spice temple]

Fancy

[pliant canopy]

/pls-help

[lilac edge]

will take a look

[drowsy zinc]

I already told him what's wrong

iterator prototype functions don't exist in node 20 which the CI runs

so it needs to be bumped or he needs to wrap it in an Array.from()

[pliant canopy]

Like 10 minutes ago. Thank you scorpy

There we go, now it works nicely https://strokkur424-feat-adventure.papermc-docs.pages.dev/adventure/version-history/adventure/

[hard quartz]

XD
Maybe can be good add the node version in package

[cloud zephyr]

new docs look amazing 👍

[elfin arch]

I agree!

[hard quartz]

why

[still apex]

codestyle issues package.json

[hard quartz]

[neon epoch]

Classic

[near ice]

That’s me

I would do that

[weak valley]

Is there documentation for the https://docs.papermc.io/paper/dev/userdev/ I can find somewhere? I'm struggling with NMS implementation for a lot of things without knowing the fields specific to my version. (1.21.4)

[drowsy zinc]

nms is not documented, no

[zealous totem]

Hi! LifecycleEventManager's link in the Note section tries to find the class from Java javadoc instead of the Paper one https://docs.papermc.io/paper/dev/registries/#mutating-registries

(it should be https://jd.papermc.io/paper/1.21.5/io/papermc/paper/plugin/lifecycle/event/LifecycleEventManager.html instead)

[drowsy zinc]

fixed, thanks

it should appear correctly once the build goes through

[hard quartz]

looks its not the only one...
i make a few test for a remark check for that but i get a few false positive...

[eager plover]

effective

[hard quartz]

@drowsy zincsorry the ping... but because im dump with web things xd
the whole jd things are a thing generate when build not? its not a thing generated when visit the page?

[drowsy zinc]

yup

it's a remark plugin

[hard quartz]

okay... the rule i try to use to detect invalid url not works in that... its good try to make a check when build the url and try to reach if not throw issue? that can make build fails but more easy notice things like that

im checking with the rule the false positives.. maybe i have a virus but one of them not make sense... https://gradleup.com/shadow/ in the docs are show like imperceptiblethoughts (.com) but that url... well i think now redirect to another place...

[drowsy zinc]

imperceptiblethoughts was the original domain for shadow's docs

uh shouldn't you be able to just put the lint plugin after the javadoc one and it'll get the correct links?

[hard quartz]

i try but dont see any checks.. or maybe i bad add the plugin... let me finish a few renames and open a PR if wanna check

ready.

[hard quartz]

@drowsy zinc by lint you mean remove the two plugins? (Lint and url)

[drowsy zinc]

yea

[hard quartz]

Okay, and how about try to check in the javadoc thing a fetch for the url and throw if not exists?
Also remove the plugin call and dependency or can keep that for allow run the local cli check?

[drowsy zinc]

javadoc verification yea sure; for me the cli checker gets a ton of urls as 403 forbidden even though they actually work, so it's not very useful

[hard quartz]

Ok then im gonna try to refactor the parse jd url for throw an error for notice thats

[trim glen]

Is this pr a good idea? https://github.com/PaperMC/docs/pull/592

[drowsy zinc]

I have no idea about velocity so someone else needs to check it out

[trim glen]

Alright. Thanks

[random pollen]

@jagged pecan @spice temple https://github.com/sveltejs/svelte/issues/15182#issuecomment-2905140106 💀

why is ag grid including a second copy of svelte

[jagged pecan]

I heard you didn't like disk space, so your node modules have node modules

[spice temple]

Version mismatch? Consult the package-lock or whatever package manager you use

Try an override to make it use the same version

I bet 4 vs 5

[random pollen]

yeah copying the dev dependency to overrides and reinstalling fixed it

kind of odd since the ag grid pkg specifies ^5.0.0 which should be satisfied by the dev dep

[spice temple]

Try a proper pnpm update to update transitives then

They don't update transitives most of the time to safe time

[random pollen]

patch roulette uses bun

hopefully bun update is good enough

ok well thats fixed now, but the svelte update that added attachments also added a new check failure 💀

https://discord.com/channels/1194404141389336576/1375521402668716082

[spice temple]

Where does that link go?

[random pollen]

https://discord.gg/fdXy3Sk8Gq

[spice temple]

Mmh, that looks like it's missing some times

I would try to inspect the type on the old version to see what it's supposed to be vs what it is now

[random pollen]

// eslint-disable-next-line @typescript-eslint/no-explicit-any
export type RestProps = Record<PropertyKey, any>;
``` using any instead of unknown fixes it...

[spice temple]

Heh

[random pollen]

god knows why it worked before

they probably had to do some funny tricks to make attachments work

(attachments is the replacement for actions which are more composable since you can spread them onto components and elements in the same way)

https://svelte.dev/docs/svelte/svelte-attachments

[lean venture]

how tf did I not know about this

oh welp

I am now a vue/nuxt fanboy anyways

(not actually, svelte my beloved still)

[random pollen]

there's also a svelte discord but I posted there since I'm using some bits-ui utilities in that component

(mergeProps)

[lean venture]

ah I was in huntabyte's own discord

(creator of shadcn-svelte and bits-ui)

[random pollen]

yeah I'm also using svelte-toolbelt and runed

some of the toolbelt stuff should probably be merged into runed, I opened an issue about that

the box utilities are very useful

[lean venture]

are you using sv5?

or 4?

[random pollen]

5

[lean venture]

how has the experience been?

I have my own opinions, but always enjoy listening to others

[random pollen]

I haven't used other frameworks very much so I don't have many reference points

but generally it has been nice, of course there's less prebuilt stuff than for react

but after looking at react hooks I feel like it's worth it (and I need a bunch of custom components for PR anyway)

[lean venture]

yeah, I agree with you

although from the limited time working on hangar, I have been falling in love with vue/nuxt

ecosystem is nice, and the framework itself does remind me slightly of svelte which I am used to

last time I used vue, it was way worse

they improved it a lot

but one thing I keep mentioning that is an absolute bless from heaven compared to svelte are v-if, etc.

[random pollen]

does that work for components too or just elements

[lean venture]

by elements you mean normal html elements?

either way, works for "everything"

you can use them in template elements, components, html elements

if I am not mistaken, you can use almost all v-(x) in any of those

it's a way better system than how svelte does it honestly

way cleaner

auto import for components also helps a ton

[random pollen]

auto imports works for me with svelte except for 'nested' ones in intellij

like, it gets confused by bits having Button.Root and needing to import Button

[lean venture]

yeah

[burnt ice]

https://docs.papermc.io/paper/dev/component-api/i18n/
this warning can be removed fyi

[pliant canopy]

Is that information being send with the initial login packet now or what changed?

[burnt ice]

Configuration

[pliant canopy]

Ah nice checks out

[rancid yacht]

Maybe that message could again be useful with dialogs sent in config stage?

[opal flare]

@small crane hey! appreciate your work on https://github.com/PaperMC/website/pull/133 - thinking of merging it. we're looking to get the migration done soon, wondering what else needs to be done to get it done. is your todo up to date?

[small crane]

Sorry, I suddenly got super busy and didn’t have time to commit on that anymore, but that PR was complete and ok to merge. I don’t really super remember what my next plan was. I think I was working on one of the heavier components but hit a Hangar CORS error and stopped, but that’s fixed now. If I end up getting time I’ll come back to it but atm it’s in a state that anyone could pick up where I stopped

[fair river]

Thank you @small crane

[slate arrow]

simplify-remote-item-matching: false didnt this option get removed in some commit?

its still present in the docs

https://github.com/PaperMC/Paper/commit/358e72ec0d353fddce4ac99944d668cbee3e8b6c

https://github.com/PaperMC/Paper/commit/04ffca0b6ba65b0c711ffc3ac592c2302a48c0a9

[pliant canopy]

Lol

Epic logo load fail

[random pollen]

https://github.com/PaperMC/patch-roulette/pull/50 https://papermc-feat-directory-compa.patch-roulette.pages.dev/diff previews working

[winter verge]

Heyo,

I use the jsdelivr devins minimal badges on my github to show what I support like I see many others do. I noticed that they did not have one for Folia, so I figured I would take the backdrop they use and slap the Folia leaf svg onto it. I was curious if there would be any issue with me doing that. I am messaging here because the docs website has an assets branding page that mentions that I can use the PaperMC logomark to represent a project in blogposts and other places to bring attention to the project. It does not mention anything about Folia in particular, though. I was also curious if there was any specific issue with jsdelivr 's paper badge also shown in the screenshot since it does not use the exact logomark provided on the asset page. Is that considered a "modified version of the PaperMC logomark or derivative work of it?"

If I am overthinking things, or in an obviously wrong spot to ask this, feel free to just tell me so 🙂

[neon epoch]

I’m not aware of any issues but I’ll CC some people :)

[opal flare]

I'll get back to you soon.

[winter verge]

Thank you

[tepid rapids]

as i understand it, because of the hardfork spigot plugins aren't garunteed to work with paper anymore right? even if for now most of them will probably function fine?

because i dont think this article mentions that https://docs.papermc.io/paper/adding-plugins/

unless im understanding it incorrectly

[lilac edge]

all plugins made before softspoon happened will continue to work for the forseeable future. there will be proper communication/documentation/replacement api done to accomodate users of spigot api when it gets ripped out farther down the line. any plugins that are based on any spigot api that were introduced into spigot after softspoon happened will not work on paper. some new spigot api may work if paper cherrypicks spigot changes at the maintainer's discretion

[tepid rapids]

ok

but the softspoon has happened already right? so the article should mention it but it does not at the moment

[steady fulcrum]

Yeah, most likely no one got around to it yet

[vagrant olive]

hi all, does PaperMC have some kind of plugin requests channel? I'm looking to write a plugin

[scenic gull]

I don't see how this question relates to our documentation, but: Do you mean you're looking for help writing a plugin, or you're looking to hire someone? If the former, #paper-dev , if the latter, no. I suggest helpchat (see #community-guilds ) for that.

[vagrant olive]

Sorry, I was imagining you might have something like the old spigot, there used to be a some kind of spigot forum for this. I'm pretty sure it's dead.

Thanks for the pointer!

[random pollen]

https://github.com/PaperMC/docs/pull/602

[gritty orchid]

https://youhavetrouble-main.papermc-docs.pages.dev/paper/reference/world-configuration/#anticheat_anti_xray_hidden_blocks

I feel like this is way more clear which section I linked to :3

[random pollen]

need to figure out how to make the new compare files gui not ugly (old on top of white line)

maybe one of these would be better than checkbox

[fair river]

🪤 Summoning the webdev lurking

[random pollen]

this seems a bit better

(ignore the uneven gap around the arrow, some debug element is there)

[lean venture]

That indeed does look a bit better

I can hangarify it if ya want xD

[brazen moon]

the github auth doesn't seems to works anymore since the migration and the app button still redirect to patch roulette instead of diffs-dev

[random pollen]

I have the link fix for the button locally, let me push it

and I'll test the gh auth

[random pollen]

classic github

perfect text wrapping

[twilit fog]

D

[ocean void]

. . . is this the correct channel/server for this message?

[random pollen]

look at the pinned issuein the screenshot

[ocean void]

Ahh - the image thumb was too small for me to see without clicking.

[random pollen]

fixed

[brazen moon]

thanks

[random pollen]

anyone know how I can get the CORS error to display to the user instead of 'TypeError: Failed to fetch'

ideally in a browser-agnostic way

    private readonly urlResolver = $derived.by(() => {
        let url = this.url;
        if (!url.startsWith("http://") && !url.startsWith("https://")) {
            url = `https://${url}`;
        }
        return lazyPromise(async () => {
            let threw = false;
            try {
                const response = await fetch(url);
                if (!response.ok) {
                    threw = true;
                    throw new Error(`
Failed to fetch from URL: ${url}
Status: ${response.status}
Body:
${await response.text()}`);
                }
                return await response.blob();
            } catch (e) {
                if (threw) {
                    throw e;
                }
                throw new Error(`
Failed to fetch from URL: ${url}
Error:
${e instanceof Error ? e.message : e}`);
            }
        });
    });

[random pollen]

I gave up: ```ts
throw new Error(Failed to fetch from URL: ${url}\nSome errors, such as those caused by CORS, will only print in the console.\nCause: ${e});

[random pollen]

interesting that tailwind doesn't include ```css
textarea {
vertical-align: bottom
}

with the default setting textareas have extra margin on the bottom in chrome and ff

[neon atlas]

gonna emit the average data converter diff

[random pollen]

https://github.com/PaperMC/diff-viewer/pull/7 https://papermc-multimodal-input.diffs.pages.dev/

I should create an action that comments a link to view the PR diff in the preview viewer

like https://papermc-multimodal-input.diffs.pages.dev/?github_url=https%3A%2F%2Fgithub.com%2FPaperMC%2Fdiff-viewer%2Fpull%2F7

[rain ocean]

does no one else have this issue? been a couple of weeks like this. the entire site jitters if u click something on the sidebar

not an issue in safari, but it is an issue in chrome even an incognito tab w/ no extensions

[drowsy zinc]

looks like chrome doesn't want to preload the font or something

[rain ocean]

it makes it quite the horrific experience

Other astro/starlight docs don't have this issue w/ chrome

[drowsy zinc]

we don't use the system font unlike the rest so

[rain ocean]

are you able to replicate on your end?

[drowsy zinc]

on mobile chrome yea

[drowsy zinc]

could you see if it's fixed?

[random pollen]

triangle or no triangle for popovers

[pliant canopy]

With triangle

[vocal halo]

took me a while to see what you mean but now that i see it i prefer with

[random pollen]

https://github.com/PaperMC/diff-viewer/pull/7/commits/a7f33e2b2989f7c7bc176df94c379709fc0a3e50

Added the triangles

[acoustic wing]

Where to Download Paper 1.16.5?

[fair river]

Also wrong channel @acoustic wing

[random pollen]

I really need to do some refactors but adding features is more fun

https://papermc-file-type-override.diffs.pages.dev/

[lean venture]

looks pretty neat

so this website is mainly to show differences between commits?

or uh

?

[spice temple]

It's a better diff viewer since githubs got increasingly worse

And then jmp escalated from there

[lean venture]

Some criticism: Found it a bit confusing to use when I first used it

not sure if it's just the layout or idk

but after you understand it it's pretty easy

[bitter heron]

Maybe a drop down to select which option you want to use

Like from patch, comparing two files directly, etc

[lean venture]

hmm

maybe

[pliant canopy]

@random pollen what is this?

lmao

[hard quartz]

where? xd

[pliant canopy]

Own patch

I actually use the extension to look at non paper related patches

[random pollen]

a bug

i have it written down already since i ran into
it before

but you can open an issue if you want

[pliant canopy]

I mean, sure, why not

[uneven ocean]

bro what is this discord

is this for roblox?

@hard quartz

@pliant canopy

@hard quartz yo bro help me out for a bit

[fair river]

@uneven ocean you can read #welcome if you want to know what this server is about. (though idk how you end up here without knowing)

[uneven ocean]

my bad...just that this discord hs alot of channels

[ocean void]

It does, yeah. But randomly pinging people is generally considered rude in any Discord. Just FYI.

[uneven ocean]

Yeah, my bad for that...

[vapid shadow]

Sorry if this is the wrong place to ask, am I doing something wrong here?

[eager plover]

I guess it failed to load the perms list or something, #hangar-help too

[random pollen]

removing padding around the main content yes or no

personally I think I should have done it sooner

still need to figure out how to place nicely with mobile safari's dynamic viewport... not sure if I can or want to use the body as the virtual list

right-side file tree also coming for all the weirdos who like that :p

[random pollen]

better in some ways but overall I think it's worse just making that single change

[vocal halo]

I think the top bars location shouldn’t change at all. Only the dynamic content should affect the height. Otherwise it looks way to jumpy somehow.

[vocal halo]

Seeing it only on the screenshots on mobile with padding looks better imo

[pliant canopy]

With padding is clearly better

It doesn't remove much space, but it nicely seperate the parts from eachother imo

[eager plover]

Docs team hates this one simple trick

[neon epoch]

We resent all authority

[pliant canopy]

@drowsy zinc I took the liberty to rebase my feat/adventure branch to the latest main commit, hope you don't mind. Thought I should notify you of this before you change something locally and get a fat merge conflict

[drowsy zinc]

you're one second too late

[pliant canopy]

Damnit

[drowsy zinc]

I cherry-picked one commit of mine because it fixed a bug with velocity's JD social icon

[pliant canopy]

Should I rebase again?

[drowsy zinc]

yeah if you don't mind

[pliant canopy]

sure

Yay, conflict

(Nothing I cannot handle though, dw)

Also, uh, cloudflare pages broke?

[drowsy zinc]

needs some time to propagate probably

[pliant canopy]

Weird, I have never seen that before

Anyways, I might as well ask you directly @drowsy zinc: Did you remove the quick fix I had here?

Just need to verify that my rebase didn't break something

[drowsy zinc]

yes

[pliant canopy]

Okay good

There were quite a few conflicts for some reason lol

only in the astro config and that one version util though

There, all rebased

[drowsy zinc]

fyi, I'm already working on the documentation changes for the maven central stuff

[pliant canopy]

aight great

I couldn't have looked into it until sunday anyways

[random pollen]

you should make mention of the region-specific mirrors:
NA: https://maven-central.storage-download.googleapis.com/maven2/
EU: https://maven-central-eu.storage-download.googleapis.com/maven2/
Asia: https://maven-central-asia.storage-download.googleapis.com/maven2/

[broken flint]

Is Hangar down rn?

I just get a fatal error whenever I try to visit it

[twilit fog]

#hangar-help and as Mini said, keep reloading if you get 400

[broken flint]

Oh, I dont have that channel lol

[twilit fog]

hey I have a suggestion, have you ever considered adding like a search bar for configs? I always forget which config a setting is in and would find it much easier to just search for it

[drowsy zinc]

hmm, I could try to get it to be indexed in the main search

no guarantees it'll work out though

[tight shard]

Is this still the case?

[pliant canopy]

Why wouldn't it be 💀

Oh you meant the theme, I thought you meant the code example

[tight shard]

the theme lol

[pliant canopy]

No the theme changed since the Astro migration. @drowsy zinc which one is it now?

[tight shard]

i may be blind but can't find it in the code

apparently if none is specified it defaults to github-dark, but it doesn't look like github-dark

[lean venture]

tbf that could simply be replaced with Player#sendRichMessage()

that util method is essentially useless

[pliant canopy]

I think it is night owl

That's the default starlight theme

[tight shard]

oh what

i read it was github dark

[pliant canopy]

There are more uses for components than sending a message 😭

[lean venture]

that is true....

and a very good point....

[pliant canopy]

I asked in the astro discord a long while back

[tight shard]

but you are right yes

https://starlight.astro.build/reference/configuration/#themes

thank you

[pliant canopy]

👍

[lean venture]

although idk

// Constant holding the mm instance named MM

MM.deserialize(""); // Can receive placeholders too!
// or
Components.mm("");

[pliant canopy]

Static import mm()

And placeholders can be done with adding a TagResolver varargs param

Anyways, not the point

[echo canyon]

wow, ok I haven't look at the docs page in forever. it looks pretty nice.

I like the dropdown to switch entire projects, that's smth I've been wanting for a while. But tied in with that, I kinda wanted the buttons on the top to be Admininstration, Development, Contributing...

Because I want the sidebar to remove one level

how often are you wanting to read docs pages inside multiple of the top level categories? I feel like the extra level is just unneeded and takes up space

Making it so everything in the box was all that's visible if you are in the administration section of the paper docs

[neon epoch]

Yeah I know what you mean. Odds are if you are opening docs you are either a dev or an admin

[eager plover]

jetbrains mono font

[zealous sinew]

theres an incorrect value in the docs

https://jd.papermc.io/paper/1.21.6/org/bukkit/spawner/Spawner.html#getMaxNearbyEntities()
the default value is actually 6, not 16

[vocal halo]

It also says „set the new…“ but I think it’s meant to say „get…“ no?

[spice temple]

open an issue and/or PR pls 🙂

[pliant canopy]

@random pollen (no clue where else to ping you for this lmao) but two things:

1. If the new papermc-pr-publishing workflow, do you think it could make sense to temporarily set the message contents to "Still building" or something during builds so that one doesn't accidentally grab an outdated artefact while the PR is still building? I know it says "Last updated for", but I doubt everybody just knows the commit hash of the latest one or cares to check lol
2. Do you plan to add a way to download PR versions on your run-task gradle plugin?

[burnt ice]

I can prolly look at the first one

[last bear]

https://i.imgur.com/ApgQjwr.png and it's not even the right page

[drowsy zinc]

¯_(ツ)_/¯

[spice temple]

Is Kenny using bing?!

[last bear]

no LMAO

[echo canyon]

lol what? Is that the gist I originally wrote the spec for it on?

[still apex]

https://docs.papermc.io/velocity/configuration/ looks really ugly

https://i.imgur.com/SjDS7RR.png

[spice temple]

love tables

[drowsy zinc]

more importantly the page doesn't even have all entries from the configuration file

[unborn steeple]

wait you are german?

[opal flare]

who wants to update the website to use fill v3 endpopints

[steady fulcrum]

scorp voluntold me to

then regretted it

[steady fulcrum]

I am actually working on it though

Or well the examples

[opal flare]

can we also look at https://github.com/PaperMC/docs/pull/573 soon @neon epoch

[neon epoch]

80 files changed 🫠

Yeah lemme take a look now if not between starting a job and moving house it’s gonna take a bit

[steady fulcrum]

ollie I apologize in advance for my cursed jq

[opal flare]

reminds me that we really need to finish the website migration too :(

[neon epoch]

Ok i reviewed it - 1:30 am and i just got home from a flight but 😂

It’s really nice, good work. I skimmed the platform implementation stuff cause frankly I don’t know enough to comment on half of them

Was just a couple places where, as I say in the review, we are mentioning specific versions and idk if that’s intentional

I’m not gonna be able to test the fill code examples - but they look good. Happy for someone else to test and merge :)

[vocal halo]

To be fair, most of it is just directly copied from the current adventure docs

There are just some unavoidable formatting changes no?

[neon epoch]

Likely

[meager herald]

Am I overlooking something or does Fill v3 no longer give you the project Name?

[still apex]

an ID

[meager herald]

Yeah you still get the id but no longer the name. I guess I could hardcode the names but it's not that clean.

[still apex]

most of the time it’s just the id but with the first letter uppercased right

[opal flare]

it will soon; I have a few changes locally that are coming

[meager herald]

[opal flare]

should be live now

[still apex]

what all changed besides adding 'name'

[opal flare]

just that for now; delaying the rest

[still apex]

what's the rest :p?

[opal flare]

🙊

[jagged pecan]

new rest endpoint? :O

/s

[steady fulcrum]

riley changing the api to make my life hard

[drowsy zinc]

ollie

greatest PR merge velocity

[neon epoch]

Great minds think alike

[pliant canopy]

Finally, this is moving along

Though this is blocked on that one bug report by scorp anyways

I have a feeling one of us will have to fix that because the issue isn't moving along

[drowsy zinc]

I already tried looking into it but no idea where to start

the easy solution would be to disable the stylesheet inlining, but that makes for ugly layout shifts

so that's a no-go

[pliant canopy]

I mean, something something include the css stylesheet, but I also dunno

that's all I could read out from the issue comment

[ancient quiver]

much faster merges than Paper ;)

[spice temple]

@steady fulcrum the javadoc link to folia broke, folia only hosts major version javadocs, so link would be https://jd.papermc.io/folia/1.21/

velocity is broken too, idk why yet

ah, velocity needs to be https://jd.papermc.io/velocity/3.4.0/

[spice temple]

@opal flare thoughts on https://github.com/seiama/javaducks/pull/22 ?
I have seen multiple 3rd party websites to workaround this now.....

[steady fulcrum]

Cursed but fixable. I’ll get a PR rolling for it.

[spice temple]

Guess no other way than to special case it all

Or, mmmh, the API could include that info?

But right now java doc hosting is all manual

[steady fulcrum]

Yeah I'll have to special case it.

I think before the website always just used the version group listing

[opal flare]

merged, I thought I merged it ages ago

[spice temple]

Ty

[wispy crypt]

folia can go on the website now

[pliant canopy]

A rare leaf message

Hello leaf!

[still apex]

get ignored

[trail jasper]

Build explorer on folia dl page opens the paper build explorer

[steady fulcrum]

scroll down

[spice temple]

It would still be a nice feature to have like query param for that

It's a frequent issue

Oh, also, wanna gift the commit view some work wrap?

[vocal halo]

One is a single commit the other is multiple commits within one release

[sharp pivot]

the commit view including the whole commit description looks weird to me, having some buttons to expand the full description would seem nice

[steady fulcrum]

yeah it feels a little verbose (especially for leaf commits)

[steady fulcrum]

Fwiw, 1 commit would word wrap, the others then wouldn't

[steady fulcrum]

Any chance we can get pngs of Velocity' logo on https://docs.papermc.io/misc/assets/?

[drowsy zinc]

cc @opal flare

[still apex]

why are we now going to switch to nextjs 15

[opal flare]

It exists already - https://assets.papermc.io/brand/velocity_logo.256x128.png
/cc @steady fulcrum

[steady fulcrum]

mf hiding it from me

[opal flare]

you'll also find folia_logo.256x139.png useful

[trim dirge]

just wondering does papermc.io use wordpress or smth like that? or does it use custom code like react, node.js, etc

[still apex]

nextjs

https://github.com/papermc/website

the plan was to switch to astro but that apparently(...?) got scrapped

[spice temple]

we are reevalutating the options, yes

[still apex]

well the pr got closed and an issue assigned to jroy for nextjs 15 got opened

[trim dirge]

I see

if I like build my own site somethign similar to papermc.io what should I use?

i know it is a bit offtopic

[spice temple]

whatever you want really

it depends on a ton of things

[random pollen]

for the docs side we don't change too much from the base starlight setup

but for the landing page and downloads/etc you can use whatever

[steady fulcrum]

Btw I did fix a few of the issues mentioned in https://github.com/PaperMC/website/pull/143 alongside some stuff to make commits less pain to look at when they're long.

Mostly parsing out commit links and turning newlines into <br>

Folia for example, although idk how to feel about the buttons being vertically centered with text that tall

[random pollen]

yeah probably want top align with a thin divider

[grand dew]

do you also wanna fix the seo i accidentally forgot to swap when i copied papers page

[steady fulcrum]

The SEO looks fine

[grand dew]

check minis review on my pr

that is papers seo but called folia not folias actually desc/seo

[steady fulcrum]

ah

[grand dew]

https://github.com/PaperMC/website/blob/main/src/pages/software/folia/index.tsx#L21 actually seo should prob be this

[steady fulcrum]

I think having Download Folia first is fine

It'll hit when you google how to download folia

[grand dew]

that bit yeah, mostly the rest of it

[grand dew]

lgtm

[steady fulcrum]

that's better

[brazen moon]

i don't think the commit message should be included or maybe only in a hover popup

[steady fulcrum]

Fair, it's been something I've been fighting on my webhook too. Since leafpeople like to make long commit descriptions.

yeah that makes it a lot better

[random pollen]

long proper

[steady fulcrum]

[random pollen]

yes I am calling myself out

[steady fulcrum]

looks betterer

[random pollen]

the padding around the button is uneven somehow

[steady fulcrum]

[random pollen]

bottom shorter than top

[steady fulcrum]

fuck me it is

[random pollen]

margin*

guess it could be either, not sure what the "correct" term is

[opal flare]

Fill gives builds certain colours based on the channel, should likely do that too instead of just the red for the build number. I can give you the colours I use

[steady fulcrum]

Ya pls do

[opal flare]

public static final Color COLOR_BLUE = Color.of(0x4fc3f7);
  public static final Color COLOR_GREEN = Color.of(0x4db6ac);
  public static final Color COLOR_PINK = Color.of(0xf06292);
  public static final Color COLOR_ORANGE = Color.of(0xffb74d);
  public static final Color COLOR_PURPLE = Color.of(0x7e57c2);
  public static final Color COLOR_RED = Color.of(0xff6f61);
  public static final Color COLOR_YELLOW = Color.of(0xfff176);

Alpha is red, Beta is yellow, Stable is blue, Recommended is green

[steady fulcrum]

holy flashbang

[serene vault]

pretty poor contrast with the text

[opal flare]

You can use colorhexa.com to get a darker version of the colour

Just make sure they all match

(If you need to make it darker)

[steady fulcrum]

COLOR_GREEN
dark teal

Darkened them all by 1 shade and they're readable now

[grand dew]

is riley colour blind

[steady fulcrum]

https://www.colorhexa.com/4db6ac this is apparently green

It's kinda green

[opal flare]

In a RGB color space, hex #4db6ac is composed of 30.2% red, 71.4% green and 67.5% blue.

[steady fulcrum]

[serene vault]

its ugly. reminds me of the color of operating rooms

[steady fulcrum]

good thing only velocity uses it

[opal flare]

If you want to make a new palette of colors that look good on the website in both light and dark, you can do that

Just give them to me so I can update other things with them

[steady fulcrum]

I ain't no artist

it may be my last name

I'll make it easily configurable so someone else can go fuck withi t

[opal flare]

lmao

[steady fulcrum]

consider yourself colored

[south thunder]

How about a color that's more similar to the link color?

[random pollen]

@burnt ice your regex change PR feels kind of fragile, couldn't a diff theoretically still false positive for a signature?

it's probably good to move away from a regex there in general, but for now maybe we could do something like preprocess the string to remove a signature before passing it to the original regex from before your other PR

[burnt ice]

I completetly agree, regex is yucky

[random pollen]

can maybe substr(lastIndexOf(-- ), end)

and then check if theres any more plus minus or header lines in that

if not assume it's a signature?

or idk, that kind of just moves the issue to the last patch

need to take into account the last hunk header...

maybe I'll just merge the PR as-is for now, and then look into using the methods from diff

@burnt ice

[burnt ice]

Yeah that sounds good. We defo need to cleanup the state handling for that.

I am surprised there are no libraries for that already, however.

[random pollen]

there is, diff

we use it for parsing individual patches

it can parse multi-file, but I don't use it because I need a way to inject the image diffs for github

[burnt ice]

I see

[random pollen]

it might be possible, I haven't tried yet

I hope it is

the github website has been so glitchy for me lately

every 5th reload causes half the css to be broken

on every page

I also swear it used to use the PR title by default when squash merging

[burnt ice]

I had a bug where the same file kept showing up repeatedly

(in github viewer)

[random pollen]

yeah the file browser especially has been bad

((I don't use the diff viewer))

I guess you can use the in-browser vscode view

thats probably better

you press . iirc

[lilac edge]

in-browser vscode view is goated

i use it everytime i need to search for keywords in repositories

[pliant canopy]

So it's not just me? I've been having issues with certain elements being on top of each other

[drowsy zinc]

what's the benefit of migrating the website from node+pnpm to bun?

[opal flare]

It’s just bun for building as far as I know, and it builds so much faster

[spice temple]

Bun is nice and fast

[opal flare]

like, unreasonably fast

[drowsy zinc]

doesn't it build faster just because it uses turbopack now?

[opal flare]

no, bun itself is like extremely fast

even installing deps

Etc

[spice temple]

If you want you can still install them with whatever package manager you like of course

[drowsy zinc]

oh okay

I just didn't want to install another runtime just for the website if I end up needing to contribute something

[spice temple]

We don't use bun exclusive features, you can use whatever runtime and package manager you want

Sadly corepack doesn't support bun yet

And it seems like corepack in general is dying? So idk if it ever will

But that solved the package manager stuff for me in the past

[jagged pecan]

Bun poo-poo ed it's inclusion it looks like
https://github.com/nodejs/corepack/pull/307#issuecomment-1718747222

Although that's not the only thing

[mighty dirge]

would it be ok to cleanup some of the item meta usage?

e.g. item.editMeta(meta -> {
-> item.editPersistentDataContainer(pdc -> {
-# https://docs.papermc.io/paper/dev/pdc/#adding-data

[pliant canopy]

I don't see why not, go for it

[mighty dirge]

Currently, the Last updated tag seems to be mostly useless because it tracks all commits editing the page. What if it only tracked commits that have some sort of indicator they updated that page?

[pliant canopy]

I am not sure I follow?

[mighty dirge]

that was not a page update

the page was just modified but the content wasn't really updated

[pliant canopy]

I mean, the page was still last revised in the commit

I think it's fine

It says "Last updated" and not "Last updated content"

[mighty dirge]

Sure, but it could be improved and could be a nice indicator in what state this site of the docs is in and currently it does not serve that purpose

[drowsy zinc]

no docs website ever does that

that commit fixed broken links on that page, which is a meaningful change

our main indicator of up-to-date-ness is the version tag

[mighty dirge]

Altough that seems unused (at least I can only remember maybe 1 page with it)

[drowsy zinc]

for now

[mighty dirge]

-# will add that to my pr

[drowsy zinc]

if you could note the version those read-only views were added in, that'd be great

[neon epoch]

Can y’all stop blowing up my emails on my watch whilst I’m swimming 😭

[still apex]

no

i'm going to send you more emails

[left egret]

Needs triage

[vocal halo]

If it’s apple just put in a focus lol

[burnt ice]

just stop swimming

[rigid bluff]

At the moment to drown or get out before ? :D

[random pollen]

I just transferred the diffs.dev github app to papermc

so if you run into issues with github login ping me

[brazen moon]

works fine for me, only thing i noticed is when you search a text in the diff and nothing is found, then click the arrows you get a NaN / 0

the new dialog registry is missing in https://docs.papermc.io/paper/dev/command-api/arguments/registry if you want to update it Strokkur

[pliant canopy]

Ohh we have a dialog registry now? That's exciting

I'll play around with it a bit and update the page 👍

[spice temple]

I think dialogs even deserve a general page for how to use them

[vocal halo]

Yeah

[pliant canopy]

Well, I guess I have just been given yet another side quest, darnit xd

I will never be able to escape the cycle of projects

[vocal halo]

You like it, admit it.

[pliant canopy]

B.. but I wanna get to my own projects too, which have been piling up 😭

[vocal halo]

Then do it?

[drowsy zinc]

less talking, more shipping

[vocal halo]

No one is forcing you to work on paper docs

[drowsy zinc]

I am

[pliant canopy]

Yes, scorp is

[vocal halo]

Ah

In that case

[pliant canopy]

welp, now I gotta learn dialogs

Is there documentation on it for me to read?

Oh wait, there isn't

[drowsy zinc]

one person had a test plugin on their paper fork here

and mm had it as a part of his branch before merge

[pliant canopy]

I mean, it's not hard to figure out

I grasped the basics already

[neon epoch]

That intuitive huh? Guess we don’t need docs

[drowsy zinc]

so true

self-explanatory code

[vocal halo]

Let’s just link to strokkurs repo

[twilit fog]

mbax made a demonstration

[pliant canopy]

Well, one thing I would like to know: Is it possible to resend the dialog registry to the player via code? Like for commands we have Player#updateCommands. Is there a Dialog-variant of that?

I cannot seem to find any method for that in the JavaDocs

@opal flare are the module names supposed to be cut of like this in the jd?

Feels like somebody screwed up a String.substring

[pliant canopy]

Okay first things first, this is incredibly incredibly verbose and hard to find a good way to format for

And I don't even have any body content

Just format it like this ```java
ctx.getLifecycleManager().registerEventHandler(RegistryEvents.DIALOG.compose(), e -> e.registry().registerWith(DialogKeys.create(Key.key("strokkur:cool_dialog")), builder -> builder.empty().base(DialogBase.create(Component.text("A dialog title", NamedTextColor.GOLD), null, true, false, DialogBase.DialogAfterAction.CLOSE, List.of(), List.of())).type(DialogType.multiAction(List.of(ActionButton.create(Component.text("Cool button", NamedTextColor.GREEN), Component.text("Cool tooltip", NamedTextColor.LIGHT_PURPLE), 250, DialogAction.staticAction(ClickEvent.callback(audience -> audience.sendMessage(Component.text("You're cool!", NamedTextColor.RED))))))).build())));

Hm, maybe like this

[twilit fog]

Use builders

If you want I can send you how I did it

[pliant canopy]

I am aware, don't worry

[twilit fog]

Kk

[pliant canopy]

I am just playing around with the possibilities right now

[twilit fog]

[pliant canopy]

first things first, the size width/height attributes on the item ones are a bit weird

[twilit fog]

They dont work at least if you add plain message

I havent tested if they're standalone

[burnt ice]

We dont have api for this. Technically its possible but not without crazy server refactors

the server doesnt support it, the client does

[pliant canopy]

Okay got it

Btw do you know how I can keep a dialog open during config phase?

It shows for me, but just closes instantly

Or do I have to block the AsyncPlayerConnectionConfigureEvent until I get a response from the dialog?

[echo canyon]

Wait, doesn’t config phase resend it?

[burnt ice]

No its not resent by default when bumping them back to the config phase, but its possible to do, yes.

[echo canyon]

Oh I thought it was resent by default.

[burnt ice]

No, bumping them back to config phase is basically a noop, all their data is retained.

Its alot different than i thought it was

[pliant canopy]

Wait so is this considered the "right way" to do stuff like that: https://pastes.dev/qXl7vkFLeu

Feels awfully hacky

(And adding a processedConnections.remove(event.getConnection()); so it gets cleared lol)

[echo canyon]

You can use a completable future too

[burnt ice]

Yeah I would rather use a future

[pliant canopy]

Okay but that works, nice. So how would I make this use a future instead?

(also wth is this)

[stable copper]

player texture information, base64 encoded

[pliant canopy]

Well duh

But why is that there

Feels like an oversight

[burnt ice]

I thought we had a fix for that kek

[pliant canopy]

Anyways, what would be the future approach?

[burnt ice]

CompletableFuture<Boolean> future = new CompletableFuture<>();
instead of Boolean

In the AsyncPlayerConnectionConfigureEvent do
boolean res = processedConnections.get(event.getConnection()).join();

in PlayerCustomClickEvent
processedConnections.get(event.getCommonConnection()).complete(false/true);

something like that

[pliant canopy]

Ahh

Okay that's smart

Yeah works wonderfully https://pastes.dev/oQWDT7tdgx

[pliant canopy]

Is there any API for modifying the quick actions or custom options dialogs?

I cannot seem to find anything regarding that in the javadocs

[brazen moon]

you can use the tags

[pliant canopy]

Could you elaborate? I assume you mean the DialogTagKeys, but how can I work with that

[brazen moon]

yeaa i guess this part is not documented yet on the website

[pliant canopy]

Also more general question, but is there any reason why the pause(boolean) method is exposed at all in the dialog base builder?

Since it only works in singleplayer anyways

[pliant canopy]

Oh no, I smell more documentation work

[brazen moon]

probably just to match vanilla generally this api is experimental and can change

[brazen moon]

that's only documented on the javadocs rn (and a PR) see Pre/PostFlattenTagRegistrar

iirc there was a reason to not put it directly on the docs

[pliant canopy]

Ohhh

So that's what the TAGS lifecycle event is for

[brazen moon]

so you put your custom dialog or built-in inside those tags

so they appear on the client/works on the keybind

[pliant canopy]

Mhmhm

Considering this whole tag registrar api is not documented yet at all, does it make sense for me to show that you can edit the quick actions/etc this way

Or should I leave it out and it will be added as part of the general tag registrar docs pr?

I suspect that this specific part of the dialog api will not survive the next version anyways considering the interest of switching most dialog api to adventure

So I am a bit reserved, eventhough I am only speculating

[brazen moon]

i think a smaller example should be taken for the "general tag registrar" but probably better to do it after rather than introduce a whole thing

[pliant canopy]

Fair enough

I doubt many people are in need of it atm anyways, and if they do, they can ask in #paper-dev

@drowsy zinc does your JavaDocs link parser/validator not allow for nested classes? I have a jd link declared as jd:paper:io.papermc.paper.registry.data.dialog.DialogBase.Builder, but it complains to me: ```ansi
20:40:44 [ERROR] Javadoc link "jd:paper:io.papermc.paper.registry.data.dialog.DialogBase.Builder" is not valid
Stack trace:
 at C:/Dev/Web/paper-docs/src/content/docs/paper/dev/api/dialog-api/introduction.mdx
 [...] See full stack trace in the browser, or rerun with --verbose.

[drowsy zinc]

you need to use $ to denominate inner classes

[pliant canopy]

Ah

Smh

[pliant canopy]

@burnt ice can I somehow either get an audience or a player from a PlayerCommonConnection?

[burnt ice]

instanceof GamePlayerConnection -> getPlayer()
instanceof ConfiguringPlayerConnection -> getAudience()

[pliant canopy]

Fair enough

thx

I assume it is either the one or the other? So I cannot just get an audience in general without instanceof-ing both?

I don't assume the player to be in the configuration phase when they use this dialog, but it's still an edge case in my understanding

[pliant canopy]

There you go: https://github.com/PaperMC/docs/pull/621. Side quest completed

Only took me an entire day

Better be grateful

@neon epoch would you have a problem with it if I were to rebase your mob-goals branch and continue the PR for you?

[neon epoch]

Enjoy it

I just started work so yeah I’m busy lol

[pliant canopy]

Alright, fair enough

Expect a superseeding PR somewhen soon™

I just think we should merge the entity pathfinding docs pr soon, as it really is overdue

[neon epoch]

Calm, I, or probably whoever gets to it first, will close the other lol

IIRC it was blocked waiting on someone who actually knew the API to review

But that was a hot minute ago

So yeah can be done now

[pliant canopy]

Yeahhh okay I see

[neon atlas]

@random pollen https://github.com/orgs/community/discussions/163932 you scared github with diffs.dev (still ass tho)

[jagged pecan]

Tangentially related (found it using that thread)
:shipit: on GitHub

https://github.githubassets.com/images/icons/emoji/shipit.png

[spice temple]

That existed for absolute ages

[neon atlas]

oh

I just got it

how is it still so slow tho

[burnt ice]

It’s just the PR diff viewing, not generic commit diff I think

[neon atlas]

did test that on the 1.21.7 PR

was still slow af

[burnt ice]

true

also ik theres this vibe coded app which has been popping off

https://github.com/yoshiko-pg/difit

its on command line tho lol

likeeeeee

[drowsy zinc]

love claude

[burnt ice]

@neon atlas opinion if you started seeing that in my commits?

[neon atlas]

I always wanted a reason to ask for org ban permissions

[drowsy zinc]

at least the code will be good

[neon epoch]

lol

[spice temple]

is this something we want? https://github.com/PaperMC/website/pull/90

[ocean void]

IMO it's not necessary

[spice temple]

@steady fulcrum pls rebase https://github.com/PaperMC/website/pull/143, I think some of that was done, like the overflowing commit messages

@neon epoch do you remember the reasoning behind https://github.com/PaperMC/website/pull/108?

[neon epoch]

Ummmm no tbh

Lemme have a quick think

^ @spice temple think it’s that

I never did get that $5

[spice temple]

Ahh

Can you rebase and add it to all pages? Can prolly find and replace for the string of classes

(I really don't like utility css shit, scss + vars so much nicer)

[steady fulcrum]

Done, it was fully done, just Riley wanted to review it I think idk

[last bear]

only when it's merged you have a new sponsor

[spice temple]

reeee I fixed this yesterday, how did you broke it again? 😄

ill fix

huh now it doesnt happen again

uff, the website repo has no action to check that stuff passes lint and tsc and shit

love how I get another error locally

love react

https://nextjs.org/docs/messages/missing-suspense-with-csr-bailout

what do you mean!?

I have a giant "use client" on top, I know that its csr....

react really has the most complicated ssr/csr division possible

its live

I also fixed the legacy build page flashing when you opened the build explorer

ok, onto fixing my fix for fill-gradle/waterfall

[spice temple]

idk who needs to heard this but waterfall 1.21.7 is now available on the website

[neon epoch]

I did rebase it, didn't do that though lol, didn't see this

[neon epoch]

🙏 Thank you

[echo canyon]

so what is the plan with the adventure docs? @pliant canopy 's pr, is that gonna just flat out replace the advntr docs page, or what?

[still apex]

what would happen with the existing adventure doc links? get redirected to the proper pages..?

[echo canyon]

yeah, idk. thats why I'm asking. I'm sure its been discussed somewhere

[drowsy zinc]

the path structure should be the same minus the /adventure prefix, so it should be easy to redirect

[pliant canopy]

I think riley can tell you more on that, so you should poke him if he doesn't respond here

[fair river]

I believe that is the plan. But yeah we can ask riley to confirm later on the day when he pops up

[spice temple]

somebody wanna play with this? https://delucis.github.io/starlight-llms-txt/

[steady fulcrum]

I'm down and bored lol

[opal flare]

yes

[steady fulcrum]

nvm gave up when it started throwing

[pliant canopy]

[twilit fog]

Has a particle guide come up to discussion yet? I feel like there should be something since the spigot forum one is outdated a bit (at least in the sense of the particle list) https://www.spigotmc.org/threads/comprehensive-particle-spawning-guide-1-13-1-19.343001/page-2

I'd be happy to make a draft if you like the idea :D

[velvet gulch]

pull requests are always open ^^

[neon epoch]

Yeah draft something up

[twilit fog]

I know, I just like to ask before doing some work which gets turned down not because of crap code, but because of the idea

[velvet gulch]

i think everything in paper should have a docs page one day

[ocean void]

You could make an issue that outlines the proposal, and if that gets greenlit then draft the actual PR. Probably the best way to go about it to have a conversaton about what that would look like.

[fair river]

Contribution is always welcomed! Make a small piece draft and have a discussion here before sinking too much time in it!

[frail valley]

I just wanted to say that I love how much more useful and detailed the docs are these days, big thanks to everyone who's contributed to it!

[spice temple]

the team and contributors have been doing an amazing job, yes!

[rich heron]

what is this option for? saving scoreboard-teams? or removing it?

[vocal halo]

if it’s true it saves them, if it’s false it removes them

[spice temple]

feel like that can be a big clarified, somethng like "if set to false the server will remove those empty teams automatically" instead of that second sentence

[twilit fog]

So I dove into internals (paper server and NMS) and found out why my flame particles startet shooting everywhere. Apparently not setting the speed (extra parameter) defaults to 1 due to paper's implementation?

    @Override
    public <T> void spawnParticle(Particle particle, double x, double y, double z, int count, double offsetX, double offsetY, double offsetZ, T data) {
        this.spawnParticle(particle, x, y, z, count, offsetX, offsetY, offsetZ, 1, data);
    }

I guess Mojang decided to use speed all of a sudden for some particles

[burnt ice]

all the sudden
Are you sure this hasnt been the case for a while?

(Like when did you start seeing this)

[twilit fog]

yeah I worded that poorly, it's been like this for a year now I think

maybe less

I opened an issue

[burnt ice]

Cause maybe there was a change in that value at one point

[twilit fog]

I made a draft PR https://github.com/PaperMC/docs/pull/625
Comments are welcome

[twilit fog]

Just one question, is it fine to add the exception for typo in COSY?

[drowsy zinc]

yeah

you can make it into a .mdx file and encase the identifiers in {/* spellchecker:on */} and {/* spellchecker:off */} lines

[fair river]

This can be very helpful to other and as long as Mojang doenst change much, it shouldn’t be too hard to maintain. My main concern is that, it doesn’t seem to have much stuff on it yet.. is it due to this being a draft and more is planned to be added? Like examples and illusions etc?

[twilit fog]

Just a draft yes, I will add all the data types, was focusing more on directional particles and how they work

[twilit fog]

did I do it wrong? IT doesn't seem to work

anyways I added more chapters, still not done, I have way too many types to lol

also TIL, note particles are actually HSB and don't approximate to nearest index of a preset color

[twilit fog]

Should I include the list of parameters with their description from the javadoc or should I leave that to the users?

I already linked the first method for spawning particles (not the full one with all of the parameters)

[hollow sky]

<@&748618676189528155>

[twilit fog]

Trying to find out how to color Particle.EFFECT and I come across this:

Basically meaning I'm cooked on how Mojang colors them

[twilit fog]

getting there eventually

[twilit fog]

My PR is now ready for review. A few things I'd like to specifically point out:

• How do I fix the typo checker to ignore the contents of <summary>
  -# Turns out it was an example using cosy smoke throwing the error, I'm just legally blind and can't read error logs
• Those of you who are native american english speakers or speak better american english than me, to check the wording as english is my 2nd language and am not sure if I worded something nice or butchered it.

Comments are very welcome :D

[neon epoch]

British English is better anyways :D

[pliant canopy]

Denglisch is the best language smk

[drowsy zinc]

🥀

[echo canyon]

https://github.com/PaperMC/docs/pull/626
idk the best way of doing this, it's not liking what I try to do

[vocal halo]

You are missing the default key

I think

[echo canyon]

yeah, but its default is a whole config section.

[vocal halo]

Is the minecraft:place_recipe just an example or part of the default config?

[echo canyon]

that is there by default, but you can add more obviously

[wintry flint]

shouldn't it be 1.21.8?

[trail jasper]

yes

[pliant canopy]

@opal flare docs website needs a rebuild

For 1.21.8

[spice temple]

Just push something, smh

Huh deployments for docs are paused in CF

No clue, did retry the last deployment

[pliant canopy]

Give me push perms 🤝

[spice temple]

Why is it that every time I touch cf workers shit I end up in pain

I hate everything about this

[pliant canopy]

What did you break 😅

[spice temple]

It used an outdated node version, so Astro complained. I changed "build system version" to update node, because screw you if you want actual control, now it fails with some other dum issue

[pliant canopy]

Ah yeah we updated the node version

[spice temple]

14:30:03.019 12:30:03 [ERROR] [astro-d2-integration] An unhandled error occurred while running the "astro:config:setup" hook
14:30:03.426 [AstroUserError] Could not find D2. Please check the installation instructions at https://github.com/terrastruct/d2/blob/master/docs/INSTALL.md
14:30:03.427 Hint:
14:30:03.427 See the error report above for more informations.
14:30:03.427
14:30:03.427 If you believe this is a bug, please file an issue at https://github.com/HiDeoo/astro-d2/issues/new/choose
14:30:03.427 Location:
14:30:03.427 /opt/buildhome/repo/node_modules/.pnpm/astro-d2@0.7.0_astro@5.10.2_@types+node@22.15.16_rollup@4.40.2_typescript@5.8.3_yaml@2.7.1_/node_modules/astro-d2/libs/integration.ts:6:9

[pliant canopy]

Ohhh D2 missing

Yeah you cannot rebuild the site that way

You have to do it with GH actions

So yea, push to main

[spice temple]

Why do we still have this worker shit then,m

[pliant canopy]

No clue, but building manually don't work anymore

You might also be able to just retrigger the gh workflow

That should work too

Scorpy added D2 for some of the graphs and since it had to be installed separately, just a build doesn't work

[spice temple]

Whatever, fixed

[pliant canopy]

Nice

Lol

I gotta sit down and find a better solution for the D2 issue because I cannot build the project manually either

Or pretend to find a solution so that scorp gets angry at me for it not being good and fixes it himself

[trail jasper]

because cf workers and pages are awful
everything about it sucks

[pliant canopy]

No, this time it is not a cf fault

[trail jasper]

still, my point stays 😄

[pliant canopy]

The build just genuinely doesn't run out-of-the-box

I've never had issues with cf pages

[trail jasper]

i really tried to use it
i really really tried
cf just didnt want me to
wrangler is the biggest peace of sht the world has ever seen
you run your builds locally it works
you publish from local it works
you let cf pull the changes from gh everything breaks
you fix the auto publishing now locally it doesnt work anymore
nothing about that is reproducible
it might work in prod but not on one pc but another

[drowsy zinc]

deployments are done in gh actions

not the pages ci

there is a workflow trigger

[drowsy zinc]

install it

you only need it for production builds though, dev works fine without it

[pliant canopy]

No 💀

[drowsy zinc]

then you can't do production builds locally

what's the big deal

[pliant canopy]

I want to do prod builds without having to install external software

It worked fine with the prev graphing integration, idk why we have to use an external one

[drowsy zinc]

mermaid generates the diagrams client-side, which frankly doesn't make sense when we're using astro

and it's a giant PITA to set up

[pliant canopy]

But generating the graphs on build time every time feels unnecessary too

Not like they gonna change that often

And it works fine in dev

[drowsy zinc]

they're not regenerated if the output is committed in the repo

[pliant canopy]

So unnecessary. Might as well generate them fully externally and just drop in the done .svg files

I just dislike that it breaks cf pages and simple 'pnpm build' setups

[drowsy zinc]

you're welcome to think of a solution to not make it require d2 if it doesn't need to generate anything

I really don't think it's a big deal to install one dependency for production builds though

[twilit fog]

Thoughts? https://github.com/PaperMC/docs/pull/625#issuecomment-3139332509

[lilac edge]

more doesn't hurt it's still in scope

[brazen moon]

I agree could do something like the gradle doc do for kotlin/groovy dsl

[twilit fog]

alrighty, I like the feature of starlight with synced tabs, will use that

[twilit fog]

Done

[twilit fog]

https://github.com/PaperMC/docs/pull/625#issuecomment-3145485011 thoughts?

[brazen moon]

i'm reviewing the PR but yeah you missed some of them

[twilit fog]

Just one question, some of you have been pointing out to me and the docs say that falling leaves have an alpha channel, but is that true? Because I can't reproduce it not via the particle command, or API. The Client code doesn't even use any alpha channel

[zealous locust]

hey i need help how do i make a command block clear only a certain item

[ocean void]

This isn’t really the right channel for this question.

Also please don’t cross post the same question to multiple channels.

[twilit fog]

also I just realised cherry leaves apparently don't swirl

[twilit fog]

Would it be okay to create a custom Note particle color picker and import the svelte code?

[pliant canopy]

Do we really need to do that? What is the reason for it

[twilit fog]

I think it'll be useful to some people, it's rather simple. Just a slider and a div with a background color changing

Instead of users "guessing" a color

[drowsy zinc]

sure

[twilit fog]

👍

I hope I won't mess up the code, I haven't touched typescript in a few months now 😅

[drowsy zinc]

you can reference the slider logic in the start script generator

[brazen moon]

yeah it's an oversight, the doc needs to be updated and generally your note can just say that only the entity_effect particle support that channel instead of listing the one that don't

[twilit fog]

kk

[twilit fog]

I'm guessing I can't use the particle texture since it's Mojang's property?

anyways the color picker is completed and documented

[twilit fog]

Okay so I've gone through ALL particles in the client code and wrote down everything (except really specific offset modifications which the end user shouldn't really care about). I've also applied most of your review comments.

Therefore, I feel like the PR is now in it's final stages. I kindly ask for final reviews and comments on the PR, as I most likely won't be making major changes anymore.
-# Also I'm still wondering if I can use the Note particle texture or not :D

https://github.com/PaperMC/docs/pull/625

[lilac edge]

@twilit fog where are you wanting to use that texture?

[twilit fog]

Note particle color picker, currently it displays the raw color, minecraft multiplies the grayscale texture with the raw color to achieve the displayed particle

[lilac edge]

IANAL but i'd assume it's fine

[pliant canopy]

-# For anyone wondering what IANAL means, it means "I am not a lawyer".

[lilac edge]

i'd just make a note somewhere that you're overlaying the official minecraft particle texture

[neon atlas]

?

[scenic gull]

Please do not ping random users, thank you.

[neon atlas]

(legitimoose now with a ghost ping )

[twilit fog]

Is it just me or does this look really cool

[neon epoch]

Nice work!

[lilac edge]

damn it really is that small huh

[twilit fog]

8x8

[burnt ice]

@twilit fog Nice work. Yeah particle stuff is fun.
https://github.com/Owen1212055/ParticleHelper/blob/main/api/src/main/java/com/owen1212055/particlehelper/api/type/Particles.java#L52

I had a plugin I used to work on which was meant to expose all of those behaviors. Its a few updates out of date at this point, but maybe you could also double check it.
This also documents the count particle behavior, as that has its own behaviors.

But this was back when entity effect for example used offset for the color, which caused it that the more green a particle was, the higher up it would go, somthn like that lol

[lean venture]

Damn hella nice, didn't know docs had previews for PRs

wish hangar had that

[mighty dirge]

Should you just not link adventure javadoc or what? Because I can't seem to find any in the adventure docs pr.

[pliant canopy]

Those pages are mostly just a 1:1 copy of the Kyori Adventure docs, which doesn't have any JavaDocs. Before it gets merged, JavaDocs should probably be added though, yeah

[mighty dirge]

those are afaik not yet setup with the papermc jd system though

[pliant canopy]

Luckily the system is very modular

If you need to place down any javadocs currently, feel free to just do the links directly. The current pages with Adventure JD do it that way

Those will definitely be replaced if we move to the docs jd system for Adventure stuff

[drowsy zinc]

the main point of the docs javadoc link syntax is to make up-to-date links, since stuff from jd.papermc.io doesn't have a latest tag

so it's fine to just use the full links for jd.advntr.dev, since it has a latest tag

[pliant canopy]

(it does now)

[mighty dirge]

the thing is though I thought it would make sense to use the version that latest paper is using

[drowsy zinc]

time to remove it

I'm concerned about plans to make the adventure docs versioned, our current setup doesn't work with starlight's versioning

the current setup is just all projects ("topics") in one collection and the sidebar switches based on the "topic" associated with the page

gonna have to figure something out for that, but that is a problem for the future

[timid cedar]

More than Paper uses Adventure - the adventure docs should be independent of paper versions or anything

[raven onyx]

hm, that doesn't seem to work yet, for example https://jd.papermc.io/paper/latest/org/bukkit/entity/Player.html redirects to https://jd.papermc.io/paper/1.21.8/latest/org/bukkit/entity/Player.html

[pliant canopy]

Don't do latest

Just omit the version entirely

[raven onyx]

oh I see

[pliant canopy]

https://jd.papermc.io/paper/org/bukkit/entity/Player.html

[mighty dirge]

Well yea, but in the PaperMC docs the version that Paper and Velocity use should IMO be used instead of latest but ig you could say that they try to keep up anyway

[pliant canopy]

@drowsy zinc shush it happens every time 😭

Each time I do stuff with patches I commit the patch file too

[pliant canopy]

@drowsy zinc I could need some help with #626. Basically the astro component for rendering the configs does not support having a description on a node which has subnodes. Aka it just ignores the subnodes. I am not well acquainted enough with that component to figure out how to fix that though

And I think it would be unfortunate to drop either the description or the subnodes, as they both add important value

[drowsy zinc]

make it into a message block

see the help.yml stuff

[pliant canopy]

I mean, that forces it to always show up, but that's good enough I guess

[pliant canopy]

@smoky granite PR is out (#629): I would appreciate a review from your side to verify that the information is correct: https://papermc-feat-menu-type-api.papermc-docs.pages.dev/paper/dev/menu-type-api/

And if I might have missed something

[smoky granite]

Looks good to me as far as actual information goes

[onyx vector]

Is the fill project open source?

[pliant canopy]

Please don't crosspost

[neon atlas]

don't think fill is yet

mostly because it is still not in a state where it is like, done done

[wintry flint]

Hey the particle pr could use some video examples https://github.com/PaperMC/docs/pull/625. similler to the command api docs https://privatech38-particles.papermc-docs.pages.dev/paper/dev/command-api/arguments/registry/

[pliant canopy]

Could you give some examples on what parts could use video examples?

[wintry flint]

Mainly for some types of particles

Nono - I mean, example for Directional particles, Colored particles, Converging particles simething and such

cause there are a lot of words and not much visual examples

As well as the entire docs could use visual examples for "complex' (?) things when needed

This can make reading comprehension easier

[wintry flint]

I noticed this feature added to new docs like the data components

[wintry flint]

https://f44cb79e.papermc-docs.pages.dev/paper/dev/dialogs/
and the upcoming (🙏 ) Dialog Api docs

[pliant canopy]

You know what, fair enough

[twilit fog]

fine, @pliant canopy do you have any tips on how to make these videos?

[pliant canopy]

Install OBS, go in game, find a nice spot in the world (I usually go with cherry biomes), film your particles, and then make sure that the output file is relatively small as scorp hates big video files

(a CLI tool like ffmpeg is very useful for the latter, as you can just reduce the size of the video rather quickly)

[drowsy zinc]

if it's short and doesn't need the user interaction, i.e. pausing, like the command examples, it's also better to just use an animated image than a video

it can be optimized by astro (sharp)

[pliant canopy]

Which, tbh, when showing that particles look cool, that is definitely prefered

[twilit fog]

okay, thanks

but honestly, you should make this part of the CONTRIBUTING.md

[pliant canopy]

I mean, it's not really an official requirement

[drowsy zinc]

it should be

[twilit fog]

so make it official :D

[pliant canopy]

Well, sure

[pliant canopy]

Btw has anybody worked extensively with either the entity pathfinders api or the mob goals api before? If so, I would appreciate if somebody could look over PaperMC/docs#289 and verify whether the information there is correct, up-to-date, and not missing crucial information. The relevant pages are as follows:

• Mob Goal API
• Entity Pathfinder API

[serene vault]

only ever used NMS for these, but the pathfinder page should probably mention that the pathfinder path length is limited by the follow_range attribute. Notably the path length itself, so if the mob can't go in a straight line its a lot less than you might expect.
imo a common cause for "why won't my mob go where I want it to go" - its probably too far away

can be overriden with setRequiredPathLength (e.g. Happy Ghast babies do), but that does not seem to be exposed in the API

[pliant canopy]

Ohh interesting, new API yet to be added

[cloud haven]

Entities also get stuck on rails, they can't go through them or jump over them, kind of silly stuff

[twilit fog]

Okay so here is the thing, the cut video (1080p60 at 10 seconds length) is 2.5MB while the GIF is 4.8MB at 720p15 10 seconds

Do I just use the video?

[pliant canopy]

Tbh yeah just use the video

@twilit fog some more info btw:

• If you check the files of my recent #632 PR, you can find the Video component and how to use it. I also have a small patch in that PR, which, if you want to, I can push directly to main rq so that you can use the play-always property on the Video tag

Which yeah sounds like a good idea as it might be useful to you, lemme quickly push something

[twilit fog]

I think I'll make gifs for the particles that take less than a second or two to display

[lean venture]

did ya try webp?

[pliant canopy]

Yeah so if you rebase your PR to main now you should be able to write <Video play-always> for that

[pliant canopy]

WebP and WebM are horrible lol I've only had bad experiences with that. Very slow to encode and big file size

[lean venture]

wut

how

I have always had lower file sizes

lol

[drowsy zinc]

video autoplay is usually blocked so that's no good either

[pliant canopy]

Hm, is it now. Well that's inconvenient

[twilit fog]

I mean i've tried encoding into webm and it was 0.5MB bigger

[pliant canopy]

It works fine on my machine

[twilit fog]

not on mine

[pliant canopy]

Hmm

[lean venture]

just converted a 514.11KiB GIF to 167.54KiB Webp without any loss in quality

[twilit fog]

alright lemme try

[drowsy zinc]

push the gif or webp first and see sharp's build log

[pliant canopy]

Idk, but I've only had bad experiences with it, but that could also be a ffmpeg issue

Scorp, should I revert the commit I just made if it doesn't work for some people anyways? That's kinda annoying

[lean venture]

it probably is, I have had issues with webp in ffmpeg in the past

[pliant canopy]

What do you use for web(m|p) conversion then?

If not ffmpeg

[twilit fog]

hmm

[drowsy zinc]

leave it

[pliant canopy]

Yeah that's what I mean....

mp4 files are just superior

[lean venture]

can you send me the video?

[twilit fog]

sure can

[pliant canopy]

Smh no 32 rd

[twilit fog]

hmm?

[pliant canopy]

Paper sucks, can't even do 32rd /s

[twilit fog]

whats that

[pliant canopy]

render distance

[neon atlas]

okay but where is the secret paper reference in the image

[twilit fog]

oh it was 10, i set it to 16 so its not just fog

[twilit fog]

like I would ever use Sp*got

[drowsy zinc]

the secret paper reference is the friends we made along the way