#docs-website
1 messages · Page 11 of 1
so on the top one
dropdownArrow?.addEventListener("click", () => {
dropdownContent?.classList.toggle("hidden");
dropdownContent?.classList.toggle("block");
console.log("Line 107")
});
``` always gets called
but the other ones only call
document.addEventListener("click", (event) => {
if (dropdownContent && !dropdownContent.contains(event.target as Node) && !dropdownArrow?.contains(event.target as Node)) {
dropdownContent.classList.add("hidden");
console.log("Line 113")
dropdownContent.classList.remove("block");
}
});
is it really a good call to do everything in vanilla js?
¯_(ツ)_/¯
I mean, it looks like you have to do vanilla js for astro?
no
hm?
Well, any recs? as astro's docs are just like "scripts and event handling? oh, script tags!"
it uses an island architecture, you can have interactive islands that use whatever lib you want
i.e. we could use something small like solid
its like 4kb gzipped iirc
neither have I, just mentioned it because its small, svelte is sub 10kb too iirc
maybe this is an excuse for me to finally try qwik
been meaning to do that ever since I drank a beer with the creator at devoxx couple years ago, lmao
he randomly mentioned he created angular, that was fun
oh they released 2.0 today
like this looks super simple https://github.com/QwikDev/astro
Yea, for what we're doing there that looks pretty simple
Could that also be used for the build explorer page then?
Yes
why are the astro error messages so bad
idk this seems a mess
revert
we can just write vanillajs that isnt stupid
Well, the issue there is that we're using ids, and getElementById will just fetch the first one
i.e. using IDs is stupid
The quick ez fix would be to just have some form of generated id
https://i.imgur.com/GeLdWJi.png works a little, the "copied" from the sha only appears on the top one thanks to copilot
let me do it
holy hell I make the dumbest copy paste errors
oh just saw you have similar code to me with the nextElementSibling
meh my animation doesnt work
SoftwareDownloadButton.astro:131
Uncaught TypeError: console.log is not a function
wat
lol
I tried to svelte and it just blew up in my face
and then complained about react jsx missing
can only use console.log in the script section i believe
no webstorm imported console from node
I ponder if react really would be so bad
yes
psh
I wonder if astro is the best choice for what we are doing
I mean, that's what I was starting to question
Well, given that 90% of our site is static, it would work; but, we really need a component framework of some form
like, when we start pulling in react or any framework, then those components cant import astro stuff
so we would move more and more stuff away from astro anyways
can't you at least pass in props into them?
Well, yea
or, well, idk how it handles that
-# || would have been done eons ago with just html and two .js files, change my mind ||
we had that before
Well, yea, I mean, you can do that
I can't xD
and I mean, astro is basically that
It's the longevity of maintaining it which is the headache when you do that
But you are now our upcoming web dev sooo
fix it 
its a templating engine with now frontend logic part
ah
Well, yea, "templating engine" is great if we were just templating
interactivety, at least for certain parts, is bleh
Couldn't you just have done this? https://pastes.dev/Ki4YDkoz2J
I mean, that is basically what he did?
with a bunch of more code ig
wanted to have accessibility
doing this with JS is generally kinda bleh anyways, no matter what the way you navigate stuff is less than ideal
so my pr can probably be closed then..?
got it
@spice temple @still apex I'm fine with adding in some SolidJS, we just didn't need NextJS.
If it makes it easier to have some components in Solid, I'm all for it.
It's still going to be eons more optimized than how NextJS was performing.
I tried adding qwik to the component but then I saw it uses some other component and stuff would basically cascade
Software download button
It uses skeleton
Or something
Idk something was using skeleton that I wanted to turn into a qwik component
Skeleton is literally just <div class={clsx("rounded-md bg-gray-400/40 animate-pulse h-6 w-auto", className)}></div>
So you can just push that into qwik
.
Probably
If you have a base setup with qwik integrated, can you push that to another branch and draft PR it to the astro branch? I'll pull and play with it.
Or if it's not worth those changes, I can just look at adding in solid myself?
I reverted it
Prolly still in my local history somewhere, but I am in bed
Stuff was strange anyways, svelte or solid are prolly better choices
Just was interested because of the no hydration part
I had a look at svelte and it blew up because jsx
Oh?
I might just need to mess around with some settings or something
Do you care if I give solid a shot?
I mean, I don't mind whatever is used
I'm not really somebody who looks after the site, so, doesn't entirely affect me 😄
I would prefer to have something without jsx but I realize I am prolly a minority
What would that be, Svelte?
Well, I added svelte and it blew up about some react-jsx-somethingorother not being there
I might just need to mess with the config files
Did you do pnpm astro add svelte?
Yes
I mean, I did abrupty try to change that entire download button into a functional return, so, I might have induced something through that
I did say that tonight I was going to not be on the PC as I just finished my bootcamp and am dead from the final week project
which page?
the all downloads thingamabob
To be fair, as I said, I might have just screwed up converting the thing to a component
Not much experience with that but yeah, I prefer a traditional syntax like that to jsx
But don't value my opinion too highly
svelte is also jsx?
okay, I have done something cursed, never mind me, I guess; I think I misread the docs
Got Svelte working for the downloads page kinda.
No, svelte is a traditional template engine
Like, it looks close to free marker
But it's really powerful and has some features I miss in Vue rn
(some of that was only added with svelte 5)
I think I was being partially cursed because I made it a .svelte.ts file
Ahh
That is meant for util files I think
So you can use runes
They need compiler magic, so they aren't as flexible as vues reactive system
god, this takes me back to php
Okay I lied I don't have time to fiddle with it tonight
+1 for svelte
no (just disagreeing with scarp)

22:43:37 [ERROR] Cannot find module '@/utils/api.js' imported from '/Users/shane/work/paper/website/src/components/SoftwareDownloadButton.svelte'
Stack trace:
at nodeImport (file:///Users/shane/work/paper/website/node_modules/.pnpm/vite@5.4.9_@types+node@22.7.7/node_modules/vite/dist/node/chunks/dep-Cyk9bIUq.js:53036:19)
at eval (/Users/shane/work/paper/website/src/components/SoftwareDownloadButton.svelte:4:37)
I have it on the page, but, I can't get the click button thing to work
Okay, I have the click working
I have no idea why I spent so long working on this
why even a drop down? only for the hash copy?
Well
We have things like waterfall which offers multiple assets
older versions of paper which has multiple assets
astra
oh true, I knew about the paper things but i forgot its still relevant
I'm going to crack open a can of lager and plant my ass on the couch
AAA
There's a pr that supposedly has a workaround but either i'm doing something wrong since I have no idea how to 'publish' it locally or i'ts not working
Just lmk how I can be of help. It's looking good so far Cat.
yeah fr
So it looks like Purpur went that route that I had kind of talked about. They have a URL parameter for each version to pull all the builds.
Docusaurus 3.6 has released with 3x quicker site builds. I have a couple job interviews this week but will look when I have a chance
Good luck on interview!! 
Good luck 
I'm skeptical that it'll be 3 times faster
Are you calling ollie a liar
what was the reason for going with astro-icon?
it's astro
I'd imagine that that was purely it
i.e. they flat converted from next and didn't consider we'd need to deal with interactivety there
Real
But that’s what they said
middle approximate value
React native is 3.04
Rspack
and some other faster tools in the pipeline too, so, doesn't sound unreasonable
Pretty much cat's response. I know there's the unplugin icon package. We could probably use that instead as long as it tree-shakes and allows local icons.
@dense acorn Not sure what that was but it doesn't appear to be related to the documentation website.
Oops
olive oil 0n goes hard
I think google translate was still trying to translate the page after I came from another non-english gh page 
Danish apparently
https://papermc.io
is this normal?
hold on let me make a video to show you cause its hard to explain
your browser probably blocked the CSS or something stupid
css?
yes, the styling info for the website
do you know any way on how to fix it?
First step would generally be to check the dev tools and see if the request failed or something
I can't really advise there for random browsers
sorry im not really that good with these type of stuff can you give me step guide?
No
Try a private browser tab
- it's not a trival thing to write a step by step for
- I have no idea what that browser is
- I'm far too full of covid to wanna deal with that
this?
probably
let me try edge
that looks like that stupid opera gaming browser with that sketchy marketing
it is 😭
oh sorry ping
but you want to open the network tab of the dev tools
on edge works but not Opera Gx
reload and find the red request
and now randomly it starts to work, i did not even reloaded the page
🤷♂️
Well i guess thank you for helping and sorry if i wasted your time
np
vanilla or Vanilla what was it
the latter
i’ll ask the same question next time lynx does it wrong
VaNIllA
Interesting. The latest Astro beta does automatic responsive build-time images. Going to play with it.
Hmm. On the latest beta, the teams page broke with __vite_ssr_import_4__.default.map is not a function. Idk what would've changed with importing that json data. I can confirm it loads, it just doesn't want to map it. No IDE errors other than the runtime.
Vanilla ice cream?
I fixed this. It was a brief vite issue, just had to override the dependency until Astro bumps it again.
Pushed it to a separate branch for the time being - https://github.com/PaperMC/website/commit/879eb720bab3bf44b2c4ebc549b4c85fe7709fb9
(6741ad936ed5010734cf19f8) // @crisp inlet (@lochefmassi / 241888213004713985) has been banned by @radiant summit (451779815415218177)
Reason: Quick-banned for sending a message in #docs-website
I believe I may have found a possible typo in the docs, but want to make sure I'm not wrong before PRing.
The Scheduler page for paper mentions ..., you can simply use the this.cancel() method: for Using BukkitRunnable, but the example has task.cancel(); which by the looks of it wouldn't make sense...
So I assume this may be wrong?
Link to the section in question: https://docs.papermc.io/paper/dev/scheduler#using-bukkitrunnable
yeah that seems like a copy-paste error
Paper copy pasting?!
My world is in shambles. D:
It's more related to the forum post: https://forums.papermc.io/threads/paper-velocity-1-20-6.1152/
The links is redirecting to "https://docs.papermc.io/paper/dev/commands-api/commands" instead of "https://docs.papermc.io/paper/dev/command-api/commands" which results in "Page Not Found"
should be fixed once the CI goes through, thank you for reporting that
The Javadocs page for Velocity at https://papermc.io/javadocs is still linking to Velocity 3.0/3.3 (https://jd.papermc.io/velocity/3.0.0/) and gives you the warning that it's outdated and to go to Velocity 3.4.0 Javadocs
Also should the Waterfall docs indicate the EOL status like the downloads page does?
We are aware of that
Fixing that is slightly annoying
And the website is currently being rewritten in another framework, so idk if anyone cares enough to fix it in the old stuff
I personally am fine with how it is since the javadocs page has that banner now, but I can see that it can also be annoying
If it's in the middle of being redone anyway it's not a big enough deal to bother with, it's one more click to get to the 3.4 docs
I think the banner could be made a bit nicer
I always miss click, lol
(and I was the idiot who wrote the banner...)
same 😭
was actually going to ask, can you make the whole banner, or, maybe even just the whole text, clickable
this was nicer than I remember
yes that should be easily possible, I am about to run tho
found 5 minutes
good enough?
was literally just clicking fork, but, yea
am sorry 😄
my eye is irritating me af anyways, so, doing me a favor;
tests
didnt actually see that they failed, lol
fixed now
@opal flare when you have a minute please take a look at https://github.com/seiama/javaducks/pull/24, makes the banner less confusing
@drowsy zinc how do you link https://jd.papermc.io/velocity/3.4.0/com/velocitypowered/api/event/Subscribe.html#priority() the #priority() part in the javadoc component thing
exactly the same
nah i mean like https://i.imgur.com/2Mr747L.png
can't be bothered to open up ij to test
yes
done
Best version for paper?
1.20.1 or 1.20.4 or 1.21.1???
You want #paper-help
Can someone tell me which theme is used to highlight code on the paper docs website?
vscode light & dark
annoying bukkit components, is it possible to hide them?
In the Javadoc? No?
I mean it's possible to hide them from the javadocs, but it'd be for everyone. You can use an @hidden tag in the javadoc itself, but idk if we want to do that since they aren't deprecated
Thank you. 🙂
Some help on how to migrate to gradle might be helpful for those that need paperweight and are still on maven
like a simple example and links to gradle/other docs and the auto migration command (though gradle docs specifically can be really overwhelming, I honestly don't like using them)
I thought we pointed to that somewhere
I've got a free weekend so I could do it later if necessary
There's a link in the project setup, but its own page with some more details/handholding would be good, I sure know I would need it 
maybe also mention how to get IJ to recognize it as a gradle project afterwards too, I found that the hardest part in my experience 
Lmao
LOVE SCORP
who
"What to do when the auto migration fails" would be helpful. Such as maven shade to gradleup shadow.
I think that's the most likely issue people would run into
fabric's documentation pages have this in the sidebar
perhaps we could do something similar?
reminds me of the NHS website where they have like a "last checked" thing
We kinda already have that?
not really the same
take e.g. https://docs.papermc.io/paper/dev/pdc a example
this page was maybe written in 1.18 and hasn't updated in 1.21
I don't think a lot has changes with pdcs
it's a random example
Sometimes, api doesn't change and doesn't need much other than a glance over
the point being is that it'd a be good indicator of when the document has been written
as maybe a page shows outdated info by accident
Like, I'd imagine that if somebody gave it a skim over for .21, they'd update the version?
right
it's basically more of a user friendly indicator of when the page has last been updated
I'd worry that we'd get users complaining our docs are outdated for any stable API
yeah that is a fair concern and is something i'm thinking of aswell
Would it be automatically updated or something we do manually
i'm all for manual - could just be a thing that's defined at the top of the file in the front matter
if its automatic we kina have something already:
but unfortunately this also updates when updating docusaurus for example
manual seems nice
I'd also be in favor of ditching the current versioning system, since it's just bad
people add new stuff to some pages that also applies to older versions, so you have people looking at the older versions' pages and they don't see the possibly helpful additions
See, last edited only accounts for "last time the file was changed", as said, fixing a spelling issue or updating a link would mark the file as just changed
Yep
https://github.com/PaperMC/docs/pull/508 threw this together
No
bad
Add a comment on GH and I’ll do it another time
why arent we using the version badge?
anyways, i would make it more visible. I don't think the end of the page is the best spot for that.
Version badge?
Hmm.
meh
I feel like this may be too strict for our usecase
evan did remove it because "its ugly"
i agree with evan
Yeah
i mean it doesnt have to be this exact thing
Yeah, could be designed like that tho
(67613c756ed5010734cf1a44) // @flat moss (@osblaze / 237788729467600898) has been banned by @rugged bay (404882575543238656)
Reason: Quick-banned for sending a message in #docs-website
I'd like to get some feedback on the overview for the build-in Brigadier argument types from Paper. If somebody could take a few looks and tell me things they notice that are negative, make sure to ping me:
https://strokkur424-improve-brigadie.papermc-docs.pages.dev/paper/dev/command-api/arguments/minecraft-arguments
I don't know about the "quick link" column
i definitely like the idea of having them but idk if a column is the best way to do it.
Do you have an idea what it could be instead?
i think it's fine for now in the initial PR. if someone thinks of a better way to format it then it can be PR'd in the future
Amazing PR, great work and super well documented. The only thing I fear is that some of the explanations are quite wordy, and I know an ADHD brain like mine is gonna see some of these blocks of text and panic
yeah, its a great PR i was surprised how much it adds to the current version.
great job
only suggestion is having return value linkified to javadocs too
oh also entity argument selectors are not just behind op, the permissions minecraft.command.selector works too
current seems undocumented on the permissions page tho
I think having guide like those are nice as hardfork is coming 
remove older command api /s
maybe for https://strokkur424-improve-brigadie.papermc-docs.pages.dev/paper/dev/command-api/minecraft-arguments#adventure-key-argument
show the validation in the video?
so an invalid key could have 2 colons, or a "/" in the namespace or smth
(or uppercase characters literally anywhere iirc)
edit: nvm thats shown in the namespaced key video a couple more down.
Maybe a beginner guide to setup everything for someone who wanna make paper plugin? 
also https://strokkur424-improve-brigadie.papermc-docs.pages.dev/paper/dev/command-api/minecraft-arguments#named-color-argument
breaks the pattern of all i have seen so far, of only having the single argument.
you could have a hardcoded message that gets coloured instead?
idk
I think that streamlining the initial setup for the basic plugins would be good; paper plugin stuff would be really great too, but, I feel that there are still some sore points
Would be nice if we had a means of collating this kinda stuff; I say that as if I'm ever going to get around to working on documentation
yeah like a guide that someone new (but know Java) can follow and set up their intellj to start making Paper plugin 
i mean currently, even with all the ide plugins etc, there is still a bit of friction, right?
yea
Somebody, please, courage to add kotlin build templates
my encouragement
that's just good ol family entertainment
I think @pliant canopy is very good with those detailed stuff so he would be perfect to make the beginner's guide on how to set up ide for making Paper plugin 
well, I have a whole bunch of VMs to migrate by EOD thursday, weekend, might be some recovering, but, I might be able to look over stuff, I'd probably just need a reminder

You know what, that sounds like a thing I would have done some time anyways. Maybe after this PR is done.
Thank you all for the kind feedback, it means alot to me
In fact, I was thinking of maybe adding a whole tutorial series right into the paper docs. Usually, the only resources you can learn plugin dev on are on random forum posts or 10 year old (still great) youtube videos. Having some readable, unified beginners tutorial might be extremly useful, also for the softspoon that is happening, as the API drifts further and further apart
There is a guide already?
It could be extended. It doesn't cover any of the IntelliJ stuff except for that one plugin
what intellij stuff
Yeah what IntelliJ stuff?
Just a quick overview on the UI of IntelliJ, maybe a new page that deacribes the installation and setup of IJ for various platforms, maybe also mention Toolbox, mention the importance of keeping your IDE up to date, that sorta stuff
I've been thinking about that, and do you think it could help seperating these arguments into seperate pages? Like I'd have /commands/arguments/ and in there I'd put everything argument related
Does jetbrains academy not cover IDE basics?
And then also have like entity arguments, registry arguments, etc
No idea, never looked at that tbh
we should refer to JB first if they have it
This word is without the link 😮
https://strokkur424-improve-brigadie.papermc-docs.pages.dev/paper/dev/command-api/command-tree
Well aware. I haven't setup links or JavaDocs yet
I am not entirely done with the structure of the pages, so setting up links would be pointless
It even says (WIP) next to it 
As resident party pooper and grumpy old man, my concern is that we get a lot of great contributions and then end up with a bunch of outdated material better updated by other resources online that have the person power
Yeah very valid
We might link the JBA for learning about the IDE though
If we don't do that already
Yea, JBA is to my understanding pretty darn good
They also have a YT channel which I've picked up some cool little tricks here and there over the years
(Like, I don't have editor tabs enabled in my IDE cos keyboard go brr)
speaking of outdated
would anyone update this to say 1.20.5-1.21.1
(don't want to make a pr for such a nitpick)
if you can PR it I can just insta-merge it
Time to let it sit for a couple of years
I don’t really understand why we need to explain the IDE in the first place. Earlier we said „people should assume how to code java“, for me that either includes IntelliJ knowledge or they use another IDE they are already familiar with.
Yes, Paper recommends IntelliJ but I don’t think this tutorial should give the impression that people have to switch to IntelliJ which imo will be implicitly said if we include a tutorial from JBA.
I also think the docs need improvements for Velocity before trying to „10 star“ the Paper section.
Somebody got to it before me, I got distracted because cloudflare is refusing to cloudflare for me
I just wanna rewrite the velocity docs
their flow is just, inexistent
lol real
Well, not me that's for sure. I suck at velocity. I will probably stay with the Paper section 
Based on the possibly problematic attention span issues, I have decided to split up the pages into seperate categories. For the arguments this means that they are grouped together in a slightly more logical manner. The Minecraft Arguments page now only serves as a kind of introduction and quick-links page to find arguments quicker.
The change is not public yet, as I first wanna sort out some issues. Just wanted to document that here
well now you could add something to the "Arguments and Literals" page under the "Minecraft arguments" section that shows all available aruments and have links to the respective page#header
I am actually thinking about converting the now empty "Minecraft Arguments" into a sort-of argument overview and even splitting up the Arguments and Literals
I have started to toy around a bit with the docs around the whole "paper plugin" issue. we talked internally and want to rebrand them as lifecycle plugins, as that makes way more sense, our future goals isnt to replace the existing plugin loading with the lifecycle loading, its something you can opt into to take more control over your plugins lifecycle, which is required for datapack stuff for example.
ill put up a draft shortly, but what do people think about moving the lifecycle plugin page and the userdev page into an advance directory?
Sounds fair
organisation is always one of those issues that we've had with the docs
(reasons why I wanted a wiki 103)
I hate docusaurus/react/webpack
lifecycle plugins sounds good
smh.... time to move to astro docs (idk if they do it)
nice joke
It would definetely give more "fullness" to the whole lifecycle thing
i think scorpio was looking into astro but for docs
They do. At least for internal links. Only on build, not on dev though
starlight
I am not touching the versioned pages, that seems stupid
i think the plan is to ditch them anyways so 🤷
Yea, they make no sense as they're not maintained
Like, they're mostly just a "point in time" snapshot of then, they're not really maintained as such
wouldn't care so much if we had a better way of dealing with that such as on a branch basis, but, it seems like an annoying overhead
nor do we want to support stuff like that
i mean for internal use maybe but it shouldnt be public
I mean, maintaining a history of stuff is good for prosperity
its a git repo
"shouldnt be public" means it shouldnt be accessible via the docs website
We offer older resources for a lot of other things
not supporing something doesn't mean that we need to cast it 100000 miles off the face of the earth
though, migrating all of this stuff to a better means of handling it would be a pita
so, I mean, I agree on the front of not putting effort into shuffling it around, but, idk, maybe in the future worth doing it, just for stuff like the config ref type things
fair point
does this also intend a rename of the paper-plugin.yml file?
Btw, wouldn't brigadier commands count under the advanced tab as well?
Because they are also somewhat complex and also use lifecycle events
i dont think so
its under API, thats a different category and more like a reference
should we add some hardfork admonition? That Spigot API past 1.20.3 might not exist on newer paper verisons?
meh
we dont need to keep references to spigot forever
its mentioned in the announcement, thats enough, no?
I mean
idk it was just an idea
There is probably a few plugin devs who are not in the discord
Who just use the docs
and suddenly, they plugin stop working on spigot, and they have no idea why
i mean if we want to reach the highest amount of userbase we need to add a warning log in the server console 

and even then people would come into the discord and ask why their spigot plugin doesnt work on paper
ill add something similar to the docs, gimme a sec
ill link mbaxs goodbye bukkit post I guess
Eh, unnecessary reading link for a docs page IMO.
I guess this would be a better one anyways https://bukkit.org/threads/bukkit-its-time-to-say.305106/
That one causes extra confusion with luke's edit at the bottom, made before 99% of the drama occurred. I think it's easier to just say what you say there, without elaboration.
ye
gib feedback plz https://github.com/PaperMC/docs/pull/511
no rush tho, it will most likely sit for a while until the changes in paper have been made
what needs a bit more work I think is the difference between https://papermc-lifecycle-plugins.papermc-docs.pages.dev/paper/dev/advanced/userdev and https://papermc-lifecycle-plugins.papermc-docs.pages.dev/paper/dev/internals
the whole "What is a Bukkit plugin?" section is weird
Weird why? It's the last in a chain of clarifying naming that naturally would extend back chronologically to what a lot of people still call their plugins.
I don't know. A Bukkit plugin can also be a spigot plugin.
I mean, that's a bit of needless disambiguety
the bukkit plugin loader is the topic there, not the "what variant did you build against"
a bukkit plugin could be called a paper plugin. I think it's useful to clarify that a plugin page you see that says "bukkit plugin" can still probably run fine on paper, and that section does a nice job of it.
maybe
also why "but plugins that use Spigot API introduced after Minecraft 1.21.4 may not work correctly." What if Spigot adds api in 1.21.4, at this point paper already split. Or will you pull everything that they add until 1.21.5/1.22 into hardfork?
I mean, spigot no longer supports 1.20.4, it would be pretty weird and out of character for them to go back and add API to it
yea, was just about to say
We've already announced that we plan to pull API for a while
ah i see
I really appreciate to change the naming of that topic, especially with the soft spoon and the growing attention to difference (and well, it was a confusing naming from begin on :/)
Changing the name of paper-plugin.yml to lifecycle-plugin.yml is not planned as far as I can see?
It would be an unnecessary break I guess but would terminate the last confusion (I would hope it at least).
i mean you could still keep the paper-plugin.yml naming for compatability but the whole Lifecycle plugin stiff is experimental and thus subject to break anytime.
I am barely convinced lifecycle-plugin is a good rename but that shall be internally fought out
Yeah, I mean more streamlining it with the commonly used name, although when it sounds not great as file name
DING DING DING, ROUND 1, FIGHT
owen plugin
Stuff will be changing, ye
Call it Extended Spigot Plugin With Access To Paper-Internal Functionality
I think that's a very good and memorable name
Since when is the bootstrap api internal?
Well, it is for spigot plugins 
But plugin.yml != spigot plugin
Well, yes
but what else to call it
What is a way to say both spigot and bukkit plugins?
Traditional Plugins?
Legacy Plugins :P
Yeah that's a good one
I mean, I generally just tend to confer on the base platform
Extended Legacy Plugin With Access To Paper-Only Functionality
There
fixed all issues
plugin.yml is the bukkit plugin system way of loading plugins
But they already exist, I mean legacy plugins as without any api version in the plugin.yml
has my new name recommendadtion been noted internally????? ||(#docs-website message)||
yes, we are considering the correct punishment rn
This is going in your file.
how many pages is it
A project towards a universal library. By this art you may contemplate the variation of the 23 letters.
maybe a good idea to add the beta version of PW now?
since it’s required for .4 afaict
and the required gradle version
stop playing gta and fix it powercash 
i died
finally money
(676ae0c76ed5010734cf1a6c) // @earnest umbra (@jiajun9028 / 884967776907837450) has been banned by @neon atlas (202850073812402177)
Reason: Quick-banned for sending a message in #docs-website
Someone needs to shoot me if I don’t review PRs today
🫡
Sure
/s
Regarding the many questions, I did a quick PR to address userdev usage on 1.21.4. If somebody could look over that really quickly, that'd be great, https://github.com/PaperMC/docs/pull/513
Once userdev goes out of beta, I will make another PR to remove that warning again and to just say that it requires 8.11.2 of gradle now. But until that is not the case, I think this should suffice
my review still stands
the fact that 8.11 or older doesn’t work should be self explanatory
what’s your address
Why is it self explanatory?
considering we've had one person in paper-help earlier who tried to use gradle 8.11 instead of 8.11.2
doesn't hurt keeping it in there
I agree with Michael. It's short enough to not be worth worrying about.
Surprisingly amount of plugin dev who doenst read the error message lol
its a big error message and it requires interpretation
its not "this is wrong, do this". its "you requested this but we only found this" and you need to diff it yourself
4 hours left :whip:
😛
oh I guess you just merged one
👀
I think season 2 started slow but I’m really enjoying it now
What’s this that you use?
He saw you typing 
Ive given the brig docs a read through
im a fan but they are still a WIP
once they are complete ill review it
👍 fairs
also for redirects with removing versioning, do we reckon this:
/paper/:version/* /paper/*
/paper/:version /paper
s2 been a banger till now
Why?
some of them are pretty big
Think that’s all the stuff I can review
and gifs are used elsewhere in the docs already
the argument for having mp4s is the ability to pause the video, which you can't do with gifs
invent a new format like gif but with pausing
small docs sidequest
Can you pause animated webps?
no I don't think so
But you can also change the fps on mp4s, no? To make them smaller?
and compress them yes
Ah ffmpeg can change the frame rate and reencode
i saw that
Yeah sorry i did not read all messages
finally
(67706fac6ed5010734cf1a82) // @high pulsar (@zetaaxc / 1240818657274105940) has been banned by @neon atlas (202850073812402177)
Reason: Quick-banned for sending a message in #docs-website
What custom component
For compatibility, GIF would be best option, but it fails at any other standards, like functionality, file size etc.
MP4 is better option, but I think WebM is also good.
Nowadays, most of web browser supports WebM, and the file size of WebM would be smaller than MP4 because of higher compression.
Compatibility: GIF wins (but not a big deal)
Functionality(pause, playback speed etc.): MP4, WebM wins
File Size: MP4, WebM wins
Sometimes, file size of MP4 is smaller than WebM file. So I would suggest encoding the video to both MP4 and WebM to see which one has smaller file size.
I don't know the condition of MP4 become smaller than WebM 😦
Well, mp4 and webm are both containers, what matters is the codecs
Many (or, most) MP4 file uses H.264 as video codec which is supported by (almost, except for Opera Mini) every web browsers.
MP4 file that uses H.265(HEVC) as video codec exists, but not that 'popular' because it requires hardware support which may not be available in many video cards.
WebM uses VP8 or VP9 video codec which are supported by (almost, except for IE and Opera Mini) every web browser.
From what I know about WebM is, it is basically Matroska that uses VP8/9.
Yea, but while H.254 is mostly used for MP4, it does support 265 and even AV1; ofc, I would probably just go for VP9 on my relatively primitive knowledge; We could maybe consider messing with something like youtube if we wanted to embed videos on the docs, rather than dealing with that front ourselves
I also support on using WebM/VP9 for this.
But I think using YouTube will create 'security concerns' like securing YT channel account. I've seen many famous YT channels got compromised and changed into cryptocurrency billboard or famous electronic car company billboard.
Which is generally done through click stealing of some form, we're not a high risk target with a bunch of staffers logged into privilaged accounts as such
So like i send a message containing a component of a docusaurus docs. I later read at the top that this discussion was about the brig docs.
Recap of message:
I have been using a custom player component with webm videos(Message contained an image).(Or something like that)
(Sorry for later response)
isn't there a v2 literally in the url?
Different version
I think the userdev page should just recommend the latest beta in general - the bit pointing people to beta 8 is outdated already, and v2 also includes fixes for older dev bundles breaking due to the old repo endpoint closing
Very well. I will do a PR to remove the removal of beta versions and just generally recommend running the latest gradle version
thanks
ollie....
merge a pr without proper review
solve
Get in the bin
good thing it wasn't your pr
That one I did read more thoroughly…
Well, the better that it had no issues
yet
Yea. I messed up a one line pr :)
(67785e386ed5010734cf1a9d) // @junior burrow (@leishybo / 951592203610882110) has been banned by @neon atlas (202850073812402177)
Reason: Quick-banned for sending a message in #docs-website
#bot-spam
Damn. That's crazy
consult the contributing guide for the only good english flavor
English not British pls
Alright, the PR is done. I'd love some feedback (and as many nitpicks as possible!!) for the current state of the improved Brigadier command api documentation: https://strokkur424-improve-brigadie.papermc-docs.pages.dev/paper/dev/command-api/basics/introduction. Otherwise, this is pretty much mergable from my side
If some stuff is missing that's because this PR is big enough as-is already, so most of the more advanced stuff (forks/redirects, custom arguments, and similar) will be content of their own PRs
just finish it so you're happy with it and I'll review it when it's mergeable
I don't want to go over the PR more times than I have to
I mainly want people to look over it to find some weird inconsistencies or bad grammar or stuff. Or if somebody thinks that the structure needs a change. Not forcing you specifically to read it 
What I had read before was good, but like scorp I’ll wait until it’s complete before leaving reviews. One thing I will say though is to keep unnecessary rambles out. Some paragraphs got quite lengthy
Very well, I think I am happy with it now @drowsy zinc @neon epoch
no
no
Ok! As I said I’m on holiday but I’ll look at it when I get a moment
Yeah no worries! It's not like there is a ticking bomb on that PR and it deletes itself after 3 hours

just enter a random office building
True why didn’t I think of that
You’re just smarter than me I guess
yeah
@pliant canopy the file size is very large. 20mb assets take a noticeable amount of time to load on a page. I thought we discussed this in the past when we were talking about gifs / mp4
Might have missed that part in that case
Sure. These can be compressed though
I could reduce it by like half or even a quarter
Just want to make sure it remains readable
@neon epoch I tried decreasing the file sizes a bit, and here is how it looks:
I feel like the 50% compressed one looks alright enough that it can be put there. the 1/3rd one might be a bit too much.
With this, a 12 megabyte video is compressed down to just 3.4 megabytes. Do you think that's enough?
(File sizes for reference)
I guess the question is also, do we need the full screen for the demo?
Can we get away with the chat box
Just a thought, but yeah. Cut it down as much as you see reasonable
I mean, there is only like 2 or 3 full screen demos
the rest are just text boxes
What happened to the idea of using gifs?
The way I have it currently is so that it is pausable. You can't exactly pause a gif
That's why it has to be a video
Are there videos that need to be paused?
I find it more convenient so.
I would crop them to just the chat though, otherwise the actual important part is really tiny
It is cropped to just the important part lol
They are usually cropped to about this, sometimes slightly smaller depending on how big the tab completion actually looks like
for the databases page, is there a reason we're showing a really old version of hikari?
a 4 year old version
No
I don't think we need to do automatic latest versions for every random library mentioned in the docs, just place a placeholder and a link where the latest one can be found
Remove the config since the issue has been fixed by mojang and there's no check anymore there.
Has the webhook killed itself?
Who knows
Anyways, gone thru all prs that needed looking at
done
Absolute legend
skipping yours
huh
disgit successfully executed webhook
wait checked wrong repo
no I am ocnfused
@neon epoch which webhook doesnt work?
stuff like the review on there is missing from the gh posts channel
or, that's request changages, nvm?
https://github.com/PaperMC/docs/pull/520 - has a comment on a request
the preview
wait that's a paper pr
not docs
none of the review comments or the regular PR comments on the brigadier PR went through
I dont get it
None on my PR either
its def configured properly
is disgit doing some filtering or other?
then the response is different
restarted it, no change
and since paper is going thru it cant be a rate limiting
no clue what else I can try
@slate scarab ever encountered this before? a webhook is showing no error, response to github says it was successful, but nothing shows up in discord even tho the I tripple checked the webhook id and secret and thread id and stuff?
does "fetch" throw if status != 200?
. A fetch() promise does not reject if the server responds with HTTP status codes that indicate errors (404, 504, etc.)
maybe thats worth catching and logging
on paper PR description are not shown anymore too
Got 400 from discord: {"message": "Webhooks posted to forum channels must have a thread_name or thread_id", "code": 220001}
wat
ooooooh
yeah ok I had /github?thread_id=blah, need just ?thread_id=blah
my fault when moving from discord to disgit
lemme know if you notice anything else
@slate scarab this should help idiots like me to not misconfigure stuff https://github.com/JRoy/disgit/pull/10
lemme know if anything else is wrong with the hooks
lol will take a look later
@drowsy zinc fixed all your issues. I hope I didn't overlook any of them, but thank you so much for the reviews. You spend a ton of time looking through all of it, so I am very grateful for it
It should be mergeable now, as Ollie previously commented
hmm so I should read it huh
Oh no
Then you will merge-block it 
But hey, I am on 109 commits now
That's probably the most commited PR I have ever done
(full commitment here 😂)
if whoever merges this does not squash, I'll haunt them
Ahahahahahahah
i’ll force ollie not to
}
else {
also like, early returns pls, we gotta teach semi good code style xD
also adventure components
Literally using those elsewhere
Fine whatever. Wanted to make sure users understand the flow
Will "fix" it when I get home
Quick note for https://strokkur424-improve-brigadie.papermc-docs.pages.dev/paper/dev/command-api/basics/arguments-and-literals#string-arguments - underscore isn't the only allowed character, allowed are 0-9, a-z, A-Z, _, -, + and .
Force push my beloved
it's going to be squashed on merge, squashing locally and force pushing makes it harder for me to incrementally review your changes :)
I mean, yeah. I had it updated in the table but I overlooked it in the more extensive description
Didn’t know people do that. But I guess it’s fair with such a large PR.
GitHub only showing „this file has changed since last view“ doesn’t really help either.
Not sure what adventure components really make sense here, since it is just very simple String concats. If it were slightly more complicated, which in some other place it is, where I btw do use sendRichMessage and tagresolvers, I'd definitely use sendRichMessage. Just in this one specific place it makes little sense I think
I mean, just generally components are nice to use consistently
just wrap it in text and call it a day pretty much
Yeah but
but also, CommandSender#name exists
^
wouldn’t it make sense to do that
Yeahhhhhhh ahhhhhh
That is the executor page
I dont want to mix concepts
Its mentioned here though https://strokkur424-improve-brigadie.papermc-docs.pages.dev/paper/dev/command-api/basics/requirements#defining-more-advanced-predicates
I could be down for haunting
👻
Great. Then time to yolomerge this before I have time to do whatever lynx is saying

Thank god I can push straight to main 
Also yes, screw you lynx, let me update that one example, you have won
but yea, pls switch to components. Even if you technically don't need them here.
Just so we have a consistent usage of components across the wiki
no PDA!!
Thi make you happy? ```cpp
if (sender == executor) {
// If the player executed the command themselves
player.sendMessage(Component.text("Successfully set your flight speed to " + speed));
}
else {
// If the speed was set by a different sender (Like using /execute)
sender.sendRichMessage("Successfully set <playername>'s flight speed to " + speed, Placeholder.component("playername", player.name()));
player.sendMessage(Component.text("Your flight speed has been set to " + speed));
}
This fine?
(cc: @neon atlas)
Yea 👍
Yay
pushed
I guess it is now actually mergable, assuming no one else has anything to say?
i have something to say
your opinion doesnt count
I got 20 mins before work, no time to check today
Maybe tomorrow morning around packing for uni
Aight. Take your time
No need to stress it. PR has been in development for just about over a month now after all, so one or two days longer will not matter
can you remove the NotNull annotation and migrate to jspecify? you also use the obsolete CommandSender#sendMessage method if you only want a string you can use sendPlainMessage
(6787d1326ed5010734cf1adb) // @ionic finch (@eusoufelipefg / 937426544954056834) has been banned by @neon atlas (202850073812402177)
Reason: Quick-banned for sending a message in #docs-website
alright, I migrated to jspecify and replaced all of the obselete sendMessage
thanks i will review the PR further now
Nice, very much appreciated
lol
I fixed your insertion error and updated the branch
lol
Yeah uh no. I am not moving the elses to the closing braces
yes you are
+1
@still apex the _redirects dont work
not even the predefined ones
(this is a build instance, run with pnpm run serve)
that's annoying smh
cant even test on local
@still apex why is cloudflare broken
there’s this great button
The thing compiles just fine on local, and gh actions also compiles successfully
it says view logs
This great button shows all green
smh
The logs themselves seem to be clean as well
Cloudflare just broke
@opal flare (idk who else has cf access) could you look at logs?
Probably Mini too, no?
I'll peek
Cloudflare API returned non-200: 503
API returned: upstream connect error or disconnect/reset before headers. reset reason: connection termination
not your problem
re-running in case it was a one-off error
That'd be great, thanks!
it deployed is what I mean :3
But yeah, it's up-to-date now
Thank you very much
Riley for the win!
yeah okay redirects work fine over cloudflare
that's handy
I absolutely refuse to, which is why I got rid of all elses entirely
still here https://i.imgur.com/9KpGEAo.png
perhaps vsc is lying to me then
but there is no other instance of else in the entire documentation
Probably out-of-date deployment
it's really too big for a pr, split them more if you can next time so it's easier to review
I hate prettier
Yeah I am planning to do that for the next ones
I already cut out a lot of stuff I wanted to add anyways
The next ones will all have a PR each
Looooong disclaimer
if you dont like the limit, change the limit?
Alright, I have addressed all of the issues you have pointed out @brazen moon. Thank you so much for taking the time to review my PR. It definitely helped a lot
great and what happened to the second video there https://24ad8e40.papermc-docs.pages.dev/paper/dev/command-api/arguments/registry#resource-key-argument why you have an error?
The only difference is that using ArgumentTypes.resourceKey does not provide error checking. We can visualize this using RegistryKey.ITEM
oh wait nvm misread
When trying to retrieve the registry value, it cannot do it, since yk, that key doesn't exist. That might be fixed by that RegistryResolver thingy, I am not sure. Will test that tomorrow
i don't think that will solve your issue Registry#get is nullable and will be null for invalid keys but it's the plugin responsibility to just not NPE like that (which you do fine in the next snippet)
I feel like it is fine to show in the preview that not properly checking the registry will result in an NPE and it should be handled by the developer
Perhaps I should add a note regarding that, if I haven't already
The duality of this stuff is "how much do you go full beginner"
I mean, these parts of the docs are aimed at fairly novice developers
At least that's how I tried to write them
yea it's fair but people can't guess the issue since it throw the generic exception message rn but still feels weird when the next snippet properly check that
I will just add something below the video to clearify that something like that should definitely be checked
Somehow smooth it over I guess
it's the year 2032 and paper still has no brigadier documentation as that's still stuck in review

contributing a 2.5k line change in a single PR usually gets you thrown in prison and then tortured
unless your name is Spottedleaf
The new commits are not even showing up
Somebody please merge this
GitHub might be sending out assassins right now
FUTURE!!!!!

I WILL BE COMMENTING
WHAT THE HELL
Does that happen for anybody else?
AHHH LEAKAGE
you have duplicated the painting variant overview
fixed
can you remove all the trailing whitespaces too?
what i mean for entity argument was more that for all methods you have "<argument type name>Argument" except for the entity arguments. Also noticed you dropped the experimental warning of the api
I mean, I fixed that, didn't I? Or did I forget to modify one. Also I will readd the experimental warning, though to me it seems pretty much stable already. Afaik somebody from the paper team also noted that it might as well be stable
Well very thankful to not have stabilised brig yet given our upcoming alias changes
that's fair but still not the case in the api.
alias changes? Well color me surprised
i mean you have singleEntityArgument and playerArgument for the names
but yeah, will re-add the warning
I mean, that's what you wanted me to fix? Because before it was just singleEntity() and player
Or I misunderstood you again, in which case I am very sorry 😅
yeah but why you have single and multiple before only for the entity ones
Because entity and entities has so little difference, so I had originally added it. I probably just forgot to add it to the player ones, though I do agree that it sounds a bit silly with single and multiple, so I might just remove those parts
can also probably add it to the comparison page since it's a little bit standalone
nope everything should be good
Great, very glad to hear that
If you notice anything afterwards, we can always fix it anyways
That's not the docs page and just PR a fix
It wouldn't be accepted
why
Paper doesn't currently accept random spelling or style fixes
Please refrain from simple cleanup PRs 👍
Javadocs are fine, a spelling mistake isn't a cleanup. Cleanup is like, removing a public from an interface method or adding a final to a variable
Things like, don't have any effect
Smh lynx sending me mixed signals
You're just saying that because you hate me 
then free PR for me xd
my 3 javadocs pr would have definitely got a comment if they didn't accept javadocs prs at this moment
Yes, I know now
. I had assumed "cleanup prs" to also mean jd prs
The general gist is that such cleanup PRs are often relatively valueless and get in the way of working on other things; changing a JD is generally not going to cause such conflicts, so it's not a huge concern; ofc, if we had 20 dozen people submitting such PRs, it would become a sustainability issue
(678fc61c6ed5010734cf1afb) // @jolly marlin (@v1mkss / 541507485471735810) has been banned by @eager plover (126975485493248000)
Reason: Quick-banned for sending a message in #docs-website
Magic
i think you should revert it
Lesgoooo
Thank you kindly @neon epoch
I can finally sleep in peace and continue writing more documentation hehe
Was large path? XD
Took 5 weeks and like 140 commits 😂
Not sure why i feel this is not break any record for docs PR xd
I think it is the most commented on PR with the most co-authors
Whether it is the one with the most commits, I do not know
Lol, @mighty dirge, because you co-authored on the Brig docs PR, you also got attributed the 2.6k line additions. So you have a lot of changes now 
(This is also true for Ollie, cas, and scorp, but they already had a ton of changes anyways)
That seems kinda weird though. I would've expected that GitHub actually goes by changed lines, not by PR. But maybe because it was squashed, git doesn't actually know who wrote which line?
ye, git just sees the commit and all the authors
I mean those people reviewed your 2.7k PR so see it as credit for that.
Not complaining
If I didn't want co-authors, I wouldn't have commited their suggestions. But I did want this to be a community project, so I am completely fine with it
For #530 I want to collect some more points that one could put into the FAQ or Common Issues pages. If somebody has questions or ideas, make sure to submit them so they can be covered there
I mean, it’s not a competition 😂
You were busy with issues
I mean... there is a leaderboard... 
Not really a leaderboard but more statistics
wtf is this discussion 
I don’t know
smh it's all Ollie's fault
Why?
No reason
usually is ollie's fault
Welp, this became off topic, so I just want to bump this in case other people look into this channel
btw do br tags still dont work for the same stupid reason?
<br> tags work fine, idk what you mean
Make sure they are self-closing
Ie
<br/>
Interesting, what is weird about this is that when I myself squashed a pr (which was dumb) where Scorp co-authored, he didn't get any credit for it.
https://github.com/PaperMC/docs/pull/466/commits/b5c353e9ebc53c3270882a2b037f5a816132a9e9
-# And why do I get attributed the lines, when they only show the commits I authored in the commit overview?
co-authored-by needs to be at the end of a commit message for it to be "parsed"
Strokkur AdventureComponent is part of internal, the recommended way is to use the right serializer (MessageComponentSerializer)
Oops, my very bad 💀
The project I am testing it on has userdev, so I have internals access. Didn't realize it was internal, since it is under the io.papermc package smh
Will fix it
Fixed 👍
this was a "I am blind issue"? https://github.com/PaperMC/docs/issues/538
PLEASE DON'T MENTION IT
I have a big monitor


