#docs-website
1 messages · Page 5 of 1
It's almost always people doing stuff they don't understand, so we should guide them away
whos this guy lighting up my inbox
haha, dont worry. Realistically i would think it would group them together under a single chain but 
I've addressed all comments I know how to, there are some broader comments at the end that I may not have addressed like about calling the pathfinder every tick
What have they done to my code good lord
embed scare
lmfao
<@&748618676189528155>
(65b6bd8f2d5bd6481a064801) // @vapid tundra (@drew7651 / 680256296871788560) has been banned by @ancient quiver (1098722699116810331)
Reason: nsfw ad spam
merci
https://paste.ollieee.xyz/yufivafigo.sql
I started working on something but my painfully lacking knowledge of the parsing commands is kinda making it really jank. if anyone wants to have a look youre welcome to :)
Why not write it all in JS
Half of it is js already, just rewrite the jank stuff in JS too
gotta be an action that already does this... can't be a unique thing. Could even be some general spellcheck thing
Looks good except the comment at the end:
If it's not immediately fixed, every change to the pr will make a new comment
And since it also triggers on push, there might be an error trying to create a comment on the commit in the same way as a pr? Idk
Would be better to hook into whatever GitHub uses for checks
i created this since i was board: the ::error:: thing creates annotations on files after they are detected
https://github.com/456dev/github-app-test-spell/blob/main/.github/workflows/spell.yml
https://github.com/456dev/github-app-test-spell/pull/2/files
the biggest issue atm is probably when pring not on a fork, as it gets run on push and pr, so the annotations sometimes get duplicated
https://github.com/456dev/github-app-test-spell/pull/1/files
and also it not showing the column, even though i gave it that information
Ah thanks I’ll read this later
On the How Plugins Work page in the Permissions section, I have a few points to adress:
- I think the way sub-permissions are explained is a bit misleading, because unless Paper implemented the perm-system of Sponge where
permission.agrants permissions likepermission.a.1is it afaik not like that? I recall it always being hardcoded in the sense that a permission is only for this particular check and does not include other perms, unless defined so in a plugin's plugin.yml... I talk about the default perm behaviour here when Permission plugins like LP aren't installed (Which allow sponge-like permission inheritance behaviour). - Should the
*behaviour be mentioned (That it grants access to any sub-permission, if included in one likepermission.*and grants all if defined as-is)? If mentioned, maybe as a warning that it is not recommended behaviour to grant permissions like that, as it opens a large security risk.
Also, the configuration section implies that plugins only use YAML, even tho they can utilize other file types (or even non-files like databases) to store and access data.
Maybe change this to mention that the server offers a configuration API that allows the plugins to create and manage files using the YAML format?
Sounds logical, PR it!
Yeaaaaa, whoever wrote that, darn
So, also, * is not a bukkit/paper thing, that’s something that perm plugins provide
re: the "no british spellings" thing, something I wanted to look at more a few years ago was using some proper linter like https://vale.sh/
never really got far with it though beyond deciding that spending any time on tooling wasn't worth it then, but taking another look now still looks promising
i love branching from a branch
ive run it on the paper admin docs and after about 5 minutes it still has not completed... idk if this is me being dumb but its just like hung 
oh rip
@drowsy zinc made a config which made it performant but there were 800 errors in the grammar of our docs 
the base styles are very opinionated
yeah I would def not use their provided base ruleset, or at least minimize it a bit
Yeah I was not a huge fan of their docs but I was tired and my eyes weren’t working so I’ll look another time
Not my project anymore lol and I didn't look too far into vale at the time either so by all means don't use it
(65bdf4ad2d5bd6481a064812) // @terse nimbus (@gaethanos__ / 349156006414123019) has been banned by @fair river (177150983258767360)
Reason: spam bot
experimenting with languagetool (the same stuff IJ uses), seems promising, but it'd need to be tuned to work with mdx
plus it has a way to automatically apply the rules unlike vale
Ah cool
Ollie check the latest issue plz
probably lost in notifications or you're intentionally ghosting me
im ghosting
or is it ghousting in British
yeah the notification came thru whilst i was in the shower just forgot
why do you.. nvm
hey @neon epoch
the entire point was to yeet the 12+ section from Aikar's flag
read the linked cat's message lmfao
ik that I no longer recommend aikars 12G+ tweaks because that pretty much WILL cause issues these days
There is not enough red ink in the world to cater for the hole in your IQ
||I'm just, british, I'm a nice person, I swear||
i believe you. I've been called a lot worse for a lot less
||yeah, when your bipolarity turns to the other personality||
||Sometimes I just get a little tsun 😦 ||
||I don't know you enough, but still want to give that comment, so.... "Sometimes"?||
Ollie gonna close your PR if that's fine
how about we ditch the short command entirely
and just link to the script generator
Sure, up to you
oh it's already there
Use our Startup Script Generator to generate a startup script for your server.
nvm
meant that one yes
hey @eager plover
are the explanations, etc. on that page still up-to-date
[no-ping]
idc ban me
The flags intend to do the things the do there
The only sorta like “long term” question is more on the size tweaks in those flags
The adoptive algorithms of the jvm has progressed substantially in the past few years, how often do we see somebody without his flags having noticeable gc issues without them?

not often
but maybe we don't have many people with GC issues because most do use the flags and they do help still
I marked it as done and approved
le merged
(65bfd0852d5bd6481a064819) // @still apex (@powercas_gamer / 292616490958454795) has been banned by @ancient quiver (1098722699116810331)
Reason: crypto scam
OH FUCK
hes a notable member of the community
he pinged mods to report the other guy
hahahaah
XD
tsk tsk tsk
I blame discord's terrible autofill
SAVING IT
rude
bro lost his roles
i just keep getting abused here first eternity now noah 😦 ||/s|| #spicy-general message
LOL 
You know I’m not actively trying to be number 1 lol 😂
Sulu put a lot more legwork in at the beginning
Just want to make sure my workflow didn't get lost: it uses ripgrep, which I've found is practically instant, with the result parsing / formatting taking longer as-is (and less time than GitHub takes to set up a run (12 seconds for whole run, 5 seconds for running the stage)
https://github.com/456dev/github-app-test-spell/blob/main/.github/workflows/spell.yml
Oh yes sorry I forgot, it did get lost
Does it work
Oh and things like when PR’ing does it run doubly and stuff like that
If we can avoid
So currently the case it runs doubly is a pr from a branch in the paper/docs repo itself
But that could be prevented by just running the branch check on the main branch, and rely on pr for all others
Yeah I got it working, with the results published as annotations on the pr and stuff
Actually looking, it looks like since the pr triggers on more events, you can have multiple runs on the same commit, so multiple annotation sets
Looks like the cause of this one is triggering the pr event on "edited" which I think is pr title
Ok cool. I’ll take a look when I have time. @drowsy zinc is working on a proper spellcheck action but yours will work for just British for now 😁

yeah thats fair. probably the other issue is since it uses 1 regex pattern, it doesnt know exactly what matched, so cant eg suggest fixes without extra parsing
https://pastes.dev/ZvufQ3bNKH
is this a bit too much to ask from docs, or should i just create the issue and see where it goes

it'll be closed anyways if it's too much so created
I mean material is deprecated for removal imminently
It’s being split into item and block types
@rancid yacht hey just wanna say I appreciated you taking an interesting in doc writing 
then it will be useful for migrating, since there are no Material -> namespace etc
I can’t lie it was about 1am when I read it, I’ll read it again later 😄
@neon epoch Your are the docs manager right ?
Is there planned to add a page for quick start of paperweight ?
What exactly do you mean? There's a userdev guide already
https://docs.papermc.io/paper/dev/internals
I never saw that. I searching everytime for the specific "string" named: paperweight
?
Maybe i am Stupid
Yeah there is one already :P
Yeah I’m not completely against this idea I’m just not sure if the docs are the place for it. It’s gonna be one helluva page
Maybe input from others here?
i thought about adding a data generator in paper for this and other data which is dumped into json files
there will be a default of the latest version of those files in assets, and the tables would be autogenerated for this file
also since datapacks (and soon paper plugins on bootstrap) can change the data (right now by adding their own tags or changing what items/blocks are tagged), players may put the file generated by themselves if they really need to see the custom data from their server, but this only if it can be done without uploading but just parsing the file on the browser
inb4 tools.papermc.io
so this was already planned? 👀
no
No, I was joking about us basicaly having to setup our own tools site instead of tryna litter the docs with various tools
idk, if you don't consider this fit for the docs then ok
otherwise I would even learn properly the front end just to make this myself
as i think this will be used by a lot of people, as javadocs are too scary for many novices
I never said no
I'm just wondering how much in terms of "magical tooling" makes sense to be on a docs site, which is generally more intented to be a static thing
having some automation on there is nice for trivial aspects like the presentation of the version in the "how to add paper to your project", but, I question where the line comes in terms of "this really would be better off on a more dedicated tool where we're not bound to the whole 'it's a documentation site'"
I’m also hesitant to buttfuck vercel with this json
Like it’s gonna be quite expensive to serve
I still don’t understand what this could be useful for
Eh it would be nice to be able to search thru them and filter for like items and blocks and see namespace keys but it won’t be as useful once they are split which I hope will be .21
I think the effort would be too high compared to its usefulness. The POTATOES / POTATO issue could easily be avoided by checking if the user accidentally typed POTATOES and replace it with POTATO during runtime. And as you said. As soon as Material is split this problem most likely doesn’t exist anymore.
At runtime? Nah that sounds horrible. That should just throw an error
Or that. Depends on the use case.
I mean, items are items and blocks are blocks
I think that I can see the utility of this, but I don’t think docs is the place for it. I also think this is something to consider in the future once we are able to pipe stuff directly from a Paper CI run into docs
once the seperation occurs it's generally gonna be harder for people to "mess up" once that occurs, as a lookup against the items list will fail
^^
- javadocs are very confusing for newbies, even after the split sending them to javadocs will continue to be a bit overwhelm with info, it's a very great place for code documentation, but not really for administration and configuration of the server
- the transition won't happen instantly
- other data of the material (in future block types and item types) like the namespace is just unknown and people just need to lose time guessing and testing
The transition will happen instantly. Material will be removed and Commodore will handle legacy conversion based on my short read of the PR a long time ago
and maybe in future plugins will finally use more namespaces and not names from the code, as the namespace is even seen in the client with f3+h, but then it will be hard to migrate
What Im saying is that documentation is generally intended to be something that is fairly static
"be able to upload a file to from the server and have it parsed", is anything but
i mean in configs etc, i don't think the transition will allow instant guessing if the plugin reads blocks or items from config
I have nothing against such tooling existing, but I'm not sure that the site which is intended to have static documentation is the best place for that
well, that may be omited, it would just be cool to see the tags that are on the server, but it could be just vanilla built data
I think the bottom line is though that unless we can automate this somewhat this ain’t gonna happen as the data is gonna get stale pretty quick
It’s not something I would merge without automation, but I won’t close your issue right now
That's kinda like, the odd area of this
Because I do believe that a page of some sorts that achieves this would be useful
I want to be able to populate the documentation page with data somehow
Like, I'm not 100% set on everything being static, data generation is something we've wanted for a while for certain aspects, i.e. see the config page
Yes, I am completely happy and honestly look forward for some aspects being driven by automatic fetching however I am not the person to design it
Would require a fair amount of work to probably bibliothek or whatever
And that’s without actually generating the JSON
It’s somewhat difficult due to comments and default values
Well, the idea wasn't so much of fetching as, we put a json file somewhere and the page reads it or we dump out the pages content from the json
Go away please
I shall be writing all typescript on my phone going forwards
@jagged pecan ive been messing around with your script and it doesnt seem to work if i commit thru web UI
is that directly to a non-main branch vs pr? do you have a commit link?
to reduce duplication when pring, it only runs on "push" to the main branch
it can be run on non-main branches, but it would duplicate the checks
if you open a pr for that branch, it would then run the checks
yes. that seems kinda silly tho no? Idk.
i dont normally touch actions if i can avoid it
yeah, ideally id have it run on push for any branch, but the pr files tab shows checks from both "push" events and from "pull request" events
actually maybe i can fix it by only running for prs if the target branch's repo is different to the source branch repo
ok, done, but it might show the inital check twice for local prs, but one will be skipped
https://github.com/456dev/github-app-test-spell/commit/51ab825df155845c7578c633824e6d001ab714dd but its easy enough, and seemed to work well
Oke I’ll take another look later
we have the PR-atahon tonight anyway
That sounds fantastic.
👀 whats the goal?
200 is the copium answer
why not… 180? 
The War against the Machine.
lol or even better: a Machine pr marathon. just merge/close all the machine prs
I can imagine that many of them are outdated or even still work in progress
@jagged pecan can you approve this workflow run please https://github.com/456dev/github-app-test-spell/pull/6
huh ok, done, but it seems my skip check was broke
looking into it now
fixed: missed an object in the event path, so they were both null
just checked push only (local branch) and it seems to work there too https://github.com/456dev/github-app-test-spell/pull/8
Are we expecting that skipped thing there?
yeah, since it also triggers the pr event for local branches
oke cool
specifically, https://github.com/PaperMC/docs/pull/303/files#diff-203285d1aa85b30b7cd672df926f6239c419f1308a5b1cbe3c5df4a0497f1dafR14
checks to only run if its
- a pr event (so it always runs for pushes)
- or the repo names from the source and target are different (so the push event doesnt trigger)
Cool thanks
Actually @neon epoch , would it be possible to add me as a co-author for the spell check action?
also, does that account for regioniser x regionizer?
the code uses the former, so might wanna add the latter as unwanted spelling to keep it consistent
Yeah I was planning on when I merge it
Why does leaf use British

If you update that to include it, you probably want to remove the message about looking up an American version in a dictionary
uhh does permission-message even get used anymore? when you try executing a command you don't have permission to it just shows the Command doens't exist message (that behaivour is already mentioned in the docs). Soooo is there just some weird case where it does get executed or is that just outdated docs?
No, the permission message stuff was broken when spigot moved over to using brigadier for commands
i guess it should be removed from the docs then (or atleast mentioned)
There was some intent to fix it
honestly seems better not to. what the player isn't allowed to execute he mustn't know about
That command does not exist in this server.
i've tried a more conventional spellchecker (typos) here and expanded its wrongs list with the spelling from the existing PR, lemme know what you think
planned to use languagetool, which provides more styling suggestions, but its CLI is unusable for MD(X) without lengthy preprocessing and I didn't want to roll my own
This looks interesting
by the way, that flagged Paper.FilterThreshhold property, I can't find it in paper's repo to check if it's just a doc mistake or it's actually like that
it used to be in this patch (you can see it if you go to an old enough commit and yes, the documentation is not wrong, it is written like that), but doesn't seem to be the case anymore
yeah sure
it doesnt seem to find the neighbour in config-specs/paper/server-properties.yml
unless im blind
I’m blind
we know
@still apex
PR and supercede @drowsy zinc
Oh no hurry
What's the timing per run like?
I think it said about 4 seconds
like 4 secs until it actually fails yea
Nice
Unfortunately this is more powerful that your solution
Yeah, I knew mine was only temporary really
Thank you regardless
huh, apparently folia spells it regionized on the overview and regionised on the region logic page
and the api seems to use regionized as well, so we probably want the AmE spelling then?
it's kinda all over the place
Consistency is not Leaf’s strong point.
fixed stuff accordingly to match AmE spelling, the classes referenced in the docs ended up being AmE as well in the patch
if anyone could check whether its usage was really removed or just moved into a different patch, I'd be grateful - having to ignore the spellcheck error pains me if it could be removed
<@&748618676189528155>
(65d169202d5bd6481a064845) // @safe bone (@amneedamnee / 506459855490383872) has been banned by @opal flare (105923848263753728)
Reason: Quick-banned for sending a message in #docs-website
(65d169502d5bd6481a064846) // @safe bone (@amneedamnee / 506459855490383872) has been banned by @proper steppe (80853648891977728)
Reason: advertising spam
Hope everyone is well - I’ve been a bit AWOL for the last little bit with life. I’m gonna try and get back to docs at some point this week. I’d just like to mention that I’ve seen a few comments floating around about translating docs into different languages. I’ve had a think about this and I cannot see a way to do this nicely without:
- making maintaining existing pages a PITA for keeping different languages at the same level of completion
- increasing the bar to entry for new contributors
For this reason this is something I am not going to be focusing on unless someone proposes a way that I hadn’t thought of. Thanks ☺️
I think that's everybody's stance so far
And please never feel obligated to do anything, the team and community appreciates your work, no matter how much or regular
Haha, yeah I know - just thought I’d give a little update 😝
Yeah paper doc in its current state is already way better than what we had years ago
Having outdated docs in other languages feels worse than directing to use a translation tool on the up to date ones.
We can tell people to dm eternity, who will then respond with chatgpt translated docs 
especially with modern translation tools just generally being much nicer, the only way such a document works is if you have like a wiki where people can just maintain stuff in different languages, but, that's how you get divergences, etc
You could in theory provide a "built-in" auto translation system, would probably make life easier for people that have no idea how to translate a page.
But that could easily lead to issues of terms that shouldn't be translated, being translated.
Which of course would also happen if the user manually translates it, but at that point it was the user translating and not the website itself, which sounds worse?
Many browsers offer translation automatically.
I mean if Paper do start to popping off even more, we can setup crowdin officially and actually have a workflow for it. In theory that would work, no? 
Unironically, I think with chatGPT doing the initial translate and have human review it later should probably do a pretty good job already 
Exactly. I’m gonna update a doc in English and it’s gonna lag behind in other langs. it just wouldn’t work, hence i wanted to summarise earlier
I don’t mind like a button that auto-translates using deepl but it wouldn’t know what to ignore and also paper would then need like API keys and pay and it’s just not worth ti
yeah nvm then.
@neon epoch excuse me? It’s actually your duty as British and doc writer to get these done SMH
Good one mate
in the future you just need a paper-gpt. On the docs page you give it a prompt about ignoring bits and it translates the rest ;)
paper model when?
You can train gpt on paper docs I think
I’ve seen something like this for like gitbook
I can easily train one
the thing is... it would become expensive af 😭
like 0.002 cents per request, with a project as big as paper can become expensive
but if enough funding is there, you can make a discord bot that answers questions when needed.
By the way, training won't work for this, unless you do smth particularly clever with the dataset
why wouldn't it?
Finetuning is not great at incorporating knowledge, especially domain-specific one
It'd be possible to hammer some knowledge into the model like this if you have a massive and very high quality dataset, but that's not very viable to build out of docs
This is a very hard to solve problem in natural language processing, if you check e.g. OpenAI's own support bot in their server, they haven't quite figured it out either, the bot doesn't work that well
However, it's pretty viable to make a cheap bot that does bqsic contextual searches, or responds to very common questions if they exactly match something in the database, I'm currently working on this
I mean, it's not the best, but it's decent in assisting.
(This was only by providing the "Getting started" section of the docs)
What's that UI
It's not difficult to make something that answers direct questions such as this one, it's just that once you try to ask it anything moderately complex it'll go downhill really fast
mind giving an example? I will try it
(preferably one that can be answered with knowledge from the docs)
How do you handle questions that can't? That's the harder/important part
"I don't know." xD
There are pros and cons on both sides yet.
Since, like you said, finetuning is still not good enough.
But listening to specific questions with a predefined answer can work and can't... what about spelling mistakes? Totally rephrased sentences?
You could potentially use some type of binary tree style thing with words, but it's also not that efficient.
Although, unsure if they use AI or not, Pterodactyl has done a decent job at this.
That's simple, GPT/embeddings will catch these
Ask "why is my world migrated from Forge to Paper not starting up?", for example, see if it brings up the reasons outlined in the docs
yeah doesn't properly work lol
so a mix of ai with normal FAQ?
My workflow currently is to have a little pipeline for each common question (FAQ)
It's as follows:
- Have a list of keywords, of which the user's message must must have at least one, even if it's very general, as this already eliminates a lot of candidates (optional)
- Make a database of a few very different rephrasings, could be 5-10 as long as they're fairly different
- Embed them and throw into some sort of vector database
- Prepare a model that will classify whether a given query matches the exact question, e.g. asking ChatGPT something along the lines of "is the following query asking about migrating from Paper to Forge/Fabric/Vanilla? Answer Y/N", (The prompt won't be as simple, it's just a simplification) and this is sometimes worth making into a finetuned model, for this kind of classification they work pretty well with not much data
- Now, when a user sends a message in chat, run through the three filters in order: keyword filter -> embeddings filter -> the confirmation model above (as the first two steps are very very cheap)
The embeddings filter will just compare the user's query with the reworded questions in the database, we can just calculate similarities with all of them and average everything out. Then, check if it's above some threshold before passing it into the selection model
It works pretty well, and if we use a finetuned models we can bias it into producing more false negatives (err towards not answering the question at all if unsure if it's a match, which imo is better than false positives), I just haven't implemented that one into my bot yet
makes sense, I had this insane idea some time ago which I wonder if it would work, which is using SoundEx.
You input a text like "Afonso" and it outputs how it sounds, so "A152", "afomss/oos" would return the same input, which can be useful imo, I wonder if AI uses it in some way.
Nah, language models work with tokens
yeah makes sense
(65d6695b2d5bd6481a064855) // @winter pecan (@abbyaqua / 727092698133168158) has been banned by @small harbor (215448923681062913)
Reason: Quick-banned for sending a message in #docs-website
Heya, as much as I love merging prs without any approval, would anyone mind looking at https://github.com/PaperMC/docs/pull/293 - I have the intention to be in VC later working on docs. Hopefully at a more reasonable time 🤭
I really really really hate css
I had to explain the flex box algorithm to my apprentice today and I am not sure she understood a thing 
Like... Ugh
Mmh?
The scroll bar? Guess you have an overflow scroll somewhere
Scrollbars aren't the default
ugh yeah thanks there was a lurking overflow-x
why? it's so cool lol
I mean, I don't work with vanilla css since idk when
Tailwind on top
but yeah
(65d9373f2d5bd6481a064860) // @wraith sand (@babasss44 / 375407210353786880) has been banned by @eager plover (126975485493248000)
Reason: Quick-banned for sending a message in #docs-website
(65da20af2d5bd6481a064861) // @stuck grail (@whiron06 / 764944995157016636) has been banned by @eager plover (126975485493248000)
Reason: Quick-banned for sending a message in #docs-website
what about only providing translations for config files since that is probably the thing users look up the most?
Config file will be in English though. Maybe the text that explains them? I can see that being manageable
not sure whether translating only config descriptions is worth it
might as well translate everything, but then you run into the translations being outdated way too fast
I think the idea is nice and since we have decent amount of people willing to contribute
It would increase the accessibility for everyone
I agree with this personally
is that worth the effort tho when most browsers already have translation tools built in
Yeah. I thought that was obvious sorry.
Translating everything would be 100x more effort than only providing support for the config file section.
yeah but imo it doesn't make sense to translate only that one thing
I think it makes at least as much sense to justify it.
Once. They'll contribute once. Then we'll have some great work and a bunch of [insert translation here] for new content.
Love the joy of having one enthusiastic person who does everything beautifully for a language and then disappears and nobody else speaks the language but you're afraid to just delete its now half outdated self because it was so well done.
Yeah I can see the concern of maintenance burden but the setting wise largely don’t change much I feel. We may just need more slaves contributors 
or someone just comes up with a usable machine translator integration
But browsers do that already ...
Well, to be fair, browsers usually use Google Translator which is pretty bad
Even though I'm not sure about automatic translation, do docs have a note of using DeepL or something? I feel like that's the most viable thing, redirecting the user to a better tool
firefox has a homegrown one which is even worse
at least it works offline
Ouch lol
If anyone is interested in a simple PR,
we want to include Velocity branding in our asset page https://docs.papermc.io/misc/assets
From https://github.com/VelocityPowered/Branding?tab=readme-ov-file#usage-directions
I mean I know I could maintain pretty much all of the docs translations for Polish
just probably not that many people to dedicate a significant amount of time translating regularly
So imagine you get hit by a bus just as I merge docs for new tooling on a major release and no one else speaks polish very well. What happens then. This is my problem… i think it’s better just pointing people at some sensical translator
The bus factor is a measurement of the risk resulting from information and capabilities not being shared among team members, derived from the phrase "in case they get hit by a bus". It is also known as the bus problem, truck factor, or bus/truck number.The concept is similar to the much older idea of key person risk, but considers the consequenc...
went down a rabbit hole looking at that and found this cool website https://deviq.com/
Any way to have some built-in translator tho
The quality of built-in translators seems questionable to put up on a website
I haven't seen a browser prompt to translate a page for years but that may just be me
The fact that this is a thing, I love it. I didn’t realise when I wrote it lol
It's in the url bar
Plus it also depends on your settings, I have set my os to English and German so it will never offer to translate that
wdym you cant test it cat? You can just apply a translator to the preview
I have no idea how to force it to do that and my eyes are ouchy
or, well, I guess now that it's on vercel I can shove it through google translate
if you wanted, you could swizzle the code block components as well
@neon epoch what do you think about converting the plugin.yml page to the same format as other config files?
Hmmm I’m not opposed to it I just don’t know how it would look
seeing adrian's PR, canceling x cancelling should be added to the spell checker
idk whether the single-L variant should be the wanted one; bukkit has Cancellable and stuff, but /shrug
either way, while you're at it ollie, add cancellation, cancelled and cancellable
the checker is a bit too dum
That will probably make the CI fail tho for the event docs
I just made that change from my phone
since i've seen it mentioned a few times recently i think it would be nice to add a "to singleplayer" section to the migrating to/from page that would cover using an nbt editor to move the data from playerdata/uuid.dat to the data.player in the level.dat
Fr, I can make a rq app to do that if ya wish, can do it tmr probably, just gotta first find out how to migrate to singleplayer haha
once I know what needs to be done, I can easily automate that with a java app.
Sounds brilliant 😂
Although if ya want a web tool then a bit harder on my side
just lmk if I should do that app. Will probably use the good old swing xD (we already use it at school so, why not, that + a dark UI makes wonders)
wait that exists?
there's https://toolbox.helpch.at
which funnily enough contains some cool tools but sadly their website is laggy af
no idea why
I can read, was mentioning the fact that helpchat also has a tools website
ah ok
Anyways, lmk if I should do this or nha
I don’t know about doing web nbt manipulation lol
me neither 💀
That may or may not be possible
hence why I said if it would be web based, I wouldn't be able to help xD
certified spring backend moment?
I mean, if you are doing a tools website, then at least add more than just one tool xD
I figure this would require some extra work to make rather than just having the script in the docs or a java app to migrate to singleplayer
Welp, someone can make that. I ain’t doin it
i mean, the startup script generator and the conversion tool would be a good start i guess
someone wanted to add a table of the Material enum a while back but idk if thats happening or not
conversion would require backend work... unless it's possible at the frontend level?!
but yeah idk im not a team member and dont know about whats planned or whats not planned
helpchat's toolbox works at the frontend level if I am not mistaken?
but helpchat doesnt modify nbt does it?
Ngl, I would love to work on a few prototypes for a website like this if I am allowed.
it does not, hence why ^^
same tbh. Sping was quite fun in class
I wanna try to do it at the frontend level ngl
We gonna need a tools.papermc.io now 💀
ngl not the worse idea... Having some parsers, validators, etc. there would be nice
scripts, converters
only requirement is that it uses Svelte 😎
weird way of spelling react
both cant spell flutter
💀 react, imagine... just imagine using react in 2024.
Vite + Svelte + SvelteKit + Tailwind = Best thing in the world
imagine using the 30284th js frontend framework

Svelte 🔛 🔝
fake
developer popularity is a rather useless stat
I know HelpChat has one
Shit, Astro catching up....
@fair river 
that's actually interest xD
if we go to usage, ofc react is at the top
quite normal
But in terms of interest and the one devs are happier with
Svelte is on top
Go look at docs.papermc.io
nah lets just do all on backend and just use thymeleaf for the few buttons and upload fields
Silly guy
that should do the job
svelte still better ngl
react is just the current standard
smashing together flavour of the month tech in our org is useless
like angular once was
we are already using one tool set
ik
I assume next.js for backend?
although isn't hangar done with vue?
hangar is technically on another org
oh?
so papermc overall uses react and next, hangar uses vue and nuxt? Understood
interesting actually
yeah me can't help, very limited knowledge with react, never had the interest.
where does paper use next lol
idk, I am assuming things
ah website
spring is just goated
ngl it's on my list of things I want to get into
but still gotta improve knowledge with SvelteKit
better have one good techstack than several average ones
Let's not turn this into a tooling debate
The website being react is just what the person writing it was comfortable with and we were happy we were getting a new website
It wasn't a conscious decision
It was not meant to be a tooling debate, I am not screaming for you to switch to something else, after all, react is standard for a reason...
Although this convo started with we speaking about a tools website and yeah, still inconclusive xD
React isn't a standard
React is just powerful and very popular
according to statistics, I would say it has become "standard" in some way.
maybe not standard like, idk, python for ai stuff, but it has gained quite a big chunk of ground
That number is misleading since the total is greater than 100%
But again, this isn't the place
meant to represent that 82% of the people that answered use it.
And yes, true.
I use it too, if I have to, but I also use others
That's why it's misleading, it's not asking for the primary or favorite
Fair enough.
Yeah I was massively scrolled up. 
bump

true
the sunsetting message's padding looks pretty odd on mobile
So it does
It's a quick POC I threw together, hopefully I'll have time to refine it tomorrow
Oh, it's the docs, didn't even see that yet
Oh yeah, didn’t think about that
Waterfall is now cancelled
As well as velocity
Thoughts @twilit pivot ?
waterrise when.
But yes, no more waterfall, probably after the next minecraft update
good.
couldn't we make the sunset message into a danger block?
Btw, you did some incorrect grammer in the pr @neon epoch, I’ll add a review when I wake up tomorrow
holy hell why is the format script so fucked on docs 😄
it just destoryed my whole workspace
Yep…
dont have passion to fix that today but I can help with that some day
Probably remove the border left @spice temple
ye looks better
Needs a comma: For more information**,** see the announcement
also and of life lol
looks good, but the contrast between the red bg and the blue link on dark mode is killing my eyes
submitted a PR for fixing the prettier config, lemme know what you think
wasn't sure about what quotes to use, but double ones as default had a smaller diff so /shrug
Is double a German thing?
I use double
maybe im mixing it up with " " and „ ”
bet u guys feel dumb you dont have the papermc docs app
what do you even need to do to have chrome ask you to install an app for it?
honestly, no clue :P
It really really isn't
For the docs it may, but for a proper pwa it makes super sense
I.e. at work I do all the cloud shit in a vscode instance that runs in the cloud too, that works at a PWA I just add to the desktop and it becomes as smooth as the real vscode, since it stores all files locally
(65ecaac42d5bd6481a06489f) // @cursive ivy (@srlinpar / 618408433858904065) has been banned by @eager plover (126975485493248000)
Reason: Quick-banned for sending a message in #docs-website
There's no command api docs, is there?
Wouldn’t it make sense to wait for brig api for that?
just wanted to ask regarding this issue i made like a couple of months ago https://github.com/PaperMC/docs/issues/251 (i know i am pretty late on this
). but i don't really get your point "That included a lot of information about the patch system which isn’t used here". Someone wanting to contribute to paper, would probably want to learn about exactly that right? (considering this kind of patch system is pretty unique) i was just thinking of some kind of "for info on the patch system and general rules to follow when contributing see our Contributing.md"
The thing is that the papermc docs shouldn’t document how patches work (I think that’s what Ollie is talking about). I think the contributing guidelines is at a good position where it is right now.
Yeah, patches aren't a uniquely Paper thing.
Just like the Paper docs shouldn't explain to someone what git or Java are.
I also tried moving it to the paper docs but it just didn’t fit and would mostly be a copy paste anyway
yeah that's why i was thinking about referencing it and not copy pasting the entire Contributing.md into the docs
i mean yeah but java and git are way more known then patch systems
A patch is just a commit tho
It's how git worked before we had GitHub, lol
You would mail around patch files
i mean yeah but when i first saw papers repo structure i was like wtf is this?
i guess it also just depends on who is looking at it. a dev with 10 years of experience of just 1 year. though if we already have docs regarding contributing to paper, i don't see a reason why not to reference the file where all the important commands and formating things are listed.
Do you mean contribution md? In the end embedding works.
yes.
Deliberately wasn’t doing that until the Brig stuff.
And if that was what you were thinking about, no I haven’t written them yet
I think I might have misunderstood your issue, however the paper repo structure is gonna be changing in the short to mid term so I don’t think this is worth committing time to
is there an issue on the downloads page or is it just me? https://papermc.io/downloads/paper it doesn't seem to load properly / allow me to download
the API is having issues
U make good points @fair river - I linked your guide for optimisation but really not an area I’m familiar with too much - it is very opinionated tho I agree
Yeah I point it out on the GH to bring discussion here. I am not entirely oppose the idea but I want to mention it still. 
Are you talking to someone specific with this:
Contribution is always welcomed; however, it is always good to communicate via Discord with our maintainers before proceeding to avoid doing extra work.
Yeah I want them to run though with @neon epoch first 
And discussion on GH can get spammy on people’s inbox.

Do we? I only see that we mention "do backups" and not how to do them.
Ah. Maybe needs rewording then haha
Documentation always wraps on mobile - how do we feel about renaming Documentation -> Docs so that we dont have this weird wrap
^yes
Definitely
someone forgot to run prettier on the configuration page PR 
Why is prettier not running in actions?
PR'd, not sure whether the action is fast enough though
looks manageable when the deps are cached
<@&748618676189528155>
(65f588ec2d5bd6481a0648ba) // @weak laurel (@koitheblub / 1076817962213838879) has been banned by @ancient quiver (1098722699116810331)
Reason: scam spam
finally, no collateral damage
Fast enough for what?
nvm I thought it took a few minutes, a minute without cache is good enough
What about including some basic backup options to point people in the right direction without being overly detailed?
Ex. Just ZIP the server, use a backup plugin, use something like Borg or Rsync
Because providing options generally implies vouching for them
and using plugins for backups is always stupid
(though, if you're running on windows or limited environments, you probably have little choice)
That's fair, if you list it you'll be expected to support it even if you say otherwise.
I assume a lot of cases backup plugins are the choice between doing that or not having any backups, so even if not a great solution better than no backups
Use your host's backup system is probably the most general advice, but there's no way to cover how to do that besides "go ask your host"
See, I'm not 100% against adding examples of them and maybe linking to high quality guides elsewhere, like, DO has a few good articles on general sysadmin stuff
That would be me 😁😁

ollie forgot to run prettier...?
Yes
Proposal:
- Make clear that paper-plugin.yml is entirely optional in that warning.
- Make clear that paper-plugin is not required to unlock paper api.
Approved
do paper-plugin.yml automatically make it a paper plugin or...?
or is the usage of loaders, etc what makes it?
the yml does trigger another load path
Couldn’t we just load the plugin.yml and check for an option like “paperPlugin” and trigger the load path later?
I don’t really understand why there is a different file anyways
The paper-plugin.yml is exactly the option you described.
It is another file because it is experimental and the syntax is in important parts different. It also has different options: It accepts class loaders, The definition of dependencies is different and most important you can't define commands there.
It is just better to not make a mess with an existing working file but do it in a different place.
I know what paper-plugins can do. And the default paper-plugin.yml can also be loaded by Bukkit. It just ignores the paper-plugin specific options.
Because spigots plugin loading sucks
?????
Papers plugin loader behaves differently around dependencies for example, which would be a behavior change we didn't feel comfortable doing
I think you generally misunderstand my comment, so let’s stop this.
If you have 0 dependencies, yes.
except it can't
Like, yea, sure, you can rename it and the common properties will load, but, part of the entire reason why we went for a seperate file name is fallback-ness
i.e. you can have two entry points, one for paper, one for spigot
part of it is also so that we could break the format, i.e. see how dependencies are declared
2 different entry points is a good argument yeah
In the end, I am sure you had your reasons when deciding this.
<@&748618676189528155>
Why might I not be able to access any paper sites? I get this when I try using any paper site https://gyazo.com/f0f1ee9c5ab207e00eaeb9277e7a4fd2
I had this issue yesterday as well, I am on a school network if that makes any difference?
it should work while redeploying as well
hopefully, eh I mean, invalid cert authority is sus
scuffed school computers might be using be an outdated browser?
It's my own computer, tried with different browsers and got the same result
Ehh, school might have an SSL proxy then?
I'm not sure, I came here for anwsers, but guess it's got something to do with my schools networking then?
Well, if you can access the pages outside the school network, the school network seems like the likely culprit 😅
at least in my experience, school networks tend to have some rather scuffed setups
You would need to look at the certificate
many institutional networks have website blockers which generally act sideways on https websites
Not lynx coming for my job

@drowsy zinc mind giving it a final review?

Do i get any role for contributing the docs
Nope, just the joy of being a contributor.
No, but thanks! 

I am getting ERR_CONNECTION_TIMED_OUT when trying to access the javadocs, any suggestions on how to solve this issue?
What javadocs?
Paper's jd
Well, the latest JD page is working here
Yeah, it is likely something to do with my internet connection
When I switch to a VPN the issue goes away 😩
Like, we proxy through CF, so all I can guess is network configuration type stuff
I see, thanks for the assistance!
Yeah, turns out it’s because they are blocking “bad” websites 😟
@cursive portal your CI failed
do we have a migrate to velocity page that outlines a few differences in how you do stuff?
or is stuff so easy that it doesn't matter?
if we had it would go here I guess, lol https://docs.papermc.io/velocity/migration
I have no idea about proxies so not sure if something like that is even needed tho
Honestly just need to rm -rfv the velocity docs one day
I think yeah this is the migration stuff, not really comprehensive though. someone like @fierce adder maybe wants to write some more?
let me check this out
why?
I just find that they're hard to reference
and if they're hard to reference I can only imagine that they're hard to follow
I guess, it seems like a lot of people don't understand the player forwarding section or can't find it?
maybe thats something we would improve on
There is no "follow through" for the guide
Like, you can't be sent to one page to walk through it and end up with a secure proxy setup, it's like a confusing offbranch to two seperate pages, for ex
Yeah, that page is old lol. I would honestly love to but I'm not sure on exactly what needs to be added, I also gotta study for a test tomorrow so I can't really do anything right now.
this is not urgent, good luck for your test
yeah, thats a good point. When I first read the docs, I was basically just reading every page because I didn't know where to go. And even then I had to re-read a bunch of things.
but, if you wanted to add some random stuff about conversions then that would be great 
ofc, I'll check it out tomorrow :)
I think a revisit and maybe a follow-along style guide will be awesome. (Just for the proxy section) as I found many people who don’t even understand what a “proxy” is
Most common one is just people thinking that “velocity” is the lobby server.
yeah, for sure
For the waterfall eol PR. Have you thought about changes to the front page of the docs website. eg. Moving waterfall all the way to the bottom, adding a eol badge next to it, or something like that?
Changes to red on hover
could swizzle the navbar if people were interested
Like this is what i mean
any icon for archived...? That actually does the oppsite to bring attention to it 
yeah its only visible for the first 10 seconds of the page visible
and then it dissapears lol
Make the Waterfall text 95% transparent
maybe just a simple greyout or apply some transparent would do...?
Strikethrough perhaps?
oHH MISC
@neon epoch, why don't we remove it entirely 😉
strikethrough def looks best
Whatever changes you do, please also do them to the website for consistency
Or leave a review on the website PR do I don't forget
Sure, so far I’ve only copied you
I don’t think it looks good to be honest
It doesn’t fit into the Navbar
Not really a fan of any of them tbh. gray looks like its selected, strikethrough just looks weird.
Maybe a small red "EOL" badge instead?
yeah but I mean, its better than a random gray color
I mean, its hard doing that in a nav bar though ;(
yeah, that's what I said lol
but thats not the only place it comes up too, so we'd have to make sure to remove it from all navbars
removing it from the navbar isn't that bad anyway because it makes sense to only show software thats maintained unless you go to some other page
like for example right here it would be easier to add a EOL badge and just remove waterfall from any navbar https://papermc.io/downloads
I thought you meant it as a joke cause 😉
ig it kinda was and at the same time not.
something like that looks good yeah
I like the red buttons but the EOL header is really red and bright. Shouldn't it be the same as the docs?
maybr ad a hover over it? That explains that it's deprecated / eol
I honestly don’t know if the box at the top is really needed. Have you tried making the text under “Get Waterfall” red?
this looks fine other than the top header maybe being a little much, it is needed to drive peole away though
for sure
(65fd7d9e2d5bd6481a0648d6) // @proud sundial (@jamdzu / 771828196898242580) has been banned by @ocean void (188417437295706113)
Reason: NSFW spam
I'm not sure the box is big and flashy enough!
I think just fixing that padding would be enough
for visibility https://github.com/PaperMC/docs/issues/325
oof here's an old one for it from before the new era of docs https://github.com/PaperMC/docs/issues/21
B.D.
Before docs
ye sulu dug it out, there is enough into to get started on a page I guess
well, we had docs, we just had no ollie ^^
Looks good, I’m not gonna get a chance to look at this for a little while if anyone else would like to write this as a PR!
Thanks @vocal halo I’ll read through at some point later probably
wait why is waterfall being removed
it sucks and isnt really maintained anymore
i still use it
besides upstream updates
people still use 10 year old bukkit plugins but that doesn't mean you should
velocity has existed for quite some time
wdym
then i think everyone here will recommend you switch to velocity
exactly what michael said
it's not bungeecord based
yeah but its good
velocity solves most of the performance issues bungeecord/waterfall suffers from
and a better developer experience
what does a proxy do
you run waterfall but you don't know what a proxy is
- doenst it have a more secure connection?
and that
the server runs it
what
im just wondering what does a proxy do
well im the co owner
connecting multiple servers basically
oh that would make sense
u see this one
why doesnt alice directly ask bob for the time
instead of having the proxy act as a middleman
what..
@sacred socket I suggest you to read the docs of Bungeecord when you want to know what it is.
how does it not make sense
This channel is for updates to the docs. If you have general proxy questions you should go to #waterfall-help or #general
do we actually even have a what is a proxy docs page
don't think we actually explain what velocity is anywhere
So no, not currently.
I feel like it would be helpful to write docs about how velocity is a proxy and what a proxy is but I feel like, people using it, should kind of know what a proxy is. If I was writing docs for a vpn, I wouldn’t write a whole page about what a vpn is or why it might be needed. If you’re looking at documentation for something like that, you should already know what the software does.
I mean, it's a matter of context; people looking for a VPN generally learned what a VPN is for already, at least in the proper sense of a VPN
not everybody looking for velocity knows what velocity is, they just heard that setting up a network is easy and fun and woo, servers! or something
True but I guess we could just have a short introduction then, maybe link that Wikipedia page as a Learn more here
We really just need to fix the velocity docs as a whole
I mean, a proxy misses the "multiple servers" aspect unless you show an LB proxy
it's probably worth it just to add a paragrapth or two as to what a proxy is and why you'd use one
"used for letting you hope between servers in one session, and can also be used to host ingress elsewhere (i.e. VPS, but, might want haproxy for that instead)
Yeah I think this would be nice and if anyone willing to contribute that will be epic 
The introduction page should make assumption that one doenst know what a proxy is and a brief explanation for like one paragraph should be enough imo.
Is there a specific reason haproxy is the TCP proxy of choice? Is it just because they made the proxyv2 protocol?
Basically, yes
I mean, not that they made it as such, haproxy is basically just the minimal solution here
you could use anything that implements the proxy protocol, like, nginx supports it
The description is mostly for beginners, the idea is maybe to add a not to something more appropriate for a specific usecase in which has a relatively simple configuration that you can google for
Yeah
I’m sure exactly how it should be explained but essentially velocity is just a gameserver proxy specifically made for Minecraft. We could explain that in the gameserver world, you need to hop around servers and stuff.
is everything supposed to be red when hovering over it? i thought that should only be for waterfall
welp ig all PaperMc projects are eol now 
ollie bad
that's what happens when you merge prs without approval lol
(oh nvm mini acutally merged that)
Lol
also did the project names always redirect to the github?
yes
Project.eol && "whatever" is a more concise way of saying that
Iirc react also has an object syntax, {{"something": true, "some other class": project.eol}}
does that work correctly when concatenating to a string though
typing "" + (false && "test") into a console yields "false"
I thought that that one was provided by some util method thingy?
My eye hurts and I kinda wanna finish off this 'tech challenge', so, if somebody else wants to improve it or whatever, go ahead
Am on the couch, I can try it tomorrow
In Vue it works without special lib shit
But react do be stupid, so idk
I would just use a backtick fstring like everywhere else
I was using tailwind merge for the thing I was thinking of
I’ve merged it for now, if anyone comes up with a better way we can just change - thanks for the report and fix ❤️
was it this? https://www.npmjs.com/package/classnames
that looks like one of the options available
that is already used in other places
or at least was
ah just with clsx instead of classnames - seems to accomplish the same thing though
ok, className={"projectGitHub " + (project.eol && "archivedProjectTitle")} turns into projectGitHub undefined and projectGitHub archivedProjectTitle, lol
className={{projectGitHub: true, archivedProjectTitle: project.eol}} turns into [object Object]
right
className={clsx({projectGitHub: true, archivedProjectTitle: project.eol})} works as expected
insane that you need a lib for such basic stuff in react
its 228B minimized, lmao
there's a "lite" version that's 140 bytes lol
can even do className={clsx("projectGitHub", project.eol && "archivedProjectTitle")}
def looks more clean than the object syntax
some more chores regarding clsx:
- inline style could be turned into a css class that uses clsx
- ternary operator here
- another inlined style
- another ternary operator
- and another ternary operator
- inline style
- inline style
- inline style (although it's dynamic so could be skipped)
- ternary operator
dont have time to do that rn, so feel free to PR or open an issue
(battling gradle rn, wooooo)
i do not envy u
I also plan on looking at paperweight patcher later today so you dont have too ^^
the lowerCamelCase class names are also pretty concerning, should be kebab-case like the rest probably
where do you see that?
Everywhere
I think I use kebab but a lot of the original CSS uses camel so there’s a bit of a mix
there's some snake_case as well lol
lol
you and your usage of anchor tags instead of Link components smh
basically; Link component = instant loading, anchor tag = refreshing the whole page
and some other fancy stuff behind the scenes
Ahhh cool
@neon epochwas there supposed to be styling in the dropdown of the script generator flag for a selected flag type? there's no dropdown-item-selected or selected class, so i'm guessing it's an oversight https://github.com/PaperMC/docs/pull/331/files#diff-2f3095b4ce3d91ad1945b2b62d5ecda708034aad3757be558773943f2e7832b5R211
It just doesn’t exist tbh
Like, you click on it, you set those flags. If you reopen it, you can set new flags. Didn’t bother having a selected at the time
If you wanna add it, be my guest :)

Awesome

ok enough testing, time to set my phone back to uk time
Nah
maybe we should consider having the start script gen add a pause at the end for when windows is selected (considering most people insane enough to run a mc server on windows are going to just double click the bat file instead of running it from a terminal) since without it the terminal will just close when the server stops for whatever reason.
@fierce adder
is the docs site accessibility compliant?
pretty sure it is, april fools? maybe not
@lilac edge This slider is pretty innacurate now, it was not great before, but it was centered so it was minimised. didn't know if you wanted to have a look at that?
i found a stack overflow answer that i wanted to experiment with but i can't find it now
Yeah I am talking about that one. I can read git commits. It was a response to them.
Yeah, I made sure it generated colour combinations that were different on light and dark theme and clamped so that they would be somewhat readable, if a user cant read it then they may refresh the page to get a new combination or just disable it using the button.
found it https://stackoverflow.com/a/71041703 i can mess with it in a bit
how can i set the HANGAR_API_TOKEN ? apiKey.set(System.getenv("HANGAR_API_TOKEN"))
by setting an environmental variable named HANGER_API_TOKEN with your api token
set HANGAR_API_TOKEN=token in a cmd?
gradle.properties or somewhere
HANGAR_API_TOKEN=token inside gradle.properties?
probably
probably not, it doesn't work
Who knows anymore
what?
i don't know where to put the token
i was able to do it
now i have this error `Execution failed for task ':publishPluginPublicationToHangar'.
class com.google.gson.JsonNull cannot be cast to class com.google.gson.JsonPrimitive (com.google.gson.JsonNull and com.google.gson.JsonPrimitive are in unnamed module of loader org.gradle.internal.classloader.VisitableURLClassLoader @688fdb67)`
send your build script in #build-tooling-help
it's too large to be sended
or just hangarPublish?
should be better now
heya
does anyone know a plugin
that allows people to claim land but it only shows up on dynmap
and lets non members of the claimed land still break blocks?
whats the point of claiming that land then?
well you can see it on a map
okay... but why
what relevance does it have
i have ways to revert griefing but the no more block breaking approatch isnt that good since imagine your in a mine you find diamonds and then you cant pick them up
for some reason everyone on my server loves the map a lot and ive been filling it with the area commands but that gets a bit annoying at some point
so i hoped there was some kind of plugin for that
You should keep this to #general
Please don’t ask the same question in multiple channels.
i got a bit confused with the new channel names
@fierce adder
Plesse stop randomly pinging people
if anyone wants a free PR you can use https://docs.papermc.io/paper/dev/internals#getting-the-current-minecraft-version for auto-versioning, otherwise i may do it at some point
Guys you think AMD Ryzen 9 5900x or Intel ® i9-10850k is better for paper server
Could you please make your way over to #paper-help !
no

yea I did test it
we all need to pray!!
PR material for anyone: #paper-help message
to clarify, this is the permission section https://docs.papermc.io/paper/dev/how-do-plugins-work#permissions
a space that shouldn’t be there
yea, the note tag was messed up
would be nice if markdown had some of this stuff as a standard feature, rather than basically everybody hacking their own features into it, maybe then we could have useful markdown tooling
May be that the docusaurus update was more strict with this, not sure how long it’s been there. Thanks cat
yeah, I don't have time to make a PR since I'm in school so I just decided to highlight it, maybe I should have made a github issue but whatever
I already nailed it
it was a commit from a few days ago, so, eeerrrr
sure, ollie, sure
||<3||
nice
womp womp
hey there i was thinking i could implement one of the the page requests which is addressed here: https://github.com/PaperMC/docs/issues/261
are there any general tips aside from reading the contributing guidelines which you mock a lot of people for?
I think the velocity page linked in that PR looks like a good model
right
Following the velocity page sounds like a good start! Thanks for showing interest in contributing 
also if you have any questions or get stuck do not hesitate to ask, we can help you
thanks!
is there any standardized format for commit messages? looking at history it doesnt seem so, should i just stick to format that i find readable & usable then?
no, yes
I’ve been trying to stick to a fix/feat/style etc recently but name it whatever and it’ll be squashed anyways 😄
@chrome aspen I am making the page already.
I have a little work to do and it's done
Ah i see
Havent done much yet so its not a problem
Maybe next time it would be worth making atleast draft pr? (If it exists mb, i didnt see it lol)
I forgot to make it.
Sorry
No problem, atleast you told me now 👍
I'll make a draft PR tomorrow cuz i haven't committed anything. It's saved on my pc.



