#code-discussion
1 messages Β· Page 154 of 1
it's just an organization thing you can do whatever you want
it does affect caching though so having more screenguis is slightly better for performance because an update to a UI element would only re-render the screengui that contains it
never tested this fact though
After 3 years, I finally discovered how to make things react to music/audio
after 6 months i learned how to mske a frame visible
well forgive me for taking breaks and the Roblox dev forum not being straight forward, having every time you google it just being "How to add music to your game!!!!" and this server just insulting you when you genuionely need help and not knowing how to remake ChatGPT in Roblox
guys help, my game logic and everything is ruined after trying multiple times.
If you don't want to code noone is forcing you to. More money for us π€·
that match is absolutely terrible π what the hell
@oak verge
?
You mean just based on volume?
Or like pitch too
RAAAAAAA
yes
Fyi everyone does this idk who u talking to btw. Also if you ask smth basic expect to be insulted
,I mean you can search it up on youtube or mess around in studio and figure it out. Coding fundamentals you can watch yt for or smth.
Was it just reading PlaybackLoudness
Join the challenge or watch the game here.
If smth is too complicated start smaller or do it in parts. I think I saw smth earlier that said make systems not games.
But I thought you do that for games anyways idk.
e.g you make a gun system or a physics system and can use them for different game ideas.
which is better for safety:
(large table data)
- json encode (over 100k str len) and store it into datastore
- json encode (over 100k str len), then buffer.fromstring it, then json encode it again and store it into datastore
from my test, that method 2 reduced the string.len upto ~3-5x but both still have the same byte size.
does it have to be json
because you can make it much smaller without json
so just serialize it into buffer instead?
yeah there's open source libraries for that
if you know what the structure of the table looks like beforehand you can make it very small
it contains vectors, cframes, numbers, string, arrays
im trying to find something i made using this a year ago
https://github.com/kampfkarren/protoc-gen-luau
but seems like it's disappeared
I ain't pulling up π π π
idk whats popular today but you can use this
https://data-oriented-house.github.io/Squash
guys i spent like half a DAY just trying to comprehend the BASICS of CFrame, am I a bit slow to learn scripting TvT
hello guys i am making controls similar to the football game FIFA, and I was wondering, should I make a gameplay server sided, wich means if your network is weak its not gonna affect the game fundamentals but mostly just delay your actions, or else doing half gameplay server half gameplay client, taking the risk?
and I dont know much about powers, nor physics, Innitialy I am an UI Designer / frontend scripter
depends are you ai π€¨
I think you anwsered your own question
keep it server-authoritative
what's the basics for you?
I'm 2 years in and still don't fully understand CFrames
Wait what π Iβm not AI
susssss
Cframes are not intuitive unless you are familiar with linear algebra
Which most people aren't
- and if you hate math
^ this is why most people aren't familiar with linear algebra
Algebra and knowing cframes are a position in 3d space are two different things. you can js say its math which usually you can solve using parts ingame.
Unless ur tryna make smth with math then idk chatgpt it or smth. π
Nah cframes are pure linear algebra
x,y,z
They are 4x4 transformation matrices
anything you do with CFrames that isn't moving an object or rotating it, is a bit more complicated
give an example
I need bigger brain
Transforming from global to local coordinates
Or vice versa
There's functions for it but still
bruh what does that even mean. Are they not the same location. You could just clone it as a client object or smth.

Me when local can mean things other than client:
β
I refus
lol
π
Local coordinates as in, the same cframe but with respect to a different origin
this is a problem someone posed in 2022
bro... I don't understand it to give examples
but the functions I see and don't understand are: Dot, Cross, CFrame.fromMatrix
easier functions: ToWorldSpace, ToObjectSpace
and example code that I don't fully understand:
local function getRotationBetween(u, v, axis)
local dot, uxv = u:Dot(v), u:Cross(v)
if dot < -0.99999 then
return CFrame.fromAxisAngle(axis, math.pi)
end
return CFrame.new(0, 0, 0, uxv.x, uxv.y, uxv.z, 1 + dot)
end
local function getSurfaceCF(part, lnormal)
local transition = getRotationBetween(
Vector3.new(0,0,1),
lnormal,
Vector3.new(0,0,1)
)
return part.CFrame * transition * EXTRASPIN
end
it used to be pretty fun solving things like this but now mfs just use chatgpt
Dot and cross are just vector operations
Cframe.fromMatrix let's you construct the 4x4 matrix yourself
I hope I can use gpt π it doesn't spit anything useful for me
even the code above I got it from EgoMoose on the devforum
is that like animation with math or smth
my solution
||
local function projectOntoUnit(u: Vector3, v: Vector3)
return u:Dot(v) * v
end
local newPoint = target - projectOntoUnit(target - intersection, normal)
||
No
I get the theory, but not how to use it
thats what programmers do lel
this one isnt hard but it does require a solid understanding of vectors
is that not the same word
maybe
there are harder problems but id have to dig pretty deep to find them
to position a part using the raycast normal rotation? ( can't explain )
Im sure theres easier solutions idk
Newpoint = result.Normal * (target-result.Position):Dot(-result.Normal) + targetPosition
that looks about right
Vectors and Cframes are extremely easy if you know how they work
Linear algebra is so useful
β
Especially in my physics classes i see it everywhere
I mean I havent dived deep into it but I get the basics ig. It just seems like understanding the question and than getting what methods do like dot cause I haven't used that yet.
please donβt say itβs easy π
lol
this one isn't vectors but it was interesting (i think)
everythings easy once you understand it π€ π
I see Pascal's triangle
π
thats a forumla though they doesnt explain anything lmao
I think if u just know the basics of linear algebra its pretty easy
Not sure what he's trynna do though
Me when person know math, Ah yes the photosynthasis forumla
The mathematical phenomenon of photosynthesis
lol
he just wants to distribute the squares evenly around the 0 point
In a circle?
no, just a line
like
yea I'm trying to learn it online
I didn't learn shit from public school they just told us to remember x, y ... even the gravity formula I learned it afterword
Theres tons of really good books online
Linear algebra done right is really good
Yeah they ain't teaching you linear algebra, that's curved arithmetic
π
I thought linear algebra was taught in college
Ok i think i kinda get what they mean
he has a given number of squares to distribute evenly along this vertical line (i.e. the distance between any 2 squares must always be the same)
and the center of the distribution has to be at a given point, in this case it's 0
Given a number N
Oh that's just a basic mapping function
the answer:
||
y(i) = d * (i - (n - 1) / 2)
where:
i is the zero-based index of a square in the list, starting from the bottom
y(i) is the position of the center of a square at a given index i
d is the desired distance between the center of each square
n is the number of squares
||
Isn't linear algebra about matrices and stuff
Ty, I'll give it a read
Yes
what grade
math.map(i,1,n,-n/2,n/2)*40
They get really theoretical down the stretch though atleast for me, things like vector spaces, eigenvectors, eigenspaces, diagonalization
what does this function do exactly?
It maps from 1 range to another
idk mate π I just know that there's stuff that I learned from the internet that I should've already known in mid-high school
so it creates like a region of possible values per 1 input? or what
For example if you wanted to convert a range of 0 to 1 to a range of -2 to 3
Rip Vro
Nah
It's just a mapping function
It takes a range and converts it to another range
so it just codes a range of values..
sure thanks
I think they only use the basics of it in gamedev?
not the advanced stuff
Yep
The most advanced vector math I know is just dot products
Cframes are 4x4 matrices so yes linear algebra
what would anyone even need math.map for
Honestly, those stuff are all you really need, you dont need to anything else
Atleast from what I know
Very common
@somber vault @stuck vortex
don't look here
https://discord.com/channels/211228845771063296/1308882894302875768
Like for doing rng
nope, alot of other important math calculations are necessisary in gamedev
Looking at my physics classes, like we mostly just used matrices, dot products and cross product
Of course theres a lot more but i don't really think you need all the theoretical stuff from linear algebra
Most of the math other than basic arithmetic in gamedev is infact just basic vectors
Yeah
not necessarily
just finished a $200 comm! (usd)
was for scripting lol
js so happy they did not scam
I had to make a probability algorithm for my obby game π
i think math.map didn't exist when i answered this
You don't need to know how cframes work under the hood to use them, but it will be way easier if you do
don't remember though
Maybe
I only found it recently
U made ur own algorithm ?
I would just make my own mapping function when necessary
any of you guys heard of Khan Academy? what do you think of it?
j
Loved it but now that I look back at it, i only like the stuff where they explain concepts and not solve problems
I think when people look online to solve problems, it's plagiarism.
local map(value, r1start, r1end, r2start, r2end)
local alpha = (value-r1start)/(r1end-r1start)
return r2start+alpha*(r2end-r2start)
end```
^ this function is equivalent to math.map
Is just a quality of life thing
The reason why i think this (for people in STEM) is that you go into STEM to learn how to become a scientist, I think you should learn on your own and figure out the problem via textbooks, you needa learn how to research when looking at a problem, you shouldn't look up a similar problem online to solve it, it is plagiarism.
I think it's good yea
There's also openstax textbooks which I find to be okay
What's it for
for learning math
Oh
Just learning in general
huh
They have other subjects too
yes i arranged obbies into teirs, and used Binomial and Cumulative distribution to arrange it into a bias if its in a specific significance level, so pretty much, if I want teir 1 obbies, i would make a significance level (acceptance value) of 10, give teir 1 obbies a value of 3, teir 2 and 3 obbies with a value of like 1.7, then pretty much use a randomised value and the probabilty of the obby value being in the significance range, then multiply it by the value assigned to the teir, then interpret the range results, check if its above or equal to the acceptance value, and then it accepts the obby, then it clones that obby and positions it correctly in the game, so you get a range of obbies, for my game theres like over 20^7 variations for each match.
Just watch 3blue1brown π
He does have a good series on linear algebra ig
I did, but when I'm too stupid I have to try many different sources :D
this seems to give incorrect results
local PROMPT_DISTANCE = 40
local function f1(i, n)
return PROMPT_DISTANCE * (i - 1 - (n - 1) / 2)
end
local function f2(i, n)
return math.map(i, 1, n, -n / 2, n / 2) * PROMPT_DISTANCE
end
for n = 1, 5 do
local prompts = {}
for i = 1, n do
prompts[i] = f1(i, n)
end
print("f1", table.concat(prompts, ", "))
prompts = {}
for i = 1, n do
prompts[i] = f2(i, n)
end
print("f2", table.concat(prompts, ", "))
end```
Pretty much, when someone looks at a problem, and then go online to watch someone do a similar one or the same one and do it, then they do it like them, i think it's plagiarism.
Oh yeah the linear algebra stuff is great
my brain is too turned off to think about why it's wrong but it is
I'm talking mostly about STEM majors
Oh so you're supposed to bang your head against it while teaching yourself misinfo
Off by one error π₯
The animations of the vector space are really helpful
yea that came to mind hold on
?
this little text transition comprised of just over 100 lines of code is why im still awake at 2:49 am
Hello everyone, is there someone who can help me set up the voice chat player and the hearing of player sounds in Roblox?
whats the difference between researching on a book and online π
okay my brain is still disabled i dont think its off by one
how much ad creds do you have
or is it
Where are the 100 lines
Nah it is
lines of code*
Oh I thought it was crack, mb
probably just type writer effect, and tweening day / night

There are three transitions and a text effect
When you study with textbooks, you teach yourself, a page can hold a 10 min video worth of information
yea thats atleast like 50-90 lines of code
How
the typewriter effect is based in the center on the text instead of left -> right or right <- left
How does your code look π
It is dense, it is filled with information and with little tips and advices from the author
Not only that, it shows TONS of examples and practice problems
do you wanna see it
Yes
how would i fixi t then
I dont why im getting questions marks lol its true
Because it's not true in a broad context?
isn't this a bad thing?
How so
i agree but disagree, youre talking about watching a video as research, most videos dont give you as much information rate as reading and interpretting the information yourself, as videos interpret the information for you but most likely simplify it making it too vague understand. I like researching WAY MORE online, i think what you meant was reading for research and information rather than watching videos. Online research will always be better than just reading a research book
When u go into STEM, you learn to become a scientist, you need to learn how to research
How do you learn to research?
too much information in a small space can confuse more than help
maybe that's a good thing if you're reading to learn 100% of the subject
Thats why you need to learn how to read a textbook
I mean researching online is a valuable skill
Just need to verify your sources too
research in STEM is accepted more viably from online sources
You need to learn to read a text book 
guys i can code and can make graphics for your games dm me if you have any work
When i was talking about online, i was talking about people doing problems where when you watch it, you just copy their steps
there are plenty of online textbooks
You guys went sideways and made it broader and talked about RESEARCHING in general
yea ofc thats dif
click view whole file
math.map(i,1,n,0,(n-1)*d)-d/2
You fell flat when talking about the entirety of STEM and ALL BOOKS and ALL ONLINE SOURCES
but there's no learning in this π
they're just digging their hole deeper
Thats just me, i think if you look online to do a problem, you watch someone do something similar OR possbile THE SAME one, I think its plagiarism, thats just me
;compile lua
Oh yeah that doesn't work here
You're talking too broad for it to make any sense
π After 2 months of intense solo development, I just launched SuperbulletAI , for free . Every users now gets 1M free tokens/month to use a purpose-built AI assistant just for Roblox game creation. I have over 9 years of experience making Roblox games now. Coincidentally, 9 months ago, I wasnβt even into AI. Iβve always been a Roblox g...
that statement is just wrong, so just say "when you watch a video of someone solving a problem and copy their exact steps"
I had to ask... why this?
Module.Init = function()
local FadeElements = function(Time, Target)
local Transition = function()
and not this
function Module.Init()
local function FadeElements(Time, Target)
local function Transition()
local Area = require(AreaModules:FindFirstChild(AreaValue.Value))
Handle this a little more gracefully
were absolutely cooked
preference
Does anyone know this aforementioned veteran dev
is there a difference
still wrong
local PROMPT_DISTANCE = 40
local function f1(i, n)
return PROMPT_DISTANCE * (i - 1 - (n - 1) / 2)
end
local function f2(i, n)
return math.map(i, 1, n, 0, (n - 1) * PROMPT_DISTANCE) - PROMPT_DISTANCE / 2
end
for n = 1, 5 do
local prompts = {}
for i = 1, n do
prompts[i] = f1(i, n)
end
print("f1", table.concat(prompts, ", "))
prompts = {}
for i = 1, n do
prompts[i] = f2(i, n)
end
print("f2", table.concat(prompts, ", "))
end
true
Slight
wat is the differenc
Its never too late for us to boycott π
fair
is there a difference
don't think so, but most ppl find it easier to read the second one.
honestly true, the local x = function() kinda clashes with any other variables around it
I don't know if boycotting will be necessary
You can't define a method with the 1st way
compiler optimizations (not sure if they changed it so that both are equivalent)
and in the first one you would get a warning from your editor if you tried to call the function inside itself
π₯
It's
β€οΈ great for him
I guess kids will finally be allowed to make low earning games and buy gag gamepasses
math.map(i,1,n,0,(n-1)*d)-n*d/2
until it gets saturated because it can't do much
And uh
also that ai model is definitely just a repurposed model
It's still gonna give nan for n = 1 i think
thats not the problem . If its easy for him to make games with scripts and earn 5k rbx per week, then the game development market will soon become oversaturated and roblox's cost per effective games will go up due to low effort slop being produced
yea
Since it's not designed to handle ranges where start = end
Yeah but he probably just got lucky
or he's just lying
he sounds like he's 12 and I don't know where he'd get money to market his game
most likely "connections"
5000 robux a week is only $70 a month
you don't have to worry about the market being saturated with people making $70 a month
Yeah that means he makes 700R$ per day
with 1R$ per player
that's 700 visits
that's mostly 0-4 ccu
to make 5k a week from a game atleast requires 30-40 ad creds, and a pretty addicting game, and for its possible ccu it has to do with a trend.
this is unrealistic, what ccu consistently spends money on a game if it isnt trend slop
That's what I'm saying
It doesn't make sense that it's even possible to make so little consistently
Your game is either dead or not
if its trend slop the demographic is a kids audience, kids audiences have less robux, likely 30-150 rbx
so the passes and developer products have to be like 5-50 rbx
π
so if its 5-50 rbx the average should be like 27 rbx spent, so 5k / 27 - > 185 players paying 27 rbx average per week
and since not all ccu would do that
its most likely a game with like 300-500 ccu
so he gets average 400 ccu
Yeah that's not right
Depends on the PCR
PCR?
Conversion rate
as in the quantity of players willing to purchase products in a game?
thats high for trendy slop games
CCU is based on how many DAU you have and how long your playtime is
its atleast like 15-30% for slop games
and it heavily depends on the price of the pass also
30% you oughtta be next Jandel
Nope, depends on the price of the pass, if its towards a kids demographic and the average price of passes is 5-50 rbx then it would make sense for the PCR to be high
for example look at these games and this genre
Obby?
the target demographic should be kids 12 and under, and the gamepass prices are often very cheep
You're giving really conflicting information
people get them confused for static obbies which are like
Do you know what the usual robux per player is
Just take this game for example: https://www.roblox.com/games/16646849903/W11-Pogo-Obby
[UPD 27/1/25] - Added WORLD 11 π΄
Beat the obby on your springy little pogo stick.
π€ This game is DIFFICULT it takes time to get used to
πΎ Your progress saves, return to finish worlds any time.
π LIKE the game & JOIN the group for FREE in-game rewards
π Make sure to FOLLOW the DEVELOPERS for more games!
π Thank you for playi...
Check its demographic and average pass price
You're making assumptions based on prices π
You can't even check its demographic
you can literally judge its demographic by seeing the users per each server
There are 800 players
the way their avatar is , their typing patterns its good enough assumption
by checking one full server, you've covered 4%
how could i make like a drone be able to do flips its like a fpv drone it truns left right up and down but cant do flips?
Not true either
Ive played the game plenty of times and been through plenty of servers, so it would be more like 30-40%
A lot of purchases actually come from people who have free avatars
Yeah but your numbers are still bogus
so youre just proving my point π
exactly they have robux likely in the range of 5-100, becuase of roblox's cheap mobile option for 80 rbx
also your point is completely wrong statistics
and for some countries 40 rbx
and that this game has noob avatars
Yeah
that doesn't make your PCR 15-30% π
can you imagine how rich you'd be if that was true
imagine you get 100 concurrent, for 10 minutes
throughout the entire day
14400 visits
or players in general
at a
sub par rate
of 1 robux per user
you made 14400 robux
15-30% = 22.5% average, 800 x 22.5% = 180, players average spending 5-100 rbx, lets say per day. ~ 48 rbx x 180 = 8.64k robux a day, which isnt far off for a game like this https://www.roblox.com/games/16646849903/W11-Pogo-Obby
[UPD 27/1/25] - Added WORLD 11 π΄
Beat the obby on your springy little pogo stick.
π€ This game is DIFFICULT it takes time to get used to
πΎ Your progress saves, return to finish worlds any time.
π LIKE the game & JOIN the group for FREE in-game rewards
π Make sure to FOLLOW the DEVELOPERS for more games!
π Thank you for playi...
Those are completely made up stats π
Even if theres some margin of error it should ofcourse be making atleast 3.5k+ robux a day
1 robux per user is just so unrealistic and such a fake stat
its better to use robux per percent of players in its demographic, then break that down into robux per average user
thats simply the most basic and vague average, so much so its margin of error is huge
people have games with 500 visits and have made atleast 3 robux
real games with different demographics, different PCR, different amounts of gamepasses, different style of games
so its a vague average that barely applies to your game unless its like theirs
1 is sub par but it's not completely uncommon to have below 1 but then it's considered bad and you should work on it
ITS ACTUAL STATS
FROM ACTUAL GAMES
my statistics make sense and is likely
Btw those actual games include these π
your statistics are vague, taken from roblox's top games
no?
and if you know about statistics and average, the games with the largest user base heavily effect the average, especially considering their target demographics
I mean top games probably don't have better stats I think, they just have a larger amount of players. But I don't know about that
and to even aquire 1 robuck per user average youd need a fairly large amount of ccu, an interested demographic and a good enough PCR
huh?? π
large amount of players literally messes with the average measurement π
no?
are you ragebaiting
the most common thing in statistics is that bigger sample means more accurate result
just because your game jumped from 100 concurrent to 1000 concurrent
doesnt mean users are any less incentivized to spend
bigger sample also means more averaged results, meaning less accurate sampling, the other most common thing is massive sample sizes are awful and smaller sample sizes like looking at smaller games are better, as they can reflect the larger statistic
its called quota sampling
you realise that if out of 100 people, only 5 players donate 20 rbx, then the average taken by the sample would be 1 rbx per person
yeah
with a smaller sample you get more accurate results, that arent determined by a vague average, and you get more information per the statistic who donated in the 100 sample
thats the point of quota sampling
PCR is 5% in your example btw which is still way too high but it makes sense because avg paying is low
any reliable statistic uses quota
"this doesnt make any sense!! his annalogy that describes exactly what ive been saying says something i didnt want to say!!"
per the annalogy the PCR is 5% becuase the sample size is small
No you just got closer to what I'm saying with your example
Yeah
you are aware you previously said PCR is 20%
and then proceeded to claim that more players = worse stats
youre proving my point as i literally stated for the average statistic of 1 rbx per person to match the average statistic of PCR 1% ish, the sample size of ccu would have to be at a threshold and way higher than the average newly released game, to get statistics like that the game would need to be of these statuses
thanks master
im saying that per the stats you mainly accept (PCR ~1%) its too vague and doesnt apply to the majority of games, only unless it has high enough ccu
except it's not made up
wah? I thought you were gonna block me for this
... I mean, I would
I couldn't block my belovered
the average is completely wack and i already disproved it π
proof please
im done w/ this
Would this be a good roblox game?
What is this?
A game called dare or doubled dare i found inside of my favorite television show.
Yeah no but what's the game like
First it pciks two random words that you have to do hten if u dont want4 to then you can do a double dare
which is just a different one ut its two words
the 4th segment is a longer word almost like a sentence.
I mean
no idea how you'd implement this
but I think it could be a pretty good social experience
I would probagbly implement it inside of a hangout game.
Definitely but how would you make dares that are actually interesting
and also still follow TOS
Something like "Collect 5 coins areound the map"
or "Eat hair"
Yeahh I don't know about those
No
Lower percentile average of obbies is 44 Robux per paying user.
At an assumed 10 minute playtime and assumed avg ccu 800 (due to peaks) that is roughly 115200 players in a day.
If the PCR was 22.5% there would be 25920 paying users at an average of 44 which makes 1140480. A DAY.
Do you see how that PCR makes no sense?
for proof, that is the Roblox given benchmark for 50%-90% of games in the obby genre
flawed stats, one like i mentioned before there are different types of obbies, that yield different stats, two your argument assumes the rate of paying users is every 10 minuites, thats completely flawed. I mentioned that the rate of paying users on LOWER BOUND average was around 47 rbx per day for each paying user,and the PCR being 22.5% would overal result in 47 x (800 x 22.5%) = 8.46k robux a day
You're a fucking idiot
argument please?
How am I meant to make an argument when you're just stupid and wrong
π and no argument
good you conceid π
the misconception lies in the fact that they do not have 800 players in one day
they have 800 players RIGHT NOW
and if 800 players RIGHT NOW play an avg 10 minutes
that means they will cycle through
You're not accounting for the fact that most of those players come back
roughly 144 cycles
And aren't unique players
I believe retention is counted into those stats
also no
most players do not come back
Depends on the game
most would mean 50%+ retention
so mind you your argument was that the lower percentive average of obbies in general is 44 rbx, once again a vague average statistic π , and 800-500 users is the games DAU, hence why i literally said
For something like deepwoken most players are returning players
I don't think you can hide behind irony
Id also assume grow a garden has a lot of returning players
π sure show your next argument
But I wouldn't assume a obby game to have a lot of returning players
You genuinely believed completely wrong statistics and now are trying to sound stupid on purpose
I can check if there's a becnhmark for it
π youre genuinly melting and not adressing my claim
D1 4,75% retention
on a lower percentile
Factor that in the calculations then
it already is
Oh k
also 5% wouldn't make a significant difference
compared to the completely wrong PCR
also you could probably check the avarage playtime of players instead of assuming
I'm preety sure there's a tool for that
Today and yesterday btw , 23 hours timelapse
Its an estimation too but
still on average 800 DAU
10 minutes could be generous
Do you know what DAU is?
since if they have a lower time and still maintain 800, that means they're cycling more users
daily active user
mhm
Daily
active users
not average concurrent
or concurrent
great and i showed you little standard deviation in the users π
Hmmm
are you a spammer
infact if i send a timelapse of its user count per hour for another day it would be in the range of 500-800 users like i said π https://discordapp.com/channels/211228845771063296/1138164783967899678/1401520800745525258
Same
You don't have to
Dude there's tools to check the playercounts of games
He's not trying to be right
He's trying to be stupid to ragebait
Sure bud, so once i get the tool and show you youll conceid π
where do i search up the game
Here's the link for the game
You can see the avarage playtime is about 6-8 minutes
makes that PCR look even more outlandish
The playercount only drops to 710 on the lowests
And there's about 1k new visits every hour
yeah that one
Sometimes it dips to 400 CCU
Well visits do include players that come back
Yeah it does
Preety sure a visit is every hour
I dont think its per hour because 1000 new visits per hour would be really weak
I mean put it into perspective
8 minute playtime, only 1000 users per hour
they wouldnt even be putting up 100 concurrent
and the DAU like i said was 800-500 π , even considering 7.3 min playtime, that doesnt mean that per the average of 7.3 minuites there is a perchace per the PCR i mentioned π , you automatically assumed i meant that per the 7.3 minuite playtime, approx 180 players would play and spend robux, despite how i reiterated that it is a likely 180 players within the spanned duration of 1 day, meaning per each session of 7.3minuites average a likely of one of those players will join and spend the proceeding 47 rbx
also the description says
new visits every interval
so every little bar
is how many new ones
We don't know how long the interval is
Dang
I mean do you think they would just make up the bars in-between the hours?
Also 8 minute avarage playtime doesn't mean everyone plays for 8 minutes, players that quit immediately is counted into that statistic
I wish there was an indicator for each unique visitor
its great for abstracting data while keeping it accurate
what for
For gamesn
yeah but like
It would help a lot
what purpose
I mean it would tell you how fast a game is growing
If your game is a paid game it shows how many have bought it
Making a badge for it gets old
oh im pretty sure you can see this
in the analytics tab
it seperates new from returning
Using AI to do test cases for you program is goated technique
but honestly you're within 8%~ if you just assume theyre all new
Low-key w color scheme
I need to make the camera collide with a tool that has collisions off in my game, does anyone know how?
I usually use a.i to clean my code up
Ai doesn't write good code but it sure knows how to clean it up
i put a lot of thought into choosing the colors π
yea nobody replying bru
make a part, track the parts position to your camera pos, and then make them collide
ai is honestly so good at theoretical stuff π
good conceiding puppy
like back then programmers had to find textbooks
and relearn what they once new
but know i can just type in a prompt and relearn eveyrthing
yea but the cam will still clip through right?
the part just wont
then set up a collisiongroup for the part and camera
Yeah I mean that's what it was built upon
but really do watchout
it will hallucinate if something isn't as common
or somewhat new
and no, you can't just give chatgpt the docs
you can, but it's not actually going to learn it
you'll just point it towards the correct thing
you can gaslight chatgpt its insane
if u just say somethings wrong it will say you're right without thinking
even though its correct
i mean yeah its a word predictor
you cant expect it to be good at proof reading logic
code formatting at best
Ill try
or creating common code from scratch
yo, anyone know an estimate on the price for a sports game script
yeah i always try to tailor my prompt with wording that ive seen in literature/textbooks
really good strat for getting it to use the correct sources
just make the part's cancollide true again, then make it so that part has the default collision group off, and then for the part attached to the camera make it so that that part has its collision group off for default aswell, so the cam will only collide with the part
It's really interesting how prompting chat bots work though
eh
It's like language crafting
depends on the sports game
afl
Ill just put the camera into a different collision group than the tool
yo anyone got a good Tensor library for roblox?
get my full Roblox Studio course here (50% OFF):
https://linktr.ee/ByteBlox
photos of my pet pigeons: https://www.instagram.com/subnautica_man
discord: https://discord.gg/ay5gc9XcAh
wondered how to make a main menu in roblox studio? or how to create a shop which has working GUI? ive made lots of 2024 roblox scripting tutorials about all the d...
tryna make it really good not just bad scripts
No π
whats the script about though, if its basic framework like movement and passing it should be minimum 27k
handballing, tackling, stamina, diving, kicking
not tensor the math
yea 35k minimum so β£ 35,000- β£50,000
You can't really get a price like this since it really does vary from scripter to scripter
Reach out to several scripters to get several references
im j tryna find a range not a set price
that will only make the part collide bruh. not the camera
hence why i say β£ 35,000- β£50,000 .
about "handballing, tackling, stamina, diving, kicking"
This Fooly guys advice has been absolutely awful today
I'm talking simulated ball physics for rollback
the camera will still clip through and the part will just follow it
custom character controller, stamina management
then youre gonna need to make regions in which the camera cant be in, so if a camera is in a region CFrame, it gets pushed back from the vector direction its moving in
what is this even for bro
gun system
I made...a scoping system but when you look too far up, the camera starts glitching through the tool
u ever make a solid framework for combat then have to redesign it cuz it sucks after u made it

Youβll have to insert some constraints in your camera cframe update logic
And by constraint I mean algebraic constraints
I...actually just used the roblox default camera logic
And changed the subject
So was it solid or not
an illusion
It sucks π£οΈπ£οΈπ₯π₯
what does solid mean
In this context it practically means good
dude fooly i completed my combat system
then i realized i hardcoded some parts
lol
how do you realize that after??
wew

my whole system is kinda hardcoded rn
atleast it was good while it lasted
But it should be changable
now we make something cleaner
Hardcoding is for tough guys like us anyway πͺ
What kind of exploiting
then it aint up for public release bro
I mean you can get away with quite a bit
Im just trying to make my own fun game, it pmo that I have to think about exploiters
ammo is stored on the client
and youre pretty much good
not that bad
but also a really easy fix no?
Ok of all things I feel like thatβs the easiest to make server sided
yeah u just need to add sanity checks
phew
for the ammo n all that stuff u got on the client
like who gives a fuck if an exploiter flies around an obby for example
that means checking if the ammo is a reasonable amount?
so anyways
yea
just keep track on the server
Here is my camera logic
This is better than sanity checks @last hawk
yep u should leave the ammo calc on server
You can model everything analytically on the server as on the client for ammo
and just display it on client
its more responsive to have it on the client as well
You should do calcs for both server and client
actually Im not making it displayed lol
So itβs responsive too
for realism
hardest part about this is deciding what to do once you detect a discrepancy
hmm i havent played with ammo yet imma find out whats best
Well thereβs rlly only one optionβ¦
for combo i just use sanity checks rn
Sync to server..
melee tho
yeah
I'd probably try something like a remote function from client to server
then spawn another thread for the shooty part
and return the correct ammo
yea so just make it so that the camera cant be in the region of the gun. I recommend making the CFrame region this ```lua
local gun:Model = --[[your gun ]]
local region = workspace:GetPartBoundsInBox(gun:GetPivot(), gun:GetExtentsSize())
for _,v in pairs(region) do
if v.CollisionGroup==""--[[you desired collision group]] then
--now push out the camera from the camera vector its facing--
end
end
to match it on the client
hmm
im using packet module instead of remot functions so my server to client is fast ngl i'd use that to my advantage and design the combat around it
Remote functions r highkey scary
How come
wait let send a video how the scoping is
I mean you could also make another
client can fire them fake
remote event
I also frogor to mention that aiming isnt view modelled, but the arms actually move into the direction of your camera
but a remote function would make the use really nice
More overhead than remote events and unreliable
Idk if they have more overhead than remote events
all you need is the region which the camera cant enter, and a part which tracks the cameras location, then if the part is in the region where the camera cant enter, you can just push it back by overwriting the cframe of the camera, for testing purposes you can just reset the cframe of the camera back to the head's cframe.
or to the cframe before where it starts to glitch out
Well the server doesnt need to tell the client anything until the player stops firing the gun, rather than sending data every shot
local function Gun:Shoot()
self.Ammo -= 1
self:UpdateDisplay()
local serverAmmo = ShootFunction:InvokeServer(self.Id)
self.Ammo = serverAmmo
end
something like this
So I think itβd be better to just use remote events
hmm
Doesnt need*
yea but Ill finish the other mechanics first
the server NEEDS to validate the hits
yeah but I dont think thats too different
if you let the client decide the hits it's basically exploiters' paradise
from just firing a remote back
yea so when that happens just offset the camera by ```lua
Camera.CFrame*CFrame.new(0,2+(gun:GetExtentsSize.Y/2),0)
Well for semi automatic firing there will be no discrepancy unless youβre using unreliable, or if the client is cheating
yea but wouldnt I need to set the camera type to scriptable for that?
Might be firing to the server too much
Yeah no remote events would be fine
just that this is really easy to use
It's better to check if you can do everything instead of only the ammo
anybody up to playtest some volleyball
nah scriptable is useless, the camera can still move around and change cf
so you dont have to change it
You could even use attributes, there'd be a slight delay but doesn't matter since I wouldn't expect client desync to happen often
No what Iβm saying is that the server doesnβt need to communicate to the client at all in this case
Only the client sending the server signal to fire
It's more of a protective measure
how do you make these fancy ahh code messages btw
so the client doesnt get stuck with a softlocked magazine
Right now Im only working on semi auto firing
literally everything can go wrong on the client
There will be no desync tho
Unless client is exploiting
Remote event firing is reliable
if I cut my wifi and fire a remote it probably might
the whole thing is consisting of remote events
It wonβt

how come
ok lets say you also rate limit the gun
Because the client doesnβt stop sending data until it has received from the server affirming that the server receives it
No all 20 will go through unless you disconnect
It will just take a while
No remotes are dropped
That's not what I said
@verbal coral
because obv youre not going to allow inf fire rate
so I replace "gun" with the mesh name here?
oh yea it allows inf fire rate too at the moment
π
hopefully should be an easy fix
π
The server can send a signal to the client that it has rejected the firing command in that edge case, but for all normal cases the server can simply not reply
hardcoding or vibe coding
To save bandwidth
no with the gun model
yeah that would be a remote event way
a single number isn't really an issue
but yeah I guess you'd save a few bytes of bandwidth
No you would only use remote if you know youβre getting a response
But the server dictates that, the client doesnβt know
Oh I read it wrong mb
and what does getextentssize do?
pretty much returns the vector3 size of the model
Holy moly atp I just refer to remote events as remote
its like using Part.Size
so its either send a little more packets for your convenience or don't
but for model
Like remote function doesnβt exist in my mind
I mean theyre really cool for interface or reloading
Dont you need () at the end of a function?
yea you dont need it for gun:GetExtentsSize
π
who
why is bro still adressing me
the person who told you you don't need to use brackets
couldve swarn you blocked me bc you melted
AAAAAAAA
Itβs okay there is a replace all feature in Roblox studio
yeah make sure your settings are corrected
fellas
promise module
good or bad?
I accidentally switched every event name to the same one by accident
bruh
Im still learning lua
The module itself is fine but tbh itβs redundant, just manage your signal connections properly
Lf someone who wants to make easy games together (I am a modeler)
thanks for the feedback
its GetExtentsSize().Y/2 mb

si
I mean I dont really use promises
but if it manages your signal connections for you
why not
bro i forgot to add it π i literally added it here https://discordapp.com/channels/211228845771063296/1138164783967899678/1401531880616235008
you said I dont need it
I mean they would definitely help me if they were cross context
you do 
whenever I need to wait for a response from the player
does it work
but also time out
and also account to the fact they could leave
like 30 lines extra and several connections
it says the function doesnt even exist
model:GetExtentsSize() does exist, unless what you added to it just isnt a model
oh
OH
yea no I think I dont even need it if it shall only collide with the receiver lol
non β!strict user detected
show me if it works im curious
strict is so annoying because I have to keep adding that it's an instance AND it has these elements
The humble ::any
Wait if itβs an instance you can lowk cast it to any descendant of instance
whats the actual point of ```lua
@native
And it doesnβt error
pretty sure it compiles it to machine code
Hey, can someone help me make sense of why scripts work in solo play test in Studio but not in the game, team test or "Local server" test?
Itβs faster especially w math and sh, supposedly
it doesnt help with api use but if youre doing a lot of math it might help
Yes native code generation
doesnt it help with the processing speed of complex code
Why is solo play test in Studio different from all the other options? I don't understand
it helps by compiling to machine code
but the API doesnt get any faster
mostly just math operations
uhh
I canβt tell for certain but for me itβs usually race conditions, as in your code only works if certain things load faster than others
And solo play test usually facilitates that
I need to put it into an if loop to only set the offset incase the camera actually clips through right?
yes
how would I check that?
Math
Weird.. I've made a money, XP and leveling system and want the level to be on the players overhead GUI aswell but for some reason it works correctly in studio solo play test but not for any other way (in game, team test, local server test) and I've tried to make it work by adding more waits and stuff but it doesn't change anything sadly :(
Hey Guys i am new scripter i started yesterday
Nice man good luck
Thanks
#outdoorcooking #camping #outdoors
zirvede kiremit tavada menemen - Menemen in a tile pan at the summit
KANALA ABONE OL ΓCRETSΔ°Z β€β€ https://bit.ly/2YLToO1
πΔ°NSTAGRAM TAKΔ°P ET β€β€ https://www.instagram.com/karadenizli...
πVideolarΔ±n devamΔ± Δ°Γ§in videolarΔ± beΔenerek kanala abone olursanΔ±z sevinirim herkese iyi seyirle...
bro who keeps sending that π
But I prefer meth over math
crypto pics
actually yk what, theres a better, less demanding method
but its prob gonna be less accurate
Keep the region of the gun model thing.. but instead make a set of upper and lower bound cframes
for where the camera cant be
and constantly run on client with runservice
to push it out
what does that mean
otherwise youll need to loop on serverside aswell, with multipel scripts
the whole thing is already running on client
how code
is bro cooking the entire bible or wha
is this jump possibly im scared
local gun:Model = --your gun
local camera:Camera = workspace.CurrentCamera
local runservice=game:GetService("RunService")
runservice.RenderStepped:Connect(function()
local CFrameSet = {}
local gunCF=gun:GetPivot()
local gunSize=gun:GetExtentsSize()
CFrameSet[1] = gunCF.Position+(gunSize/2); --< frontofGun
CFrameSet[2] = gunCF.Position-(gunSize/2);--<< back
local CameraCFpos=camera.CFrame.Position
if ((CameraCFpos.X>=CFrameSet[1].X) and (CameraCFpos.X<=CFrameSet[2].X)) and ((CameraCFpos.Y>=CFrameSet[1].Y) and (CameraCFpos.Y<=CFrameSet[2].Y)) then
camera.CFrame=CFrame.new(camera.CFrame.Position+(gun:GetExtentsSize()/2)*1.3)
end
end)
--pr som'
it still clipping
show
debug it then, add print function at the start of the if mouse down thing and above the end) and add a print inside the if statement i sent you
if nothing prints in the if statement that i sent you then its likely juse to do with the indexes
so just switch this
Add an offset for the arms
Position the gun forward half of the arms size
The arms size would be the collective size of all the limbs
The upper arm the lower arm the hand
what yt channel or website should i use to learn scripting
Ask chatgpt with the search feature enabled
It's a better google than us
guys uhm
I've started scripting 3 months ago and I can autonomously make combat scripts w even raycasting am i doin good or am i lowkey slow
tryna get a lil insight from people who are much more advanced than me ty.
Any beginners that wanna make a parkour game
but what did u use
It was so long ago I don't remember
idk i think brawl dev tutorials
k
GPT (for theory only). Beginner videos are also good. Then there's devforums which I would just use ai for instead
does shell ejection look bad?
It's good but I think it shouldn't fly off so hard
Yeah that's good
cool
but why in the reloading anim does the players hand enter the camera
kk
started watching em around 3 months ago
what?
uh ik the animations are bad
*that's pretty cool
k
so i think u should
oh thanks
watch brawldev tutorials till you master the basics
π
Which is uh, quite hard @somber vault imma be real w ya
im doing it rn
should u do the advanced playlist too after u finish the basic one
u watch a tutorial then go code the stuff in ur own way
advanced stuff is much more fun
compared to basic
it's like, the basic playlist is fun but the advanced one really just is sum else
also @green valley @somber vault take breaks btw
how does one find talented/experienced scripters
i have so much money everyone in #scripter-hirable are so ass π
alr
u can burn urself out quickly if you try to delve too deep into the unknown
k
I mean, maybe the standards you're lookin for might be a lil high π€
though, you can probably see someone there
(might have to scroll a lot)
i dont care i have the money so PLEASEE π
why does it always go same velocity same direction
or am i tripping
it does
what's the stuff you're even tryin to make scripters do tho :
i scrolled the ENTIRE channel everyone of them is not what im expected, i have wayy too much standards
If I can ask
does it look bad
https://youtu.be/0ktS0snM32U I am not sure what could be the issue. The video file below is how the cutscene should play out but in the yt video, it does not work.
i think u shared the wrong file
The video file is how the cutscene should play out..
