I am trying to make a timestop ability

[hasty umbra]

    if not Timestop then return end

    overlaps.FilterType = Enum.RaycastFilterType.Exclude
    overlaps.FilterDescendantsInstances = {Timestop}
    
    
    local canfreeze = Timestop:FindFirstChild("CanFreeze")
    if not canfreeze or not canfreeze.Value then return end
    

    local partsInBox = game.Workspace:GetPartBoundsInBox(
        Timestop.CFrame,
        Timestop.Size,
        overlaps
    )

    for _, part in pairs(partsInBox) do
        local char = part:FindFirstAncestorOfClass("Model")
        if not char then continue end
        if char.Name ~= Timestop.Player.Value.Name then continue end
        if not processedChars[char] then continue end
        
        
    
                local streak = char:FindFirstChild("Streak")
                if streak or streak.Value < 100 then continue end
            
            local hum = char:FindFirstChildOfClass("Humanoid")
            local hrp = char:FindFirstChild("HumanoidRootPart")


            if hum and hrp then
                processedChars[char] = true -- prevent repeat processing
                

                for _, bodyPart in pairs(char:GetChildren()) do
                    if bodyPart:IsA("BasePart") then
                        bodyPart.Anchored = true
                    end
                end
            end
end
end)```

the player/entity are not being anchored and even if it did, it not very accurate/consistency. what am I'm doing wrong here?

[hard belfry]

this is overcomplicated, i personally wouldnt do this and instead would use GetPartsInPart() along with a for loop to anchor all the parts then countdown after to disable the effects and can simplify this

local function timestop(timeAmount)
  local parts = GetPartsInPart(--[[clone a part or make a sphere]]);
  local anchored = {};
  for i, v in parts do
    if v.Anchored == true then table.insert(v, anchored); continue end
    -- anchor parameters
  end
  task.wait(timeAmount);
  for i, v in parts do
    if anchored[v] then continue end;
    --unanchor parameters  
  end
end```

if this doesnt work its because im too tired to write working code rn and you canj prob solve it yourself from this

[tardy tide]

Making a new part for every single time stop would be performance unfriendly especially if it's the main ability and that's why people use boundinbox (they practically do the same thing without having to make a new part)

personally if I had to do a time stop ability I'd just do it clientsided where it would use os.clock so you can't modify the timing w exploits and have the timestop be stored in an array to just be stopped (as in you can cancel the time stop) later

though making a whole object for time stops may be inefficient if you have time constraints

[tidal cypress]

game.Workspace you probably should have stopped there, but you didn't, so your code stopped you for you at if not processedChars[char] then continue end instead.

this probably gonna stop you too local streak = char:FindFirstChild("Streak") if streak or streak.Value < 100 then continue end

[tidal cypress]

"just do it clientsided" and "so you can't modify .. w exploits" do not belong in the same sentence.

[tardy tide]

it really depends

[tidal cypress]

if that's how you would do it hey dont let anyone stop you, but this is generally bad advice

performance of creating one part i.e debris, or a hitbox for that matter, is negligible

if you were spawning 100's of parts per frame then ye maybe performance would be an issue

but one?

bruh

in the course of a single vfx many parts are created and destroyed in very rapid succession, that's what vfx is, and performance never been a problem there, why would it be a problem here?

[tardy tide]

You can assume multiple people would be using it as things like abilities are usually against other players

and it's still not worth using partsinpart as you have to also know timestop abilities will also anchor anyone who enters the area

[tidal cypress]

bro can't get his timestop working in the first place and .. performance is your priority? why

legit question btw i see it a lot

[tardy tide]

y not

[hasty umbra]

I decides to use Magnitude

[tidal cypress]

sure for vfx it should probably be on the client depending on what it is since you may not need physics sync and one small remote is cheaper than networking a bunch of part creation, but that's also saying that moving those purely-visual effects from server to client is not trivial in the first place. the actual timestopping would be server regardless to start with ;o

[tardy tide]

also it does fire every frame btw

[blissful wigeon]

** You are now Level 6! **

[tardy tide]

since you do need to watch for anyone going into the timestop radius

[tidal cypress]

i can understand your reasoning but perhaps it is supposed to persist for some time. for one frame you'd be right, we dont know if that's the case here

still, trivial to move it to a one-shot when it is working

[tidal cypress]

yes, this 👍

[tidal cypress]

so what i'm really wondering here is why there is such an obsession with performance when those changes are usually pretty trivial and insignificant compared to "it doesnt work". why oh why is performance and optimization a priority over simply working as a first step?

it confuses me greatly

perhaps it is some unwitting device or strategy to eliminate competition so people get into performance (a rather difficult subject specially for beginners), get overwhelmed, and quit ?

[tardy tide]

A lot of people are addicted to having silver performance in their games and I am no exception

[tidal cypress]

a lil offtopic but this opportunity presented itself nicely 🤷

[tardy tide]

It really depends on if you want the bare minimum or the most optimum solution

[tidal cypress]

minmaxing is a final step, not a first step, wouldn't it be?

for a beginner who hasn't ever pulled off a particular effect before... isn't bare minimum the more desirable goal? why does it have to be perfectly optimal on a first attempt?

[tardy tide]

I willingly script modules in strict because I hate myself (not actually but I just do --!strict for the fun of it)

[tardy tide]

human nature to be abnormal

[tidal cypress]

it's like chucking a toddler into an olympic swimming pool and asking them to compete for the gold medal, and i can't tell if the expectation for said toddler to actually win is satire or sarcasm or whatever it might be, unwitting or not -- it just makes no sense to me

[tardy tide]

win or lose, you wouldn't have chucked the kid into the pool if you knew he would lose immediately

[tidal cypress]

exactly 💯 so why are we chucking beginners into the gold medal standard of optimal performance when they don't know how to even keep their head above water so to speak?

just something i realized and it's hard to put it into words. maybe its unintended, or just wasn't recognized thats what was happening 🤷

maybe its just easy to talk about performance and sound like you know what you're talking about, when really its just all flash no photo

[tardy tide]

To make others suffer but also because I just don't like when people set their own self standards higher than what they can attain but I still would want to help

idk man I learned luau in a REALLY short time but still managed to make some decent stuff

adhd powers I suppose

[tidal cypress]

hyperfocus is a helluva drug

for this timestop tho i was actually wondering about how you would approach a tweened timestop, where stuff slows down while the world around it continues

if something happens to collide in the middle of that 'tween' and changes direction - i can't quite figure out how to translate that into the speedup

i'm sure there's a way but i aint even bothered about performance til working theory there ;p

[tardy tide]

just make a whole physics system atp

[tidal cypress]

as they say;

1. make it work.
2. make it right (do it properly)
3. make it fast (optimize).
   in that order.

[tardy tide]

roblox physics are just hard to control

[tidal cypress]

nah i think it can be done with roblox physics, it is more optimal and performant

[tardy tide]

well then for debris and unanchored parts you can just add bodyvelocity or some type of force to slow their descent/ascent

for characters just have the animation slow down until time gets stopped where their animation is frozen

[tidal cypress]

antigrav is relatively simple, the collisions between timestopped stuff and external stuff is relatively simple (make the timestopping stuff tween their mass as well so if you jump into it, it'll move but much slighter than it normally would)

and just tween assemblyvelocities down

but returning to original speed, not sure on that one..

particularly after a collision where the velocity reverses direction entirely

[tardy tide]

maybe just cache their speed into an array? who knows

[tidal cypress]

as small bumps are unnoticeable

[tidal cypress]

ye that's the starting point

[hard belfry]

if theres not a billion timestops at once its fine

id use it here because its simpler

[tardy tide]

like it identifies the part and just stores it like

{[Instance] = {AssemblyLinearVelocity, AssemblyAngularVelocity}}

[tidal cypress]

i think the difference in speed lost to intentional slowdown vs speed lost per frame needs to be recorded every frame, and the velocity direction on the final frame, then you average the total intentional slowdown as the target velocity (so you maintain 'natural' physics-based slowdown i.e friction) apply this to its current trajectory as the mass and gravity increases back up, or something.

you'd want to get that math juuust right so it doesn't slam into the floor due to gravity, or maybe you intentionally calculate what the acceleration due to gravity should've been so you can then cancel that out.

gets a bit complicated trying to do this in a generic way

now this, this system would be incredibly expensive at around 100 parts so you'd want to optimize this almost from the outset after first working proof of concept

...but not before

[tardy tide]

if only roblox actually had those physics calculations as a function library

[tidal cypress]

hah yeah game.timescale=0.1 when

• at the very least -

if you wanted to localize it... yeaa more physics api would help... but i still think it can be done

localize as in timestop a part of the world but not all of it

[tardy tide]

That's usually how people do time stops

[tidal cypress]

rather, how people would expect it to function

[hard belfry]

i mean its an annoying system if your not doing a full stop but i think a stop would be fine yeah

a slowdown would be horrible

[tidal cypress]

i might make something for this later on, physics time slowing and time stop is pretty cool effect

[tidal cypress]

slowdown for like a second towards a full stop

or even simply bullet time

deffo not horrible

[tardy tide]

if it's serversided then yeah

[hard belfry]

?

you would always do it on the server here

[tardy tide]

not if it's things like slowdown effects

[hard belfry]

yea ig

[hard belfry]

you would do slowdowns on the server

so everyone can see it

if its a single player game

you do it on the server

incase of a low end device

[tardy tide]

nah id :FireAllClients and make a clientsided stopped time area

[hard belfry]

you do realize with exploits you can just block fireallclient calls and you can change anything client sided

it would be so easy to bypass

[tardy tide]

you can block invokes but not FireAllClients

[hard belfry]

yes you can

if your recieving anything on the client its blockable

[tardy tide]

exploits can only block remote events called from the client itself (which highkey is a bad idea if your remote event isn't an input logger

[hard belfry]

if something is sent to the client for code to run on the client you can stop the code from reaching the client

[tardy tide]

if you could find the source of the connection

[hard belfry]

its easy to

getconnections() is literally one of the most common exploit functions

[tardy tide]

and even if you do destroy it you'd practically be damning yourself as that also holds input responses so

what's the point of exploits if you can't even enjoy the game

[hard belfry]

thats not hte argument here

[hard belfry]

you do realize you can just

send input

the client shouldnt be used for anything important because it is all changable from exploits

exploits run on the client, if something is ran from the client you can change that from the client

[tardy tide]

why would one make a whole input logger js to send inputs outside of a connection made by the game

[hard belfry]

its 10 lines dont say "a whole input logger" 💔

you could also juist

disable the freeze effect

fromt he client

instead of doing all taht

[tardy tide]

the thing is most people who use exploits are just googoo gaga babies

and the people who do use it with experience are just prob testing their own anticheat

[hard belfry]

yeah and the 1% who are actually developers are good at bypassing anticheats

people make a lot of money selling scripts

for popular games with strong anticheats

[tardy tide]

byfron is actually good now so the people who do exploit have the likelihood of just never getting their accounts back

also those team of anti-exploiters that make scripts to log peoples users and send them to roblox

which I forgot the name of

[hard belfry]

💔

your just dodging your point of using the client for important things

you should never use the client for important things

its easy to bypass

[tidal cypress]

cheat baiting is hilarious to watch

[tardy tide]

depends on what you describe as important but yeah

[hard belfry]

things that gameplay and enjoyability depend on

core systems are always in the server no question

things that gameplay and enjoyability depend on are things like (for example) a time stop system right

and if you do that on the client right

you can just disable the entire system right

because your setting it on the client not the server

[tidal cypress]

or just anchor the player's character on the server which also happens to strip them of network ownership at the same time then they cant do nothin

[hard belfry]

yes

[tardy tide]

and also

[hard belfry]

theres no point in doing fireallclients for that

its vulnerable

[tidal cypress]

but also roblox by default trusts the client for their character under normal circumstances so like... wat you gonna do, make your own prediction model?

[tardy tide]

exploiters can just unanchor their character and move freely again

this argument has no point of being an argument

but yeah

[tidal cypress]

na not if the server anchors it lol

[hard belfry]

😭

no they cant

they can on their client and it looks fine

but they are still anchored on the server

its not replicated

[tidal cypress]

ye idk how this turned into exploiting

[tardy tide]

someone just trying to get help on their timestop logic and then it becomes an exploit debate

😭

[tidal cypress]

coz im doing exactly this anyway, i'll probably need it for advanced timestop

🙂

full server authoritative + predicted movement, just like how source engine does it.

idk why roblox doesn't, its dum for not doing it

maybe too hard for kids to develop for 🤷

as if this is any easier..

[tardy tide]

I know a lot of people that exploit use dex explorer and dex explorer can't see tags as of forever since its old as hell but

you can probably just use tags as a sort of detection

[tardy tide]

how most multiplayer games do it

which is kind of annoying

but yeah

[tidal cypress]

lol 'most'

[hard belfry]

they are switching to it soon

iirc

[tidal cypress]

na when you hear an announcement for auroraservice then you'll know it's "soon".

[hard belfry]

there was a dev forum post about it a while ago and the exploiter community started shitting themselves saying "IS THIS THE END OF EXPLOITING"

[tardy tide]

since they don't want their players to look like they're teleporting and accidentally triggering an anticheat

or something

[hard belfry]

i think september or october is the switch

i forget

[tidal cypress]

auroraservice in sept/oct? bs. cite sauce

[hard belfry]

lemme find it

[tardy tide]

aurorascripts when???

[tidal cypress]

silksong will be out long before auroraservice drops

[tardy tide]

apparently with flags you used to be able to use aurorascripts (using internal) but I think they patched it

I saw some guy use aurorascripts without it being public yet

[hard belfry]

@tidal cypress found it https://devforum.roblox.com/t/creator-roadmap-2025-spring-update/3624895

server authority

or smth

[tardy tide]

if it does happen it will ruin a lot of games for the McDonalds wifi community

[tidal cypress]

Server Authority (Beta): The Server Authority beta is coming this Summer, to designate the server as the single source of truth for game actions, logic, and data. This is a huge effort, and we want to make sure we get this right, but our goal is for a full release by the end of the year.
"our goal"

[hard belfry]

oh wait

"our goal"

coming next year

or year after that

[tidal cypress]

Range limit for lights: We’re increasing the range limit for lights to 120 studs.
i like this one

[hard belfry]

mb

[hard belfry]

it was lower?

i thought it was like 1000 studs

i dont use lights that much tho

[tidal cypress]

no i think its like 70

server authority gonna get delay, calling it now

[hard belfry]

yeah probably

the only strong thing on the roblox engine is the anti-tamper

probably one of the best usermode ones out right now

its not kernel though ✌️

[tidal cypress]

even if it were it will never be good enough

[hard belfry]

that means its forever undetected to use external cheats on roblox, but its not really a game where external cheats would be preminent so its probably fine

[hard belfry]

the team working on it is suprisingly skilled but yes i agree

they stopped all internal cheats for the time being

its been about a month with no new public ones i believe

then again most of the public/paid executor devs are skids who cant code for shit ✌️

[tidal cypress]

either way im still making my own, and if roblox releases a good one, should be trivial to refactor my code to use the new stuff since the design patterns for prediction are all the same

[hard belfry]

lowk the light thing got me thinking im probably gonna try to make a raytracing system in luau