Implement VL scaling for Simulation #90

GigaZelensky · 2025-05-17T20:20:38Z

(Title)

Makes it so you can set grim to add more levels for simulation per violation depending on the offset in the prediction/actual comparison.

Fixes the punishment decay issue that was present in WHY-HAVE-YOU-SUMMONED-ME's PR (GrimAnticheat#1961).
Nothing else is changed.

Also includes translations for the configs which I think are pretty solid but anyone can feel free to change them later.

(I just PR'd to upstream Grim but this makes it easier for you to merge since it already has everything lightning has)

- Add basic runtime reloading support. - Will only actually reload if current database is different/has different setting than the one requested in config. - Added [log] statements matching [alert] statements

…ierce) - Reduce unnecessary reach vectors used for calculations in versions

Notes: - Mark actions builds automatically as alpha on Modrinth - Shorten Fabric version string to be same as Bukkit on Modrinth

- Update gradle buildscript to not include branch-name if branch is called "lightning" - Update GitHub Actions to automatically include changelog and to rename the jar LightningGrim-${platform}-${version}.jar - Mark Paintings as entities that cannot be hit to prevent EntityPierce (formally HitboxEntity) falses.

- 2.2x reduction in queued task overhead - reduction in size of queued single block changes - up to 8x reduction in size of queued multi block changes - fixed blocks disappearing in B73 caused by incorrect shifting in new block udpate handler

* "Wall Hit" -> "WallHit" * "Entity Pierce" -> "EntityPierce" * Fix modeling experience orb as unhittable entity * Fix sending hitbox with /grim hitboxdebug in the hitboxdebug handler * Update upstream to 35811fe

- Add basic runtime reloading support. - Will only actually reload if current database is different/has different setting than the one requested in config. - Added [log] statements matching [alert] statements

…ierce) - Reduce unnecessary reach vectors used for calculations in versions

Notes: - Mark actions builds automatically as alpha on Modrinth - Shorten Fabric version string to be same as Bukkit on Modrinth

- Update gradle buildscript to not include branch-name if branch is called "lightning" - Update GitHub Actions to automatically include changelog and to rename the jar LightningGrim-${platform}-${version}.jar - Mark Paintings as entities that cannot be hit to prevent EntityPierce (formally HitboxEntity) falses.

- 2.2x reduction in queued task overhead - reduction in size of queued single block changes - up to 8x reduction in size of queued multi block changes - fixed blocks disappearing in B73 caused by incorrect shifting in new block udpate handler

* "Wall Hit" -> "WallHit" * "Entity Pierce" -> "EntityPierce" * Fix modeling experience orb as unhittable entity * Fix sending hitbox with /grim hitboxdebug in the hitboxdebug handler * Update upstream to 35811fe

This reverts commit 4d6c9c6.

…s-for-config-options Codex/add translated comments for config options

GigaZelensky · 2025-05-19T23:09:28Z

So I added a millisecond delay option which was requested by a user (heckergato), the objective is to have a smoother command execution when multiple VLs are fired at once due to VL scaling. I also added an option which I will explain here:

TL;DR: The punishment system tracks how many times a command is executed (executeCount) for a given violation level (VL). The problem is: this counter never resets, not even when VL decays.

Example:

You set a command to run at 20:20.
A player cheats, hits VL 20 → command runs.
VL decays (player stops cheating for a bit).
Player cheats again, hits VL 20 → command does NOT run again.

Why? Because the Grim thinks, “Ah, I already ran this command once for VL 20, no need to do it again.” Even though the VL dropped and came back up, which should be treated as a new violation threshold being reached.

Why This Sucks:

No repeat punishments or alerts, even for repeated behaviour.
Mods might miss the first alert, and the system gives them zero visibility after that.
Commands at specific VLs become useless if they only ever fire once.
Encourages a cheater to just chill for 5 minutes, then repeat the same thing forever without triggering anything again.

What Should Happen:

executeCount should reset when VL decays past the previous threshold.

This is exactly what I have also implemented as a config option.

Axionize and others added 27 commits May 7, 2025 17:02

Runtime startup check of PE version

ac576bb

Add /grim history command to see historical alerts

245d239

- Add basic runtime reloading support. - Will only actually reload if current database is different/has different setting than the one requested in config. - Added [log] statements matching [alert] statements

Add Inventory checks

bad73fa

- Re-add HitboxBlock and HitboxEntity (renamed to WallHit and EntityP…

6aed9df

…ierce) - Reduce unnecessary reach vectors used for calculations in versions

Add support for Hitbox Debugging

a8bf6ac

Cull dead entities from canHit()

32070a3

Update buildscript

da952f8

Notes: - Mark actions builds automatically as alpha on Modrinth - Shorten Fabric version string to be same as Bukkit on Modrinth

Heavy memory use optimization

d4a77a8

- 2.2x reduction in queued task overhead - reduction in size of queued single block changes - up to 8x reduction in size of queued multi block changes - fixed blocks disappearing in B73 caused by incorrect shifting in new block udpate handler

* Remove space in names of new checks:

706e3ca

* "Wall Hit" -> "WallHit" * "Entity Pierce" -> "EntityPierce" * Fix modeling experience orb as unhittable entity * Fix sending hitbox with /grim hitboxdebug in the hitboxdebug handler * Update upstream to 35811fe

Optimize Bukkit PistonEvent() usage

9d8aaa4

Runtime startup check of PE version

cf3c78b

Add /grim history command to see historical alerts

f54d265

- Add basic runtime reloading support. - Will only actually reload if current database is different/has different setting than the one requested in config. - Added [log] statements matching [alert] statements

Add Inventory checks

8e9ad26

- Re-add HitboxBlock and HitboxEntity (renamed to WallHit and EntityP…

d434c4b

…ierce) - Reduce unnecessary reach vectors used for calculations in versions

Add support for Hitbox Debugging

9b97ebe

Cull dead entities from canHit()

1d5ed18

Update buildscript

a444ffe

Notes: - Mark actions builds automatically as alpha on Modrinth - Shorten Fabric version string to be same as Bukkit on Modrinth

Heavy memory use optimization

6374878

- 2.2x reduction in queued task overhead - reduction in size of queued single block changes - up to 8x reduction in size of queued multi block changes - fixed blocks disappearing in B73 caused by incorrect shifting in new block udpate handler

* Remove space in names of new checks:

2f08694

* "Wall Hit" -> "WallHit" * "Entity Pierce" -> "EntityPierce" * Fix modeling experience orb as unhittable entity * Fix sending hitbox with /grim hitboxdebug in the hitboxdebug handler * Update upstream to 35811fe

Optimize Bukkit PistonEvent() usage

eba4105

Temp stash

0880b5e

yay

a667d8d

Merge branch 'Axionize:lightning' into lightninggg

9ba9156

Track scaled VL separately

6cc1878

translate

ec893c5

GigaZelensky marked this pull request as ready for review May 17, 2025 20:35

GigaZelensky and others added 2 commits May 17, 2025 21:44

Don't scale by default

3f2c776

Update PunishmentManager.java

38a71ab

GigaZelensky and others added 7 commits May 19, 2025 11:22

Fix duplicate comment in English config

4d6c9c6

undo vl scale delay

2f51a64

This reverts commit 4d6c9c6.

Add VL scale delay support

69a0d7f

Add translated config comments

0a3d73e

aesthetic thing

9f5bb99

fix translation

628ced4

Merge pull request #25 from GigaZelensky/codex/add-translated-comment…

2a788e2

…s-for-config-options Codex/add translated comments for config options

Axionize force-pushed the lightning branch 2 times, most recently from 2582e6c to 4ff4a27 Compare May 26, 2025 02:09

Axionize force-pushed the lightning branch from 4ff4a27 to baa3d91 Compare May 27, 2025 05:11

Axionize force-pushed the lightning branch from baa3d91 to d83dd54 Compare June 10, 2025 18:55

Axionize force-pushed the lightning branch 5 times, most recently from 0763048 to 326e200 Compare July 7, 2025 03:56

Axionize force-pushed the lightning branch from 326e200 to 293bede Compare July 13, 2025 10:35

Axionize force-pushed the lightning branch from 71bd30a to 0315338 Compare August 1, 2025 23:56

Axionize force-pushed the lightning branch from 0315338 to 98f9417 Compare October 1, 2025 06:54

Axionize force-pushed the lightning branch 2 times, most recently from 037abab to 71d93c9 Compare November 8, 2025 22:13

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Implement VL scaling for Simulation #90

Implement VL scaling for Simulation #90

Uh oh!

GigaZelensky commented May 17, 2025 •

edited

Loading

Uh oh!

GigaZelensky commented May 19, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Implement VL scaling for Simulation #90

Are you sure you want to change the base?

Implement VL scaling for Simulation #90

Uh oh!

Conversation

GigaZelensky commented May 17, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

GigaZelensky commented May 19, 2025

So I added a millisecond delay option which was requested by a user (heckergato), the objective is to have a smoother command execution when multiple VLs are fired at once due to VL scaling. I also added an option which I will explain here:

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

GigaZelensky commented May 17, 2025 •

edited

Loading