SilicatYT
@silicatyt.bsky.social
Professional Minecraft Datapack Maker | BotW and Mario Maker 2 Glitch Hunter | Glitch Showcaser on YouTube | Maker ID: GF1-BD7-YWG | He/Him
All the calculations are so much easier with mods 😭
Calculating the corner pos takes literally 2 lines (or 6 if you count the for-loop).
And whenever the orientation changes, all derived data (rotMat, cornerPos etc) automatically updates. No desyncs! Just with minor overhead.
Calculating the corner pos takes literally 2 lines (or 6 if you count the for-loop).
And whenever the orientation changes, all derived data (rotMat, cornerPos etc) automatically updates. No desyncs! Just with minor overhead.
November 27, 2025 at 10:32 PM
All the calculations are so much easier with mods 😭
Calculating the corner pos takes literally 2 lines (or 6 if you count the for-loop).
And whenever the orientation changes, all derived data (rotMat, cornerPos etc) automatically updates. No desyncs! Just with minor overhead.
Calculating the corner pos takes literally 2 lines (or 6 if you count the for-loop).
And whenever the orientation changes, all derived data (rotMat, cornerPos etc) automatically updates. No desyncs! Just with minor overhead.
[Physics Engine Update | Fabric Mod]
Linear and angular velocity calculations are done.
It's unbelievable how much easier everything is with Java as compared to Datapacks. No fixed point scaling, no hardcoding every cross product for optimization, multiple ops per line... ❤️
Linear and angular velocity calculations are done.
It's unbelievable how much easier everything is with Java as compared to Datapacks. No fixed point scaling, no hardcoding every cross product for optimization, multiple ops per line... ❤️
November 26, 2025 at 10:24 PM
[Physics Engine Update | Fabric Mod]
Linear and angular velocity calculations are done.
It's unbelievable how much easier everything is with Java as compared to Datapacks. No fixed point scaling, no hardcoding every cross product for optimization, multiple ops per line... ❤️
Linear and angular velocity calculations are done.
It's unbelievable how much easier everything is with Java as compared to Datapacks. No fixed point scaling, no hardcoding every cross product for optimization, multiple ops per line... ❤️
[Physics Engine Status Update]
I just started developing a server-side mod that I can use as a reference (for figuring out the scaling and other things more easily).
Here's the entity I set up in the past 2 days.
More detailed information:
bsky.app/profile/sili...
(1/6)
I just started developing a server-side mod that I can use as a reference (for figuring out the scaling and other things more easily).
Here's the entity I set up in the past 2 days.
More detailed information:
bsky.app/profile/sili...
(1/6)
November 25, 2025 at 7:43 PM
[Physics Engine Status Update]
I just started developing a server-side mod that I can use as a reference (for figuring out the scaling and other things more easily).
Here's the entity I set up in the past 2 days.
More detailed information:
bsky.app/profile/sili...
(1/6)
I just started developing a server-side mod that I can use as a reference (for figuring out the scaling and other things more easily).
Here's the entity I set up in the past 2 days.
More detailed information:
bsky.app/profile/sili...
(1/6)
Minecraft snapshot 25w41a won't let us grab the result of a stopwatch query, so I wrote a small datapack that uses binary search to get it.
(It'll definitely be a regular feature in the next snapshot)
(It'll definitely be a regular feature in the next snapshot)
October 10, 2025 at 7:37 PM
Minecraft snapshot 25w41a won't let us grab the result of a stopwatch query, so I wrote a small datapack that uses binary search to get it.
(It'll definitely be a regular feature in the next snapshot)
(It'll definitely be a regular feature in the next snapshot)
Minecraft spears in a nutshell (I love it)
October 9, 2025 at 4:45 PM
Minecraft spears in a nutshell (I love it)
It's a tiny bit more stable now. The 2 core problems are:
- Due to calculation inaccuracies, impulses overcorrect, reinforcing the next problem:
- Impulses make it hard for all 4 corners to touch the ground simultaneously. If not all 4 corners do, it can wiggle back and forth
- Due to calculation inaccuracies, impulses overcorrect, reinforcing the next problem:
- Impulses make it hard for all 4 corners to touch the ground simultaneously. If not all 4 corners do, it can wiggle back and forth
August 30, 2025 at 10:06 AM
It's a tiny bit more stable now. The 2 core problems are:
- Due to calculation inaccuracies, impulses overcorrect, reinforcing the next problem:
- Impulses make it hard for all 4 corners to touch the ground simultaneously. If not all 4 corners do, it can wiggle back and forth
- Due to calculation inaccuracies, impulses overcorrect, reinforcing the next problem:
- Impulses make it hard for all 4 corners to touch the ground simultaneously. If not all 4 corners do, it can wiggle back and forth
[Disclaimer: A lot of bugs were temporarily resolved with ugly code that only works for testing]
Found a TON of bugs and temporarily fixed them, but it's still very jittery and a bit buggy. Not sure what's causing that.
Found a TON of bugs and temporarily fixed them, but it's still very jittery and a bit buggy. Not sure what's causing that.
August 29, 2025 at 3:25 PM
[Disclaimer: A lot of bugs were temporarily resolved with ugly code that only works for testing]
Found a TON of bugs and temporarily fixed them, but it's still very jittery and a bit buggy. Not sure what's causing that.
Found a TON of bugs and temporarily fixed them, but it's still very jittery and a bit buggy. Not sure what's causing that.
Made a temporary ugly fix for a bug in contact generation, and it looks a lot more stable now. I can be pretty sure that any bugs in this clip are from velocity resolution now.
August 28, 2025 at 9:41 PM
Made a temporary ugly fix for a bug in contact generation, and it looks a lot more stable now. I can be pretty sure that any bugs in this clip are from velocity resolution now.
I don't know why, but I think there might be a calculation error somewhere.
Call it a hunch 🤔
Call it a hunch 🤔
August 28, 2025 at 7:40 PM
I don't know why, but I think there might be a calculation error somewhere.
Call it a hunch 🤔
Call it a hunch 🤔
Made a rudimentary penetration resolver.
(Please be a sign error in velocity resolution)
(Please be a sign error in velocity resolution)
August 28, 2025 at 7:06 PM
Made a rudimentary penetration resolver.
(Please be a sign error in velocity resolution)
(Please be a sign error in velocity resolution)
Finished for world_axis_y and fixed some bugs. It can now resolve multiple contacts per tick, too!
Will work on penetration resolution next, so the object doesn't clip into the ground xD.
Will work on penetration resolution next, so the object doesn't clip into the ground xD.
August 28, 2025 at 6:20 PM
Finished for world_axis_y and fixed some bugs. It can now resolve multiple contacts per tick, too!
Will work on penetration resolution next, so the object doesn't clip into the ground xD.
Will work on penetration resolution next, so the object doesn't clip into the ground xD.
Velocity resolution for world_axis_x is finished!
(It currently doesn't update other contacts' values based on the resolved contact. It also doesn't resolve penetrations yet.)
(It currently doesn't update other contacts' values based on the resolved contact. It also doesn't resolve penetrations yet.)
August 26, 2025 at 1:36 PM
Velocity resolution for world_axis_x is finished!
(It currently doesn't update other contacts' values based on the resolved contact. It also doesn't resolve penetrations yet.)
(It currently doesn't update other contacts' values based on the resolved contact. It also doesn't resolve penetrations yet.)
Fixed a sign error that caused the object to accelerate when colliding with the left block
August 26, 2025 at 1:09 PM
Fixed a sign error that caused the object to accelerate when colliding with the left block
Started working on the resolver.
Linear velocity change is calculated correctly for this specific axis (VERY early test).
Angular part and penetration resolution are still missing, and ofc making it work for all other axes. This is just a test to see if the formulas are correct.
Linear velocity change is calculated correctly for this specific axis (VERY early test).
Angular part and penetration resolution are still missing, and ofc making it work for all other axes. This is just a test to see if the formulas are correct.
August 26, 2025 at 11:51 AM
Started working on the resolver.
Linear velocity change is calculated correctly for this specific axis (VERY early test).
Angular part and penetration resolution are still missing, and ofc making it work for all other axes. This is just a test to see if the formulas are correct.
Linear velocity change is calculated correctly for this specific axis (VERY early test).
Angular part and penetration resolution are still missing, and ofc making it work for all other axes. This is just a test to see if the formulas are correct.
Optimized the ray intersection logic so it can stop after finding a single valid intersection per object. This change makes this part of the code ~3-4% faster, but it's nigh unmeasurable due to it already being very fast.
Also, exam season is over and I've prepared some resolution pseudocode 👀
Also, exam season is over and I've prepared some resolution pseudocode 👀
August 6, 2025 at 7:10 PM
Optimized the ray intersection logic so it can stop after finding a single valid intersection per object. This change makes this part of the code ~3-4% faster, but it's nigh unmeasurable due to it already being very fast.
Also, exam season is over and I've prepared some resolution pseudocode 👀
Also, exam season is over and I've prepared some resolution pseudocode 👀
Currently working on:
- Get the TPS (Without worldborder shenanigans)
- Detect pausing/unpausing (And how long)
- Get the TPS (Without worldborder shenanigans)
- Detect pausing/unpausing (And how long)
April 19, 2025 at 3:14 PM
Currently working on:
- Get the TPS (Without worldborder shenanigans)
- Detect pausing/unpausing (And how long)
- Get the TPS (Without worldborder shenanigans)
- Detect pausing/unpausing (And how long)
⏰ TimeLib - Major update (v2.0.0)
- Support for lag, different tick rates etc
- Added a daytime storage with leading 0s
- Added event tags for "date changed" and "time changed"
... and MUCH more!
(Full changelog on Modrinth)
Download link in the replies 👇
- Support for lag, different tick rates etc
- Added a daytime storage with leading 0s
- Added event tags for "date changed" and "time changed"
... and MUCH more!
(Full changelog on Modrinth)
Download link in the replies 👇
April 18, 2025 at 8:34 PM
⏰ TimeLib - Major update (v2.0.0)
- Support for lag, different tick rates etc
- Added a daytime storage with leading 0s
- Added event tags for "date changed" and "time changed"
... and MUCH more!
(Full changelog on Modrinth)
Download link in the replies 👇
- Support for lag, different tick rates etc
- Added a daytime storage with leading 0s
- Added event tags for "date changed" and "time changed"
... and MUCH more!
(Full changelog on Modrinth)
Download link in the replies 👇
I've spent the last 8 hours without breaks on reworking TimeLib from scratch. I'm tired, but I'm also very happy. It's a lot better than before, in my opinion.
Here's a sneakpeek:
Here's a sneakpeek:
April 18, 2025 at 4:12 PM
I've spent the last 8 hours without breaks on reworking TimeLib from scratch. I'm tired, but I'm also very happy. It's a lot better than before, in my opinion.
Here's a sneakpeek:
Here's a sneakpeek:
[Minecraft Datapack Update]
In an upcoming update for my TimeLib library, time will stay synced even at incredibly low or high tickrates.
You won't be able to tell from the timer that the game isn't running at 20 TPS!
In an upcoming update for my TimeLib library, time will stay synced even at incredibly low or high tickrates.
You won't be able to tell from the timer that the game isn't running at 20 TPS!
April 17, 2025 at 8:30 PM
[Minecraft Datapack Update]
In an upcoming update for my TimeLib library, time will stay synced even at incredibly low or high tickrates.
You won't be able to tell from the timer that the game isn't running at 20 TPS!
In an upcoming update for my TimeLib library, time will stay synced even at incredibly low or high tickrates.
You won't be able to tell from the timer that the game isn't running at 20 TPS!
Hitting the object now works with the new system, though the calculations are still a bit janky (Will need to replace some of them later on).
April 15, 2025 at 7:54 PM
Hitting the object now works with the new system, though the calculations are still a bit janky (Will need to replace some of them later on).