[DiscordArchive] Hi, im who opened the PR and thanks for the discussion on git, what would be the best or accurate wa
[DiscordArchive] Hi, im who opened the PR and thanks for the discussion on git, what would be the best or accurate wa
Original source
No, the upside down pyramid would be more expensive in terms of performance than the radius.
rektbyfaith
09-23-2025, 06:04 PM #11Archived author: Pursche • Posted: 2025-09-23T18:04:27.492000+00:00
Original source
No, the upside down pyramid would be more expensive in terms of performance than the radius.
Original source
Hang on, time to bring out Paint again.
rektbyfaith
09-23-2025, 06:04 PM #12Archived author: Pursche • Posted: 2025-09-23T18:04:38.600000+00:00
Original source
Hang on, time to bring out Paint again.
Original source
This is the issue you were having before, when the client treated collision tests as just a position with no width.
Red is the clientside collider, green is the serverside "collider" and how the server might decide to place the character if there is no +0.5f hack. Notice how the slope is clipping into the pyramid, so if the server forces the player to this position, the clientside physics might decide "we're on the other side of this triangle" and just fall straight through.
![[Image: image.png?ex=690c3ae9&is=690ae969&hm=55e...71449202c&]](https://cdn.discordapp.com/attachments/284323424032129024/1420109099860758578/image.png?ex=690c3ae9&is=690ae969&hm=55ed1e9cfd29a06c1f963dcdabe18c5f0a4ff212e4f505a1220b87271449202c&)
rektbyfaith
09-23-2025, 06:06 PM #13Archived author: Pursche • Posted: 2025-09-23T18:06:33.830000+00:00
Original source
This is the issue you were having before, when the client treated collision tests as just a position with no width.
Red is the clientside collider, green is the serverside "collider" and how the server might decide to place the character if there is no +0.5f hack. Notice how the slope is clipping into the pyramid, so if the server forces the player to this position, the clientside physics might decide "we're on the other side of this triangle" and just fall straight through.
Original source
The sphere vs triangle check is instead going to do this:
![[Image: image.png?ex=690c3b26&is=690ae9a6&hm=ccb...86358b89f&]](https://cdn.discordapp.com/attachments/284323424032129024/1420109356170481896/image.png?ex=690c3b26&is=690ae9a6&hm=ccb1dfd6ac5a8d0df90ecf79a5e8b311f4d84b9e8a365b0c14226c286358b89f&)
rektbyfaith
09-23-2025, 06:07 PM #14Archived author: Pursche • Posted: 2025-09-23T18:07:35.014000+00:00
Original source
The sphere vs triangle check is instead going to do this:
Original source
Yes it's inaccurate, but it's inaccurate in the "safe" direction, of pushing the player further out from the problematic area. Assuming the radius is set correctly that is.
rektbyfaith
09-23-2025, 06:08 PM #15Archived author: Pursche • Posted: 2025-09-23T18:08:02.393000+00:00
Original source
Yes it's inaccurate, but it's inaccurate in the "safe" direction, of pushing the player further out from the problematic area. Assuming the radius is set correctly that is.
Original source
The fact that the client and server doesn't perfectly agree on what the collider looks like is not inherently an issue. The issue is when the server decides the collider is smaller than what the client thinks, places the player too close to a slope, and the client goes "I have no idea how we got here, but I'm slightly under the ground so lets fall"
rektbyfaith
09-23-2025, 06:09 PM #16Archived author: Pursche • Posted: 2025-09-23T18:09:42.877000+00:00
Original source
The fact that the client and server doesn't perfectly agree on what the collider looks like is not inherently an issue. The issue is when the server decides the collider is smaller than what the client thinks, places the player too close to a slope, and the client goes "I have no idea how we got here, but I'm slightly under the ground so lets fall"
Original source
Now, the golden question (don't laugh at me)
Do we even know if blizz uses the upside-down pyramid server-side?
rektbyfaith
09-23-2025, 06:09 PM #17Archived author: Elitia • Posted: 2025-09-23T18:09:45.939000+00:00
Original source
Now, the golden question (don't laugh at me)
Do we even know if blizz uses the upside-down pyramid server-side?
Original source
I have no idea, and frankly since I'm doing Novus stuff I don't want to know either.
rektbyfaith
09-23-2025, 06:10 PM #18Archived author: Pursche • Posted: 2025-09-23T18:10:19.370000+00:00
Original source
I have no idea, and frankly since I'm doing Novus stuff I don't want to know either.
Original source
So, all the way down to the actual approach i did?
rektbyfaith
09-23-2025, 06:10 PM #19Archived author: Elitia • Posted: 2025-09-23T18:10:59.892000+00:00
Original source
So, all the way down to the actual approach i did?
Original source
I think your approach is fine. I do think that there should be a lot of testing, especially what happens if you wedge yourself into a corner first and then blink (both into the wedge and away from it). And you might have to slightly adjust the radius you use a bit. And double check so the center of the sphere isn't at foot height of the player, then you need to move up the sphere a bit.
rektbyfaith
09-23-2025, 06:13 PM #20Archived author: Pursche • Posted: 2025-09-23T18:13:10.574000+00:00
Original source
I think your approach is fine. I do think that there should be a lot of testing, especially what happens if you wedge yourself into a corner first and then blink (both into the wedge and away from it). And you might have to slightly adjust the radius you use a bit. And double check so the center of the sphere isn't at foot height of the player, then you need to move up the sphere a bit.
1 Guest(s)
1 Guest(s)