Forums » Bugs
jump check does not match distance meter
Summary:
Sometimes I can make an in-system jump before the distance meter reaches 3000m. Usually this is only by a little bit, but sometimes it is by a lot: for example, in Sedina H-10, when making an in-system jump to Sedina H-11, I went out to 2300-2350m and came to a full stop. I targeted an asteroid (the hud said it was 2600-2700m away). Then I pressed the activate key and successfully jumped to Sedina H-11.
I do not believe this is lag, because it occurs with a large distance (~700m too soon) and occurs when I am at 0m/s (so I don't think there would be a big difference between where my computer thinks I am and where the server thinks I am). My ping is about 300ms+-20ms. I performed this jump a few times in Sedina G10, G11, H10, and H11 in a Centaur, on 2023-05-16 at 7-8pm PST.
Some jumps were performed with plugins installed, but the H10-H11 jump I detailed above was performed with all plugins deactivated (by renaming the plugins folder to _plugins and restarting the game).
Device: Dell laptop running Windows 10 Home with NVidia GeForce GTX 1050 Ti graphics card.
Sometimes I can make an in-system jump before the distance meter reaches 3000m. Usually this is only by a little bit, but sometimes it is by a lot: for example, in Sedina H-10, when making an in-system jump to Sedina H-11, I went out to 2300-2350m and came to a full stop. I targeted an asteroid (the hud said it was 2600-2700m away). Then I pressed the activate key and successfully jumped to Sedina H-11.
I do not believe this is lag, because it occurs with a large distance (~700m too soon) and occurs when I am at 0m/s (so I don't think there would be a big difference between where my computer thinks I am and where the server thinks I am). My ping is about 300ms+-20ms. I performed this jump a few times in Sedina G10, G11, H10, and H11 in a Centaur, on 2023-05-16 at 7-8pm PST.
Some jumps were performed with plugins installed, but the H10-H11 jump I detailed above was performed with all plugins deactivated (by renaming the plugins folder to _plugins and restarting the game).
Device: Dell laptop running Windows 10 Home with NVidia GeForce GTX 1050 Ti graphics card.
While this could be a bug, it could also be that the asteroid you targeted was not large enough to be considered a large gravitational object that would interfere:
https://www.vo-wiki.com/wiki/FAQ#How_do_I_Jump.2FWarp.3F
https://www.vo-wiki.com/wiki/FAQ#How_do_I_Jump.2FWarp.3F
There is some known variance to this, because of the difference in how the game client displays distance, versus how the Sector itself enforces jump availability (the sector is authoritative in whether the jump is allowed, we do not "trust" the client in these situations).
High-precision distance checks are more computationally intensive, because they factor in the geometry of the object to find the actual closest point. This is what the client is doing, to get a high-precision measurement of your distance to a remote point.
The sector, however, only aims to be "roughly" correct, by using the center-point of the object. But, for a large asteroid, the center point could be pretty far from the surface, or from the closest-point of the object's surface to you, particularly if the object is kind of "oblong" or something.
It's something we'll probably need to look at in more detail as time goes on, particularly if we get larger-and-larger individual objects, which induce a greater perceived distance offset.
The difference in computational intensity of high-precision vs low-precision isn't really that big of a deal, on the server (particularly not for a single player trying to jump); but it's one of those things where, if for some reason it happens a hell of a lot, that can become more of an issue. VO is designed to be pretty efficient, to be able to scale to unexpected situations if necessary.
High-precision distance checks are more computationally intensive, because they factor in the geometry of the object to find the actual closest point. This is what the client is doing, to get a high-precision measurement of your distance to a remote point.
The sector, however, only aims to be "roughly" correct, by using the center-point of the object. But, for a large asteroid, the center point could be pretty far from the surface, or from the closest-point of the object's surface to you, particularly if the object is kind of "oblong" or something.
It's something we'll probably need to look at in more detail as time goes on, particularly if we get larger-and-larger individual objects, which induce a greater perceived distance offset.
The difference in computational intensity of high-precision vs low-precision isn't really that big of a deal, on the server (particularly not for a single player trying to jump); but it's one of those things where, if for some reason it happens a hell of a lot, that can become more of an issue. VO is designed to be pretty efficient, to be able to scale to unexpected situations if necessary.
Thanks for the explanation!
That makes a lot of sense. And explains why it seems to occur only in certain sectors at certain places - those sectors/jump points are where the closest object is big enough that ship-to-surface distance is noticably different than ship-to-center distance. And also why I don't see it occur in reverse, because the center of an object is virtually always inside of it, ensuring the center is always further away than the surface. Neat.
Thanks for the detailed post, it's really cool to see a glimpse into how the game works inside!
That makes a lot of sense. And explains why it seems to occur only in certain sectors at certain places - those sectors/jump points are where the closest object is big enough that ship-to-surface distance is noticably different than ship-to-center distance. And also why I don't see it occur in reverse, because the center of an object is virtually always inside of it, ensuring the center is always further away than the surface. Neat.
Thanks for the detailed post, it's really cool to see a glimpse into how the game works inside!
Happy to help :). We appreciate the bug reports as well.