Guild icon
Teeworlds
discord.gg/teeworlds / development
For discussions around the development of the official Teeworlds
Between 2020-05-22 00:00:00Z and 2020-05-23 00:00:00Z
Avatar
GitHub BOT 2020-05-22 13:46:43Z
Author icon
BannZay
[teeworlds/teeworlds] Issue opened: #2630 Mouse pointer dramatically affected by V-Sync
Any fps limitations dramatically affects your ability to control mouse pointer. The game is unplayable for me with V-Sync/low fps limitations I believe, game cursor should have much higher priority over other things when on game renderings Maybe there is any way I can use "hardware" cursor ingame ?
Avatar
GitHub BOT 2020-05-22 14:46:45Z
Author icon
Sonix-
[teeworlds/teeworlds-maps] Issue opened: #21 ctf2 background (mountains) empty corner
![](https://cdn.discordapp.com/attachments/407308363031117834/713367793797693460/screenshot_2020-05-22_14-25-08.png) reported by mind on discord
Avatar
GitHub BOT 2020-05-22 14:53:54Z
Author icon
Zwelf
[teeworlds/teeworlds] Pull request opened: #2631 Remove unused declaration
Avatar
Deleted User 2020-05-22 16:10:21Z
hello, i worked on a skin converter from 0.6 to 0.7(https://github.com/Jupeyy/teeworlds_skin_converter_06_to_07) and noticed, skins on 0.7 have some buggs: the feet position is changed, the whole body(including eyes and rest) is around 2 pixels too far to the right Even if the feet position is discussable, the body thing seems more like a bug, i add a gimp file where u can change the opacity of the layer to see the difference from 0.6 to 0.7 i ofc could be wrong
16:10
tee.xcf
1.48 MB
16:10
from my findings its just programmatically not in the images itself
Avatar
fokkonaut 2020-05-22 16:44:35Z
Its pretty much due to the new skin system
Avatar
Deleted User 2020-05-22 16:50:37Z
i'm pretty sure its caused by float x2 = (x+w-1/32.0f)/(float)cx;
16:50
https://github.com/teeworlds/teeworlds/blob/c013d9d14b103f32706b31f18679222c6f0e0bd2/src/game/client/render.cpp#L40
teeworlds/teeworlds
A retro multiplayer shooter. Contribute to teeworlds/teeworlds development by creating an account on GitHub.
Thumbnail
16:50
the extra offset makes the image smaller
16:50
causing it to shift to the right
16:52
texture bleeding is a problem since tw exists, 0.7 uses a 3d texture for tilesets and still tries to fix it by hand for non tiles which is bad
16:54
also adding 1/32 is too much anyway
16:54
a texel is 0.5 of a pixel not 1
16:55
or better: trapping the image in the texel area needs to add a 0.5/32 offset to the x and y coordinate
16:56
and also the offset has to be outside of the cx variable which already is the grid size
16:56
float x1 = x/(float)cx + 0.5f/(float)(cx*32); float x2 = (x+w)/(float)cx - 0.5f/(float)(cx*32); float y1 = y/(float)cy + 0.5f/(float)(cy*32); float y2 = (y+h)/(float)cy - 0.5f/(float)(cy*32); (edited)
16:56
this would be the ddnet way todo it (edited)
Avatar
Deleted User 2020-05-22 17:03:37Z
but that's still wrong as far as i see
17:03
its dependend on the actual image size
17:05
float x1 = x/(float)cx + 0.5f/(float)(ImageWidth); float x2 = (x+w)/(float)cx - 0.5f/(float)(ImageWidth); float y1 = y/(float)cy + 0.5f/(float)(ImageHeight); float y2 = (y+h)/(float)cy - 0.5f/(float)(ImageHeight); smth like this would be correct (edited)
17:07
but it would be better to add no offset than a wrong one like 0.7 does
Avatar
Ravie 2020-05-22 18:06:30Z
If you flip a 0.7 screenshot horizontally back and forth you will see that feet are actually pretty centered, but the body moves
Avatar
Deleted User 2020-05-22 18:08:24Z
anyway the issue i just mentioned here affects everything that uses this function
18:08
so also feets
Avatar
louis 2020-05-22 23:31:07Z
feet are so off centered
23:31
Embedded image
23:32
red lines are centered on body
Exported 28 message(s)