Time  Nick      Message
00:13 lhofhansl sfan5: Serializing without compressing worked beautifully. I'll have a PR soon which handled map save in a separate map saving thread on the server, and compression is done without holding the global lock.
00:30 lhofhansl #10709
00:30 ShadowBot https://github.com/minetest/minetest/issues/10709 -- Allow map save to compress blocks asynchronously. by lhofhansl
06:51 m42uko    The deadzone is supposed to work, but the default was way too low. I think I had to go up to 6000 for my XBox Wireless controller. Also, I think I used the standard profile instead of the specific XBox one.
06:53 m42uko    The back button behavior is expected like this. It's the same if you old ESC on they keyboard, right?
06:54 m42uko    "walking left or forward can happen without even touching the controller" - That's the behavior I saw without my patches (due to the hardcoded button thresholds), but now it's stable for me.
06:55 hecks     Has someone been messing with joysticks? because I'm getting some weird deadzone behavior with keyboard controls...
07:09 m42uko    I have, but it's not merged yet.
07:09 hecks     hm, okay
07:10 m42uko    What do you mean "deadzone behavior with keyboard controls"?
07:10 hecks     Basically, when I walk in a direction that's really close to a cardinal axis, my velocity gets snapped to said axis.
07:10 hecks     This is what would happen with a naive deadzone implementation.
07:10 hecks     But it could be precision gremlins too
07:11 m42uko    Sounds more likely. The joystick deadzone code is completely decoupled from that logic
07:13 m42uko    sfan5: I just tested my PR again (also rebased against the latest master), and it works properly for me.
07:19 m42uko    rubenwardy, thanks for the link. Controller identification was something I wanted to look into as well, so I'll keep that in mind.