Minetest logo

IRC log for #minetest, 2020-01-09

| Channels | #minetest index | Today | | Google Search | Plaintext

All times shown according to UTC.

Enable nick filtering
Time Nick Message
35 more elements. Show/hide.
09:25 BuckarooBanzai !tell sfan5 hey there ;) quick question: if i build minetest with "CMAKE_BUILD_TYPE=Debug" i get stacktraces on segfaults, right?
09:25 MinetestBot BuckarooBanzai: I'll pass that on when sfan5 is around
09:25 sfan5 .
09:25 MinetestBot sfan5: Jan-09 09:25 UTC <BuckarooBanzai> hey there ;) quick question: if i build minetest with "CMAKE_BUILD_TYPE=Debug" i get stacktraces on segfaults, right?
09:25 sfan5 you get stacktraces either way
09:26 sfan5 if you build with debug info you get stacktraces plus function names (and parameters)
09:26 sfan5 you can also use -DCMAKE_BUILD_TYPE=RelWithDebInfo to build in release mode (code optimization) but include debug information
09:27 BuckarooBanzai sfan5: currently with "CMAKE_BUILD_TYPE=Release" the engine crashes and leave a one-liner in the kernel-logs :P
09:27 BuckarooBanzai *leaves
09:27 sfan5 since "full" debug is potentially slower (though I don't know by how much)
09:27 BuckarooBanzai i have a few of these every few weeks: [10284076.384135] Server[20933]: segfault at 7fc7477e0580 ip 00007fc7477e0580 sp 00007fc745f6cb10 error 15
09:28 BuckarooBanzai RelWithDebInfo sounds good, i'll try that...
09:29 BuckarooBanzai thanks sfan5, i'll open an issue if i get to the cause :D
2 more elements. Show/hide.
09:55 rubenwardy BuckarooBanzai: you'll only get stack traces when running with gdb
5 more elements. Show/hide.
10:53 BuckarooBanzai rubenwardy: yes, i'm looking into core_dumps right now ;)
34 more elements. Show/hide.
15:26 deltasquared hmm, set a server tick rate to 100hz and then played an explosion noise on a globalstep... my ears regret everything
17 more elements. Show/hide.
17:32 deltasquared has there ever been entertained the idea of allowing blocks to dynamically generate meshes to send to the client at runtime?
17:33 deltasquared protocol and server aside, I think the client side part might be the easiest bit. it ought to fit right into the current mapblock meshing routines.
17:33 deltasquared protocol encoding... anyone's guess. server would need some sort of callback that has to be called immediately on loading a relevant node though.
17:34 deltasquared (unlike say lbms which I've observed to be somewhat relaxed in when they're called after the nodes themselves are loaded)
5 more elements. Show/hide.
18:10 MinetestBot [git] kostya13 -> minetest/minetest_game: Replace Russian translation of Pickaxe (#2556) 7e9a54a https://git.io/Jvvn3 (2020-01-09T18:09:06Z)
3 more elements. Show/hide.
18:49 MinetestBot [git] Wuzzy2 -> minetest/minetest: Add missing entity documentation (#9297) 3189e44 https://git.io/JvvcY (2020-01-09T18:48:16Z)
22 more elements. Show/hide.

| Channels | #minetest index | Today | | Google Search | Plaintext