Time Nick Message 05:53 hmmmm christ this channel got big 05:53 hmmmm just came to visit 05:53 hmmmm sofar I was in hillsboro last week, was debating whether or not to hit you up and hang out 05:54 hmmmm interviewed with VPE group, graphics driver && hardware optimization 08:10 nerzhul hi hmmmm how are you ? 17:04 Hijiri reminder tha #7269 has been updated, mainly with respect to behavior of set_physics_override 17:04 ShadowBot https://github.com/minetest/minetest/issues/7269 -- Player physics modifiers by raymoo 17:04 Hijiri otherwise it is mainly bug fixes I came across while writing the code 17:24 Krock oh man. so much discussion to follow 17:24 Krock will look at the code instead to get an impression 17:51 Hijiri next time I make updates I won't amend commit, so it's easier to see changes 17:52 Krock The diff is not too big, but separate commits between reviews are appreciated, yes :) 18:05 Hijiri Krock: how do I print a script backtrace? 18:06 Hijiri or is there an example in the code you know of where I can just look at it 18:10 Krock script_get_backtrace(L), probably. 18:10 Hijiri thanks 18:10 Krock defined in c_internal 18:17 Hijiri when I print traceback it only shows one level 18:19 Hijiri like this paste.debian.net/1041382 18:19 Krock ugh. I already thought it would come that way. The backtrace code needs some work; especially it sometimes does not provide helpful information 18:20 Krock the debug functions from Lua might help 18:20 Hijiri should I print the backtrace anyway since the code may be improved? 18:21 Hijiri If the problem affects other uses of it then I think it should be a separate item 18:21 Krock maybe you should move the warning to the physics override function 18:21 Hijiri I could have it in both places 18:21 Krock so that warnings are thrown when modifiers are active 18:21 Hijiri in case the override happens first 18:21 Krock well, the new modifiers would be added and removed too often -> spams the logs 18:22 Krock ugh right 18:22 Hijiri but then you wouldn't get any warnings if a mod sets a physics override when the player joins and never again 18:25 Krock Only displaying the warning per source script line once would be great to filter out duplicate situations 18:26 Hijiri that seems like a lot of work 18:26 Krock like some copypasta 18:27 Hijiri fine with static mutable variables? 18:27 Krock you might just put some of this into a generic function, taking a vector as argument: https://github.com/minetest/minetest/commit/49509d2f746b7a8d50e685cfd0e0b391676b9466#diff-f49b73df912f38d8e5dffa40c208a42aR102 18:28 Krock downside is that the stack getter index might need to be adjusted for your case 18:29 Hijiri ok 18:29 Hijiri I'll do it a bit later since I need to pick up my bike and buy food 18:29 Krock no hurry :) 18:29 Hijiri I'll upload my current changes so you can see them 18:29 Krock sure thing 18:32 Krock the helper function will not even need to take a vector as argument. it could just return the backtrace string to simplify things 18:33 Hijiri imo should be a helper class that contains the already reported set 18:33 Hijiri then I don't need to pass anything in 18:34 Hijiri I'm not totally convinced though that it should only happen once per line 18:34 Hijiri if a server operator is looking for errors at some certain time point then they won't immediately see the warning 18:34 Krock well, sprint mods would then update it with each change and so -> trigger a warning each second 18:35 Krock or every few seconds, but multiplied by the online players 18:36 Hijiri I think it's better that the warning is close to the point buggy behavior is observed 18:36 Hijiri ideally there would be some way to filter out those messages but still have them 18:37 Hijiri deprecation doesn't cause visible bugs so there's no buggy behavior point to have th warning message near