Time Nick Message 00:37 [MTMatrix] #14436: why are calling "backwards compatibility" the fact of keeping a bug? If modders don't code properly, why should we be at fault? 00:40 MTDiscord Break that thing break it 00:40 MTDiscord That bug goes into the hole 00:41 [MTMatrix] (Note: I'm at fault as well) 00:44 MTDiscord If it's attached to the entity, when the entity disappears, kill the entity 00:44 MTDiscord Or it should be a setting like "die_with_entity" 00:44 MTDiscord I mean, kill the sound, the entity already died 04:46 MTDiscord @Anti Null Pointer Administrator this would actually be a really good way to kill audio streams, right now the best solution is having two separate instances of the world map and positioned mono audio. That could be simplified if sounds died with the entities they were attached to. 05:08 MTDiscord Yes 05:09 MTDiscord Will it happen when it's fixed? 05:09 MTDiscord No 05:44 MTDiscord @Anti Null Pointer Administrator some other ideas could be a stop function, or a callback that let you know when the audio stream naturally stopped. 12:01 MTDiscord Zughy: First off, note that there is no blatant "contract violation" by modders here. The behavior on entity removal was simply unspecified. 12:01 MTDiscord It is not unreasonable for modders to then assume that the observed behavior is the expected behavior. (Though I would encourage modders to seek API guarantees.) 12:05 MTDiscord My PR poses no significant reduction in functionality: If modders want to stop sounds on entity removal or deactivation, they can still do it themselves. 12:06 MTDiscord As an additional bonus, my PR stops the sounds serverside, meaning clients don't have to update to get the fix for #8094 - only servers have to update. 12:06 ShadowBot https://github.com/minetest/minetest/issues/8094 -- Audio attached to object stops moving when object leaves 12:08 MTDiscord There are however admittedly two slight downsides I should point out: 12:08 MTDiscord (1) having to store the sound handle may be slightly inconvenient (though it really amounts to just +1 line of code, +3 if on_die / on_deactivate would otherwise be undefined) 12:09 MTDiscord (2) if the "stop sound" packet from the server is delayed vs. the entity remove packet, the sound may continue playing slightly longer than expected 12:09 MTDiscord For these reasons, we might want to add a new field for sounds that are supposed to be removed on object removal, as jordan suggests. 12:15 MTDiscord Presently, 5.9-dev breaks backward compatibility ("a property of a software that allows for interoperability with an older legacy system"): Death sounds that worked fine in 5.8 and earlier won't work in 5.9. Realistically, many mods and games won't be updated in a timely manner, leaving users frustrated. 12:17 MTDiscord 5.9-dev presently also harms the use case of wanting a short sound to not be cut off when the object dies. 12:18 MTDiscord I think my PR can keep everyone happy, both in terms of available functionality and in terms of keeping backward compatibility. 16:22 MTDiscord Needs to be a new setting because it was just showcased that this can be wanted behavior 16:49 sfan5 merging #14419, #14369 in 10m 16:49 ShadowBot https://github.com/minetest/minetest/issues/14419 -- [no sq] Fix some common SAO methods to not generate useless update packets by sfan5 16:49 ShadowBot https://github.com/minetest/minetest/issues/14369 -- Try to preserve metatable when exchanging data with the async env by y5nw 17:36 sfan5 https://dev.minetest.net/Code_style_guidelines#C.2B.2B_features I revised this section a bit 19:25 appguru i'd like to merge #14436 in 10m 19:25 ShadowBot https://github.com/minetest/minetest/issues/14436 -- Stop sounds if objects get out of range, but not if objects are removed by appgurueu 19:30 MTDiscord "Stop sounds if objects get out of range, but not if objects are removed by appgurueu, but do stop them if they're removed by anybody else" 19:31 appguru heh 19:31 appguru do you have a good suggestion for how i could reword this? it's a bit lengthy already 19:33 appguru I think I'll go with "Do not stop sounds if objects are removed serverside" 19:33 MTDiscord I'd probably use punctuation instead of " by " for the separator. It'd break the flow a bit better, so "not if objects are removed - appguru" or "not if objects are removed [appguru]" or something. 19:33 MTDiscord Ah, if you mean the PR itself ... heh 19:35 appguru Actually I think I'll go with "Fix attached sounds stopping if objects are removed serverside" 19:36 appguru done 19:38 MTDiscord OIC, this doesn't add stopping sounds out of range, that's already a feature, this just fixes a related nuance? 19:42 MTDiscord Stopping sounds when an object is removed client-side (including when it goes out of range) was added to 5.9-dev by #14341. 19:42 ShadowBot https://github.com/minetest/minetest/issues/14341 -- Remove sound attached to active object when active objects is removed by sfence 19:42 MTDiscord And is now partially reverted. 19:44 MTDiscord lol, I didn't know what "OIC" means, so I asked Duckduckgo and the first result was "what does it mean when a girl says OIC" 19:44 MTDiscord lol 19:44 MTDiscord for anyone else who didn't know, it means what it sounds like: "oh, I see." :-) 19:45 celeron55_ sfan5: i quickly read through the diff for that wiki page and it looked ok 20:52 sfan5 👍 20:52 sfan5 merging #14434 in 10m 20:52 ShadowBot https://github.com/minetest/minetest/issues/14434 -- Reduce translations log spam by sfan5 21:33 sfan5 spent way too long refactoring some code today. in hindsight not sure if this was a good use of my time 21:34 MTDiscord Did you enjoy it? 21:35 sfan5 maybe 21:35 MTDiscord Then it might have been a good use of your time