Time  Nick       Message
11:06 MTDiscord  <josiah_wi> #14565 should be easy to review given that it has concept approval and sfan5 has previously reviewed most of the changes.
11:06 ShadowBot  https://github.com/minetest/minetest/issues/14565 -- Upgrade client active object mgr tests to Catch2 (reviving #13609) by JosiahWI
11:07 MTDiscord  <josiah_wi> Seeing that sfan5 has PRd some benchmarks for active objects (thank you, sfan!) it seems like a good time to work on these unit tests as well.
11:43 rubenwardy #14616 seems like it could be merged under the one approval trial
11:43 ShadowBot  https://github.com/minetest/minetest/issues/14616 -- Add benchmarks for server::ActiveObjectMgr by sfan5
12:08 sfan5      it could be
12:22 rubenwardy I'm not familiar with the benchmarking library used but it appears to only run the function once in that callback. Shouldn't it run loads of times to get an average? Or does the library do that
13:33 sfan5      the library does that
17:58 pgimeno    sfan5: re https://github.com/minetest/minetest/issues/14611#issuecomment-2094364044 the problem is not the jump range per se, but the failure to allocate memory that is within that range of a couple certain symbols, and that *is* OS-dependent, not just arch-dependent.
22:00 sfan5      apparently std::fstream classes only return error information if you enable exceptions and it's also implementation-defined whether there will be anything useful in them
22:00 sfan5      this is terrible, who designed this?
22:19 sfan5      it doesn't even return anything other than "basic_ios::clear: iostream error" on libstdc++
23:04 MTDiscord  <luatic> just commented on another post on the forums effectively complaining about free not returning (enough) memory to the OS, mislabeling this as a memory leak (https://forum.minetest.net/viewtopic.php?p=435760).
23:04 MTDiscord  <luatic> should we consider tweaking M_TRIM_THRESHOLD via mallopt? should we perhaps consider a setting to let server admins conveniently tweak it themselves?