Time Nick Message 01:34 paramat well, first i'll bisect to confirm if it is that commit 02:00 p_gimeno it's faster to check that commit and the one before :) 02:08 paramat heh, yes that's what i'm doing :) 02:09 paramat bisecting from a range of 2 commits 02:11 p_gimeno lol 02:26 paramat confirmed the cause is the suspected commit, i might attempt a revert PR 02:30 paramat doing that now 03:08 paramat not too bad. done #8288 03:08 ShadowBot https://github.com/minetest/minetest/issues/8288 -- Revert CSM particles commit to fix bug for 5.0.0 by paramat 04:24 paramat merging #8288 build fails are due to translation thingy 04:24 ShadowBot https://github.com/minetest/minetest/issues/8288 -- Revert CSM particles commit to fix bug for 5.0.0 by paramat 04:32 paramat nerzhul (for when you are around) are you able to help with #8284 ? 04:32 ShadowBot https://github.com/minetest/minetest/issues/8284 -- Can’t compile due to weblate, translation issue 04:33 ANAND paramat: Updated #8260 with your suggestion 04:33 ShadowBot https://github.com/minetest/minetest/issues/8260 -- README: Remove references to Minecraft and InfiniMiner by ClobberXD 04:35 paramat ok 07:53 nerzhul seriously 07:54 nerzhul 01cd63bd3bca0192dab2834faf414b022706a77e is unacceptable. 07:54 nerzhul re commiting. We should find a fix. 07:54 nerzhul it should not be difficult to find the problem the client part is not very difficult to understand 07:54 nerzhul the problem is in that part 09:06 nerzhul the PR fix #8289 09:06 ShadowBot https://github.com/minetest/minetest/issues/8289 -- Fix particle spawners not visible since CSM spawner implementation by nerzhul 09:07 nerzhul this is the real way to fix it. 09:07 nerzhul dropping a whole feature is not the solution when the problem is just the registration of a id... 15:02 jas_ i would agree, if i were in a position to 19:09 ANAND #8034 is no longer WIP, and is ready for testing and review 19:10 ShadowBot https://github.com/minetest/minetest/issues/8034 -- Disable auto-setting server as favorite on connect; display "Add Favorite" button by ClobberXD 20:27 paramat nerzhul reverting a commit by yourself is not acceptable, especially at the 'We should find a fix.' stage. the first revert should stay until your PR is tested 20:28 paramat we are very close to a release, your fix may cause unforseen issues, the safest solution is a revert, then you can make your fix after release. the feature is not important and CSM is unfinished 20:31 paramat we have reverted a few other things that caused bugs, due to doing this being the least risky solution, this situation is no different 20:32 paramat you could have based your PR on the reverted code, it would have been no more difficult, then the revert would have been a backup solution 20:33 paramat i will recommit the initial revert then you can rebase your PR 20:33 VanessaE paramat: so you want to revert the reverted revert? 20:37 paramat but anyway, we all make impulsive mistakes, i hope you admit your mistake and apologise ;) 20:38 paramat sfan5 could you look at nerz's PR? i remember you considered a proper fix non trivial 20:39 sfan5 that was a guess 20:39 sfan5 but sure 20:45 paramat heh yes 20:46 paramat i'm concerned this is too risky and we won't have enough time to test. 5.1.0 will likely be soon after so the proper fix won't be far off. ugh 20:47 sfan5 not every code is buggy, I don't think risk is a concern here 20:47 p_gimeno even 5.0.1 20:59 paramat i'm not saying the PR is buggy, but we have decided several times recently to apply the low risk solution due to how close release is, this is no different. so yes risk is a huge concern at this stage, obviously 21:12 paramat but, i won't revert yet, as that would then break this latest PR, even though we 'should' 21:41 paramat should a build warning be considered a blocker? #8292 i'll make a PR 21:41 ShadowBot https://github.com/minetest/minetest/issues/8292 -- Warning about “writing to an object of type 'struct NoiseParams' with no trivial copy-assignment” in file src/noise.cpp when compiling 21:54 paramat i can fix one warning but have no idea about this one #8291 21:54 ShadowBot https://github.com/minetest/minetest/issues/8291 -- Warning about out-of-bounds offset in src/util/srp.cpp when compiling 21:58 paramat hm suddenly lots of blockers, i can see release slipping another week 22:07 nerzhul test my pr. 22:08 nerzhul it's just working and restoring the behaviour you got by reverting the feature PR. 22:08 nerzhul build warning are not intended to be blocker paramat 22:10 nerzhul i just registered particlespawner properly. the problem in the original pr is the mapper between server & client ids 22:11 nerzhul it's a wrong idea due to another code part which requires that ID but doesn't know about it. i just merged the two behaviour by using a uint64 with lower values (u32) for server part, like before, and upper parts for client ids 22:23 paramat it's mostly your responsibility to thoroughly test it, since you are promoting a risky situation. understandably i don't much feel like testing it 22:25 paramat and yes i somewhat agree a warning is not a blocker, removing labels ...