Time Nick Message 07:18 Krock Will bump the formspec version and protocol version for 5.13.0 in about 1 hour so that it's not forgotten --> https://pastebin.com/raw/JNdsgMsJ 08:08 pgimeno I do the same as SFENCE_arch. I really hope that symlinked mods work. I'm using RUN_IN_PLACE as well. 08:12 pgimeno Is there a PR I can track about the trusted_mods on symlinked dirs problem? (I can't track issues, they require JS now) 08:14 Krock pgimeno: I cannot find or remember any such PR 08:15 pgimeno would it be useful if I bisect it? 08:16 Krock hmm I don't think so. It's pretty clear to me which code is affected. I just haven't gotten around testing it yet. 08:18 pgimeno ok, thx anyway 08:20 Krock pushing the version bump patch above 08:27 Krock correction: I cannot reproduce this as of master, although I'm using a RUN_IN_PLACE=1 build (which should not make a difference) 09:08 sfan5 it could relate to https://github.com/luanti-org/luanti/commit/a4d1b5b155566cb90961f6b75cf97b3ef21f20f8 09:32 pgimeno I've run into two minor bugs while creating a repro for the other issue. I wanted the firefles mod disabled and ran into this: http://www.formauri.es/personal/pgimeno/pastes/bugs-when-overriding-mod.txt 10:55 pgimeno Sorry, disregard that last report. I had another fireflies active which does have a depends.txt 11:03 pgimeno I can't reproduce SFENCE's bug with 5.12.0 either 16:41 [MatrxMT] reminder: meeting in about ~1:20h 17:57 Krock Meeting in 3 minutes. Pinging those who reacted. sfan5 Zughy. 18:01 luatic hi 18:02 Krock Hello! Welcome to IRC. 18:03 Krock Okay let's go through the milestone points 18:03 Krock > https://github.com/luanti-org/luanti/milestone/29 18:03 MTDiscord I am lurking while writing the release blog post, FWIW. 18:03 Krock Great. Thank you. 18:03 Krock #16358 might need another review but that's fortunately the only blocker on the list 18:03 ShadowBot https://github.com/luanti-org/luanti/issues/16358 -- Rewrite broken logic in blit_with_alpha2 by sfan5 18:05 luatic hmm i am a bit surprised that #16221 is not considered a blocker. either way we can for now probably merge the temporary fix. 18:06 ShadowBot https://github.com/luanti-org/luanti/issues/16221 -- Camera position flickering when attached to (some?) moving entity 18:06 sfan5 pong 18:06 Krock luatic: doesn't only the parent node need updating? the others should be covered by the attachment chain or recursively in updateAbsolutePosition...? 18:07 Krock nvm I'm misremembering updateAbsolutePosition is not recursive. 18:08 Krock however, it does call to `getAbsoluteTransformation`, which does not take the parent into consideration. That would change too much, so the proposed temporary change makes sense. 18:11 Krock Feel free to propose a PR. I'd however like to see a FIXME keyword such that it's clearly marked as temporary workaround. 18:11 sfan5 fine by me but it's likely to stick around longer anyway 18:12 Krock quoting networkpacket.h: "// ^ this comment has been here for 7 years" 18:13 luatic "there is nothing more permanent than a temporary solution" :^) 18:14 luatic worst case i could redo the way we propagate transforms and try to turn it into an on-demand kind of thing with a sort of "dirty" bit 18:15 luatic i'll submit a PR 18:15 luatic (for the temporary solution) 18:19 Krock On another topic - I did see that I caused a regression here: https://github.com/luanti-org/luanti/issues/16270#issuecomment-2997113839 However, I don't know how to debug this properly. Shall I just run the buildbot on this diff and hope it works well enough? 18:20 luatic If it's better than no temporary solution, I don't see why not. Would seem to call for another FIXME. 18:29 Krock #16361 testing results are welcome 18:29 ShadowBot https://github.com/luanti-org/luanti/issues/16361 -- Android: Workaround for partially broken inputs by SmallJoker 18:29 Krock quickly giving it a shot on my Android 6 device 18:30 MTDiscord I could test if someone handed me an apk 18:31 Krock I updated the PR description. The APK will appear on the page I linked 18:31 MTDiscord Ah, yes I noticed that late 18:32 Krock Here's a master build for comparison: https://github.com/luanti-org/luanti/actions/runs/16549071995 18:33 Krock (For verbosity: You need to self-sign the APK) 18:37 MTDiscord Blog post draft https://github.com/luanti-org/blog/pull/169 18:38 MTDiscord Small release so I dont have a ton of content to write on. Ideas welcome if any. 18:38 MTDiscord Oh, I wanted to include release stats.. Ill get to that after testing the android PR 18:40 Krock hmm. The key binding prompt also pops up on Android. That's a strange thing to see. 18:41 MTDiscord ... why is my Luanti app stuck on 5.11.0? 18:41 Krock such that you can use it for comparison :3 18:41 Krock I changed the font size for testing and now I cannot change it back because it covers the input fields 18:42 MTDiscord >play strore page updated Feb 17, 2025 Did the app store never get updated to 5.12.0?? 18:43 Krock ¯\_ (ツ)_/¯ 18:43 luatic it seems possible that ruben and grorp have been busy 18:43 MTDiscord @rubenwardy 18:43 MTDiscord Guess ill go grab the .12 APK first :/ 18:43 luatic maybe i could also be added to the android "team"? 18:44 sfan5 https://0x0.st/85Ub.png 18:44 sfan5 newest release we have uploaded to google ^ 18:45 sfan5 also btw we need to update our app api level until end of august 18:45 Krock not the minimal one, I hope? 18:46 sfan5 the target sdk 18:46 MTDiscord Tested .12, reproduced issue ... now for .13 18:50 Krock works neither with master not PR from what I've tested so far 18:56 Krock > extern gui::IGUIStaticText *guiroot 18:56 Krock ... what 18:57 MTDiscord Cant install the app. Even after self-signing. Thanks Android. 18:58 Krock did you get the right ARM arch build? You might want arm64-v8a for more recent devices. 18:59 Krock grorp's code makes sense. I don't know why this doesn't work. 19:00 MTDiscord Yes I used v8a for both .12 and .13 19:01 MTDiscord Alright apparently the manual install did not like overwriting the previous version. Had to uninstall 5.12 before installing 5.13 19:02 MTDiscord Works for me 19:03 MTDiscord You sure you got the right build, Krock? %) 19:05 MTDiscord Heh, just saw your comment. I made the mistake at first before realizing the issue was with the in-game settings menu 19:05 Krock Dear god. Now that you mention it. The two builds I tested have the same timestamp 19:06 Krock you know what this means and so do I 19:06 MTDiscord lol 19:12 Krock sfan5: I remembered an issue about the SDK version. It turns out we need a new one because the last one is already 1 year old. 14855 19:12 Krock #14855 19:12 ShadowBot https://github.com/luanti-org/luanti/issues/14855 -- Android: Update to SDK 34 by August 31 19:12 Krock couldn't we just use MAX_SUPPORTED_SDK directly and hope that nothing breaks? :3 19:13 Krock (fictive name, for example) 19:22 sfan5 well you need to read the compatibility notes and possibly adjust something 19:22 sfan5 anyway to go through some issues 19:22 sfan5 #16197 is this a blocker? 19:22 ShadowBot https://github.com/luanti-org/luanti/issues/16197 -- Player can glitch up the wall by walking towards it 19:23 sfan5 #16234 does sound bad but I don't see how we can get it fixed 19:23 ShadowBot https://github.com/luanti-org/luanti/issues/16234 -- Extreme lag when moving camera and running through Steam (on Steam Deck) 19:23 sfan5 and finally #16203 is a WONTFIX/CANTFIX 19:23 ShadowBot https://github.com/luanti-org/luanti/issues/16203 -- Both Bilinear and Trilinear filtering cause gaps along the edges of some entities 19:24 MTDiscord Requiring creators to include UV margin is ridiculous, especially for low-res pixel-perfect textures 19:25 sfan5 letting users enable texture filtering without creator consent is also ridiculous 19:25 sfan5 but it's what we have 19:26 MTDiscord Fair. I support a property to allow/disallow. But thats a future problem. 19:27 sfan5 so you think it's fine to leave as is for 5.13? 19:27 Krock I don't see the player glitch a blocker. I think we've had such glitches before (see. sneak-move through ceiling) 19:28 MTDiscord Yeah I do not consider it a blocker. Release often. 19:28 sfan5 thats okay by me 19:28 sfan5 it sounded like you would consider it a blocker 19:29 MTDiscord Sorry, I was only commenting on the conversation in the issue. Just didnt like that particular solution. 19:30 MTDiscord Did @luatic intend to apply the patch for #16221? 19:30 ShadowBot https://github.com/luanti-org/luanti/issues/16221 -- Camera position flickering when attached to (some?) moving entity 19:31 sfan5 but this is a good example of why tiptoeing around "but what if I liked the old behavior", "but what about the user choice" is unproductive and ultimately doesn't work 19:31 sfan5 #13108 and #16034 only turned out this way because someone somewhere liked the old behavior for some reason 19:31 Krock speaking of old behaviour - we might consider removing the old movement code :3 19:31 ShadowBot https://github.com/luanti-org/luanti/issues/13108 -- Remove/redo bilinear filter settings 19:31 ShadowBot https://github.com/luanti-org/luanti/issues/16034 -- Handle texture filtering sanely to avoid blurriness by sfan5 19:32 sfan5 and regressing things users like is a death sentence (but it's okay if it hits modders apparently?) 19:32 MTDiscord And this is why we need direction policies 19:32 Krock there's more players and server owners than modders, I presume. 19:33 MTDiscord That's not a useful metric. There are more players playing every game than those who developed them. Its still the developers who get final say in what goes in their game. 19:33 MTDiscord The content creator should have ultimate jurisdiction and responsibility to cater to their own audience 19:35 MTDiscord The steam deck issue has a workaround and also isnt technically a priority platform right now, so that issue shouldnt be a blocker IMO 19:37 MTDiscord The wall climb glitch is unfortunate but I am not sure it is worth delaying further to fix the highly-specific cases where it occurs 19:43 Krock @greenxenith Yes, these proportions make the decision making biased towards what players want. A direction is a good idea, but I'm uncertain how it would look like. 19:48 MTDiscord Thats a discussion for a different time 19:50 MTDiscord Zughy would normally be here to move the meeting along but since he isnt I guess Ill ask: Can we get the remaining PRs merged today and 5.13.0 released? 19:51 [MatrxMT] I'm here, I just don't have anything to say :D 19:52 MTDiscord Ah, hello then 19:52 MTDiscord #16287 and #16358 both have one approval, #16361 has been tested to work. Those are pretty much what needs merging, yeah? 19:52 ShadowBot https://github.com/luanti-org/luanti/issues/16287 -- Fix: change StartupWMClass to 'luanti' on Linux .desktop file to fix WM confusion by DEM0NAssissan7 19:52 ShadowBot https://github.com/luanti-org/luanti/issues/16358 -- Rewrite broken logic in blit_with_alpha2 by sfan5 19:52 ShadowBot https://github.com/luanti-org/luanti/issues/16361 -- Android: Workaround for partially broken inputs by SmallJoker 19:53 [MatrxMT] my two cents: if we keep postpone 5.13, releasing 5.14 will be a nightmare, considering it has to be prior to the jam 19:53 [MatrxMT] *postponing 19:54 [MatrxMT] so whatever it's done now to release today (not breaking anything), it has my approval. Also people going on vacation will make releasing later even harder 19:56 MTDiscord I'd say send it on the .desktop and android fixes (they are trivial), and get someone to glance over the alpha blit fix 19:59 MTDiscord If I were to test the blit_with_alpha2 PR to confirm fix would that suffice? Or do we need a second dev review? 20:01 sfan5 I'm confident in the code but another review wouldn't hurt 20:01 sfan5 i only review-requested luatic because he pointed out the bug in the code 20:04 sfan5 different topic: do we want to merge #16339 before release or not? 20:04 ShadowBot https://github.com/luanti-org/luanti/issues/16339 -- [Take 2] Prevent MapBlocks in generation from being unloaded by Montandalar 20:05 Krock > it's done now to release today (not breaking anything), it has my approval 20:05 Krock I'm being sarcastic here but this should apply in general. Unfortunately you cannot tell whether it's not breaking anything, Zughy. 20:05 Krock *this statement should 20:20 MTDiscord Im on Ihr's server with a shared locked furnace but cannot reproduce the crash yet 20:20 MTDiscord I did manage to crash the server itself though ;p 20:22 sfan5 a normal furnace works too 20:23 sfan5 it's a memory corruption so it may or may not crash 20:23 sfan5 or crash later 20:23 MTDiscord Do I specifically also need a cat? 20:23 sfan5 or send weird data to the server 20:23 sfan5 if you want an instant crash compile with ASan 20:24 grorp hi, I'm late to the party 20:25 grorp re "it seems possible that ruben and grorp have been busy": it appears I have lost access to our developer account, it just asks me to make a new account when I go to play.google.com/console now 20:26 grorp I think it's been like that since the account was transferred. ping celeron55 20:26 sfan5 you just need to be re-invited thn 20:26 sfan5 then* 20:27 grorp yeah 20:28 grorp but I also don't think I had the ability to upload releases, since they need to be signed. it was always ruben doing that. 20:29 sfan5 i dont have the password for the key either 20:33 MTDiscord Somehow my keybinds got very broken when opening compiled 5.12 20:34 MTDiscord Cannot seem to reproduce even with the compile flag though 20:36 MTDiscord Ok weird ... git decided to put me on a different commit than I asked 20:36 MTDiscord Trying again 20:38 MTDiscord That did it 20:41 MTDiscord sfan5: PR works as intended 20:43 MTDiscord Can these PRs get merged? 20:54 sfan5 they can 22:52 MTDiscord Thoughts on making unfavorited servers only show the name field initially if split_login_register is on? Clicking 'Login [->]' will re-show the password field and move the '[->]' to the register button 22:52 MTDiscord https://cdn.discordapp.com/attachments/747163566800633906/1399162603271491584/image.png?ex=6887ff37&is=6886adb7&hm=b5e5fc09ce1967478db69e32ae8dbd2d21bd5a4f618a4feab349e671a25cd442& 22:55 MTDiscord It puts an extra click between you and accidentally logging into a (probably) new server instead of registering