Time Nick Message 00:15 MTDiscord So core devs, I have the background music PR pending, the only remaining complaints I've had is that sfan5 is unsure about sound_stop_all. If that is the only hold up I can go back and see if it is possible to track sound handles properly in the main menu Lua scripts without a crazy amount of hassle. If there are other holdups can they be voiced so I can get this merged in? I'm trying to do small, merge-able things rather than 00:15 MTDiscord standard held up PR due to small needed or requested changes 01:33 v-rob exe_virus: If the only important reason for sound_stop_all is to stop music in the main menu and can't be done well any other way, you could add it as a main menu-only function (which is possible, I think) and make a small separate PR for adding it to the main Lua API if you still find it useful, so the original PR can be merged without argument. 01:34 MTDiscord Good advice, okay I'll separate them for now. Thanks for the tip 18:12 Krock pgimeno: IIRC you had a look at #11039 .. already a while ago. do you think it would be okay to merge, or are there any concerns? 18:12 ShadowBot https://github.com/minetest/minetest/issues/11039 -- Add metatables to lua vectors by Desour 18:13 pgimeno looking 18:17 pgimeno yes, for me it's good to merge 18:20 v-rob It would also be nice to merge #10324 to complete the trio of vector PRs 18:20 ShadowBot https://github.com/minetest/minetest/issues/10324 -- Split vector.new into 3 constructors by Desour 18:22 pgimeno oh #8515 was closed, shame 18:22 ShadowBot https://github.com/minetest/minetest/issues/8515 -- Add matrix helpers to builtin by Desour 18:23 v-rob That's actually not too bad; for the 3D successor to #10801, I will make transformation matrices and quaternions that are available to both server and client Lua APIs 18:23 ShadowBot https://github.com/minetest/minetest/issues/10801 -- Add CSM 2D Drawing API by v-rob 18:28 v-rob For vectors, there's also #10328, but I'm personally not a big fan because in-place operations could apply to all manners of things, such as `vector.floor`, so applying it to everything would add a lot of redundant functions. Lua doesn't have a `+=` anyways. 18:28 ShadowBot https://github.com/minetest/minetest/issues/10328 -- Add in-place versions for vector operations by Desour 18:39 Krock pgimeno: okay thanks 18:39 Krock I'll merge it this evening 19:06 Krock will merge #11039 and #11298 in 15 minutes 19:06 ShadowBot https://github.com/minetest/minetest/issues/11039 -- Add metatables to lua vectors by Desour 19:06 ShadowBot https://github.com/minetest/minetest/issues/11298 -- Android: Do not submit text after pressing Enter key for multi-line text by srifqi 19:21 Krock merging 19:22 Krock done 19:37 MTDiscord this will be helpful thx 21:32 Pexin what's the best way to talk one on one with smalljoker? 21:32 v-rob Using /msg? 21:34 rubenwardy he's Krock here 21:35 Pexin thank 21:40 kilbith v-rob: catching you finally 21:41 kilbith are you able to inform us about the formspec rewrite progression? 21:41 v-rob Yeah, sfan5 made a none-to-subtle dig at me for not being on IRC, so I'm here. 21:41 kilbith it is your duty, yes 21:42 v-rob Yes it is, and I apologize 21:43 kilbith so your reply might give me some indications whether it's worth to keep working on the current formspec API or not 21:44 kilbith I have some code for the scrollbar which is idling 21:49 kilbith ... v-rob? 21:49 v-rob I'm here, just had to leave for a quick moment 21:50 kilbith I'm wondering if you keep this question out of sight on purpose 21:52 v-rob For the scrollbars, I would say that it's fine to keep working on it. I don't forsee a complete formspec replacement anytime soon. 21:52 v-rob Currently, there are two holdups 21:52 kilbith you didn't work on the replacement recently, right? 21:53 v-rob I've done very little, yes. I won't make any excuses for not working because I don't really have many 21:54 v-rob Anyway, I can't really work on the API proper very much anyway yet 21:54 v-rob Firstly, SDL2 is still missing. I may have to live with a suboptimal Irrlicht event API for now to get somewhere 21:54 kilbith well that's disappointing because I spoke up about you to become a core-dev, you exposed a lot of your grand ideas, and you are not able to materialize them 21:55 v-rob I certainly still do have grand ideas. I've been really wanting to get the 3D rendering PR started, but I can't since the 2D has some obnoxious Irrlicht bugs left. 21:56 v-rob Specifically, setViewPort doesn't work right, and making it work right in IrrlichtMt makes other parts of Minetest break 21:56 v-rob There's also Freetype issues 21:57 v-rob Even though I'm not doing much at the moment, I refuse to give up on the replacement, however 21:59 kilbith the key event PR is a long overdue for example 21:59 kilbith and we can certainly make a smooth transition with the keycodes once SDL2 is a thing 21:59 kilbith I mean, that's not difficult 22:00 kilbith but you have to bring it more on IRC 22:03 v-rob It definitely is overdue. I mean, it's been around for two years now. I guess to merge, I'll still need to have universal keycodes though. Slight modifications to keybode.cpp might suffice. 22:08 v-rob Say, did you ever get the map[] element anywhere? 22:08 v-rob It looked pretty good 22:08 v-rob That was you, right? 22:09 kilbith yes that was me. I often remind myself to finish it but I'm busy partying right now 22:10 kilbith scrollbar stylizing will come first anyway 22:10 kilbith * styling 22:10 v-rob Yes. I will review either PR if you make them. 22:11 kilbith maybe on Sunday if I have time 22:12 v-rob Cool. 22:12 kilbith I'll have to add more properties to the StyleSpec enum 22:13 v-rob Almost certainly. 22:13 kilbith (since a scrollbar is composed of 3 sub-elements) 22:13 kilbith but I already have the code to replace these parts by textures 22:15 v-rob Ideally, there would be something like the CSS pseudo-selector, but I'm not certain of the best way to implement that 22:15 kilbith yes but we are not here 22:16 kilbith I can imagine that the formspec replacement would grant so much abstraction that a concept of "scrollbar" wouldn't exist anymore 22:17 kilbith that we could replicate a "scrollbar" with distinct/dynamic elements 22:18 v-rob A independent scrollbar widget will exist, but it will certainly be made of sub-buttons that you can style individually. 22:19 v-rob Also, scrollbars will probably appear automatically if you make elements overflow the space in the parent element, so no unwieldy scroll_container 22:21 v-rob An independent scrollbar widget would have to exist because a server-side API + possible slow connections + dynamic elements that have to update immediately = a special element is necessary. And, it's convenient. But it must be easy to customize all conglomerate elements like scrollbars. 22:29 sfan5 v-rob: you're not the only person who hasn't/hadn't caught wind of the irc move yet, so wasn't meant just for you ;) 22:29 sfan5 btw, what is broken about setViewPort? 22:29 sfan5 it's sucessfully used for item rendering 22:30 kilbith_ ah, because it's unsuccessfully used on model[]? 22:31 v-rob Not quite 22:31 sfan5 model[] uses it too? didn't know 22:32 v-rob It's really setViewPort + setRenderTarget. setRenderTarget uses glViewport to set the viewport. So, the member variable ViewPort is not set since setViewPort was not used. Therefore, a big mess happens when trying to set the viewport and reset it while render targets are in the mix, which is quite possible (and common) with my drawing PR. 22:33 v-rob If you saw my video a while back about dynamic textures on nodes and remember the chat shrinking, that was one of the problems caused by this. 22:34 v-rob Using setViewPort instead of glViewport directly in Irrlicht causes other problems in Minetest, namely the whole screen shrinking after entering the game. 22:34 v-rob I can't find where that bug comes from. 22:36 sfan5 hm 22:36 sfan5 sounds like it needs some kind of stack to restore viewports when appropriate 22:36 v-rob So, the reason why item rendering works fine is because there are no render textures involved. However, I wouldn't be surprised if there are some 3D mode bugs caused by this 22:37 sfan5 I can think of one bug though: if you resize the window while node items show they will appear scaled to the entire screen for one frame