Time Nick Message 01:10 Calinou hence why the noise based system is novel and a decent idea 01:10 Calinou just saying, but Minetest had noise-based farmap previews back in 0.2-0.3.1 in mid-2011 :) 01:10 Calinou it worked decently, it displayed tree coverage but was fairly costly and didn't look voxel-like (it used flat-shaded polygons and billboard sprites for trees). celeron55 did another attempt in 2013-2014 where large voxels were rendered far away (say, 4×4×4) based on noise too. Original node textures were used, so it looked more accurate 01:14 Calinou you can test it by running https://www.moddb.com/games/minetest/downloads/minetest-031 (in WINE if you're on Linux) 01:14 Calinou had to dig a bit to find a Minetest 0.3.1 precompiled download, that seems to be legit 01:15 Calinou run the game once, exit it, add `enable_farmesh = true` to minetest.conf, then run it again 01:15 Calinou it looks like this: https://0x0.st/-Lw2.png 01:15 v-rob http://packages.8dromeda.net/minetest/minetest_old_releases_2010_through_2012_mostly_win32/ is a more official place to get old releases 01:15 Calinou ah, nice 01:18 Calinou and yeah, enabling farmesh increases drawtime from 6 to 30, even at small window sizes, so it's likely draw call-constrained 02:58 MTDiscord Oh interesting, I figured we would render with the billboards and 2d sprites and just haze-blur it 02:59 MTDiscord That's much more detail and yeah only something like Vulcan would be able to handle that many draw calls I suspect 04:23 kilbith scrollbar customization: https://github.com/minetest/minetest/pull/11345 05:03 kilbith v-rob: you there? 05:03 v-rob Yeah, but only for a little longer 05:04 kilbith okay, should scrollbaroptions[] be transfered to style[] and marked as deprecated? 05:05 kilbith that'd make sense for me 05:08 v-rob scrollbaroptions[] exists as a hack from before formspec versioning, since formspec elements could not have new arguments added to them. 05:08 kilbith so what's your take about it now? 05:09 kilbith well, these options aren't really about styling either... 05:10 v-rob I know at least rubenwardy thought that it shouldn't be part of style (for instance, https://github.com/minetest/minetest/pull/8530#issuecomment-531495972), and I agree since it really doesn't have much to do with style. If anywhere, I think it would be better to integrate it into scrollbar[] where it would properly belong (with or without named arguments, IDK) 05:10 kilbith except arrows= 05:10 v-rob Arrows are styling, definitely 05:10 kilbith so, transfer? 05:11 v-rob Transfer arrows to style, but nothing else I think 05:12 v-rob I guess thumbsize could theoretically be style 05:12 kilbith yep 05:12 v-rob And maybe the steps, but that's somewhat of a stretch 05:13 v-rob Min and max, certainly not in style[] 05:13 v-rob But yeah, for now deprecate the styling parts and move them to style. Other core devs can give their input when they're on 05:14 kilbith also, basically, arrow up/down should support all the properties of button[] ? 05:15 v-rob Yeah, that's where a pseudo-selector would come in handy. Such as `style_type[scrollbar::up_arrow:hover;fgimg=koolstof.png]` 05:15 v-rob It would probably be implemented similarly to states if anything 05:16 kilbith hmm, let's see 05:16 kilbith the real deal would be a new syntax for formspecs 05:17 v-rob Ugh, yeah. I'm really loathing Irrlicht's bug right now. I can't find where Minetest relies on the buggy behaviour. 05:18 v-rob Anyhow, I've got to leave now, sorry. It's very late here. 05:19 kilbith thanks for the feedbacks 16:37 MTDiscord Is the PsuedoRandom class we use in the engine faster than the PcgRandom class? I.e. does it generate numbers faster? 16:52 sfan5 it does less operations, so probably? 21:15 MTDiscord I tested on windows, 64 bit, -O3 release, got slower times with PCG so yep, psuedo is faster, all good