Time Nick Message 00:00 RealBadAngel just tried 00:00 RealBadAngel i already implemented missing stuff as helper functions 00:00 hmmmm quaternions are rotation matricies 00:01 RealBadAngel found a function to make quaternion out from axis and angle 00:02 RealBadAngel but cant get rotation matrix from it 00:02 RealBadAngel which obiously is not the same 00:04 hmmmm oh that 00:05 RealBadAngel definitely is not the same, mainly because now the code works ;) 00:05 RealBadAngel when engine will move to irrlicht 1.8 helper functions can be just replaced by native ones 00:11 hmmmm you might want to read up on how quaternions work http://en.wikipedia.org/wiki/Quaternions_and_spatial_rotation 00:11 hmmmm it's always good to know 00:12 RealBadAngel thx i will 02:24 VanessaE jin 03:34 hmmmm ha ha... 03:34 hmmmm i just got mapgenv6 working again with the new fast perlin noise 03:34 hmmmm shit's just as slow 03:34 hmmmm man.. 03:36 hmmmm well i can't really make a fair comparison 03:39 hmmmm ah, i forgot i'm doing all this in -O1 03:50 VanessaE oops :-) 05:40 hmmmm https://github.com/kwolekr/minetest/commits/master 05:40 hmmmm ugh 05:40 hmmmm feels so hacky 05:40 hmmmm I even screwed up the push 05:40 hmmmm but i mean it's usable 05:40 hmmmm if anyone wants a super fast V6 mapgen 05:41 hmmmm set mgv6_use_smooth_biome_trans = false if you don't like the dithering thing 08:21 RealBadAngel hi 08:21 RealBadAngel im pretty done with treegen code but noticed same error i saw when was using LUA version 08:22 RealBadAngel something bout run out of vertices 08:25 RealBadAngel two error messages; 08:25 RealBadAngel Too many vertices for 16bit index type, render artifacts may occur. 08:27 RealBadAngel FIXME: Meshbuffer run out of indices. 11:36 jin_xi RealBadAngel: https://github.com/celeron55/minetest/blob/master/src/mapblock_mesh.cpp#L1299 14:57 thexyz celeron55: is it ok to push pull-request with commit in which author's e-mail is @example.com (similar for name) 15:00 celeron55 not really 15:01 celeron55 either figure out something reasonable or commit it as yours (doesn't matter which) 15:03 thexyz ok 15:35 thexyz celeron55: is "Author: Your Name " acceptable? 15:35 thexyz email is real one 15:44 celeron55 wut? 15:46 celeron55 what kind of bullshit is that again... is that iqualfragile's commit? just change it to iqualfragile then 15:47 iqualfragile uhm… sorry whats the problem now? 15:47 celeron55 https://github.com/iqualfragile/minetest_game/commit/2a97794b1807ca03b3c141d2ddbb8338df8619cd.patch 15:47 celeron55 look at the From field (which is the Author field of git) 15:47 iqualfragile gdamn, forgot to add my name… 15:55 iqualfragile ok, i have now changed that, how can i replace the commit? 15:56 VanessaE celeron55: I have been playing around with RealBadAngel's C++ implementation of that L-system trees code.. The trees are kinda plain, but nice, and reasonably fast too, but they expose a bug in the way you use mesh buffers. 16 bit index vs. 32 bit (the latter is needed). 15:57 VanessaE 16:00 celeron55 it's not a bug, it's a limitation that didn't limit anything until apparently now (...and with some of the craziest nodebox stuff) 16:00 VanessaE *nod* I saw it happen with Jeija's "irregular" mod. 16:02 thexyz VanessaE: >..and with some of the craziest nodebox stuff 16:02 VanessaE thexyz: yes, I was acknowledging what c55 said. 16:02 celeron55 anyway, the way to fix is not to increase the index size (i don't even know if irrlicht supports that) but rather to automatically create an another PreMeshBuffer when the indices in one run out, in MeshCollector::append 16:02 celeron55 anyone is free to implement that 16:03 VanessaE you're probably right about using more buffers, but from what I read here, 32-bit indexes are supported via something called a dynamic buffer. 16:03 VanessaE if that means anything :-) 16:04 celeron55 that means nothing for me at least, and even if it meant, it wouldn't be anything that should be done to fix this 16:04 VanessaE ok 16:04 VanessaE just thought I'd mention it.' 16:07 iqualfragile ok, finaly (i hope) https://github.com/celeron55/minetest_game/pull/75 16:12 iqualfragile yaay 16:13 VanessaE \o/ 22:08 thexyz celeron55: nice bug you have there https://github.com/celeron55/minetest/issues/325 22:09 celeron55 wat 22:09 thexyz KEY_SPACE has its Char 'E' 22:09 thexyz that's why it equals shift+e 22:10 thexyz i assume that this check should filter all KEY_KEY_c keys 22:10 celeron55 nice code by... umm... some guy who i don't even know by the real name 8) 22:10 celeron55 "Andreas Zwinkau" 22:10 thexyz oh, well 22:11 thexyz anyway, should i commit it? 22:11 celeron55 but quite an epic bug indeed 22:12 celeron55 i guess so