Time Nick Message 00:56 paramat hmmmm, i'll push #2640 soon, i removed checks for p and pmax for reasons given in the thread, back soon 00:56 ShadowBot https://github.com/minetest/minetest/issues/2640 -- DecoSchematic: Fix missing trees in rough terrain by paramat 01:41 paramat now pushing 2640 01:43 paramat git am 2640.patch 01:43 paramat oops heh 01:46 paramat done 01:48 hmmmm wait you removed the checks? 01:48 hmmmm hello cutoffs :| 01:49 paramat yeah i know 01:49 paramat sorry, you said the check for p was redundant 01:50 hmmmm yeah the check for the point, not the area 01:50 hmmmm contains(VoxelArea{pmin, pmax}) implies contains(p) 01:50 paramat ah, my original PR checked p and pmax 01:50 paramat i misunderstood 01:50 hmmmm well that's okay 01:51 paramat still, see reasons in thread, you agree? 01:51 hmmmm i'm going to modify a few other things in that file soon 01:51 hmmmm forum thread?? 01:51 paramat #2640 01:51 ShadowBot https://github.com/minetest/minetest/issues/2640 -- DecoSchematic: Fix missing trees in rough terrain by paramat 01:51 hmmmm i thought it just said that it was slow or something like that 01:52 paramat i think speed is more important than checks for cutoffs 01:53 hmmmm do you realize the difference in magnitude.. 01:53 paramat especially since cutoffs due to rotation are not checked for anyway 01:53 hmmmm do you mind if I add it back in when I add in the rotation check? 01:53 paramat yeah small 01:53 paramat no i don't mind =) 01:53 hmmmm ok 01:55 paramat so you will be checking for cutoffs due to rotation? impressive 01:56 hmmmm it's not hard 01:58 paramat okay. now, game#310 has approval from the team, is rebased and squashed, i'm considering pushing it for them since they're so busy ..? 01:58 ShadowBot https://github.com/minetest/minetest_game/issues/310 -- Allow overriding papyrus and cactus grow functions by HybridDog 01:58 paramat game dev work is so slow at the moment 01:59 paramat i need this pushed to fix my branched cacti 02:05 paramat well i'm gonna do it, i even asked nore yesterday and got no response 02:06 hmmmm fine by me, i mean i don't do anything with minetest_game 02:10 paramat oh hang on, some of ShadowNinja's line notes have not been acted on, so it needs improving before pushing 05:00 paramat to finish mgv6 snow biomes i will push this later game#497 05:00 ShadowBot https://github.com/minetest/minetest_game/issues/497 -- Default/mapgen: Add dry shrub to mgv6 tundra by paramat 05:04 hmmmm hmm 05:04 hmmmm i look at stuff like this http://i.imgur.com/HFEftHA.jpg and it makes me wonder if we've been doing it wrong 05:05 hmmmm should we be using 2d noise as the base terrain and then 3d noise for "detail"? 05:05 hmmmm although I suppose the current algorithms already do this to an extent 05:07 hmmmm how can the 2d noise shape be made more prevalent over the 3d 05:08 hmmmm what if the base terrain was pure 2d and then for terrain detail, the same 2d noise is multiplied by 3d noise to create the detail 05:09 hmmmm so it'd be like placing mapgen v5 on top of a mapgen v6 map 05:09 hmmmm sorta 05:09 hmmmm except the terrain would be synchronized 05:09 hmmmm i gotta try this out but i'm stuck doing unit test crap 05:16 * paramat ponders 05:27 paramat now pushing 497 to game 05:30 paramat done 05:34 paramat 'stuck'? you're not at work, do what you feel 05:35 hmmmm well 05:35 hmmmm it's a priority for me 05:35 paramat ah 05:36 hmmmm I hate work right now 05:36 hmmmm i feel like i'm getting micromanaged 05:36 paramat my river valleys use 2d valley noise to multiply 3d mountain noise, for a smooth blend from one to the other 05:43 paramat something like (1 + 2Dnoise) * 3Dnoise 05:44 hmmmm that's what mapgen v5 already does 05:44 hmmmm hmm what I'm talking about is hard to explain i think, i just need to show you 05:49 paramat yeah 14:14 neoascetic hi all 14:15 neoascetic github supports contributors notice: https://github.com/blog/1184-contributing-guidelines 14:16 neoascetic could someone add CONTRIBUTORS.{md,txt} to the repo with style guides/etc? 14:17 neoascetic I think it is better to have guidelines in the repo instead than somewhere on wiki 15:22 est31 sometimes, C is very weird 15:22 est31 I have these 4 files: https://github.com/est31/csrp-gmp/tree/master/sha 15:23 est31 where is the implementation for SHA256_Init?? 15:23 sfan5 remember how i said just ripping some code out of a big project is not a good idea 15:24 sfan5 this is exactly what happens 15:24 sfan5 you use some code 15:24 est31 it works very well 15:24 sfan5 then somehow something is missing 15:24 sfan5 or you don't know where it is 15:24 sfan5 or it breaks for some reason 15:24 sfan5 or .... 15:25 est31 I just wanted to add SHA-1 support so that I can test whether we are conformant with RFC 5054 15:25 est31 (in theory) 15:25 est31 I mean we dont need that conformance, but its a good way to check whether all computations are right 15:25 sfan5 you mean whether sha1.cpp is right? 15:25 Etzos Line 33-66. 15:26 est31 Etzos, which file? 15:26 Etzos sha256.c 15:26 est31 ah I see 15:26 est31 thanks 15:27 * est31 should read more carefully 15:27 sfan5 est31: you mean whether sha1.cpp is right? 15:27 est31 no that is ok 15:27 est31 I mean I trust that 15:27 est31 I only want to check the srp implementation 15:28 sfan5 sha1.cpp is correct btw 15:28 est31 and the test vectors in the RFC rely on sha-1 15:33 est31 I know of sha1.cpp 15:33 est31 but its c++ 15:33 est31 and the test file is c based 21:14 est31 sfan5, I think you have convinced me. I now think that libnettle is no bad option at all. 21:14 est31 Right now it might only be used for hashing, but later we can use it for encryption too. 21:15 est31 However, before its a back and forth, I suggest first running with the "ripped out" sha implementation, and then adding nettle when adding encryption.