Time Nick Message 03:24 hmmmm not a very minetesty day 03:27 * est31 is trying to apply the RFC test vectors to the srp implementation 03:28 hmmmm sweet 03:28 est31 but struggling with c heredocs for binary data 03:28 est31 "heredocs" 03:28 est31 I declare static const srp_5054_salt[4] = { 0xBEB25379UL 03:28 est31 , [...] }; 03:29 est31 when I print it again, I get 7953B2B... 03:29 est31 which is weird 03:34 * est31 tries the shorter 0x variant 03:38 est31 now it works 04:54 est31 and now I can confirm it all works 05:01 hmmmm heh 05:01 hmmmm what was the problem with the constant? 05:08 est31 no idea 05:08 est31 using 0xBE, 0xB2, 0x53, works 05:08 est31 but not the UL method 05:10 hmmmm endianness? 05:12 est31 may be 05:14 est31 https://github.com/est31/csrp-gmp/commit/a65c9226d4002e53c8720ce39246b2b540526c62 05:15 hmmmm ah that's your own 05:15 est31 yup 05:15 hmmmm so you basically ended up forking csrp 05:15 est31 yes 05:15 est31 when you port it you have to 05:15 hmmmm i think the world needed this, after seeing the sheer lack of options 05:16 hmmmm just make sure you stay up on it :) this might become your 'thing' 05:16 est31 :) 05:16 hmmmm make sure you write unit tests for it too at some point 05:16 est31 that commit is basically adding unit tests 05:16 hmmmm oh derrp 05:17 hmmmm I'm going to slap that test into the new minetest unit test framework 05:17 est31 great 05:18 hmmmm there was a point at which celeron decided to say fuckit and not add any new unit tests and nobody else wrote them to begin with 05:18 est31 but not before #2620 gets merged that adds the file 05:18 ShadowBot https://github.com/minetest/minetest/issues/2620 -- SRP based login by est31 05:18 hmmmm yeah sure 05:18 hmmmm obviously 05:18 hmmmm :) 05:19 hmmmm btw 2620 is completely ready for review? it's not a WIP at all anymore? 05:19 est31 there are still points 05:19 hmmmm ahh okay 05:19 est31 see commit message https://github.com/est31/minetest/commit/bb074a1237caf4e97740f0f74e9c6ff506528c56 05:19 hmmmm keep up the good work though :) 06:03 OldCoder Feedback from a user: People may prefer main menu closing on double click 07:44 hmmmm well uh oh 07:45 hmmmm pretty nasty combination with unsigned script execution + UDP 07:46 hmmmm actually no nevermind, you don't need to hijack the session at all 07:46 hmmmm a man in the middle could easily modify the script being sent to the client to be executed 07:49 hmmmm i have concerns about media transfer being MITMed on joining a server. when client side scripting gets added, the man in the middle could trivially modify the script and possibly use it for bad things like javascript on websites already is 07:49 hmmmm like for example advertisement injection 07:51 hmmmm getting network encryption should be a priority after SRP is added 07:54 Calinou who is going to inject stuff just for ads? 07:55 hmmmm i guess minetest has a significantly smaller marketshare than javascript and http but eh :) 07:56 hmmmm what if some asshole MITMed traffic and changed all the textures being sent from the server to something like a giant cock 07:56 hmmmm that would be hilarious and frightening at the same time 07:57 jin_xi yay 09:06 Calinou yeah, pron/epilepsy pranks are more dangerous 09:06 Calinou I had that in mind too 09:06 Calinou but a server can do this already 09:06 Calinou use a server texture pack with inappropriate stuff 09:06 Calinou I'm surprised I haven't ever seen this in 3.5 years 13:36 nrzkt Hi 13:37 nrzkt i found an long time unused variable in client.h Client::m_active_blocks (this variable is in environment.h). I will remove it in the next 5 minutes 13:37 nrzkt hi Zeno` :) 13:37 Zeno` hoya 16:38 est31 Further thoughts on this? https://github.com/est31/minetest/commit/e298bb88f528904dc542cc62be92d52188333daf#commitcomment-10745294 16:39 est31 perhaps still group them but have a smaller distance? 17:06 est31 OMG chat is really sent over the network as wstring? 17:06 est31 where was that exploding shit video again? 17:07 Calinou we could encrypt chat :P 17:07 Calinou Xonotic did it, AES-encrypted game 17:07 Calinou but nearly no one uses it :P 17:07 Calinou not to mention the legal issues 17:07 est31 legal? 17:08 est31 is it illegal to encrypt? 17:08 Calinou in some countries, yes 17:08 est31 I live in central europe 17:09 est31 before they run a server people have to check whether its legal to run it or not 17:09 est31 not my fault if they still do if its illegal 17:09 est31 this is the no warranty part of the GPL 17:10 Calinou Fedora and such would probably disable support 17:10 Calinou but then you wouldn't be able to connect to servers which require encryption 17:10 est31 is https disabled on fedora too? 17:12 Calinou no 17:14 est31 so whats the issue then? 17:14 * Calinou shrugs 17:57 Warr1024 hey, did anyone ever merge #2589? 17:57 ShadowBot https://github.com/minetest/minetest/issues/2589 -- Fix fast leaves when texture_clean_transparent is enabled. by Warr1024 18:01 est31 Warr1024, sorry for not reviewing, I'll have a look after I have read through my srp patch 18:06 Warr1024 thanks 18:21 est31 Warr1024, what are fast leaves? 18:26 est31 pushing in 15 minutes: https://github.com/est31/minetest/commit/e3764410d3934726c47b8a5b2b347678c039a0e4 18:27 sfan5 est31: looks good 18:27 est31 problem was: obstructed view 18:28 est31 pushed 18:36 est31 Warr1024, I can't reproduce the bug, could you post screenshots or how to reproduce? 18:39 Warr1024 est: if you have texture_clean_transparent = true and fast leaves, you'll see all-green trees instead of the intended green leaves with black gaps between 18:39 Warr1024 with the patch, it should display as intended 18:39 Warr1024 there shoudl be a bug report somewhere... 18:39 est31 Warr1024, fast leaves is when you don't check the "fancy leaves" checkbox? 18:39 Calinou yeah 18:40 Warr1024 yes 18:40 est31 Thats what I tried, and I can't reproduce 18:40 Warr1024 ... I can't find the bug report... 18:42 Warr1024 guess I'll have to build the unpatched client and try to repro 18:42 Warr1024 I wonder if somebody fixed this already and didn't close teh pull? 18:42 est31 possible 18:43 est31 but doesnt look like that 18:47 kilbith est31, i can confirm that but that "blend of leaves" looks better than the leaves patterns + black background 18:47 kilbith i'd say it's a positive bug 18:48 Warr1024 it's actually a bug also because of the potential to affect other non-leaves textures unexpected ways 18:48 kilbith right 18:50 Warr1024 posted the screenes 18:51 Warr1024 the problem is that the clean_transparent filter is cleaning up the transparent areas BEFORE they're flattened by [noalpha 18:51 Warr1024 the patch corrects the order 18:56 est31 ok 19:03 est31 I still cant reproduce 19:03 est31 weird 19:08 est31 Its like it works 19:08 est31 very weir 19:08 est31 d 19:11 est31 ok now I can reproduce 19:14 est31 +1ed 19:14 est31 bye 23:11 VanessaE is there a way to turn on backface culling for a mesh node?