17:51 < hansihe> after that, skylight behaves like normal light, it spreads through transparent blocks and decrements by one for each 17:52 < hansihe> some blocks have a higher opaqueness value, like ice 17:52 < morfin> because on the bottom of my well i got 0 17:52 < morfin> :) 17:52 < hansihe> they decrement the light by more then 1, in the case of ice it's 3 17:52 < hansihe> yep :) 17:54 < hansihe> leaves are opaque to the initial skylight ray, just like ice 17:54 < hansihe> but it decrements light by one for every block, just like normal air 17:55 < hansihe> glass is transparent to the skylight ray, and decrements light by one per block, just like air 18:10 < morfin> F3 provides more info than in earlier versions 18:12 < morfin> you can even see block info you're looking at + variant info 18:12 < Gjum> in snapshots it provides even more 18:12 < morfin> oO 18:12 < Gjum> *than in releases 18:12 < morfin> i thought it's impossible to show more info than now 18:13 < morfin> would you recommend looking in MCP code for understanding some thing? 18:13 < morfin> *things 18:14 < hansihe> better than directly decompiling 18:14 < hansihe> although a lot of things are not deobfuscated, most high level stuff it 18:14 < hansihe> it certainly helps 18:14 < morfin> hmm 18:15 < morfin> moon gives me skylight 15 18:16 < Gjum> morfin: skylight doesnt change, it just gets weighted differently 18:16 < morfin> oooh 18:19 < morfin> i thought it changes 18:20 < hansihe> having a distinction between skylight and block light is a very elegant solution 18:20 < morfin> why 18:20 < hansihe> it makes it possible to have smooth day/night transitions without recalculating the light for the entire world 18:21 < morfin> ah yes probably 18:22 < edk> i don't know about elegant 18:22 < edk> it strikes me more as a useful hack 18:23 < hansihe> i would still say the system itself is elegant 18:39 <+Amaranth> beta 1.7 and older recalculated the skylight 18:39 <+Amaranth> You could see the transition as it iterated through every chunk, almost made it look like a dayline passing you 18:40 <+Amaranth> It also meant during dusk and dawn the server was basically dead 18:59 < morfin> oO 19:00 < morfin> beta 1.7* 19:00 < morfin> it was so long time agi 19:00 < morfin> *ago 19:08 < ecx86> oh 19:08 < ecx86> yeah I remember those days 19:16 <+SinZ> last good patch 20:02 < morfin> why so? 20:02 < morfin> i remember epic 1.7.2 pistons block dupe 20:03 < oldmanmike> Why do Minecraft servers send an explosion packet on login? 20:05 < oldmanmike> Or at least a packet with ID 0x1b that only has a payload 5 bytes long 20:08 < hansihe> oldmanmike: what stage of login? 20:09 < hansihe> also, 0x1b wouldn't be explosion 20:10 < hansihe> 1.8, right? 20:10 < oldmanmike> On 16w05b, it comes right after the Held Item Slot packet is sent and before statistics, player info, or any chunk data is sent 20:10 < hansihe> ah, right, prerelease 20:10 < oldmanmike> So yeah, I'm targeting 1.9 on this one 20:10 < rom1504> oldmanmike: you know that packets ids get changed at almost every 1.9 snapshot ? 20:10 < rom1504> ah 20:11 < rom1504> what there is an explosion packet ?? 20:11 < rom1504> ah there is, fun 20:11 < oldmanmike> The packet I'm looking at explicitly is: 06 1b 00 00 17 dd 18 20:12 < rom1504> idk why they don't use multiblock change 20:12 < rom1504> just for the sake of create more and more packets I guess 20:12 < rom1504> *creating 20:12 < hansihe> is it not for the explosion effect? 20:13 < hansihe> oh, there is block data too 20:13 < oldmanmike> 0x1b is for explosion in the snapshots, but explosion should not be 5 bytes long 20:13 < oldmanmike> at least according to my understanding 20:13 < hansihe> it's probably not an explosion packet 20:16 < oldmanmike> Well, it's client bound and taking place in the Play state, so I don't know what else it could be 20:17 < hansihe> how did you determine that it's an explosion packet? 20:17 < hansihe> if you are using the prerelase protocol page, it's not for 16w05b 20:18 < hansihe> so the packet ids might have changed 20:18 < rom1504> "0x1b is for explosion in the snapshots" not "in the snapshots" 20:18 < rom1504> in a snapshot maybe 20:18 < oldmanmike> This packet has existed in my wireshark captures for a while now 20:18 < oldmanmike> and I've put off dealing with its enigma up until now 20:19 < oldmanmike> but yeah, I'm just searching "0x1b" in the protocol.json file in the minecraft-data repo 20:19 < oldmanmike> and I've found it fully accurate thus far for the rest of the protocol 20:21 < hansihe> look at the packets that's 1/2 above and below the explosion packet 20:21 < hansihe> see if you find one that fits 20:21 < hansihe> 1 or 2 rather 20:22 < rom1504> I last updated 1.9 protocol.json in 15w40b 20:23 < rom1504> oldmanmike: http://wiki.vg/Pre-release_protocol explosion is 0x1C now 20:23 < oldmanmike> ah, entity status 20:23 < hansihe> that's not even fully up to date 20:24 < hansihe> the page is for 16w04a 20:35 < oldmanmike> rom1504: Could I update the protocol.json to the latest snapshot? 20:52 < Not-84da> [Charge] Wallbraker pushed 9 commits to master [+2/-0/±22] https://github.com/VoltLang/Charge/compare/601a4a760cf5...c57082aeea3c 20:52 < Not-84da> [Charge] Wallbraker ee24ef7 - ctl: Fix unicode encode 20:52 < Not-84da> [Charge] Wallbraker 97ed79f - ctl: Update Ctl 20:52 < Not-84da> [Charge] Wallbraker 9d64bcd - charge: More braces 20:52 < Not-84da> [Charge] ... and 6 more commits. 21:10 < pokechu22> Is it just me, or does MC|Beacon not get validated on the server side? It doesn't look like it is, but invalid changes seem to still be ignored... I tried to make a Nausea beacon... 21:11 < pokechu22> Because if they aren't, that's a bug that should be reported. But I can't quite tell if it's actually validating them somewhere and I can't find it. 21:12 < __0x277F> Someone screwed up the linking for serverbound vehicle move and steer boat packets on the prerelease wiki page. 21:12 < __0x277F> I'd fix it but I know pretty close to nothing about using mediawiki. 21:14 < pokechu22> Vehicle move (serverbound) seems to work, but it looks like the steer boat one just doesn't have a section it's linking to... 21:17 < __0x277F> Odd. Are we talking about the same link? https://gyazo.com/c5c8acb91a014004064c58ce1e2bd468 21:18 < pokechu22> Yea. Does this link work for you? http://wiki.vg/Pre-release_protocol#Vehicle_Move_.28Serverbound.29 21:18 < __0x277F> Nope. 21:18 < __0x277F> I bet it's because I did some dubious formatting in the link when I added it. :P 21:18 < __0x277F> I couldn't figure out how to make it both green and not link to clientbound vehicle move. 21:20 < pokechu22> Hold on... the other link in teh "true" table of contents is slightly different: http://wiki.vg/Pre-release_protocol#Vehicle_Move_.28serverbound.29 21:20 < pokechu22> Note the capitalization of 'serverbound'. 21:20 < pokechu22> Does that link work? 21:21 < __0x277F> Oooh, yes. 21:21 < __0x277F> Thanks. 21:22 < __0x277F> Submitted the edit. 21:23 < pokechu22> Ok... blindness beacons actually do something really interesting, on a random note. You can see through the beam since render orders are always correct(/s), even though you should be blind. 21:24 < __0x277F> Neat-o. 22:54 < rom1504> oldmanmike: yes sure, please do ! 22:55 < rom1504> the only reason why it wasn't updated is I didn't have the time to do it 22:58 < oldmanmike> cool cool, I'll send a PR sometime soon 23:12 < hansihe> does any custom server have lighting implemented? 23:12 < hansihe> all the ones i have looked at thus far it looks like they don't 23:12 < Gjum> well it would be just for mob spawning right? 23:13 < Gjum> and none of the ones I have looked at closely do that 23:13 < hansihe> well, it gets sent to the client 23:14 < hansihe> unless there is a way to avoid that 23:17 < rom1504> it gets sent to the client but the client recomputes it anyway 23:17 < rom1504> so you can just send 15 and it looks okay 23:17 < Gjum> super bright caves! yay! 23:17 < hansihe> well, it does recompute it, but only if the client gets really close 23:18 < rom1504> right, might have to do it then :p 23:18 < pokechu22> And even then sometimes it doesn't want to do it... 23:18 < hansihe> i did the basic skylight raycast 23:18 < hansihe> but without light spread 23:18 < rom1504> cuberite doesn't handle that hansihe ? 23:19 < hansihe> might do, c++ is not the language i'm best at reading though, mostly looked at java servers 23:19 < hansihe> can have a look 23:20 < hansihe> it does 23:21 < rom1504> https://github.com/cuberite/cuberite/blob/6fdd7194c81be7234a126bdc3b48f0291fce3567/src/LightingThread.cpp 23:21 < hansihe> yeah 23:23 < Gjum> they use it for endermen https://github.com/cuberite/cuberite/blob/7767acdc4bdb4bcebcc2ae3fbf84b537d1016cd1/src/Mobs/Enderman.cpp#L177 23:25 < hansihe> this is what happens when you don't lighting calculations https://usercontent.irccloud-cdn.com/file/coV7NkA0/2016-02-07_23.24.43.png 23:26 < hansihe> notice the water as well 23:27 < rom1504> yeah 23:33 < Gjum> uh, cuberite's mob spawning doesnt seem to take light level into account...? https://github.com/cuberite/cuberite/blob/master/src/World.cpp#L1046 23:34 < Gjum> or is that done by a plugin not in the main repo? 23:44 < hansihe> i don't know if i like the way cuberite does lighting calculations 23:45 < hansihe> they light 3x3 chunks at a time, then use the light values for the center one 23:46 < hansihe> i mean, it does make the code simple 23:46 < hansihe> it does seem very heavy --- Day changed lun. févr. 08 2016 09:01 < Not-84da> [Charge] Wallbraker pushed 1 commit to master [+0/-0/±1] https://github.com/VoltLang/Charge/compare/c57082aeea3c...b0830be88fad 09:01 < Not-84da> [Charge] Wallbraker b0830be - core: Remove workarounds 13:17 < rom1504> hansihe: so, does your server scale to millions of players yet :p ? 13:18 < rom1504> run some https://github.com/PrismarineJS/mineflayer/blob/master/examples/multiple.js on it 13:19 < rom1504> someone convinced me erlang/elixir is really the thing to use to scale well, now I need proof ! :D 13:24 < hansihe> rom1504: haven't benched it, nor optimized anything 13:24 < hansihe> can try it tho, would be interesting 13:27 < rom1504> last time I tried flying-squid was okay with about 100 bots, after that it started being too slow, not sure why exactly, didn't exactly optimize it either 13:27 < rom1504> it means using offline mode of course 13:27 < hansihe> yeah 13:29 < hansihe> in my servers every player runs in its own thread, right now things like the chunk manager could be a bottleneck though 13:29 < hansihe> server* 13:31 < hansihe> rom1504: what machine did you run it on? 13:34 < rom1504> hmm, I think it was all on my laptop. So it's a decent machine, but not a big server or anything 13:34 < rom1504> also of course flying-squid isn't feature complete 13:34 < hansihe> oh, yeah, neither is mine 13:34 < rom1504> so it's not directly comparable with vanilla or more feature complete servers 13:34 < rom1504> but it's interesting anyway 13:35 < hansihe> yep, i would imagine mobs are fairly heavy 13:36 < hansihe> what is different between offline and online mode login? 13:36 < hansihe> can't find any info on the wiki 13:37 < barneygale> hansihe, offline mode servers send "login success" immediately after the client sends "login start" 13:37 < rom1504> hansihe: http://wiki.vg/Protocol#Login 13:37 < hansihe> barneygale: no encryption then? 13:38 < rom1504> but yeah what barneygale said 13:38 < barneygale> hansihe, correct 13:38 < hansihe> right 13:38 < rom1504> the login sequences is described partially in like 4 pages in the wiki 13:38 < rom1504> would be smart to put it all in one place 13:38 < barneygale> agreed 13:39 < hansihe> there isn't really any reference to offline mode there either 13:39 < rom1504> "For unauthenticated and localhost connections" hansihe 13:39 < rom1504> unauthenticated = offline mode 13:39 < rom1504> but yeah actually writing offline mode might be a good idea 13:39 < hansihe> oh, i'm just dumb 13:40 < hansihe> i ctrl+fed offline 14:27 < hansihe> with 100 players joined it drops to 4% cpu 14:28 < hansihe> node is at 80% :P 14:29 < hansihe> does it do anything when the players have joined, or do they just idle? rom1504 14:31 < hansihe> when joining node runs at 100%, beam runs at 200% 14:31 < hansihe> the machine is 2 core i5 with ht 14:33 < rom1504> you mean mineflayer ? 14:34 < hansihe> by node i mean mineflayer, yeah 14:34 < rom1504> it parses the chunk and all 14:34 < rom1504> but it doesn't do anything no 14:34 < rom1504> well multiple.js doesn't do anything anyway 14:34 < rom1504> it just stays connected 14:34 < hansihe> right 14:34 < rom1504> parsing is probably what takes cpu 14:35 < hansihe> yeah 14:35 < rom1504> it keeps using cpu even when initial chunk loading is done ? 14:35 < rom1504> ah 14:35 < hansihe> yeah 14:36 < hansihe> keeps running at 100% for me at least 14:36 < rom1504> it's going to send the position every 50ms too 14:36 < hansihe> will try two instances with 100 players each 14:36 < rom1504> and mineflayer wasn't really optimized for multiple bots, so it's going to do 100 setInterval 14:37 < rom1504> idk if it would work better with multiple instances 14:37 < rom1504> maybe yeah 14:37 < rom1504> * 14:37 < rom1504> multiple node instances I mean 14:40 < hansihe> yeah, two node instances with 50 players each is much better 14:40 < hansihe> 5% cpu on each node instance, still <4% cpu on beam 14:41 < hansihe> idk why the node instance would use 100% cpu with 100 players when it uses 5% with 50 players 14:43 < rom1504> I think it's that setInterval thing, for some reason doing too many of them is not a good idea 14:43 < rom1504> I'm not sure why 14:43 < hansihe> probably 14:44 < hansihe> is there any reason why the node processes would stop after a while? 14:44 < hansihe> no message 14:45 < rom1504> add https://github.com/PrismarineJS/mineflayer/blob/master/examples/chatterbox.js#L116 in https://github.com/PrismarineJS/mineflayer/blob/master/examples/multiple.js#L26 14:45 < rom1504> that should get you some error message 14:45 < rom1504> but I bet the problem is keep alive aren't parsed fast enough 14:45 < hansihe> that might be it 14:47 < rom1504> might want to try setting checkTimeoutInterval option to more than the default value (10000) 14:48 < hansihe> no response, but when i think about it i don't send a kick packet on timeout 14:48 < hansihe> i just close the connection 14:49 < morfin> btw rendering code looked more cmplicated for me 14:49 < rom1504> https://github.com/PrismarineJS/node-minecraft-protocol/blob/master/src/server.js#L24 yeah you're supposed to do that when kicking the client hansihe 14:50 < hansihe> yeah, i remembered 14:50 < hansihe> i'll do that then 14:50 < morfin> when you kick client you just send kick packet with message 14:53 < hansihe> i can close the connection after sending the kick packet, right? 14:53 < hansihe> i don't want to leave it open 14:54 < hansihe> nevermind me 14:54 < hansihe> i read the wiki 14:54 < morfin> yes but why you have 2 different things: kick_disconnect and disconnect? 14:55 < hansihe> one for login protocol state, one for play protocol state 14:55 < morfin> oh i forgot 14:55 < morfin> different states => different packets 14:56 < morfin> i think i'll need to use state pattern to handle that :) 19:44 < rom1504> hansihe: oh and listen on 'error' too if you want to figure out why the bot disconnects 19:45 < rom1504> well if that was an error event killing the bot it should print it in the console anyway 19:47 < hansihe> it's weird, because it's not a timeout 19:47 < hansihe> i'll try to listen on error 19:49 < hansihe> still no error 19:54 < rom1504> does it happen with less bots too ? (say 10) 19:54 < hansihe> i'll try it 19:55 < hansihe> yep 19:56 < hansihe> i'll try to figure out what causes it 19:57 < rom1504> yeah it sounds like it might be a problem with your server, I'm pretty sure it was working fine with 10 bots with flying-squid 19:58 < rom1504> I think I even tried with vanilla with 10 bots 19:58 < hansihe> it probably is 19:59 < hansihe> when i have a mc client logged in at the same time, everything is perfectly responsive though 20:02 < hansihe> looked at all packets coming in and out 20:02 < hansihe> can't see any pattern to it 20:05 < hansihe> rom1504: is there any reason why it would fail when receiving a keepalive with a high nonce? 20:08 < hansihe> my varint implementation might be fucked 20:11 < hansihe> nope, that's not it 20:17 < hansihe> the connection does get closed by the client 20:17 < hansihe> it's not right after i send a packet either 20:18 < rom1504> nonce ? 20:18 < rom1504> you mean the id ? 20:19 < rom1504> can't see why, but if it did it would print an error 20:19 < hansihe> yeah, i mean the id 20:20 < hansihe> i tried with ids in to 0-16 range only, still the same issue 20:20 < hansihe> that should rule out problems with the varint implementation 20:21 < rom1504> it's pretty weird that the client would just disconnect hmm 20:22 < rom1504> you're using last commit of mineflayer ? 20:22 < hansihe> used the version from npm 20:22 < hansihe> can try the repo 20:24 < rom1504> yeah try it, we added that not long ago https://github.com/PrismarineJS/mineflayer/commit/ee0139369d8edd2d86da959859c95308b5721fb8 20:24 < rom1504> it's relevant if you kick the client before he is in play state 20:24 < rom1504> but you probably checked you are not sending a kick packet already right ? 20:25 < hansihe> yeah, looked at all packets to and from the client 20:25 < hansihe> there is nothing abnormal before the disconnect 20:26 < rom1504> what if you add bot._client.on('keep_alive',function(packet){console.log(packet);}) ? 20:27 < rom1504> just to make sure the bots are indeed getting the keep alive packets 20:27 < hansihe> i am getting a response on the keepalive packets immediately, don't think that's the problem 20:28 < hansihe> it seems like the problem is when the client tries to send/server tries to receive 20:28 < rom1504> how often do you send keep alives ? 20:28 < hansihe> on my end i'm just getting a connection closed 20:28 < hansihe> every 10 seconds 20:28 < rom1504> ah 20:28 < rom1504> that's the problem 20:28 < rom1504> nmp client expect to receive a keep alive every 10s 20:28 < hansihe> ah 20:29 < rom1504> (by default) 20:29 < rom1504> so if you send it at 10050 it will disconnect 20:29 < rom1504> nmp server send a keep alive every 4s 20:30 < hansihe> ahhh 20:30 < rom1504> (with a maximum answer time of 10s) 20:30 < hansihe> i can reduce it to 4s 20:30 < hansihe> the timeout for vanilla mc is 30 sec though i think 20:31 < rom1504> okay, we should probably increase the default to that in nmp then 20:31 < hansihe> client dcs if there is over 20sec since a keepalive 20:32 < hansihe> server kicks if there is over 30sec since a response 20:32 < hansihe> an yeah, it does seem to work now 20:32 < hansihe> thanks for the help figuring it out :) 20:32 < rom1504> okay we'll probably change the timeout to vanilla then 20:33 < rom1504> *vanilla times 20:34 < rom1504> so now you can do 10*50 instances :p 20:34 < rom1504> well I bet node would start to take some ram though 20:35 < hansihe> i'll run the clients on my desktop 20:46 < hansihe> 151 players, 5% cpu 20:46 < hansihe> i feel like this is a bit of a cheap benchmark though, they are just standing there 20:57 < rom1504> yeah you could make them move I guess, something fun is putting https://github.com/andrewrk/mineflayer-navigate#usage in the createBot function 20:57 < rom1504> then you say come 20:57 < rom1504> 150 bot coming toward you 20:58 < rom1504> moving could probably test sending chunk and all 20:59 < rom1504> could also just make them move forward by doing bot.setControlState('forward', true); 21:01 < hansihe> sounds horrifying 21:20 < hansihe> might try it later 22:03 < Not-84da> [mineflayer] rom1504 pushed 2 commits to master [+0/-0/±2] https://github.com/PrismarineJS/mineflayer/compare/66273c50a387...fa392864a9cd 22:03 < Not-84da> [mineflayer] G07cha ab89946 - Add auth-auth plugin Small plugin that allows to login/register as bot in servers that using chat authentication system. 22:03 < Not-84da> [mineflayer] rom1504 fa39286 - Merge pull request #368 from G07cha/patch-1 Add auth-auth plugin --- Day changed mar. févr. 09 2016 10:57 < Fenhl> __0x277F: I changed the (Serverbound) to (serverbound), was planning to update the TOC as well but then the wiki died, thanks for updating. Will add (serverbound) and (clientbound) to packets with the same name on stable [[Protocol]] page as well 17:06 < morfin> seems like leaves can be decayable and non-decayable 17:07 < morfin> so user-placed are not 17:12 < morfin> lol 17:13 < morfin> http://imgur.com/ZFFFct1 17:14 < Gjum> morfin: https://www.youtube.com/watch?v=t4rCm92hWHA 20:05 < morfin> that's what i said - used-placed blocks do not ever decay 20:06 <+SinZ> they used to, forgot which update changed it 20:07 < morfin> they used to decay? why? 20:07 < morfin> that breaks all fun 20:54 <+SinZ> because it never could tell the difference between user placed ones and generated 20:55 < hansihe> http://minecraft.gamepedia.com/Leaves#Data_values 20:55 < hansihe> it's all described there --- Day changed mer. févr. 10 2016 08:25 < Not-84da> [Charge] Wallbraker pushed 1 commit to master [+0/-0/±2] https://github.com/VoltLang/Charge/compare/b0830be88fad...7c489d6da271 08:25 < Not-84da> [Charge] Wallbraker 7c489d6 - core: Add option to disable window decorations 11:10 < morfin> sooo 11:10 < rom15043> So 11:11 < morfin> leaves without decay have different data-value from decayable 11:11 < morfin> but what does mean check-decay? 11:14 < morfin> why there is 4 subtypes - 0-3(default?) 4-7(no-decay) 8-11(check-decay) and 12-15(no decay and check decay) 11:14 < rom15043> Ctrl+f check 11:16 < morfin> oh 11:17 < rom15043> And there are several "subtypes " because there are different kind of leaves 11:18 < morfin> in thery i can place decayable blocks by command block? 11:18 < morfin> *theory 11:25 < morfin> seems like default behaviour of leaves depends on 0x8 and 0x4, if 0x4 set then i do not ever check for decaying, if not set then i check and if failed then reverse it but when block adjacent to leaves change i set it to true again 11:27 < rom1504> morfin: http://minecraft.gamepedia.com/Commands#Summary_of_commands /setblock exists, so yes 11:27 < morfin> i forgt i can set datavalue 11:27 < morfin> with it 11:31 < morfin> but how it checks adjacent to leaves blocks? as example i have massive thing made of leaves which should decay and 1 block of wood which is adjacent to 1 block of leaves - does that mean whole thing would not decay? 11:32 < rom1504> "Leaves from trees spontaneously decay (disappear) when they receive a block tick, if they are not connected to any wood block, either directly or via other leaf block, with a maximum distance of 4 blocks. " 11:33 < morfin> hm i did not see that 11:33 < morfin> i see now 11:35 < morfin> that makes impssible to grow anomalious trees with huge trunk and massive branches with leaves 11:37 < rom1504> except if you use no decay leaves 11:37 < rom1504> well 11:37 < rom1504> it wouldn't "grow" anyway 11:37 < rom1504> except if you're using a plugin or custom server 11:37 < rom1504> and if you're doing that, then you control de decaying process 11:37 < rom1504> *the 16:46 < rollappuser> hi 16:46 < rollappuser> xhreow 16:47 < rollappuser> GUYS} 16:52 < barneygale> wat 16:53 < rom1504> rollappuser: you can do it ! 17:01 < hansihe> can't be done :( 17:40 < morfin> impossible 19:06 < morfin> seems like Minecraft have lots of state-related things 19:09 < morfin> i thought monsters with AI disabled(zombies, skeletons) does not burn on sub 19:09 < morfin> *sun 19:24 < Akaibu> morfin: thats not really AI, thats more like properties 19:25 < morfin> can you explain why when i used this 19:26 < morfin> this command /setblock -316 38 122 leaves 8 0 replace it still does not decay even when decayable and check_decay are true 19:27 < morfin> oh it decayed 19:28 < morfin> adjacent block update triggered decaying --- Day changed jeu. févr. 11 2016 07:55 < morfin> why glowing tag does not work for mobs? 09:40 < rom1504> why do you expect consistency ? 09:53 < morfin> i thought it will make mobs glowing 09:55 < morfin> it seems to be used not for enabling/disabling glowing but for some technical purposes 11:15 < morfin> i guess blazers and other glowing entities have it 1 11:15 < morfin> *blazes 19:00 < morfin> and entity can have Any tag right? 19:00 < morfin> i mean i can set non-existant tag for some purposes 19:06 < pokechu22> I... think so. 19:06 < pokechu22> Let me check. 19:06 < morfin> why selector @e[customtag=value] ignored(it kills everybody)? 19:06 < morfin> i tried to kill only spawned by me zombie with custom tag and value 19:07 < pokechu22> @e likes killing everybody if you make the slightest typo :/ 19:07 < morfin> it killed even me(even when i was in creative mode) lol 19:08 < pokechu22> I don't think arbitrary arguments actually do anything. It's not related to stuff set via /entitydata. 19:09 < pokechu22> The only ones that can be configured are the scoreboard-based ones. 19:10 < pokechu22> http://minecraft.gamepedia.com/Commands#Target_selector_arguments 19:10 < morfin> crap 19:10 < pokechu22> I do agree, it's silly that invalid ones are ignored rather than treated as false :/ 19:11 < morfin> too bad they're silently ignored 19:11 < morfin> so game thinks - let's kill everybody --- Day changed ven. févr. 12 2016 04:09 < pokechu22_> ... hold on 04:09 < pokechu22_> Oh, NVM. It changed my name automatically? 04:09 < pokechu22_> Let's see... I think you do it like this? 13:09 < Not-1f77> [mineflayer] rom1504 pushed 3 commits to master [+0/-0/±4] https://github.com/PrismarineJS/mineflayer/compare/fa392864a9cd...6c9d2bd0c58d 13:09 < Not-1f77> [mineflayer] rom1504 6c9d2bd - Merge pull request #370 from PrismarineJS/greenkeeper-update-all Update all dependencies 🌴 13:09 < Not-1f77> [mineflayer] rom1504 deleted branch greenkeeper-update-all 14:09 < Not-1733> [flying-squid] rom1504 pushed 5 commits to master [+0/-0/±12] https://github.com/PrismarineJS/flying-squid/compare/ef50d796d609...f50341185805 14:09 < Not-1733> [flying-squid] rom1504 f503411 - Merge pull request #188 from PrismarineJS/greenkeeper-update-all Update all dependencies 🌴 14:09 < Not-1733> [flying-squid] rom1504 deleted branch greenkeeper-update-all 15:01 < Not-1733> [flying-squid] rom1504 pushed 3 commits to master [+0/-0/±3] https://github.com/PrismarineJS/flying-squid/compare/f50341185805...7d876b931280 15:01 < Not-1733> [flying-squid] rom1504 7d876b9 - Merge pull request #189 from PrismarineJS/greenkeeper-minecraft-protocol-0.17.0 Update minecraft-protocol to version 0.17.0 🚀 15:01 < Not-1733> [flying-squid] rom1504 deleted branch greenkeeper-minecraft-protocol-0.17.0 15:25 < Not-1733> [flying-squid] rom1504 pushed 1 commit to master [+0/-0/±1] https://github.com/PrismarineJS/flying-squid/compare/7d876b931280...ff0e7e778610 15:25 < Not-1733> [flying-squid] rom1504 ff0e7e7 - update nmp 15:27 < Not-1f77> [mineflayer] rom1504 deleted branch greenkeeper-minecraft-protocol-0.17.1 15:30 < Not-1f77> [mineflayer] rom1504 pushed 4 commits to master [+0/-0/±6] https://github.com/PrismarineJS/mineflayer/compare/6c9d2bd0c58d...65850990f480 15:30 < Not-1f77> [mineflayer] greenkeeperio-bot d8be7d3 - chore(package): update minecraft-protocol to version 0.17.0 http://greenkeeper.io/ 15:30 < Not-1f77> [mineflayer] rom1504 fe4ece6 - require node >=4 15:30 < Not-1f77> [mineflayer] rom1504 807b6b2 - update minecraft-protocol to 0.17.2 15:30 < Not-1f77> [mineflayer] rom1504 6585099 - Merge pull request #371 from PrismarineJS/greenkeeper-minecraft-protocol-0.17.0 Update minecraft-protocol to version 0.17.0 🚀 15:30 < Not-1f77> [mineflayer] rom1504 deleted branch greenkeeper-minecraft-protocol-0.17.0 15:30 < Not-1f77> [mineflayer] rom1504 deleted branch greenkeeper-minecraft-protocol-0.17.2 15:36 < Not-1f77> [mineflayer] rom1504 pushed 1 commit to master [+0/-0/±2] https://github.com/PrismarineJS/mineflayer/compare/65850990f480...8d05fae89006 15:36 < Not-1f77> [mineflayer] rom1504 8d05fae - Release 1.7.0 15:37 < Not-1f77> [mineflayer] rom1504 tagged 8d05fae as 1.7.0 https://github.com/PrismarineJS/mineflayer/commit/8d05fae89006ba3f09c08e9d592060076508b649 15:38 < Not-1f77> [mineflayer] rom1504 pushed 1 commit to master [+0/-0/±1] https://github.com/PrismarineJS/mineflayer/compare/8d05fae89006...a9d80c4d2935 15:38 < Not-1f77> [mineflayer] rom1504 a9d80c4 - fix gitter badge for npm 15:44 < Not-1733> [flying-squid] rom1504 deleted branch greenkeeper-mineflayer-1.7.0 15:45 < rom1504> bots doing my job on github ;) 15:46 < rom1504> how soon until I can ask a bot "make a PR to fix issue number X please" 15:49 <+Wallbraker> Hehe nice 15:49 <+Wallbraker> How well does mineflayer work? 15:53 < rom1504> Wallbraker: it supports most basic minecraft feature with a fairly high level api 15:53 < rom1504> moving, inventory, world update,... 15:54 < rom1504> so it works pretty well 15:54 < rom1504> someday I plan on making an even higher level api on top of it though, but that requires some thougths and time 15:55 < rom1504> I started doing some of that in https://github.com/rom1504/rbot 15:55 < rom1504> some years ago 15:55 < rom1504> high level tasks like "get block X" or "protect that player" 15:55 < rom1504> it kind of work but it's not general enough 15:55 < rom1504> there are a lot of interesting stuff to do on that topic anyway 15:56 <+Wallbraker> Ah cool cool 15:56 <+Wallbraker> A long time ago I used mainflayer to connect to minecraft with a graphical client. 15:57 < rom1504> deathcap is working on that https://github.com/voxel/voxel-clientmc 15:57 < rom1504> that's pretty cool too 15:58 < rom1504> doesn't have a lot of features, but it's a fun experiment 16:04 < Gjum> rom1504: often when I have time, I look at other bot frameworks and see what tasks they implement and what system they use 16:05 < Gjum> this recent one is fairly advanced and has some aspects I would like for such a system https://github.com/michaelzangl/minebot 16:07 <+Wallbraker> rom1504: Ah cool 16:08 < rom1504> " It simulates keyboard and mouse input to interact with the word." oh wow 16:08 < rom1504> is that true ? 16:09 < rom1504> hmm no 16:09 < rom1504> it does use the protocol itself 16:10 < rom1504> would be amazing to build a bot like deepmind did for atari games 16:10 < rom1504> but so very hard 16:10 < rom1504> ah it's installed to the client 16:11 < rom1504> anyway, the strategy and all thingy looks interesting yeah 16:12 < Gjum> yeah I would also go for a standalone client. it's just the most complete and advanced task system I found so far 16:12 < hansihe> one thing that has always interested me is making bots in video games look like they are humans 16:13 < morfin> since when stone have variants? 16:13 < hansihe> complete with overshot on mouse movement, nonperfect movement, that kinda stuff 16:13 < morfin> variant: stone, variant: diorite 16:14 < morfin> variant: andesite 16:14 < hansihe> morfin: http://minecraft.gamepedia.com/Stone#History 16:15 < rom1504> hansihe: I think fps bots are the more advanced on that topic 16:15 < rom1504> (looking like humans) 16:15 < rom1504> well and chatbots of course 16:15 < hansihe> yeah, definitely 16:16 < morfin> are not Minecraft bots very hmm obvious 16:16 < Gjum> well afk bots arent ^^ 16:16 < hansihe> i know there are some very advanced bots for runescape as well 16:16 < morfin> afk bots/ 16:16 < rom1504> deepmind did bots on atari games that use only the pixel as input and the keys as ouput. It's nice, but it's all 2D, to do that for a 3D world would be significantly harder I think 16:16 < hansihe> bots that don't even interact with game code, they use machine vision 16:17 < rom1504> especially a 3D world with a lot of rules like minecraft 16:17 < morfin> :) 16:17 < hansihe> i know there are some runescape bots that do exactly that 16:17 < morfin> so they act like legit client? 16:18 < rom1504> acting like an human and using the same input/output as an human is yet an other different thing 16:18 < rom1504> a bot would soon learn to play perfectly, and would never get distracted 16:19 < hansihe> i found it, if i understand right they take control over the mouse and keyboard, use the official client 16:19 < rom1504> if you actually mean "client", then it's easy though :p you just have to use the protocol properly 16:20 < hansihe> do, they don't use the protocol at all 16:20 < hansihe> no* 16:20 < hansihe> well, they might be lying 16:20 < rom1504> yeah I was talking about what morfin said 16:20 < rom1504> well runescape looks like a fully static world right ? 16:20 < rom1504> and almost 2D 16:21 < hansihe> ah 16:21 < hansihe> yeah, there seems to be a static world map in the corner 16:21 < hansihe> that would make things easier 16:22 < rom1504> but yeah it sure is possible to use computer vision to do such a bot 16:22 < rom1504> it's just harder 16:23 < rom1504> if a bot could learn the characteristics of minecraft blocks just by looking, it would be pretty amazing 16:23 < hansihe> well, i think using machine vision and everything is taking things pretty far, even writing a custom client that acts as much as possible as a user would be interesting 16:23 < hansihe> definitely 16:23 < rom1504> almost ready to put in a robot and make it move in the real world 16:23 < hansihe> minecraft is not the hardest game to do that in i think, as you said blocks are fairly regular 16:24 < rom1504> minecraft is an highly structured game. There are lot of rules to figure out 16:24 < rom1504> unlike an fps for example 16:25 < hansihe> oh, yeah, if you mean fully trained behavior, that would be difficult 16:27 < rom1504> would be interesting to record all actions of players, and then to try to figure out what they were doing 16:27 < rom1504> and then you could make a bot do the same kind of things 16:27 < hansihe> that would be interesting 16:32 < morfin> that sounds more like neural network 16:32 < rom1504> yeah 16:32 < hansihe> you could use a neural network, yeah 16:33 < morfin> learning from real players and trying to similar thing 16:35 < morfin> but i am not sure you can achieve "realistic" behaviour 16:36 < hansihe> if you are trying to emulate things like movement with the neural network, that would probably work 16:36 < hansihe> higher level behaviors would probably have to be done in code, or with another technique 16:37 < hansihe> i mean basic things like mouse movement, how a human player moves forward, turns, that kinda stuff 16:37 < rom1504> yeah code + some nn to help some things (making the bot look like a player mostly) would be easier 16:38 < rom1504> I don't believe that using just a nn would do much things 16:38 < hansihe> nope 16:38 < rom1504> with the current state of neural networks anyway 16:39 < Gjum> btw opencog is workinnng on a minecraft ai https://github.com/opencog/opencog-to-minecraft/ http://wiki.opencog.org/wikihome/index.php/Minecraft_Bot_Development_Roadmap 16:40 < Gjum> they do some "realistic" things like restrict the vision to visible blocks instead of loaded world 16:40 < hansihe> interesting 16:41 < Gjum> and stuff like physics and movement will be in code I believe, like you said 16:42 < hansihe> i wrote things in the wrong order 16:42 < hansihe> things like mouse movement, that kinda stuff with a nn 16:43 < hansihe> it seems like opencog mostly is about the higher level behaviors though 16:43 < hansihe> i don't know if they aim to make movement look exactly like a human would 16:44 < hansihe> but yeah, movement and physics generally would be done in code i think 16:44 < Gjum> I believe theyre going for optimizing high level behaviour, not human likeness 16:44 < hansihe> yeah 16:45 < Gjum> so if it's faster to cut a corner real close, it would do that perfectly, like no human could 16:46 < hansihe> that's where i would imagine a ml could come into the picture, you could train it on how a human would move in different conditions 16:46 < rom1504> I think nn could help at the low level movement (if you want to make movement look like an human), or at the very high level (if you want to make thinking look like an human) 16:46 < hansihe> yeah 16:46 < hansihe> you would need loads of glue in between 16:47 < rom1504> yeah, learning all the minecraft rules automatically... I don't think we can do it currently 16:48 < Gjum> that's another thing they want to do I believe 16:48 < AlphaBlend> Riding entities such as boats, horses, minecarts and saddled pigs, are now run by the client instead by the server. 16:48 < rom1504> say learning crafting recipes or learning what blocks fall when you break it, what tools you need,... would be hard with ml probably 16:48 < AlphaBlend> if you run a custom server that cancels this behavior, won't it look weird? 16:48 < rom1504> well you could probably train a model for each of these things 16:48 < AlphaBlend> or rather, behave weirdly? 16:49 < hansihe> depends if the client waits for confirmation from the server or not 16:49 < rom1504> would be a lot of work for little results though 16:49 < hansihe> yeah, not worth it at all, that's the kind of stuff you can do really easily in code 16:50 < rom1504> AlphaBlend: what do you mean by "now" 16:50 < rom1504> 1.9 snapshots ? 16:50 < hansihe> whereas emulating realistic low level movement is something that would be tricky to do in code, probably easier to do a model for it 16:51 < rom1504> anyway in 1.8 the client handle attaching completely 16:51 < AlphaBlend> rom1504: that was an update note in the latest 1.9 snapshot 16:51 < rom1504> so it's just like in 1.8 then AlphaBlend ? 16:52 < rom1504> well I doubt it 16:52 < AlphaBlend> i just want to know if it will cause strange behavior, such as attaching and then later detaching if the server says no (custom server cancels action) 16:52 < AlphaBlend> just like in 1.8? did they remove client-side in the 1.9 snapshots early on then? 16:52 < hansihe> if it's that recent i doubt anyone in here knows for sure 16:53 < rom1504> yeah I don't know, I didn't try 1.9 attaching 16:53 < Gjum> attaching is still done by the server i believe, but movement is now client side yes 16:53 < rom1504> in 1.8 you just send the attach packet, and the client compute the movement of the attached entity itself 16:53 < rom1504> *himself 16:53 < Gjum> so when a custom server changes eg boat acceleration it might conflict 16:54 < Gjum> rom1504: but movement was still done by the server 16:54 < rom1504> no 16:54 < rom1504> well the server also computes it 16:54 < rom1504> but for example 16:54 < Gjum> as in, the client sends look and steer, and the server updates the player position 16:54 < rom1504> currently flying-squid doesn't track attached entities at all 16:54 < rom1504> and yet attaching works 16:54 < rom1504> (client-side) 16:55 < Gjum> ok then the client simulates it too, makes sense 16:55 < hansihe> does the server actually normally send position updates for attached entities then? 16:55 < hansihe> or does it purely rely on the client 16:56 < hansihe> i mean, it would track the position, i just wonder if it sends it 16:56 < rom1504> yeah I don't know 16:56 < Gjum> hansihe: in 1.8 server sends client pos when attached, in 1.9 no idea. I think not, like with normal movement in 1.8 16:56 < rom1504> I think it probably send the position 16:58 < hansihe> hmm 16:58 < Gjum> 1.8 packet log: https://github.com/SpockBotMC/SpockBot/issues/155#issuecomment-170052836 16:59 < rom1504> Gjum: what about any other entity mounted on an other entity ? 16:59 < Gjum> dang, in that log I omitted the received player pos packets every gametick 17:00 < Gjum> rom1504: no idea. they could be updated like a single entity, or each by itself 17:01 < Gjum> protocol-wise it would be enough information for the client if it got an attach packet and position updates for one of them 17:03 < rom1504> yeah 17:20 < morfin> disabling AI for mob disables only moving/lookups etc 17:23 < morfin> dragon is able even to remove lava and water? 17:24 < morfin> i thought only solid blocks 17:30 < rom1504> I think it can remove everything 17:31 < rom1504> except bedrock I guess 17:31 < morfin> Otherwise, target a random point (at least 10 blocks away from the dragon's current position) in a 120×120×50 cuboid centered on 0,95,0. 17:32 < morfin> is that why it fly away from bottom of world to the surface? 17:32 < morfin> i spawned 2 enderdragons on bedrock level and both flew up to the surface 17:34 < Gjum> rom1504: and obsidian and end stone 17:34 < Gjum> not sure about iron bars in 1.9, as they spawn in the end 17:34 < morfin> obsidian and bedrock and enderstone can't be destroyed 17:34 < morfin> iron bars? 17:34 < Gjum> on the pillars http://s3wc5.goplay.com/wysiwyg/2015/09/21/ender-dragon-minecraft%201.9-02.png 19:52 < Not-1733> [flying-squid] rom1504 deleted branch greenkeeper-babel-plugin-transform-runtime-6.5.2 19:55 < morfin> you use babel? 19:57 < morfin> babel to convert ES6 code into old ES5 20:05 < rom1504> yes 20:05 < hansihe> what features do you use? 20:05 < hansihe> i saw async/await 20:07 < rom1504> yeah async/await, that's es7 though 20:07 < hansihe> yeah, but in babel 20:08 < rom1504> in es6 we use mainly destructuring and arrow function 20:09 < rom1504> babel convert es7 to es5 yeah 20:12 < rom1504> oh and let/const of course 20:28 < Gjum> are classes es5? 20:34 < hansihe> they are not 20:34 < rom1504> oh yeah and classes 20:34 < rom1504> although we don't use a lot of classes 21:39 < pokechu22> So, I've been trying to update burger, continuing where deathcap left off. 21:39 < pokechu22> I've got blocks, items, and packets working, or at least I think I do. 21:47 < Gjum> pokechu22: tktech and nickelpro made this (I believe to replace solum): https://github.com/TkTech/Jawa 21:50 < pokechu22> Hm. I'll look into that. 23:25 <+Amaranth> async/await is expensive to emulate in es5 iirc 23:27 <+Amaranth> I don't think Minecraft is using any Java 8 features yet but they ship their own copy alongside the game so any decompiler that doesn't know how to handle it is probably about to break on MC code 23:28 < hansihe> async/await compile down to a generator implementation if i'm not mistaken 23:29 <+Amaranth> hansihe: But generators are ES6, no? 23:29 < hansihe> yeah 23:29 <+Amaranth> And they are expensive to emulate in ES5 23:29 < hansihe> so they do some emulation thing i think 23:29 < ecx86> that's funny 23:29 < hansihe> yeah, i wasn't disputing that 23:29 < ecx86> javaprophet isnt in this channel anymore 23:34 < hansihe> generators seem to be a feature in node 4 and 5 though 23:59 <+Amaranth> hansihe: Does the version of V8 in those releases support generators? :) --- Day changed sam. févr. 13 2016 00:00 < hansihe> Amaranth: yes, i'm fairly certain of it 00:01 < hansihe> i'm no expert though, haven't used node in a while 00:03 < rom1504> Amaranth: yes indeed generators are a bit slow 00:03 < rom1504> so async/await is like 2x slower than just promises 00:04 < rom1504> but it's not that big of a deal if you don't use them a lot 03:56 < Akaibu> ok, so i'm not sure if my thinking is right here, but if i were to say want to strip out a minecraft client so that minecraft multiplayer works fine, but all the singleplayer shit(world generation,commands, maybe some of the client side easter eggs as well) is taken out, wouldn't that make performance better/faster due to the jar not having to load those? 05:20 < pokechu22> Akaibu: I don't know exactly, but I think that it probably wouldn't be a significant improvement. 05:21 < Akaibu> i'm sure it would, would likely take out nearly ~75-100 class files at least 05:23 < pokechu22> True, but I don't think it actually loads the classes until it needs them. According to http://stackoverflow.com/a/14728058 it's not loaded until it's used. 05:24 < pokechu22> I'm not fully sure about that, though. 06:29 < pokechu22> I asked this a few days ago, but does anyone know if it is possible to use hopper.minecraft.net as a mod author to find crashes from my mod? I am setting the client brand and such so it shouldn't be hard to identify. 06:37 < pokechu22> Oh yea, it's a clientside mod that's a base edit - no forge or such (so it should automatically use the launcher's crash uploading I think) 10:42 < morfin> rom1504, why async/await is slower? 10:42 < morfin> because they're not native? 10:44 < hansihe> i'm not rom, but i'll answer anyway :) 10:45 < hansihe> they compile into javascript generators, which are emulated with a library written in javascript 10:46 < hansihe> the library is slower then just using promises 10:58 < morfin> that's what i meant when asked 10:58 < morfin> btw v8 have support for arrow functions as i know(i tested in Chromium) and probably generators too 11:00 < morfin> this is funny - Lua had something like destructuring before JS 11:01 < morfin> chrome can't use it, but arrow functions works 11:03 < hansihe> yep, node supports generators as well 11:03 < hansihe> so you could compile to that 11:04 < hansihe> the es5 preset compiles to the library though 11:04 < rom1504> yeah I didn't test the perf of not using babel for generators 11:04 < rom1504> it would probably be better 11:05 < rom1504> https://github.com/alekseykulikov/babel-preset-es2015-node5 is interesting 11:05 < rom1504> that means requiring people to install the last version of node5 though 11:06 < rom1504> which is not ideal 11:06 < rom1504> kind of waiting for full es6 support 11:06 < rom1504> which is coming soon (last v8 has 95% support but it's not in node yet) 11:07 < hansihe> yeah, there is probably no reason to rush using it 11:07 < morfin> yes could be not pretty stable 11:07 < morfin> i think 11:35 < hansihe> rom1504: hey, i just noticed transparency is not a value in blocks.json 11:36 < hansihe> it should be a boolean, true for blocks that can be looked through, false for not 11:37 < hansihe> it seems to be on most wiki pages 11:38 < hansihe> i also need that value for doing lighting calculations, unsure what the best way to add it would be 11:38 < hansihe> i am fairly certain blocks.json would be the right place for it though 11:39 < rom1504> yes blocks.json is the right place indeed 11:39 < rom1504> I'll do it 11:39 < rom1504> it's just a simple thing to add to the extractor 11:41 < hansihe> alright, thanks! 11:47 < hansihe> rom1504: just a small pitfall, blocks like http://minecraft.gamepedia.com/Cobweb have transparency as partial 11:47 < hansihe> it should still be yes 11:47 < rom1504> ah 11:47 < rom1504> so No == false 11:47 < rom1504> everything else == true ? 11:47 < hansihe> yeah, pretty much 11:48 < rom1504> if the transparent property isn't there, should I assume false or true ? 11:48 < hansihe> false 11:48 < rom1504> ok 11:49 < hansihe> this should probably be documented somewhere, i could write a gist on it or something 11:49 < hansihe> or a wiki.vg page 11:51 < rom1504> document what exactly ? 11:51 < rom1504> how to compute lights ? 11:52 < hansihe> that, and how these values work and fit together 11:54 < rom1504> yeah it would be nice to document that indeed 11:54 < rom1504> a wiki.vg page would fit I think 12:04 < rom1504> hansihe: does that data looks correct to you https://github.com/PrismarineJS/minecraft-data/pull/101 ? 12:06 < hansihe> i believe so 12:06 < hansihe> let me test a thing before you merge 12:12 < hansihe> yeah, i believe it's correct 12:12 < rom1504> okay let's merge that 12:12 < hansihe> things are more complicated then i assumed though 12:12 < Not-1733> [minecraft-data] rom1504 pushed 2 commits to master [+0/-0/±6] https://github.com/PrismarineJS/minecraft-data/compare/be607b64ae73...56007592b76b 12:12 < Not-1733> [minecraft-data] rom1504 5600759 - Merge pull request #101 from PrismarineJS/add_transparent_property add transparent property 12:12 < Not-1733> [minecraft-data] rom1504 deleted branch add_transparent_property 12:14 < rom1504> hansihe: here's the relevant wiki extractor commit fyi https://github.com/PrismarineJS/minecraft-wiki-extractor/commit/3dcce58617d6dafee47d1a2770b35f4b0ab72396 12:14 < rom1504> pretty easy 12:15 < hansihe> oh, that wasn't much code at all 12:25 < rom1504> oh 12:26 < rom1504> I just noticed that Luminance+Transparency properties give the same information that the Light page 12:28 < rom1504> so I could extract that automatically and put it in blocks.json too hansihe 12:29 < hansihe> yeah, that might be a better solution 12:34 < rom1504> I'm wondering if transparent and filter should be 2 properties or just one 12:35 < rom1504> stairs would be 15 12:35 < rom1504> but what would stone be 12:35 < rom1504> well 12:35 < rom1504> I guess 2 properties make more sense 12:35 < hansihe> blocks all light, 15 12:35 < hansihe> yeah, it would 12:39 < rom1504> hansihe: fully transparent blocks should have filterLight to 0 right ? 12:41 < hansihe> i believe so 12:41 < hansihe> i think i might have made a mistake in my pr 12:41 < rom1504> btw water and ice are 13 12:41 < rom1504> if I understand right 12:41 < rom1504> err no nevermind 12:46 < hansihe> hmm 12:46 < hansihe> water should be 0 if i understand right 12:46 < hansihe> ice should have 2 12:46 < rom1504> "Partial (-2 to light)" 12:46 < rom1504> http://minecraft.gamepedia.com/Water 12:46 < hansihe> ah, it's 2 then 12:47 < rom1504> yeah, I'm kind of hard coding the special values of transparent 12:47 < rom1504> that's what's annoying about wikis 12:47 < rom1504> these infoboxes I mean 12:47 < hansihe> yeah 12:48 < rom1504> it seems to be structured data 12:48 < rom1504> and then... some natural language 12:48 < rom1504> Partial (doesn\'t block light) 12:48 < rom1504> wtf does that mean 12:48 < rom1504> http://minecraft.gamepedia.com/Monster_Spawner 12:49 < rom1504> I'm going to say it means 0 12:51 < rom1504> "Opaque, but lets light pass through" 12:51 < rom1504> lol 12:51 < rom1504> Barrier 12:51 < rom1504> it's opaque but... not really ! 12:52 < hansihe> hahaha, this is so terrible 12:53 < hansihe> i have a feeling this is mostly the fault of the wiki 12:56 < hansihe> okey, just looked at Blocks.java 12:56 < hansihe> it's much better expressed there it seems 13:01 < rom1504> there's a single file with data in mc source ? 13:01 < rom1504> I thought it was in the 300 blocks class in various ways 13:03 < hansihe> yeah, but the light filter values are set where they register blocks 13:04 < hansihe> rom1504: https://gist.github.com/hansihe/3dd2a3c0c06f7e41ad0b 13:05 < rom1504> I'm going to make a PR with the data from the wiki soon, it might not be perfect but it's probably okay for now 13:05 < hansihe> yeah 13:05 < rom1504> ah yeah I see 13:05 < hansihe> it's easy to fix later 13:05 < rom1504> oh well I'm getting 504 :d 13:05 < rom1504> seems I'm doing a bit too much http requests to the wiki ^^ 13:06 < hansihe> ah :P 13:06 < rom1504> must like 300 requests every time I run the block extractors 13:09 < rom1504> hansihe: https://github.com/PrismarineJS/minecraft-data/pull/102 13:12 < hansihe> does look right 13:12 < hansihe> but yeah, i do think transparency is still useful 13:13 < hansihe> when writing a renderer it tells you if you can cull the geometry behind it 13:13 < rom1504> but transparent == false is filterLight==15 and transparent == true is filterLight == 0 13:13 < rom1504> hmm 13:14 < hansihe> it's true that filterlight is always 15 when transparency is false 13:14 < rom1504> ah yeah what I said is not true 13:14 < hansihe> yeah 13:14 < rom1504> it can be transparent and still be filterLight==15 13:14 < hansihe> yep 13:14 < rom1504> okay it's useful then 13:14 < hansihe> like a slab i think 13:17 < rom1504> hmm 13:18 < hansihe> there are more materials in the game than in materials.json as well 13:18 < hansihe> not sure if this is a deliberate decision 13:19 < rom1504> it's material as in "digging material", not the materials of the wiki 13:19 < hansihe> oh, right 13:25 < Not-1733> [minecraft-data] rom1504 pushed 3 commits to master [+0/-0/±8] https://github.com/PrismarineJS/minecraft-data/compare/56007592b76b...db3a7812939f 13:25 < Not-1733> [minecraft-data] rom1504 db3a781 - Merge pull request #102 from PrismarineJS/add_light add emitLight and filterLight 13:25 < Not-1733> [minecraft-data] rom1504 deleted branch add_light 13:30 < rom1504> hansihe: merged my block light pr 13:30 < rom1504> and closed yours since it's in blocks.json now 13:30 < hansihe> awesome! 14:10 < rom1504> alright I updated the doc, these values can be seen in http://prismarinejs.github.io/minecraft-data/?d=blocks now (show/hide columns) 14:12 < hansihe> nice 14:12 < hansihe> one thing i noticed though 14:12 < hansihe> tnt has filterlight 0 and and transparent true 14:12 < hansihe> is that right? 14:14 < hansihe> also unlit redstone lamp should have emitlight 0 14:15 < hansihe> same with redstone torch and furnace 14:15 < hansihe> there is no real way to parse that from the wiki though 14:15 < rom1504> http://minecraft.gamepedia.com/TNT "Partial (cannot attach other blocks)" 14:15 < rom1504> I can just hardcode it I guess 14:16 < hansihe> i think that means the only way it acts like a transparent block is that you can't attach things like torches to it 14:16 < hansihe> it might be nice to have a json file containing overrides 14:17 < hansihe> i don't know if you are overriding anything else 14:17 < rom1504> https://github.com/PrismarineJS/minecraft-wiki-extractor/issues/12 14:18 < rom1504> yes 14:18 < rom1504> having a .json for overriding is a good idea 14:18 < hansihe> haha, sorry for saying everything here in irc :P 14:18 < rom1504> I've just hardcoded some stuff currently 14:18 < hansihe> yeah 14:19 < rom1504> it's okay, it's just I don't have time to do it right now, and I don't want to have to back logs when I have time ^^ 16:56 < OnlyQubes> Hey guys, i can't seem to figure out when minecraft uses the 'downloading terrain' screen (or how is it called), i get it when for example a server uses bungeecord and i want to go to another server from the hub. 17:01 < hansihe> OnlyQubes: I think it's the time between when the login is finished and the player spawn packet 17:02 < OnlyQubes> ahh okay P: 17:07 < hansihe> Unsure how it works in relation to switching worlds though, someone else probably knows that 17:11 < rom1504> the respawn packet do that 17:11 < rom1504> well, when you change world 17:12 < rom1504> http://wiki.vg/Protocol#Respawn 17:12 < rom1504> there's some explanation there 17:34 < OnlyQubes> yup found it, thanks :P --- Day changed dim. févr. 14 2016 04:58 < stalkerr> is this channel for minecraft hacking kids? 04:58 < stalkerr> just asking 04:58 < __0x277F> Depends. 04:59 < __0x277F> Nobody here is "hacking kids". <3 05:00 < stalkerr> k 05:01 < stalkerr> also hi 05:01 < stalkerr> so __0x277F what are you doing here 05:01 < stalkerr> what are you doing on this channel people? 05:01 < __0x277F> Idling, mostly 05:01 < __0x277F> My biggest claim to fame in here is that I documented 1 packet for 1.9. :/ 05:02 < __0x277F> So you should ask someone more impressive. 05:04 < stalkerr> i made a spigot plugin once 05:05 < stalkerr> it was even doing something 05:05 < stalkerr> kek 05:05 < stalkerr> (it was useless actually) 05:06 < stalkerr> any one of you was making cheats for minecraft? 05:06 < stalkerr> damn 05:06 < stalkerr> oh 05:06 < stalkerr> __0x277F, packet you say? 05:07 < stalkerr> like sniffing packets or how it's called 05:07 < __0x277F> Normally people here aren't making clients designed to exploit bugs in the server. 05:07 < stalkerr> normally 05:07 < stalkerr> k 05:07 < __0x277F> And all it was was looking through decompiled sources for 1.9 snapshots 05:07 < stalkerr> ok 05:08 < stalkerr> would want to see minecraft packets in real time 05:08 < stalkerr> damn 05:08 < __0x277F> They're encrypted. 05:09 < stalkerr> i know 05:09 < stalkerr> but there must be a way 05:09 < stalkerr> and it is i think 05:09 < __0x277F> Is there something specific you're looking for? 05:10 < stalkerr> ye 05:10 < stalkerr> but i don't know what 05:10 < stalkerr> what decompiler u using? 05:11 < __0x277F> I'm hesitant to support you considering you came in here asking for help making "cheats". 05:11 < stalkerr> man 05:11 < __0x277F> But Fernflower is undoubtedly the best decompiler. 05:12 < stalkerr> just asking 05:12 < stalkerr> calm downn 05:12 < stalkerr> kek 05:12 < stalkerr> ok, i'll tell you 05:13 < stalkerr> http://antycheat.pl/index.php?page=versions 05:13 < stalkerr> see? 05:13 < stalkerr> i want to reverse engineer that 05:13 < stalkerr> well it's poorly made i think 05:14 < stalkerr> but i want to know how did they made this 05:14 < __0x277F> What is it, even? 05:14 < stalkerr> a modded version of minecraft 05:15 < stalkerr> you know, you cant install cheats on it 05:15 < stalkerr> and you cant join servers using their plugin without this modded version 05:16 < stalkerr> also, how would you prevent people from using cheats on pvp server? 05:16 < __0x277F> Short answer is, you can't. 05:16 < stalkerr> damn 05:16 < __0x277F> There's no amount of things you can do to prevent a determined person from cheating. 05:16 < stalkerr> :d 05:16 < __0x277F> However, you can do many things to make it more difficult. 05:17 < stalkerr> is making a modded version that allows you to join a server a good option? 05:18 < stalkerr> but damn man 05:18 < stalkerr> you can decompile it and reverse engineer it 05:20 < stalkerr> like no way of preventing people to see underground tunnels without client mods 05:20 < stalkerr> damn 05:22 <+AndrewPH> sure there is 05:22 <+AndrewPH> fill em up when the player isn't supposed to be able to see them 05:23 < stalkerr> wow 05:23 < stalkerr> man 05:23 < stalkerr> ok maybe 05:24 < stalkerr> but it would be server intense 05:24 < stalkerr> i think 05:25 < stalkerr> hm 10:07 < morfin> i remember i was playing PvP server 10:07 < morfin> but all that anticheats did absolutely nothing 10:55 < hansihe> generally the people using things like xray are idiots, it's really easy to detect analytically 10:56 < hansihe> you just look at their ratio of stone to ore 10:56 < hansihe> obviously that's easy to avoid though 11:28 < morfin> yes i know it's easy to detecty 11:29 < morfin> but i was using Xray on server where played and my ratio stone:diamonds was looking fine 11:30 < morfin> because i was not doing stupid things 11:30 < morfin> like digging directly to the diamonds 11:33 < morfin> that's because i was alone vs glitchers(abusing /home command specifics to teleport trought walls), donaters(can fly but not when fighting, also have cool gear), other cheaters(obvious haxors were banned) 14:08 < stalkerr> man 18:49 < stalkerr> rip #mcdevs 19:02 < nickelpro> lol wtf 19:03 < nickelpro> Whenever #mcdevs has an activity spike it is always something weird 19:13 < stalkerr> oh man so weird --- Day changed lun. févr. 15 2016 07:09 < ecx_> uh oh. 09:03 < Meeeh> maybe here someone will know... what happened to netty 5? (java socket library) 09:41 <+ammar2> what happened to it? 10:00 < Meeeh> ammar2, no code on github for 5.x, nu updates for few months, there are only some issues on github 10:00 < Meeeh> and old snapshot in sonatype snapshot repo 10:00 < kashike> just means they're working on 4.x 10:01 < Meeeh> but what with 5.0? they just removed it? 10:01 < Meeeh> it will be no longer created/updated? 10:05 < Meeeh> "The 'master' branch is where the development of the latest major version lives on." 10:05 < Meeeh> there is no master branch :D 13:30 < Not-1733> [minecraft-data] rom1504 pushed 1 commit to master [+0/-0/±1] https://github.com/PrismarineJS/minecraft-data/compare/db3a7812939f...a6bbb54a257b 13:30 < Not-1733> [minecraft-data] rom1504 a6bbb54 - add McEx to user list 20:51 < Not-1733> [minecraft-data] rom1504 pushed 6 commits to master [+0/-0/±6] https://github.com/PrismarineJS/minecraft-data/compare/a6bbb54a257b...80cb5f84830d 20:51 < Not-1733> [minecraft-data] oldmanmike 501a2aa - First pass over ServerBound Play 20:51 < Not-1733> [minecraft-data] oldmanmike b22f2df - Complete full pass over protocol 20:51 < Not-1733> [minecraft-data] oldmanmike 5970f10 - Add some new packets I missed 20:51 < Not-1733> [minecraft-data] ... and 3 more commits. 20:52 < Not-1733> [minecraft-data] rom1504 pushed 1 commit to master [+0/-0/±2] https://github.com/PrismarineJS/minecraft-data/compare/80cb5f84830d...545e1ca367f8 20:52 < Not-1733> [minecraft-data] rom1504 545e1ca - fix held_item_slot in 1.9 and update version 20:59 < rom1504> oldmanmike: thanks for updating the 1.9 protocol.json ! that's really the kind of thing I hoped minecraft data could achieve, updating the data in a single point by devs coding in different languages :) 20:59 < rom1504> that means http://prismarinejs.github.io/minecraft-data/?v=1.9&d=protocol is an up to date doc for the 1.9 protocol 21:01 < rom1504> (still without all the notes and comments of wiki.vg page, but still !) 21:09 < ecx86> you know what would also be useful 21:10 < ecx86> dumping entity hitbox sizes, which blocks are passable by light, block hitboxes, which blocks are translucent, which blocks are passable by entities 21:10 < ecx86> oh, entities are dumped 21:11 < rom1504> blocks transparency has been recently added 21:11 < ecx86> in 1.9? 21:11 < rom1504> both 1.8 and 1.9 21:11 < ecx86> so all blocks can have translucent? 21:11 < rom1504> I mean in mcdata 21:11 < ecx86> oh. 21:11 < rom1504> passable by entity is available as "physics" property 21:11 < ecx86> i wonder how you dumped this so consistently ? 21:12 < rom1504> we don't have block hitboxes yet though 21:12 < ecx86> so many edge cases 21:12 < rom1504> ecx86: extracted from the wiki 21:12 < ecx86> oh 21:12 < ecx86> for entity sizes 21:12 < ecx86> i used instrumentation and i looked for all calls to setsize in the bytecode 21:13 < rom1504> yeah extracting from the jar is definitely the most precise method, but it's not very easy 21:13 < ecx86> mhmm 21:13 < ecx86> it was not...pretty 21:14 < ecx86> https://paste.whitefire.in/view/gvJMAicYXl 21:15 < ecx86> https://paste.whitefire.in/view/SAW5oK0ey2 21:18 < rom1504> interesting 21:20 < ecx86> i might dump the block hitboxes. 21:21 < ecx86> i wonder how the filterlight and transparent fields were collected for the wiki originally...? empirically? 21:23 < rom1504> I have no idea 21:23 < rom1504> but I bet a lot of manual work 21:23 < ecx86> seems like alot of work 21:24 < rom1504> yeah but then every single block, item and entity have its own page with lot of description 21:24 < rom1504> that's a lot of work too 21:26 < rom1504> sometimes it's easier to find people to write stuff manually than to write extractors 21:26 < rom1504> (that's true not just for the minecraft wiki, but also for wikipedia if you think about it ;) ) 21:26 < ecx86> so let me tell u 21:26 < ecx86> about light passthrough 21:26 < ecx86> first 21:27 < ecx86> it depends whether the block is an opaque cube. 21:27 < ecx86> also the light transparency value, (like for water, etc). 21:27 < ecx86> then it depends on a bunch of hardcoded stuff 21:27 < ecx86> in the renderer 21:48 < ecx86> there is a problem with the block list at http://prismarinejs.github.io/minecraft-data/?v=1.9&d=blocks 21:48 < ecx86> slabs filter 15 light. theyre basically opaque for lighting 21:59 < rom1504> yeah I know, the data is pretty new (a few days ago), I have some stuff to fix https://github.com/PrismarineJS/minecraft-wiki-extractor/issues/12 22:00 < ecx86> omg i just checked and the blocks have their own raytrace handler hardcoded 22:40 < hansihe> a lot of the block transparency stuff was verified by looking at the decompiled source 22:40 < hansihe> the wiki is wrong on a couple of accounts it seems 22:41 < hansihe> but yeah, a jar extractor would probably be worth it to make 22:42 < hansihe> even if you have to supply it some deobfuscated names manually, it would still be better 22:43 < hansihe> it seems like a lot of the less pretty parts in the minecraft codebase is to work around java --- Day changed mar. févr. 16 2016 00:28 < Gjum> well pokechu is working on burger again, maybe they get it working again 03:52 < ecx86> the sample code for direction to rotation vector at http://wiki.vg/Protocol#Player_Look is bad 03:53 < ecx86> im going to replace it with better code (https://paste.whitefire.in/view/zL6o8VFaKU). the results are the same and its faster 05:20 < Akaibu> Silly question likely, but are those the real Grum and Dinnerbone? 05:21 < Akaibu> Also Brandon15811 is it really needed to have 4 clients connected to one channel? 05:46 <+ammar2> yes and please avoid pinging them like that 07:17 < ecx86> someone should delete these pages... http://wiki.vg/Category:Craftd 07:17 < ecx86> project hasn't been updated since 2011 07:19 <+ammar2> meh doesn't really hurt anyone having them under a subsection 07:19 < ecx86> i guess 07:20 < ecx86> http://wiki.vg/Sandbox lol 08:18 < Not-1f77> [mineflayer] rom1504 deleted branch greenkeeper-lodash-4.4.0 09:08 < rom15043> greenkeepeer is trying to tell us something 09:33 < rom1504> I'd like it better if he was creating branches with his account instead of mine though 09:42 < Not-1733> [minecraft-data] rom1504 pushed 2 commits to master [+0/-0/±2] https://github.com/PrismarineJS/minecraft-data/compare/545e1ca367f8...d21776957c41 09:42 < Not-1733> [minecraft-data] deathcap bee52e8 - Add protocol version codes up to 16w07a (101) 09:42 < Not-1733> [minecraft-data] rom1504 d217769 - Merge pull request #110 from deathcap/versions-16w07a Add protocol version codes up to 16w07a (101) 09:45 < Not-1733> [minecraft-data] rom1504 pushed 2 commits to master [+0/-0/±2] https://github.com/PrismarineJS/minecraft-data/compare/d21776957c41...3667cb860ce3 09:45 < Not-1733> [minecraft-data] oldmanmike e936b62 - Reordered 1.9 protocol packets 09:45 < Not-1733> [minecraft-data] rom1504 3667cb8 - Merge pull request #108 from oldmanmike/reordered-protocol Reordered 1.9 protocol packets 09:47 < Not-1f77> [Charge] Wallbraker pushed 2 commits to master [+8/-5/±5] https://github.com/VoltLang/Charge/compare/7c489d6da271...2596b3e26027 09:47 < Not-1f77> [Charge] Wallbraker de25086 - lib: Update GL bindings 09:47 < Not-1f77> [Charge] Wallbraker 2596b3e - examples: Update GL example 10:14 < Not-1733> [minecraft-data] rom1504 pushed 4 commits to master [+4/-0/±10] https://github.com/PrismarineJS/minecraft-data/compare/3667cb860ce3...60857b6a3c56 10:14 < Not-1733> [minecraft-data] deathcap d8f7c28 - Update audit tests: allow versions with no items/blocks/recipes for now 10:14 < Not-1733> [minecraft-data] deathcap d20aed5 - Update schema tests to skip non-existent data 10:14 < Not-1733> [minecraft-data] deathcap 90b52c3 - Restore support for 15w40b 10:14 < Not-1733> [minecraft-data] rom1504 60857b6 - Merge pull request #109 from deathcap/15w40b Restore support for 15w40b 11:46 < Not-1f77> [Charge] Wallbraker pushed 1 commit to master [+0/-0/±1] https://github.com/VoltLang/Charge/compare/2596b3e26027...558595abc38a 11:46 < Not-1f77> [Charge] Wallbraker 558595a - examples: Update shaders 11:47 < Not-1f77> [Charge] Wallbraker pushed 1 commit to master [+0/-0/±1] https://github.com/VoltLang/Charge/compare/558595abc38a...891c01354480 11:47 < Not-1f77> [Charge] Wallbraker 891c013 - lib: Remove unneeded semicolon 15:32 < Meeeh> eeem, mojang? http://imgur.com/mczRNy4 I think this is not how end portals works xD 15:35 < Meeeh> also that cages was my favorite change about dragon/end :< 15:40 < rom1504> meh, just a small offset problem, it almost works, nothing to fix :p 15:44 < Meeeh> yeach :D 15:45 < Meeeh> also /ecnahnt command with invalid enchant strng id, returns that there is no ID with ID 0 15:50 < Meeeh> Elytra don't work on creative now :( 15:50 < Meeeh> or on survival too... or I just don't know how to use it now xD 15:58 < rom1504> I don't even know what elytra is 15:58 < rom1504> I only ever used 1.9 to test that node-minecraft-protocol proxy works on it. No crash = it works, I don't care about 1.9 features :p 16:01 < rom1504> ah 16:01 < rom1504> it's the reason why all these packets have a hand enum 16:02 < Meeeh> you have 2 hands now, so yeach 16:02 < Meeeh> also for some usless reason you can choose if you use right or left hand as main 16:03 < hansihe> vulkan is out! 16:04 < olmanmike> YAY! 16:04 < olmanmike> I'm celebrating by trying to get started on a client 16:06 < olmanmike> may the FFI gold rush begin 16:06 < hansihe> haha 16:06 < hansihe> #rust-gamedev is already on it 16:07 < olmanmike> Yeah, and there's been a dummy project in Hackage for a while now, hopefully that dude has time soon 16:09 < Meeeh> nah Grum / Dinnerbone you may try too look at that portals and enchant command, I don't see this bug anywhere on bugtracker, but too lazy to login 16:20 < Fenhl> Meeeh: you need to jump in midair to activate elytra. For some reason it doesn't always work though 16:20 < Meeeh> Fenhl, nah, so they changed it, ugh, near impossible to do on creative