21:26 <+Amaranth> They are either not loaded at all, dying soon, or stop being simulated
21:26 < gurun> well, yeah. But not all entites send positions.
21:26 < gurun> mobs are my only concern.
21:26 <+Amaranth> You seem to have a real hookup on packet performance
21:26 < gurun> calculations on the server is instant .. so i don't care about that either. Only the things that generate IO is of concern.
21:27 <+Amaranth> The only time I remember PlayerTracker even showing up on my profiling was when I put a bunch of players in the same chunk
21:27 < gurun> yes exactly
21:27 < gurun> for pvp, that is the ONLY concern part from player spawn (chunks)
21:27 <+Amaranth> Simulating 10000 entities is instant but sending the packets for them is where all the CPU time goes? Is that what you're trying to say?
21:28 <+Amaranth> If you're making a minigames server you don't care about most of this anyway
21:28 < gurun> i care about entities yes
21:28 <+ammar2> which reminds me, is there a packet to send multiple entity locations at once
21:28 <+Amaranth> If you're not then PVP is rare and players will be split up often hiding from each other
21:28 < gurun> ammar2, in PE 0.11 it is now.
21:29 <+Amaranth> Personally I don't really care about the requirements of a pvp minigames server
21:29 <+Amaranth> Because a good actual Minecraft server will handle that case as a side effect
21:29 < gurun> hmm, i wouldn't ever play MC on an custom server. It's just lame.
21:30 < morfin> why?
21:31 < gurun> because the best they can do is come at half decent distance of the real functionality of both world genration and AI.
21:31 < morfin> )
21:31 < morfin> not sure but that Cuberite have nice generator
21:31 < morfin> but AI is failure
21:32 < morfin> Wither exploded once when spawned and then nothing happened
21:32 < morfin> ender eye does not work so i think no end
21:32 < Paprikachu> cuberite has pretty fugly c++ :D
21:33 < morfin> hm?
21:33 < morfin> why
21:33 < Paprikachu> also, isnt cuberite some other minecraft server renamed
21:33 < Paprikachu> source looks familiar
21:33 < morfin> i think hCraft
21:34 < Paprikachu> it's mcserver
21:34 < Paprikachu> knew it
21:34 < morfin> oh
21:35 < yawkat> does minecraft varint encoding follow protobuf rules or does it just explode on negative numbers?
21:35 < morfin> it should
21:35 < yawkat> all protocol implementations ive seen on minecraft seem to just use sign as MSB
21:37 <+ammar2> they definetly work since sending a compression threshold which is a varint of -1 (used to?) works
21:37 <+Thinkofdeath> it works but it shouldn't :)
21:37 < morfin> not bad worldgen
21:37 <+Thinkofdeath> it ends up 5 bytes for negative numbers
21:37 < morfin> no idea why you don't like such things
21:37 <+Thinkofdeath> due to them being treated as uint32 during encoding
21:38 < yawkat> yea, thought so
21:38 < yawkat> a bit stupid but okay.
21:39 < gurun> so, hmm .. if i implement one of those AI routines .. and spawn a lot of mobs .. that should be a fairly good estimation.
21:40 < Aragas> not the best idea i had, but mayb then add CompressionEnabledinstead of using -1? No more negatives then, gight?
21:40 < yawkat> it's only sent once
21:40 < yawkat> not a huge issue
21:41 < Aragas> and becuse of it, varint ends up as 5 bytes, right?
21:41 < yawkat> only negative ones
21:42 < yawkat> or 10 for varlong, or 3 for varshort.
21:42 < Aragas> yea. really wasn't one of my best ideas
21:54 < morfin> lol on custom server i can hear how my balls rings)
21:54 < morfin> when i get hit by arrow
21:56 < morfin> hey Paprikachu
21:56 < Paprikachu> suo
21:56 < morfin> i am really looking at C++11 and hm
21:56 < Paprikachu> sup
21:57 < morfin> it's way better
21:57 < Paprikachu> c++11 is old shit
21:57 < Paprikachu> c++17 incoming!
21:57 < morfin> i know i know
21:57 < morfin> tell that VS)
21:58 < Paprikachu> yeah i dont use VS
21:58 < Paprikachu> fuck that :)
21:58 < morfin> slowpoke
21:58 < morfin> develops it
21:58 < morfin> anyway even C++11 is big step
21:58 < Paprikachu> it is
22:48 <+Amaranth> It could be worse, you could be using Intel's compiler
22:48 <+Amaranth> MSVC supports most things
22:49 <+Amaranth> http://blogs.msdn.com/b/vcblog/archive/2015/06/19/c-11-14-17-features-in-vs-2015-rtm.aspx
23:18 <+XorBoole> > not using clang
23:18  * XorBoole runs
23:38  * gurun runs with XorBoole 
23:48 <+XorBoole> I was running half an hour ago noob
--- Day changed ven. juil. 10 2015
00:24 < Not-48e3> [wiki.vg] Edit by Xlrion to Authentication -> http://wiki.vg/index.php?title=Authentication&diff=6698&oldid=5927
01:15 < Fenhl> \o/
01:15 < Fenhl> yay pronouns
01:16 < Fenhl> oh dear that article looks like it needs cleanup badly
01:18 < Fenhl> …why on earth does the wiki not have the Cite extension
01:27 < Fenhl> well, in that case I guess blog-style references are the next best thing
01:28 < Not-48e3> [wiki.vg] Edit by Fenhl to Authentication -> http://wiki.vg/index.php?title=Authentication&diff=6699&oldid=6698
02:56 < gurun> ok Amaranth i've been doing tests on that 10.000 entities thing
02:56 < gurun> i picked item entities to make sure it's ticking "something"
02:56 < gurun> but no io
02:59 < gurun> adding 10.000 on basically the same spot have a sever impact on the client (rendering), but virtually no impact on the serer.
02:59 < gurun> navigating into them, generates pickup instantly of the itmes that can be picked.
03:01 < gurun> so my little theory still holds true, computation of the server is more or less irrelevant.
13:46 < morfin> does vanilla properly load NBT with tags having spaces?
13:46 < morfin> http://minecraft.gamepedia.com/NBT_format this page says tag name can contain spaces but Minecraft does not save them with spaces 
13:51 <+Amaranth> You'd have to mod Minecraft to do something with a tag that contains a space to find out
13:51 <+Amaranth> Or poke through a decompile of their NBT handling code
13:53 < morfin> interesting how custom behaviours are implemented with Command blocks
13:54 < morfin> i saw some maps which had magic items doing some custom stuff when used
13:56 <+Amaranth> Yeah they've gotten pretty clever
13:57 < morfin> as example i saw map where you could travel between Hell and Heaven and to solve it you should do that sometimes and then you get bow which copy blocks from Hell to Heaven
13:57 <+Amaranth> Redstone was already turing complete, scoreboard give simple variable storage, and commands can modify every property of a block or entity that a mod could (unless that mod actually changed the source of the block or entity)
13:58 <+Amaranth> So as things in MC get moved from hard coded to data driven command blocks get more useful
13:59 <+Amaranth> They really are a form of mod API
14:00 < morfin> i have no idea how command blocks are working
14:04 < morfin> there is one bad thing in that block
14:04 < morfin> it can't disallow some commands  depending on permissions as i know
14:06 <+Amaranth> It sort of can
14:07 < morfin> ?
14:07 <+Amaranth> Vanilla has a built in concept of permission levels and command blocks are one level below console and op
14:07 <+Amaranth> So some commands are blocked from it
14:07 < morfin> command blocks can be used only by ops
14:07 <+Amaranth> Sure
14:07 < morfin> i mean something like hmmm
14:07 <+Amaranth> Well, only modified by them
14:08 <+Amaranth> Others can hold them in their inventory and place them
14:08 < morfin> group user can use commands X,Y,Z moderator also can use U,W,V, admin can use everything he wants
14:09 <+Amaranth> That'd require command blocks to store who placed them, for that to matter to who activated them, and for a permission system to store what you're talking about
14:09 < morfin> 1 command block == one command right?
14:09 <+Amaranth> I don't remember
14:10 < morfin> hmm
14:11 < morfin> i think it's possible to store custom metadata for block?
14:11 < morfin> of course that's not compatible with vanilla 
14:18 <+Amaranth> Yeah, vanilla won't round trip the extra information in a tile entity
14:18 <+Amaranth> You can store it using the /blockdata command but when the chunk unloads it'll be gone
14:19 < morfin> hmm
14:20 < morfin> i still do not understnd how item can change world without modifications
14:20 <+Amaranth> If you stick it in the data on disk it won't even make it in to the tile entity object on load
14:29 < gurun> Amaranth, been offline. But did you see the result of my little test?
14:30 <+Amaranth> gurun: Yep, were you doing movement and collision detection as a part of that?
14:31 <+Amaranth> Also you could probably run close to full speed with 10000 item drops on vanilla too but you'd be chewing up a lot of CPU
14:31 < morfin> lol
14:32 < morfin> teleportation done with levels control
14:37 < gurun> Amaranth, check! on the collision and that calculation.
14:38 < gurun> the client struggled heavily when the items where within viewfrustrum (or whatever that's called)
14:38 < gurun> the server maybe bumped up a % but too little to judge.
14:39 < gurun> but yeah, i'm a "bit" obsessed about IO and not so much computation in general.
14:40 < morfin> soo clock+that command block do the trick
14:43 <+Amaranth> gurun: Well, that is pretty impressive I guess
14:43 <+Amaranth> But in a full Minecraft server the CPU time spent calculating and sending packets is still usually going to be down on the list a bit compared to the world simulation
14:44 <+Amaranth> If you're obsessed about the amount of IO you're doing sure but that's out of your hands really, you don't control the protocol
14:56 < gurun> it's MCPE so nothing wrong with the protocol. It would be a challange regardless.
14:58 < shoghicp> I control the protocol >:D
14:58 < shoghicp> not RakNet though :(
15:11 <+Amaranth> Does the MCPE protocol work more like a FPS?
15:11 < shoghicp> UDP :)
15:12 <+XorBoole> but does it blend?
15:12 <+Amaranth> Sure but working like a FPS means sending the state of all relevant entities in one big packet every tick and not caring if a packet gets lost
15:13 <+Amaranth> If you're using UDP and some reliable send method on top and then doing a normal MC-like protocol you've gained nothing
15:14 <+XorBoole> somehow I feel the fps model isn't that great for minecraft
15:14 <+Amaranth> Looks like that's what you've done
15:14 <+XorBoole> given that there are potentially many, many more entities, and other things that need to be kept track on
15:14 <+Amaranth> If you're going to use a protocol like MC you should just use TCP
15:15 <+Amaranth> Because UDP just causes headaches and you're not getting any gain from it if you've still got reliable transmissions and head of line blocking (required to maintain state)
15:16 <+Amaranth> XorBoole: Well you don't send all the state you send a diff against the last confirmed state
15:16 <+XorBoole> ah, I suppose that makes sense
15:17 <+XorBoole> but I can picture scenatios with diffs just as big
15:17 <+XorBoole> as far as UDP use, that's a mystery to me
15:17 <+Amaranth> If the diff is huge then so would all the single packets in the current design
15:17 <+Amaranth> Err, terrible wording there
15:18 <+Amaranth> Now, with UDP you usually want to keep your packets small, below 500 bytes or so, which could mean you have situations were sending the diff is too much
15:18 <+Amaranth> But you should design your game to make that not possible or have some workaround
15:18 <+XorBoole> well, it doesn't help that PC does stupid things like have packets for specific minor features (titles) and mixes a bunch of things into a single packet (sleep, interact, etc.)
15:19 <+XorBoole> and yes, I realize that the former is a new packet and the later is a notchpacket
15:21 < gurun> UDP makes sense if you are mixing reliable stuff (state sync) with unreliable stuff (skip updates).
15:21 <+Amaranth> I could see your state snapshot packet just having a boolean and sequence number so you can shard the state over many packets and the client won't consider it received until it has all of them. That's half way to what TCP gives you but without the head of line blocking waiting for that state snapshot if it fails but the next one goes through
15:23 <+Amaranth> iirc the reason for doing that is that even if your UDP packet fragments correctly in your OS and manages to make it through the network without getting dropped on purpose if you lose a piece of it your kernel will hold on to the rest for a long time waiting to see if it shows up
15:23 <+Amaranth> Just for you to toss it as being outdated
15:25 <+Amaranth> The real challenge would be making your client able to deal with getting unexpected mixes of state due to the reliable data and unreliable data being out of sync
16:55 <+ammar2> Amaranth: interesting, here's a little excerpt from John Carmack on quake 3 networking about packet sizes
16:55 <+ammar2> "He thinks the notion of a 512-byte optimal MTU is pretty much at least 5 years out of date. He sees almost no problems with a max packet size of 1400 bytes, and fragmentation just isn't an issue. During the course of typical activity, the actual payload size is often smaller than the UDP header size (!), and even during hectic activity the packet size is on the order of several hundred bytes. The large packets (up to 1.4K) are 
16:55 <+ammar2> typically when a large amount of game state must be sent at once, such as at connection time when there is no state to delta against and when there is a lot of startup initialization going on."
16:55 <+Amaranth> haha, cargo cult ftw
16:59 <+ammar2> also quake3 does compress the state snapshots (beyond the diff) but I'm not sure how well they'd compress 
16:59 <+ammar2> here's the full article btw http://trac.bookofhook.com/bookofhook/trac.cgi/wiki/Quake3Networking
17:00 < morfin> wait is not Quake3 using UDP?
17:01 <+ammar2> please read more carefully :P
17:01 < morfin> oh it do
17:03 < morfin> i always thought that Quake3 just sends whole state  frame 
17:03 < morfin> which include data required to display entities etc(and allows wallhaaaaaaaaax)
17:03 <+ammar2> well sending deltas allows for wallhacks too
17:04 < morfin> HLDS/SRCDS as i know just send whole frame
17:04 <+ammar2> nope
17:04 < morfin> ?
17:04 <+ammar2> source engine does delta snapshots too
17:04 < morfin> oh 
17:04 <+ammar2> it only retransmits the whole thing if there's been major packet loss
17:05 < morfin> hmm
17:05 <+ammar2> fun fact: the original source engine, goldsrc was based on the quake engine
17:05 < morfin> yep i know
17:06 < morfin> but deltas could be pretty small and UDP have big overhead 
17:07 <+ammar2> yeah, a udp-snapshot based protocol definetly uses more bandwith
17:07 < gurun> just fyi. I have an exception thrown if the MTU is less than usual max. And i turned on no-framgentation just to be sure. And it is exceptionally rare to get anything like that. I had it running for weeks now.
17:07 < gurun> and thats usually on mobile and mobile networks.
17:13 <+ammar2> wallhacks are a really tough problem to solve on a network with latency, like aside from the fact that you have to keep track of what a client can see, the time it takes to update them about seeing the other player will cause hilarious pop in
17:13 <+ammar2> morfin: ^
18:16 <+Amaranth> ammar2: UDP only has eight bytes of overhead vs raw IP and is the smallest you can get while working outside of your own computer, how would it use more bandwidth than doing it via TCP?
18:17 <+Amaranth> I doubt a diff'ed snapshot system would even use more bandwidth than how MC works now too, assuming no packet loss
18:17 <+ammar2> Amaranth: sending snapshots x times per second as opposed to when changes happen usually does
18:17 <+ammar2> wait
18:18 <+Amaranth> It's not like that data can just be ignored in the TCP system, you have to send it anyway
18:18 <+ammar2> hold on I'm having a brainfart
18:18 <+ammar2> ok yeah I meant snapshots as opposed to using your own reliable wrapper on top of udp
18:19 <+Amaranth> A snapshot of "nothing happened" is going to be pretty dang small so even in the worst case I can't imagine much extra bandwidth used
18:19 <+Amaranth> And there is always something happening, the world is never idle
18:21 <+ammar2> fair enough
18:24 <+Amaranth> Oh and if your diff used offsets for positions they'd generally have a very tiny range which your compression system could take advantage of to get you a net win on bandwidth usage
18:25 <+ammar2> Amaranth: offset to what?
18:25 <+Amaranth> But that's math and CS that is beyond me, it's magic
18:26 <+Amaranth> ammar2: Offset to the previous snapshot the client has, of course, the same thing the diff is generated against
18:29 <+ammar2> hmm I wonder if there's a reason people don't do that right now
18:29 <+ammar2> both the source engine and quake protocol send absolute positions
18:30 <+Amaranth> They're old and the people who worked on them didn't have a CS degree so didn't know what entropy coding was?
18:30 <+Amaranth> Oh, and it's not very CPU friendly, at least in 1990s terms
18:30 <+ammar2> compression and entropy are scary scary
18:31 <+ammar2> is it?
18:31 <+ammar2> its literally just currentFrame.pos - lastAckFrame.pos
18:32 <+Amaranth> That's how you get the offset but entropy coding would be what would make it worth using an offset
18:32 <+Amaranth> Maybe worth it, anyway
18:32 <+Amaranth> In exchange for some loss of floating point precision from doing a subtract then add to get the current position
18:33 <+Amaranth> That might actually be the real reason
18:33 <+ammar2> floats are bad
18:33 <+Amaranth> You'd need to send the absolute position every so often to correct for drift due to lack of precision
18:33 <+Amaranth> I don't know if every so often would be every second or every hour though
18:37 <+ammar2> also it seems like the savings wouldn't be worth it unless you have a shit ton of position data which you don't really in fps
18:37 <+ammar2> at least for someone to consider using it anyway
19:08 < gurun> i'm old, but my head is full of entropy anyway. Does that make me a "new good cool coder" then?
19:58 <+Amaranth> Wow, I'm super rusty at graphics stuff
19:59 <+Amaranth> Just took me like 8 hours to make a camera
20:05 <+Amaranth> Writing it in a language I'm not comfortable with (Rust) using APIs that have seen three rewrites since I last looked for them probably contributed a bunch to that
20:05 < morfin> what's wrong about Rust?
20:06 <+Amaranth> Nothing, I just haven't done much with it
20:06 < morfin> i guess it even have bindings for OpenGL hmm
20:06 <+Amaranth> When I stop thinking about it and just let my fingers go I still spit out Java looking code
20:06 < morfin> )
20:07 <+Amaranth> glium is a wrapper for GL that has the goal of making it impossible to use GL wrong
20:07 < morfin> oh
20:08 < morfin> as i know whole Minecraft uses "raw" OpenGL
20:08 <+Amaranth> Ideally by blocking illegal actions with the API and type system but if nothing else it'll do its own check and panic rather than submit data it knows is junk to the driver (which might reject it, accept it (!), or crash)
20:08 <+Amaranth> Sure, there is gl-rs too
20:08 <+Amaranth> But it's all unsafe API
20:08 <+Amaranth> Raw C interface, basically
20:08 < morfin> )
20:08 < morfin> but you don't have to manage memory etc
20:09 <+Amaranth> With GL? Sure you do, you allocate buffers and textures and such and have to free them
20:09 < morfin> hmmm i am not familiar with GL :(
20:10 < morfin> well, maybe you do that via GL API
20:12 < morfin> i meant things like controlling memory allocated with malloc/new
20:13 <+Amaranth> How is that any different than allocating buffers of memory on the GPU via a GL call?
20:27 < morfin> it's not 
22:33 < morfin> hmm i am looking at things guys do with command blocks
22:33 < morfin> that's pretty interesting 
--- Day changed sam. juil. 11 2015
00:07 < gurun> shoghicp and the rest of pocket-mojangsters. I just have to say a big thanks for skins. Spending a week coding in a test-server would have been seriously dull if it wasn't for the skins.
00:13 <+ammar2> thx for da skinz volvo
01:33 < Not-48e3> [wiki.vg] Edit by Aragas to Library List -> http://wiki.vg/index.php?title=Library_List&diff=6700&oldid=6658
02:44 <+XorBoole> PE has skins now? neiss
05:10 < Not-48e3> [wiki.vg] Edit by 727021 to Classic Protocol Extension -> http://wiki.vg/index.php?title=Classic_Protocol_Extension&diff=6702&oldid=6618
05:14 < Not-48e3> [wiki.vg] Edit by 727021 to Classic Protocol Extension/Support -> http://wiki.vg/index.php?title=Classic_Protocol_Extension/Support&diff=6703&oldid=6659
13:00 < gurun> dang, that font used in PE is a mystery to me. I have to doublecheck if we are stil in 2015..
13:03 < gurun> and i love the way it changes font when input smething like a € sign. To an even more horrible font.
13:03 < gurun> i remeber when i had an ABC800 and did probably the first version of "true type" in the world. Looked exactly like that euro font :-)
13:04  * gurun just claimed credit for true type, fully aware of the implications
--- Day changed dim. juil. 12 2015
02:21 < morfin> strange
02:21 < morfin> i thought Minecraft write bytes in BE order to stream
02:22 < gurun> i think it may be a bit "mixed" .. 
02:23 < morfin> i mean plain types like int,long,double,float
02:27 <+ammar2> morfin: it does
02:27 < morfin> then it means             value <<= 8; value  |= buffer_[read_position_] & 0xff; do a trick
02:28 <+ammar2> where is that from
02:28 < morfin> because output is LE =)
02:28 < morfin> from bytebuffer.h
02:28 < gurun> i love the ; do a trick part :-)
05:23 < Paprikachu> in the player dig packet, is the format really a compressed block position? i get garbage values if i try to parse it as such.
05:35 < Paprikachu> nevermind.
07:26 < WizardCM> what would be the current go-to tutorial for getting started in clientside modmaking? thinking of trying my hand at making this https://www.reddit.com/r/minecraftsuggestions/comments/3co4p1/multiplayer_list_addition_categories_and/
09:46 < morfin> i forgot is Exposion separate entity?
10:27 < morfin> guys what happens if i send to server inappropiate packet? I mean situation when as example i am already playing but boom - send packet handshake etc?
10:27 < morfin> does server just ignore that?
10:35 < Techcable> morfin: Vanilla will throw a class cast exception, as Packet.handle will try and cast the play handler to the handshake handler when Packet
10:35 < Techcable> *is received
10:40 < morfin> silly )
10:44 < yawkat> Actually, will probably just not decode properly since the protocols have different packet id registrations. 
10:46 < morfin> what do you mean?
10:47 < morfin> that when playing it is Keep Alive?
10:48 < morfin> so if i try to decode that some shit happens
10:53 < yawkat> Yea, if you send handshake during play it'll decode as keepalive and die. 
10:56 <+Amaranth> morfin: The server will drop you
10:57 < morfin> yes i see
11:25 < gurun> wow, so the PC protocol reuses the ID's in different states.
12:19 < morfin> yes
12:22 < angal> Does somebody have list of all vanilla blocks with data: block name, protocol id, form, texture, material properties(like step sound)?
12:35 < morfin> gurun, so MCPE does not?
12:36 < morfin> in PC Minecraft 0x00 as example can mean request/keep alive/disconnect
12:36 <+Amaranth> MCPE's packet ids probably get assigned by raknet, they don't even control it
12:36 < shoghicp> lost history, what about MCPE?
12:36 < shoghicp> :D
12:37 < shoghicp> we control them!
12:37 < shoghicp> mostly
12:37 < shoghicp> they start at 0x82
12:38 < gurun> lol
12:39 < gurun> morfin, no
12:39 < morfin> i think different ids is more appropiate
12:39 < gurun> shoghicp, so you guys don't control the ID's in the PC protocol :-)
12:40 < morfin> because in case if  some retard  sends you hanshake when already playing you can just discard it
12:40 < shoghicp> pc protocol doesn't have ids anymore, they happen to be those :D
12:40 < shoghicp> afaik they are assigned in runtime
12:40 < morfin> what do you mean?
12:40 < gurun> eh, uh , ah, ooops .. hiding.
12:41 < gurun> i would love to implement a protocol with randomly assigned identifiers for .. messages.
12:41 < morfin> shogicp they're not assigned in runtime i think
12:42 < morfin> but when compiling/obfuscating(?)
12:42 < gurun> morfin, i think they are having another one of these parties at Mojang today :-)
12:42 < shoghicp> that's why it's afaik :P
12:42 < shoghicp> parties? not today!
12:42 < shoghicp> only these last two days
12:42 < morfin> so that shit like Packet00Handshake is autogenerated
12:42 < morfin> probably
12:43 <+Amaranth> Their class is just called Handshake or whatever and the packet id is assigned to it by the networking system when they register it
12:44 <+Amaranth> So the order they register them determines their packet id
12:44 < morfin> oh shi*
12:44 < morfin> wait there is no holes in numbers?
12:44 <+Amaranth> Shouldn't be
12:44 < morfin> i mean they're sequental
12:44 < morfin> hmmm
12:44 < gurun> hmm .. you guys .. behave like that's anything but implementational details of the java version and how "ID's get assigned".
12:45 <+Amaranth> ?
12:45 <+Amaranth> Packet ids are not random implementation details unless you're Mojang, everyone else needs to know what they are :P
12:45 < gurun> yes, and hence, irrelevant how they gets their number.
12:46 < gurun> it's not "generated" by runtime. As far as all you guys know, they are asssigned, EOS.
12:46 <+Amaranth> Well knowing how Mojang is doing it makes it easier to figure out what happened if they change
12:46 < gurun> yes
12:47 <+Amaranth> Like if Mojang inserts a packet in the middle it'll shift them all by one. If you know how Mojang generates the ids that kind of thing is easier to spot
12:47 < gurun> yes
12:48 < shoghicp> our ids get assigned at compile time (MCPE) :)
12:48 < gurun> yes, randomly :-P
12:48 <+Amaranth> Alright then, back to figuring out why my gbuffer only has solid colors in it and not useful data
12:48 < shoghicp> that would be fun. I'll do that for next release except the three top packets >:)
12:49 < gurun> hihi
12:49 < shoghicp> in fact, server will tell you a seed
12:49 < shoghicp> and you need to implement the rng and use that seed to determine the ordering
12:49 < gurun> and you have to figure out how to implement the random generator to match the seed :)
12:50 <+Amaranth> Hopefully they just used the mersenne twister from the C++ stdlib :p
12:50 < shoghicp> even better, server sends code that you need to execute in a VM to get the seed
12:50 <+Amaranth> Wait can you seed that?
12:50 < gurun> lol
12:50 < morfin> what you're talking about?
12:50 < morfin> oh ordering or registration
12:50 < shoghicp> morfin: next-gen MCPE protocol \o/
12:50 < gurun> lol
12:51 < gurun> shoghicp, are you starting to feel limited by the byte now?
12:51 < gurun> i mean, the protocol is growing
12:52 < shoghicp> yes! I'll send everything aligned to 9 bits! 
12:52 < gurun> in true bitstream fashion.
12:52 < shoghicp> we had a lot of bugs related to alignment and different platforms :s
12:52 < gurun> you already did that trick with metadata 3 bits .. and it didn't work you know.
12:53 < shoghicp> I did not do that, that's from PC :P
12:53 < gurun> haha
12:53 <+Amaranth> Were you assuming structure layouts and got screwed when you compiled with MSVC for Windows stuff? :)
12:53 < shoghicp> Amaranth: 32-bit vs 64-bit
12:53 <+Amaranth> Oh, a boring one :P
12:53 < shoghicp> MSVC compiler is only used for win phone
12:53 < shoghicp> so maybe there as well
12:54 < shoghicp> otherwise it's gcc and/or clang
12:54 <+Amaranth> It's not just for Win Phone though, it's a universal app now
12:54 <+Amaranth> Unless the marketing was lying :P
12:54 < gurun> marketing NEVER lies.
12:54 < shoghicp> I mean the compiler :P
12:54 < shoghicp> anyway, I'm running Linux at work so no MSVS for me :D
12:54 < gurun> shoghicp, that never lies either.
13:00 < morfin> hmmmm i don't get that
13:01 < morfin> every "context" starts from 0x00 
13:06 < KexyBiscuit> MSVC... I don't think that will work for Linux :P
13:14 < shoghicp> KexyBiscuit: exactly! I use clion :)
13:14 < KexyBiscuit> clang?
13:14 < shoghicp> it's win/win
13:14 < shoghicp> I can swap gcc with clang, I use both to test for strange bugs
13:15 < KexyBiscuit> :D
13:36 < morfin> clang is nice
13:36 < morfin> also it have scan-build )
13:41 < morfin> hmmm
13:41 < morfin> Minecraft sends packets to all users can't remember
15:24 < Fenhl> <shoghicp> pc protocol doesn't have ids anymore, they happen to be those :D
15:24 < Fenhl> what
15:26 <+Thinkofdeath> Fenhl: implementation wise they don't have fixed IDs, ids are given to packets in the order they are registered
15:27 < Fenhl> … oh
16:22 < gurun> Fenhl, is't like when you have kids. Same ID implementation.
16:58 <+XorBoole> lol
17:02 < Fenhl> o.O
18:08 < Not-a8a6> [1.8-Models] drXor pushed 1 commit to master [+9/-0/±0] http://git.io/vqjbE
18:08 < Not-a8a6> [1.8-Models] drXor f7ee528 - Iron Bars
19:31 < SopaXT> https://github.com/sopaxorztaker/gamecheck
19:40 < SopaXT> Hi UUID00
22:49 < Alan> so here's a question that may or may not be offtopic... does anybody know of a reason why opening a 1.8 world with either a Forge-1.8 or Liteloader-1.8 client would cause all the 2-high flowers to pop?
22:50 < Alan> This is without any mods loaded, by the way, just the modloader enabled... I get the same behaviour for both loaders (separately, haven't tried them together), and yet if i generate a world from the same seed with a modloader enabled, the flowers are fine
23:01 < Alan> the only related thing I can find is this: https://bugs.mojang.com/browse/MC-62140
23:01 < Alan> and it seems, if i watch closely, the world loads with the top block missing from all these 2-high flowers, and then once the game realises that fact, it pops the flowers into item drops
23:05 < Alan> how would I go about trying to track down where/why this is occuring?
23:25 <+Amaranth> It'd help to have some idea of what the actual bug was but basically you need to go through all of the diff between vanilla and forge and try to find the bug
23:26 <+Amaranth> Some logic they've removed, modified, or otherwise broken that leads to this
23:27 <+Amaranth> Alan: I'd start with diff to the World and WorldServer files then move on to diff to the block itself then curl up in a ball and cry
--- Day changed lun. juil. 13 2015
00:06 < Alan> Amaranth: maybe i can proceed directly to that last step? :P
00:08 < Alan> Amaranth: I'm wondering (knowing not much about the architecture of minecraft or mod loaders) if there's some kind of "world upgrade" step that is being performed erroneously from 1.8 -> modded 1.8, and if what is running there is the same thing that caused that old 1.8 bug that I linked
00:08 < Alan> is there documentation somewhere on getting started with running a "development" version of minecraft (i.e. MCP, using a debugger, etc.)?
00:09 <+ammar2> the only time there's a "world upgrade" is when the structure of the world files changes
00:10 <+ammar2> which has happened twice(?) in the history of minecraft versions
00:11 < Alan> surely there have been other cases where the way particular things are represented has changed though?
00:11 <+ammar2> well that doesn't really cause a full world conversion
00:11 < Alan> otherwise how do you get stupid bugs like losing the top half of every 2-high flower in the first place?
00:11 < Alan> ok, I'm not on about a world format conversion
00:11 <+ammar2> server loads chunks from file, runs physics, oh no that's not supposed to be there, plop
00:11 < Alan> I mean something that actually traverses the world rewriting blocks and entities 
00:12 <+ammar2> and then when it saves the chunks backs, no more flowers
00:12 < Alan> that have changed between versions
00:12 < Alan> does that never happen then?
00:14 <+ammar2> I can't really think of any examples like that
00:14 < Alan> hmmm
00:15 < Alan> maybe knowing what the fix for that bug was could help track down why the bug still exists in all loaders...
00:16 < Alan> ooh, that bug report may be all I need...
00:17 < Alan> the comments include how to fix it with mcedit
00:18 < Alan> although, it's still not certain why it behaves broken in modded minecraft...
00:18 < Alan> so maybe that won't work
00:18 < Alan> i'll have to give it a try ,i guess
00:21 <+Thinkofdeath> Forge is still 1.8.0 last I checked. If it is that bug you linked forge wont have the fix yet
00:25 < Alan> hang on, both forge and liteloader are 1.8.0?
00:26 < Alan> I made the mistake of thinking they were keeping somewhat up-to-date, rather than nearly a year old
00:26 < Alan> maybe versioning with the .0 is worthwhile for the unambiguity...
00:27 < Alan> I'm guessing we're going to see the same thing as last release then: once 1.9 is close, the modding community will move to the last 1.8?
00:29 <+Thinkofdeath> I think they are working on 1.8.7 since that should be the last 1.8 update before 1.9 (bar anymore issues)
00:57 < Alan> well that seemed to fix it... used MCEdit to rewrite all the flower top blocks to south-facing
00:57 < Alan> and the change doesn't break them
01:06 < Alan> so, i guess the conclusion here is (a) I've been using a version with that known bug, and (b) the fix for that bug probably fixed the loader, not the worldgen, therefore (c) stuff generated in 1.8.7 will have broken 2-block flowers in 1.8.0's loader
01:06 < Alan> although I'm still not sure about all that, so i'll have to see what it does on new chunks
01:06 < Alan> luckily this was a fairly new world, so I just pruned the world down to the bits I cared about and rewrote the flowers in that bit
15:00 < barneygale_> hey folks. does anyone know where the colour hints that minecraft applies to skylight during sunrise/sunset are stored?
15:34 < Not-48e3> [wiki.vg] Edit by SopaXorzTaker to Pocket Minecraft Protocol -> http://wiki.vg/index.php?title=Pocket_Minecraft_Protocol&diff=6704&oldid=6654
17:04 < Not-a8a6> [1.8-Models] drXor pushed 1 commit to master [+1/-0/±1] http://git.io/vmkYW
17:04 < Not-a8a6> [1.8-Models] drXor 86fab19 - Fix objects in flower bots being shaded.
19:34 < SopaXT> Anyone alive
19:35 < SopaXT> https://github.com/SopaXorzTaker/gamecheck
--- Day changed mar. juil. 14 2015
03:39 < Techcable> Is the sign nbt in json form?
10:35 < angal> Array of 4 lines of json strings.
10:46 < morfin> i don't understand
10:46 < morfin> is chat message just set of nested objects in JSON?
11:05 < morfin> how can client send colored chat?
11:09 < angal> He can't... He can receive it only.
11:18 < morfin> hm
11:19 < morfin> but what if use mods?
11:20 <+md_5> server will (or should at least) strip it / maybe even reject it
11:21 < morfin> hmmmm
11:23 < morfin> but discarding all colors styles is correct?
11:23 <+md_5> I can't even remember what vanilla does
11:23 <+md_5> It's either discard or kick
11:23 <+Amaranth> It definitely used to be a kick
11:24 < morfin> hm
11:26 < morfin> usually Minecraft server kick in any strange situation =)
11:26 < morfin> like sending invalid stance - GFTO
11:26 < morfin> sending wrong packet - GFO
11:28 < morfin> so does it mean there can't be nested JSON objects in client->server?
11:28 <+Amaranth> I thought client->server was still just a string
11:29 < morfin> oO no?
11:29 <+Amaranth> http://wiki.vg/Protocol#Chat_Message_2
11:29 <+Amaranth> Just a string
11:29 < morfin> wtf
11:31 <+Dinnerbone> Why would we accept a formatted chat object from the client if we're restricting them to plain text? :D 
11:31 < morfin> lol so it's much easier
11:31 <+Dinnerbone> So what's easier?
11:32 <+Dinnerbone> This is much much easier than taking a complex object and stripping it down to find a text object and validate that it's legal
11:32 < angal> :)
11:32 <+Dinnerbone> Also much cheaper, less bandwidth, more safe, and just makes much more sense
11:33 < morfin> it's cheaper but you can't as instance "paste" item
11:33 <+Thinkofdeath> (points Dinnerbone to signs and how they do exactly that)
11:34 <+Dinnerbone> (points Thinkofdeath to how that was a mistake and they won't do that in the next protocol change)
11:34 <+Dinnerbone> :D
11:34 <+Thinkofdeath> :D
11:35 < morfin> what was mistake?
11:35 <+Thinkofdeath> clients sending json to the server for signs
11:36 < morfin> so no colored signs any more?
11:36 < morfin> later
11:36 <+Thinkofdeath> newer 1.8 versions strip color anyway
11:36 < morfin> some servers was sending colored text
11:37 < morfin> to client
11:37 <+Thinkofdeath> server -> client is fine
11:37 <+Thinkofdeath> client -> server is what changed/changing
11:38 < morfin> yes probably unsafe
11:39 < morfin> witches are spawned even without hut?
11:40 <+Thinkofdeath> apparently
11:40 <+Thinkofdeath> http://minecraft.gamepedia.com/Witch#Spawning
11:41 < Fenhl> so commands should still be able to write formatted text on signs after that change, right?
11:41 <+Thinkofdeath> should be
11:41 <+Dinnerbone> Yes. It changes nothing vs what you already do now. It's just a protocol cleanup.
11:41 < Fenhl> cool
11:41 <+Dinnerbone> No actual changes to game, behaviour or anything like that.
11:42 <+Thinkofdeath> well hopefully signs would have their permission level reduced?
11:42 < morfin> game is working strange)
11:42 <+Thinkofdeath> because its still higher than command blocks
11:42 < morfin> middle of explosion with 3 hearts => survived
11:43 < morfin> *creeper explosion
11:43 < morfin> i know any solid block reduce damage to 0.5 but there was no such blocks 
11:43 < Fenhl> Thinkofdeath: “permission level”?
11:44 <+Thinkofdeath> yeah, theres 4(?) levels of permissions. Command blocks can't access op/deop/ban etc
11:44 < Fenhl> oh right
11:45 < Fenhl> so signs can use those?
11:45 <+Thinkofdeath> last I checked they could
11:45 <+Thinkofdeath> thats what made some of the recent exploits so bad
11:46 < Fenhl> I guess it kinda makes sense because using signs isn't automatable?
11:46 < morfin> what exploit?
11:47 <+Amaranth> Clients send json to the server for signs
11:47 < morfin> what happen if i send JSON with command when updating sign?
11:47 <+Amaranth> json can include on click things for commands
11:47 < morfin> :D
11:47 <+Thinkofdeath> morfin: in 1.8.1+ nothing
11:47 < morfin> that's what i was thinking about
11:47 < morfin> oh
11:47 <+Thinkofdeath> that was one of the many ways of doing it :P
11:47 < morfin> but click command for chat == click on sign?
11:48 <+Thinkofdeath> right click on sign but yes
11:48 < morfin> so i can trick op and he will op me)
11:48 < morfin> lol
11:48 <+Thinkofdeath> no need to trick
11:48 <+Thinkofdeath> the server runs the command
11:48 < morfin> oh
11:48 <+Thinkofdeath> The console would show "Thinkofdeath: Given op to Thinkofdeath"
11:49 < Fenhl> but that's patched now, right?
11:49 <+Thinkofdeath> yes
11:49 < morfin> i thought it send command from user right clicked )
11:49 <+Thinkofdeath> (any all the other ways found too)
11:49 <+Thinkofdeath> morfin: thats chat, signs work differently
11:51 < morfin> )
11:58 < morfin> so again server broadcast without radius check>?
12:03 < morfin> guys?
12:04 < morfin> but what about netherworld/overworld?
12:12 < morfin> oops
12:12 < morfin> i see
15:42 <+Thinkofdeath> Don't suppose anyone knows how minecraft's lighting works on a rendering level? Because I still can't it to look right
15:42 <+Thinkofdeath> (Textureless example: http://i.imgur.com/xUJgbs0.png)
15:48 < morfin> wtf is that
15:49 < ScruffyRules> morfin, Textureless.
15:49 < morfin> i see
15:49 < morfin> what makes you think that works not right?
15:50 < ScruffyRules> Thinkofdeath, Got a pic of Vanilla?
15:50 < ScruffyRules> same loc and stuff
15:50 <+Thinkofdeath> not textureless, no
15:50 < morfin> hmm 
15:50 < ScruffyRules> Mkay.
15:57 < dexter0> Thinkofdeath: I wrote <https://github.com/SirCmpwn/TrueCraft/wiki/Lighting> for TrueCraft, but it covers b1.7.3.
15:58 <+Thinkofdeath> newer versions use tinted lighting which is the bit I'm having issues with. Trying to get it to look like vanilla's shading
15:59 < ScruffyRules> Ask the bone!! :P
17:03 <+Thinkofdeath> I caved and cheated :3 (looked at minecraft's code) http://i.imgur.com/eDVKIf0.png
17:05 <+ammar2> nice
17:05 <+ammar2> whats that missing texture
17:05 <+ammar2> in the middle of the arch
17:06 <+Thinkofdeath> er, redstone lamp maybe?
17:06  * Thinkofdeath checks
17:07 <+Thinkofdeath> trap door with glowstone behind it
17:59 <+XorBoole> Thinkofdeath > implying steven was ever cleanroom
18:00 <+Thinkofdeath> More for the challenge than anything
18:17 < gurun> Hmm interesting link. Isn't it SirCmpwn always going on about that clean stuff.
18:46 <+XorBoole> like I can do anything about it. I'm almost as dirty as thinky here
18:47 <+XorBoole> imo cleanrooming anything minecraft related is silly, because mojang doesn't appear to care (or ms, for that matter)
18:47 <+XorBoole> ms is pro-modding last I checked; they released a visual studio plugin for forge
19:48 < gurun> XorBoole, did MS do that?
19:48 <+XorBoole> gurun yes
19:49 < gurun> but .. hmm , isn't that Java?
19:49 < gurun> yeah, jesus , it's java in VS
19:50 <+XorBoole> yes, VS has java support
19:50 <+XorBoole> #microsoft
19:50 < gurun> scary as hell
19:51 < gurun> i'm looking at it
19:51 <+XorBoole> I mean it's dumb, I know. but at least they're pro modding =D
19:51 < gurun> Cool, i was planning on adding some teplates and stuff via NuGET when that is possible.
19:52 < gurun> XorBoole i heard they made some "wide promises" during minecon regarding modding of MCPE. Possible not entirely shared by the rest of the crew.
19:52 < gurun> MS is very opensource friendly to developers these days.
19:52 <+XorBoole> they may or may not have made said promises to me
19:53  * XorBoole hides
19:53 < gurun> haha
19:53 < gurun> you don't care about MCPE do you?
19:53 <+XorBoole> of course I do
19:53 <+XorBoole> MCPE is the future
19:54 <+ammar2> there is no future
19:54 <+ammar2> its all in your head
19:54  * XorBoole pushes ammar2 into a dumpster full of windows surfaces
19:54 <+XorBoole> me and a friend debated "borrowing" one of the demo surfaces after minecon
19:54  * XorBoole runs
19:54 <+Grum> they all had gps tags in them
19:54 <+XorBoole> lol
19:55 <+XorBoole> his idea!
19:55 <+Grum> would have been fun to see your place raided !
19:55 <+XorBoole> Grum customs would have taken it before I even got to my house
19:55 <+Grum> nah, raids are far more fun than calling customs :p
19:56 <+ammar2> pftt no one is raiding anything for a single tablet
19:56 <+Grum> lies!
19:56 < gurun> customs work for Microsoft for sure. All evil organizations do. Like NSA, FDA, CIA and imigrations.
19:56 <+ammar2> instead they'd just execute XorBoole 
19:56 <+XorBoole> ammar2 but muh money
19:56 <+Grum> we'll send someone with a freshly sharpened banhammer
19:56 <+Grum> ammar2: that depends where he lives
19:56 <+ammar2> preferably in a really painful way
19:56 <+XorBoole> how do you even sharpen a hammer?
19:56 < gurun> .. and google not to forget. They are also evil, and they all belong to that Steve guy,.
19:56 <+Grum> taser to the melons?
19:57 <+ammar2> does that kill a man?
19:57 <+XorBoole> you really like melons, don't you?
19:57 <+ammar2> you don't?
19:57 <+XorBoole> well, you did see grummy's tweet, right?
19:57  * XorBoole runs
19:57 < gurun> melons? Is it ok to post pics?
19:58 < gurun> http://edgyentrepreneur.typepad.com/photos/uncategorized/2008/09/06/melons.jpg
19:58 <+XorBoole> gurun no, they're square. https://twitter.com/_grum/status/620967975655473152
19:58 < gurun> sic!
19:59 < gurun> :-)
19:59 <+XorBoole> very square
19:59 < gurun> hmm, https://lotgk.files.wordpress.com/2010/10/melons_2010_3.jpg?w=280&h=390 not really "Minecraftish" if you ask me.
20:00 <+XorBoole> this channel, every time
20:00 <+Grum> stop dissing my melons :(
20:00 <+XorBoole> I never dissed your melons, they're nice
20:00 <+XorBoole> shit food items, but nice
20:00  * XorBoole runs faster
20:02 <+ammar2> ur a shit food item
20:02 <+XorBoole> Grum wait, why is the left melon rotated? does left-hand flip or rotate a model?
20:02 <+Grum> my eyes are up here!
20:03 <+XorBoole> doesn't help your eyes are staring at the melons
20:03 < gurun> they turn square
20:06 <+XorBoole> inb4 1.9 breaks my entire resource pack for left-hand
20:06 <+XorBoole> echo "This pack does not support left-hand." >> README.md
20:07 <+Grum> no that will happen actually
20:08 <+XorBoole> Grum left-handed models?
20:09 <+XorBoole> actually, display: { thirdpersonleft: ... } I bet
20:09 <+Grum> nah
20:09 <+Grum> mainhand/offhand
20:10 <+XorBoole> so what happens if you're a lefty? does it flip the models then?
20:11 <+Grum> depends
20:11 <+Grum> unless you do different, yes
20:12 <+XorBoole> so by default left-hand models are flipped but you can override? 
20:12 <+XorBoole> rather, left = flip(right)
20:13 <+Grum> nah its more along the lines of there is a mainhand and an offhand
20:13 <+Grum> and a 'side' which is your mainhand
20:14 <+XorBoole> it's all moot really, we'll have the snapshot in a week or so and then I can start breaking it
20:14 <+Grum> that kinda depends on if i finish this hehehe
20:14 <+XorBoole> lol
20:15 <+XorBoole> you still planning on doing multipart?
20:15 <+XorBoole> wrote a model yestarday that would have benefited from that =p
20:19 <+Grum> yes
20:19 <+Grum> not planning, that is already in
20:34 <+XorBoole> did freenode just explode for anyone?
20:38 < morfin60> yes
20:38 <+ammar2> net split bby
20:42 <+XorBoole> felt worse than a netsplit
20:45 <+ammar2> probably because you were on the server that split
20:45 <+ammar2> clearly you just have a bad choice in servers
20:46 <+ammar2> much like everything else you do in life XorBoole 
20:51 < angal> :)
21:39  * XorBoole slaps ammar2 
22:40 < morfin> i just use random server
22:46 <+ammar2> I am a random server
22:55 < morfin> guys 
22:56 < morfin> is that nessesary to have synchronized netherworld/overworld and enderworld?
22:56 < morfin> or they can be easily splitted?
22:57 < Fenhl> splitting them will break redstone
22:58 < morfin> hmmm
22:58 < morfin> redstone can work across worlds? how?
22:59 < Fenhl> items can travel across worlds using portals
22:59 < morfin> yes i know
22:59 < morfin> but what about restone?
23:00 < Fenhl> some designs rely on the dimensions ticking synchronously for item transport
23:02 < morfin> hm?
23:02 < Fenhl> in fact I built a design like that yesterday
23:03 < Fenhl> transports items from a farm in the Nether to a storage room in the Overworld
23:03 <+ammar2> eww
23:04 < angal> I don't think i can be 100% compatible with all vanilla redstone bugs... So i don't think i should implement 100% of vanilla redstone features.
23:04 < morfin> lol hax
23:05 < Fenhl> it has to be lossless because people will soon be using it to send items manually, which will break if the relative tick speed of the dimensions is not constantly equal
23:06 < Fenhl> if the dimensions are not synced, the Nether can produce an arbitrary amount of items in any amount of Overworld time, and the hopper minecart won't be able to collect all of them, and they will despawn
23:06 <+XorBoole> today in redstone bugs
23:07 <+XorBoole> or, why it is impossible to cleanroom minecraft and remain sane
23:07 < Fenhl> and yes this is something that would happen even with tiny amounts of Nether lag
23:07 <+ammar2> yeah implementing every obscure mechanic is hard
23:07 <+ammar2> down with redstone
23:07 < Fenhl> well I agree when people say they don't want to implement bugs like qc
23:08 <+XorBoole> solution: synchronized (Minecraft.class) {} everything!
23:08 < Fenhl> but synced worlds is very much a feature
23:08 < Fenhl> *dimensions
23:08 <+ammar2> not really
23:08 <+ammar2> its just a consequence of every world running on a synchronized timer
23:08 <+XorBoole> spigot has broken bigger things =p
23:08 <+XorBoole> down with hoppers
23:09 <+Thinkofdeath> Hey! I fixed those
23:09 < Fenhl> ammar2: which is a feature, the devs have talked about that in here
23:09 <+XorBoole> Thinkofdeath did not. where's your hopper patch?
23:09 <+ammar2> its a "feature"
23:09 <+XorBoole> ammar2 +++
23:09 <+XorBoole> same as qc
23:09 <+Thinkofdeath> XorBoole: On default settings it matches vanilla now
23:10 <+Thinkofdeath> however hoppers are still slow as hell
23:10 < Fenhl> no, qc is wontfix
23:10 <+XorBoole> I don't want vanilla, I want hoppers nerfed into the ground!
23:10 < angal> What is qc?
23:10 < Fenhl> last i checked
23:10 <+XorBoole> imo qc should be fixed and replaced with a proper alternative
23:10 < Fenhl> of course
23:10 < Fenhl> angal: piston quasiconnectivity
23:10 < morfin> wtf is qc?
23:11 <+XorBoole> qc is the bug used in piston bud switches
23:11 < Fenhl> modern buds don't use qc
23:11 < Fenhl> http://minecraft.gamepedia.com/Issues/Piston_Quasiconnectivity_Bug
23:11 <+XorBoole> wontfix in this case is essentially "this is a feature now" in this case
23:11 <+XorBoole> in this case
23:11  * XorBoole hides behind ammar2 
23:12 <+ammar2> minecraft has many "features" :^)
23:12 <+XorBoole> I don't know how modern redstone works, nor do I intend to learn it
23:12 <+XorBoole> I do however enjoy breaking it on my server and watching people's lag machines break >=D
23:12 <+XorBoole> #uptime
23:13 < Fenhl> oh nvm it's closed as WAI now
23:13 < angal> ;0
23:13 < Fenhl> http://i.imgur.com/97gd4dG.jpg
23:14 < Fenhl> this is the issue btw https://bugs.mojang.com/browse/MC-108
23:15 < morfin> wtf
23:16 < morfin> wireless piston)
23:18 < Fenhl> I seriously don't understand Mojang in this regard. I mean, it might have been a sensible move to keep it in back in 2013 when the issue was closed, but by now there are better BUDs which don't use qc and it's just annoying to have to work around qc every time you're trying to build something
23:20 <+ammar2> eh its hard to fix and there's probably not much motivation to do it
23:20 <+ammar2> probably hard to fix*
23:24 <+XorBoole> ammar2 ever read the piston code?
23:24 <+XorBoole> I read it
23:24 <+XorBoole> it's not pretty
23:24 <+XorBoole> it's notchcode of the highest order (1.7 at any rate)
23:24 <+ammar2> yeah and I'd think they wouldn't wannna touch it too much
23:25 <+ammar2> otherwise they'd break obscure bug "features"
23:25 < Fenhl> there are mods which fix qc, so it should be doable
23:25 <+ammar2> and then people would whine
23:25 <+XorBoole> there's doable, and then there's doable but controversial
23:25 <+XorBoole> I've always said it, software would be best without users
23:25 <+ammar2> yeah
23:25 <+ammar2> I mean you've been to minecon
23:25 <+ammar2> you know this first hand
23:25 < Fenhl> they should have fixed it with 1.5, they promised breakage and bugfixes, we got breakage and no fix
23:26 <+XorBoole> also I love how creating a URLClassLoader, deleting the endpoint jar, and loading a class segfaults the jvm
23:26 <+XorBoole> ggoracle
23:26 < Fenhl> lol
23:26 <+ammar2> undefined behaviour
23:26 <+ammar2> ez pz
23:26 <+ammar2> workign as intended
23:26 <+ammar2> its a feature bro
23:26 <+XorBoole> I love segfaulting java
23:26 <+ammar2> relax
23:26 <+ammar2> ITS A FEATURE BRO
23:27 <+XorBoole> I need a Runtime.getRuntime().segfault() method
23:27 <+XorBoole> better yet put it in unsafe
23:27 <+XorBoole> oh wait, java 9 doesn't have unsafe
23:27 <+XorBoole> frigin nerds
23:28 <+ammar2> XorBoole: which is funny since they wanted to make it public and well documented at first
23:28 <+ammar2> and then they were like nah fuck that
23:28 <+ammar2> we'll just fuck everyone over
23:28 <+XorBoole> I know =<
23:29 <+XorBoole> oracle can go eat a bag of the finiest manure
23:29 <+ammar2> well a lot of libraries use Unsafe
23:29 <+XorBoole> yep
23:29 <+ammar2> so unless they wanna alienate a shit ton of their userbase
23:29 <+XorBoole> they say they want to provide alternatives
23:30 <+XorBoole> instead of just documenting Unsafe
23:30 <+XorBoole> of course, then everything will have to update
23:30 <+XorBoole> imo, I just think they should do what C# did, which was brilliant
23:30 <+XorBoole> unsafe {} and fixed
23:30 <+XorBoole> unsafe enters unsafe contexts with pointers, fixed pins stuff on the heap.
23:30 <+ammar2> java is a bad language
23:30 <+XorBoole> fact
23:31 <+ammar2> c# is so much better
23:31 <+XorBoole> now they're making it a bad VM, too
23:31 <+ammar2> except C# is C#
23:31 <+ammar2> its only downfall
23:31 <+XorBoole> C# is better for three reasons in my book
23:31 <+XorBoole> 1) it sucks less than java
23:31 <+XorBoole> 1) unsafe pointer arithmetic
23:31 <+XorBoole> 3) sane generics
23:31 <+ammar2> wheres 2
23:31 <+ammar2> god damnit xor did you forget how to count again
23:32 <+XorBoole> I'm a mathematician. counting is for squares
23:32 <+ammar2> math is literally just counting
23:32 <+ammar2> counting how shit of a life choice you've made that is
23:32 < humerusj> ^
23:33 <+XorBoole> but... most of the things I work with can't be counted =<
23:33 <+XorBoole> I can't count the reals. I tried once
23:33 < humerusj> I feel as though the imaginaries are harder
23:33 <+ammar2> nah
--- Day changed mer. juil. 15 2015
03:09 < rom1504> you can count them, you just need an infinite amount of time
03:09 < rom1504> easy
03:30 < Fenhl> no, that's the rationals. Reals can't be counted even in an infinite amount of time https://en.wikipedia.org/wiki/Uncountable_set
03:39 < rom1504> well it's a bigger infinity yes
03:40 < rom1504> but you can still "count" them (depending on how you define "count")
19:57 < Paprikachu> in creative, how do i know the metadata of a block the player creates? it seems like the type of the slot field is just the base type, not the metadata...
19:58 < Paprikachu> in CreativeInventoryAction that is
19:59 < Paprikachu> either that, or my slot parser is broken
20:06 < Paprikachu> mh, apparently it's in the damage field. weird.
20:07 < Gjum> damage = metadata
20:07 < Gjum> for any item
20:09 < Paprikachu> couldnt they just put it into the type field?
20:09 < Paprikachu> would be more consistent...
20:10 < Gjum> it's consistent among all inventory packets ^^
20:11 < Paprikachu> yeah, but you still store uint16 if you're saving the map to disk
20:11 < Paprikachu> idk
20:11  * Thinkofdeath stabs creative
20:12 < Paprikachu> is there a generic way to differentiate between blocks and items?
20:12 < Paprikachu> or do i have to have a list of all ids to do that
20:13 <+Thinkofdeath> <256 is blocks
20:13 < Gjum> id<255 = block, id>255 = item
20:13 < Paprikachu> oh, nice
20:13 < Gjum> but some blocks are items in inventory
20:13 < Paprikachu> my server currently lets me place items as invisible blocks, which is not that cool :D
20:14 < Paprikachu> < 255 or < 256? :p
20:14 <+Thinkofdeath> first item is 256
20:14 < Gjum> ^
20:15 < Paprikachu> alright, thanks
20:19 < Paprikachu> hm, sometimes "none" for blocks/items is represented as 0, other times as -1, what's the point of that?
20:19 < Paprikachu> especially annoying because i got signed/unsigned mismatches all over the place because of that :|
20:36 < angal> Hm... -1 are often special cases, when packet performs non default action.
21:51 < morfin> like?
--- Day changed jeu. juil. 16 2015
08:56 < Paprikachu> what happens if i send spawn player before player list updates?
15:45 < rom1504> the name of the player will be unknown
17:29 < Aikar> so, first we have PayPal chargebacks, now I have a user who filed a BBB Complaint wanting to be unbanned OR refund them more money than they spent on my server -_-
17:30 < Aikar> err meant spigot channel, not sure how I Was on here, but i guess still relevent :P
17:30 < Aikar> player used stolen password lists to login like 40 accounts to profit off their items, all on his IP, chat logs admitting its him, yet he still claims it wasnt him
17:43 <+XorBoole> Aikar and sometimes I'm jealous of your playercount, lol
18:31 <+Amaranth> People still pay attention to the BBB?
18:31 <+Amaranth> You can do the worst shit and so long as that makes you more than it costs to pay off the BBB you'll come out ahead
19:35 < gurun> BBB, Big Bad Bicrosoft?
20:33 <+XorBoole> gurun Buycrosoft?
20:33  * XorBoole runs and hides
20:33  * gurun thninks XorBoole runs and hides, a lot
20:33  * gurun runs and hides
20:33  * gurun playing blockhunt
22:04 < SopaXT> Hey!
22:05 < angal> Hey?
22:07 < SopaXT> I just want people to check out my github project and donate at my homepage
22:08 < SopaXT> What I like in this channel, you can get help on protocol stuff instantly.
23:06 < Jabbi> Hey, do you think it is worth, making a Command-Block-Script-language?
23:09 < dx> ouch
23:09 < dx> those words hurt
23:10 < Jabbi> oh sorry, i am bad in english
23:10 < dx> no that's not the problem
23:10 < dx> the problem is exactly what you meant to say.
23:11 < Jabbi> do you mean it is not?
23:12 < Jabbi> yes.. ok thank you
23:12 < dx> sigh. well. the problem is that command blocks are hacks over hacks over hacks and the whole thing is awful
23:13 < dx> but if awfulness doesn't bother you, yeah maybe some people might appreciate it (and i hate to say this)
23:13 < angal> There are already mcedit filter to convert code on some specified languige in command blocks... 
23:14 < Jabbi> ah I did not know that
--- Day changed ven. juil. 17 2015
00:20 <+XorBoole> wait are we talking about command block scripts?
00:20 <+XorBoole> pls no
00:21 <+XorBoole> command blocks are the worst thing to happen to this game
00:21 <+XorBoole> except maybe the rose -> poppy change
00:21  * XorBoole runs
00:30 < Jabbi> Why? @XorBoole
00:31 <+XorBoole> as people said above. it's a giant pile of hacks
00:31 <+XorBoole> mojang had good intentions but people lost their minds
00:32 <+XorBoole> command blocks don't need a scripting language, because such a language would be very unstable given how hacky the whole thing is
00:33 <+XorBoole> that, and /some/ people have decided to backport 1.9 stuff to 1.8 via command blocks
00:33 <+XorBoole> wat
00:34 <+ammar2> XorBoole: well you add something that's turing complete anywhere and you can bet people will abuse it
00:35 <+XorBoole> ammar2 it's not turing complete
00:35 <+XorBoole> minecraft's world is finite
00:35  * XorBoole runs
00:35 <+ammar2> pftt
00:35 <+ammar2> I'll cut you
00:35 <+XorBoole> also, there's turing complete and then there's useable
00:35 <+XorBoole> commandblocks are the definition of a turing tarpit
00:35 <+ammar2> brainfuck
00:35 <+XorBoole> see brainfuck is actually usable
00:35  * XorBoole hides
00:36 <+ammar2> XorBoole: highly, https://github.com/SirCmpwn/bf-irc-bot/blob/master/irc-bot.bf
00:36 <+XorBoole> I am aware of said bot
00:36 <+XorBoole> I wrote a brainfuck to c compiler
00:36 <+XorBoole> it also generates jni bindings
00:36 <+XorBoole> so you can hava a java program partially written in brainfuck
00:36 <+ammar2> always a useful thing to have
00:37 <+XorBoole> ofc. no one will use it, so it's the best software
00:37 <+XorBoole> no users = 100% user satisfaction
00:50 < alsob> Hi
00:51 < floxen> hi
01:11 < Fenhl> XorBoole: more like ERROR DIVISION BY ZERO% user satisfaction
03:30 <+XorBoole> Fenhl you mean Infinity
03:30 <+XorBoole> the user count is obviously an IEEE float
03:35 < Fenhl> that seems like a horrible idea
03:55 <+XorBoole> Fenhl no, it's Enterprise(tm)
03:55  * XorBoole hides under a rock
15:26 <+Thinkofdeath> Is it possible to get blacklisted from the authservers? o.O
15:27 <+Thinkofdeath> I'm getting InvalidCredentialsException from every account I have even after resetting the passwords :|
15:27 <+Thinkofdeath> (vanilla launcher or mine)
15:47 < redstonehelper> there's rate limiting afaik
15:48 <+Thinkofdeath> :|
15:48  * Thinkofdeath waits
16:21 < angal> :)
16:22 <+Thinkofdeath> The one time I decide to rework the login screen and now I can't test it :|
16:25 <+ammar2> Thinkofdeath: proxy
16:26 <+Thinkofdeath> but effort
16:26 < ScruffyRules> butt nothing.
16:27  * XorBoole pets Thinkofdeath 
16:28 <+ammar2> Thinkofdeath: you can test out your client's proxy settings too :^)
16:28 <+ammar2> two stones with one bird
16:28 <+ammar2> etc
16:28 <+Thinkofdeath> but that involves implementing them first
16:28 <+Thinkofdeath> :)
16:28 <+ammar2> well you get a new feature plus it'd be faster than waiting for the rate limiting to go away
16:37 < Aikar> Thinkofdeath: you done got hakked of course
16:37 <+Thinkofdeath> totally
16:37 <+Thinkofdeath> side note: Got in, switching to windows lets me log in
16:37 <+Thinkofdeath> no idea why
16:38 < Aikar> microsoft added anti linux/osx code obv
16:38 <+Thinkofdeath> must be
16:38 < Aikar> though isnt the windows .exe launcher some new native binary now, maybe uses diff server?
16:38 <+Thinkofdeath> no chance of me messing something up on linux, must be anti-linux code
16:38 <+Thinkofdeath> Aikar: nah
16:39 <+Thinkofdeath> it launchers the jar version
16:39 <+Thinkofdeath> *launches
16:39 <+Thinkofdeath> it just handles getting it and downloading java as far as I know
16:39 < Aikar> so its a launcher for the launcher just to help get proper JRE setup?
16:39 <+Thinkofdeath> basically
16:40 < zml> because people can't figure out how to get jdk8 on their own
16:40 < Aikar> wonder if hwaccel works on spotify linux yet
16:40 < Aikar> the fact all my icon did was bounce im gonna say no, but now how do i disable it ._.
16:41 < Aikar> hmm other launch method worked, and ui looks unbroken
16:41 < Aikar> holy crap i was wondering what this cache folder i found was, its spotify, time to move to diff drive
16:42 <+XorBoole> zml you just install the ask toolbar
16:42  * XorBoole hides under some rocks
16:43  * zml installs the ask toolbar on XorBoole
16:43  * zml pacman -S openjdk-8-jdk
16:43 <+XorBoole> > not apt-get
16:43 < Aikar> apt-get install ask-toolbar-spyware ?
16:43 <+XorBoole> Aikar ++
16:43 < zml> > superior arch
16:44 <+XorBoole> real men use debian
16:44 <+XorBoole> well, I use debathena but whatever
16:44 < zml> no real men use emacs on hurd
16:44 < zml> powered by butterflies
16:45 < Aikar> coffee or 5hr energy, 5hr takes less work and works better....
16:45 < Aikar> I really gotta give my body a break from this crap soon
16:45 < Aikar> ive been on it hard for few months now :3
16:46 < zml> go coffee or tea, don't use 5hr
16:47 < Aikar> im dependent atm
16:47 < Aikar> its been a damn stressful year work wise
16:47 < Aikar> was working 110 hr weeks on day job earlier this year, gave myself a break after, but then been hitting my MC server hard lately too
16:48 < Aikar> I now have 3 other devs working with me, so its great seeing the backlog get destroyed
16:48 < zml> uh 110 hr weeks wut
16:48 <+XorBoole> Aikar can I borrow them
16:48 < Aikar> major website redesign/new tech infrastructure and Im the only dev
16:48 <+XorBoole> also can I borrow like all your players
16:48 < Aikar> so it all lied on me
16:49 < Aikar> was a success (after pushing back multiple unreasonable deadlines) at least
16:49 < Aikar> clean launch w/o issues
16:49 < Aikar> where as other brands in our company who rushed their damn website relaunch was plagued with issues
16:51 <+XorBoole> Aikar somehow I feel you'd enjoy my custom developers against humanity deck
16:51 < Aikar> lol
16:51 < Aikar> thats an interesting idea
16:52 < Aikar> i wish i had more dev friends in local area
16:52 < Aikar> i have 1... and he doesnt work with me on anything, hes a microsoft crony ._.
16:52 <+XorBoole> I have too many at school, wanna trade?
16:52 < Aikar> (no offense to the now microsoft employees :P)
16:52 <+XorBoole> Aikar I think they said they technically aren't microsoft employees?
16:53 < Aikar> yeah i wish i hung out more with people in hs class
16:53 < zml> they're microsoft property, not employees :p
16:53 < Aikar> eh, its just labels :p
16:53 <+XorBoole> Aikar > hs
16:53 <+XorBoole> I'm in college
16:53 < Aikar> but i guess I cant disagree with that statement
16:53 < Aikar> oh right
16:53 < Aikar> I have a parent company but we dont call ourselves employees of the parent
16:54 < Aikar> hell, the original parents name we were disgusted to even be related to ._.
16:54 < Aikar> thankfully that brand is now dead :P
16:54 < zml> lol
16:54 < Aikar> been through a few acquisitions in my career.
16:54 <+XorBoole> I'm a mathematician, this is potato to me
16:55 < Aikar> hell my first job was with EMBARQ right as CenturyTel was acquiring it. I joined, they were like 'btw were in a merger', few months later 85% team cut, but I survived.
16:55 < zml> company_a + company_b = compan_{hugeblob}
16:55 < Aikar> (which is now CenturyLink)
16:56 <+XorBoole> zml this is a noncommutative addition ofc
16:56 < Aikar> so I guess let me rephrase 'no offense to the microsoft related people in here' :P
16:56 <+XorBoole> microsoft seem like nice guys
16:57 < zml> there are cool actual microsoft employees
16:57 <+XorBoole> at least, they didn't call security on me when I said "open source" at their minecon booth
16:57 < zml> lol
16:57  * XorBoole hides
16:57 < Aikar> yeah im sure employees are cool, I was purely meaning that devs on MS technology stick to mostly MS tech, so my friend never messes with the stuff I do
16:58 < zml> eh true
16:58 <+XorBoole> see-sharp!
16:58 < Aikar> XorBoole: and in my college, my class mates didnt exactly have great skill :/
16:58 <+XorBoole> I'd love C#, but I don't because the style conventions are garbage and it's microsoft
16:58 < zml> XorBoole does fancy math things and can use big math words
16:58 < Aikar> I think a majority of them really shouldnt of been persueing programming
16:59 < Aikar> I did some free lance work for one of them later down the road, she couldnt do it herself ._.
16:59 <+XorBoole> Aikar I'm surrounded by people who go to cs aiming to go to work for google and facebook. welp
16:59 < Aikar> yeah i guess that is one diff, I only cared to go for an associates, avoided having all that nasty debt :P
17:00 < zml> pretty sure a large number of people who are doing CS really shouldn't be
17:00 < Aikar> I didn't even learn anything, just did it for the paper to help me start the career
17:00 <+XorBoole> zml ++
17:00 <+XorBoole> there are too many cs students at my school
17:00 <+XorBoole> like 1/3 of undergrads
17:00 < zml> eww wat
17:00 <+XorBoole> I mean, my school has a great cs department, but
17:01 < Aikar> yeah I had one guy who I love to call "Career programmers", who only want to do it for a paycheck and have no passion for it at all/understand it well