10:24 < Grum> well it tries to; it fails often because of throttling on the server (so it pretends to be down)
10:24 < nyuszika7h> I guess protocol version wouldn't matter for the list detection, even if the old and new versions collide
10:24 < nyuszika7h> as you said it just tries twice
10:24 < nyuszika7h> if you get an error, it falls back
10:25 < nyuszika7h> weird error, or whatever :P
10:25 < Grum> it really isnt this hard
10:25 < Grum> assume the old protocol is incompatible in every single way with the new one
10:25 < nyuszika7h> I'm just curious and not explaining myself clearly, sorry :P
10:26 < Grum> this means that when we 'ping' the server using the new protocol we get some weird reply back from 1.6.4
10:26 < Grum> after that we try the 1.6.4 format ping, the problem now is that the 1.6.4 server often throttles connections 
10:26 < nyuszika7h> I see
10:26 < Grum> so you see it as 'offline'
10:27 < Grum> a 1.6.4 server would reply with a 'kick' packet when you talk to it in a broken protocol
10:27 < nyuszika7h> ah
10:27 < Grum> so we could detect 'old stuff' this way
10:27 < Grum> now, on actually connecting
10:27 < Grum> i think that just breaks hard :p
10:27 < nyuszika7h> yeah, I meant to talk about the list, not connecting
10:27 < nyuszika7h> Grum: idea
10:28 < nyuszika7h> would it be possible to ping the server with direct connect, to get MOTD, version and stuff? or would you rather avoid the hassle? :P
10:28 < Grum> ?
10:28 < dav1d> oh god variants
10:28 < Grum> varints :p
10:28 < dav1d> hopefully not notchian like variants
10:28 < Grum> wha? :(
10:29 < Grum> nyuszika7h: 'with direct connect' ?
10:29 < nyuszika7h> Grum: like, show MOTD like in server list when you use direct connect - though either you need to press connect twice or detect when the user stops typing (which might be inaccurate if the user is a slow typer)
10:29 < dav1d> :)
10:29 < dav1d> Grum: thanks for not inventing your own possibly broken stuff
10:29 < Grum> nyuszika7h: euuh not worth it
10:29 < nyuszika7h> probably better to keep it for the server list
10:29 < Grum> dav1d: i never do that kind of shit, i prefer using libraries
10:29 < nyuszika7h> I guessed so after thinking about it, Grum 
10:29 < dav1d> Grum: tell that notch :P
10:30 < Grum> we would have used the protobuf library, but we didnt want to add a useless 0.5mb JUST for varints
10:30 < Grum> so i just wrote my own implementation, then verified with how google did it
10:30 < nyuszika7h> nice to see MC using Netty, btw
10:30 < nyuszika7h> Spigot is using it already
10:30 < Grum> nyuszika7h: this just shows you have no idea what you are talking about hehee
10:31 < nyuszika7h> Grum: not much, indeed
10:31 < dav1d> Grum: little thing, why is the port a VarInt?
10:31 < dav1d> Grum: it can never be > than short.max
10:31 < nyuszika7h> yeah port is 1-65535
10:31 < nyuszika7h> or is 0 valid?
10:31 < nyuszika7h> I don't think so
10:32 < nyuszika7h> nyuszika7h@cadoth ~ % socat -d -d - TCP-LISTEN:0
10:32 < nyuszika7h> 2013/10/11 10:32:02 socat[30741] N reading from and writing to stdio
10:32 < nyuszika7h> 2013/10/11 10:32:02 socat[30741] N listening on AF=2 0.0.0.0:45870
10:32 < dav1d> a short can also represent 0
10:32 < nyuszika7h> 0 means random port, at least in socat
10:32 < dav1d> but 0 is a special value iirc
10:33 < Grum> dav1d: yeah, maybe we shouldn't do that
10:33 < Grum> makes little sense, it's almost ALWAYS two bytes long
10:34 < dav1d> Grum: yep
10:34 < dav1d> I didn't look through the protocol, but I guess you cleaned it up? like y coordinate represented as byte, short, int or double?^^
10:34 < Grum> not really
10:34 < Grum> not yet
10:34 < dav1d> (my fav)
10:34 < dav1d> ah ok
10:34 < Grum> and yeah that is horrid
10:34 < Grum> meh
10:34 < Grum> err hmmmz
10:35 < Grum> writeUnsignedShort i guess :/
10:35 < Grum> and readShort
10:35 < nyuszika7h> Grum: what type is the Y coordinate in 1.6.4?
10:35 < nyuszika7h> is it some integer type, or a float?
10:35 < Grum> many 
10:35 < Grum> depends
10:35 < Grum> double, float, int, short, byte :)
10:35 < nyuszika7h> seems legit
10:37 < nyuszika7h> Grum: btw, can you help with a Bukkit thing?
10:37 < nyuszika7h> I need to get the location of the player's crosshair
10:37 < nyuszika7h> I have some implementation done that works, ish
10:37 < dav1d> gah losing connection all the time
10:37 < nyuszika7h> but it's probably not the best
10:37 < Grum> rot of the head of the player?
10:37 < Grum> oh wait
10:37 < Grum> know, crosshair? that is trivial
10:37 < nyuszika7h> I use line of sight atm
10:37 < Grum> middle of the screen >:)
10:38 < dav1d> can't you reverse the x,y position of the screen to world coordinates?
10:38 < dav1d> reversing the matrices
10:45 < Grum> why?
10:45 < Grum> euuh
10:45 < Grum> hmmm no
10:45 < Grum> you need to do collission stuff
10:45 < Grum> nyuszika7h: i am fixing this for 1.7 btw
10:45 < nyuszika7h> fixing what?
10:45 < Grum> or ... well... might not make 1.7
10:45 < Grum> but i am fixing that ;)
10:46 < Grum> knowing where the client is aiming and clicking :P
10:46 < nyuszika7h> oh
10:46 < Grum> the client will tell the server what it aimed at in the future
10:46  * nyuszika7h will just steal from Essentials :P
10:46 < Grum> you can look at the bucket code
10:46 < Grum> but its rather expensive to do it
10:47 < nyuszika7h> Grum: if you wonder, my plugin is https://github.com/nyuszika7h/LightningKit/ (2.0.0 with refactored code and /strike + /light combined into /lightning [-e] is not pushed to GitHub yet, it's in dev)
10:48 < nyuszika7h> Essentials has that... but mine can make effects, and I do it for fun too :P
10:49 < Grum> it should be pushed to github WHILE in dev
10:49 < Grum> silly ;)
11:33 <+sadimusi> dav1d: I would love to play around with the new snapshot, but unfortunately I have a big deadline coming up tomorrow
11:43 < dav1d> sadimusi: good to hear, it's over tomorrow :). Good Luck!
11:45 <+sadimusi> dav1d: it probably won't just be over. There will be dozens of bugs to fix and I also have to deploy it sometime next week.
11:48 < nyuszika7h> I'm going to see if I can notice anything with the PA OpenAL thing in the snapshots
11:53 < nyuszika7h> Grum: I know, I'm going to push a 2.0.0-dev soon
11:53 < nyuszika7h> I don't push untested stuff though
11:53 < nyuszika7h> the current changes are tested, but was too lazy to push
11:55 < nyuszika7h> also, about untested stuff, as I'll use -dev versions pushing them won't be a problem (though in most cases I'll test before to make sure the code makes sense), I just need to make sure I don't RELEASE untested stuff
11:56 < dav1d> nyuszika7h: normally you make commits after every change
11:56 < dav1d> helps you a shitton out later on
11:56 < nyuszika7h> dav1d: I know
11:56 < dav1d> and push unttested stuff, it's fun :D
11:56 < nyuszika7h> dav1d: I'm just not used to hosting stuff with actual releases on GitHub :P
11:56 < dav1d> I always do and I don't care^^
11:57 < dav1d> nyuszika7h: use "tags" for releases
11:57 < nyuszika7h> I know, yes, found out about them in meantime
11:57 < dav1d> (internet please come back)
11:57 < dav1d> :(
11:58 < nyuszika7h> dav1d: how are you here if you have no internet?
11:58 < nyuszika7h> 3G or something?
11:58 < dav1d> jap
11:58 < dav1d> tethering
11:58 < Flemmard> *magic*
11:58 < nyuszika7h> ah
11:58 < Flemmard> so you *have* internet :P
11:58 < dav1d> but for how long :/
11:58 < dav1d> and no yt/streams
11:59 < nyuszika7h> dav1d: is your plan limited, or is it pay for usage?
11:59 < dav1d> 1GiB/month
11:59 < nyuszika7h> 100 MB/month here
11:59 < nyuszika7h> for 1000 HUF, that is...
11:59 < dav1d> ah nvm
11:59 < dav1d> 2GiB/month
11:59 < nyuszika7h> $4.6, and 3.4 EUR
12:00 < dav1d> paying 10€, 1k free minutes, 1k free sms, 1k free minutes into the same net
12:00 < dav1d> and the 2GiB internet
12:00 < nyuszika7h> dav1d: here, it's about 9000-10000 HUF/month for internet, phone and TV combined
12:01 < dav1d> so 50€?
12:01 < nyuszika7h> ADSL2+ 15M/900k (usually around 10M/768k, sometimes goes above that)
12:01 < dav1d> yeah we pay the same @home
12:01 < nyuszika7h> no, 30-34 EUR
12:01 < nyuszika7h> according to Google
12:01 < dav1d> we pay around 40€
12:02 < nyuszika7h> also we have satellite TV with almost 100 channels
12:02 < dav1d> almost 100?^^
12:02 < nyuszika7h> I think so
12:02 < dav1d> our satellite has more than 2k
12:02 < dav1d> (we don't use it no more though)
12:02 < nyuszika7h> *sigh*
12:02 < nyuszika7h> restarting i3 doesn't make it source /etc/profile
12:02 < nyuszika7h> time to source it manually
12:03 <+sadimusi> nyuszika7h: how can you survive on 768k up?
12:04 < nyuszika7h> sadimusi: I don't upload that much, for what I upload (rsyncing stuff from/to my server) it's usually enough, except for a few cases
12:04 < Flemmard> whatever bandwidth you have, it's *never* enough :>
12:05 <+sadimusi> I pay around $70 for TV (only around 50 channels) + Phone (don't use it) + Internet (150m/10m)
12:05 <+sadimusi> and for all practical purposes that's fast enough
12:05 < dav1d> switzerland...
12:06 < nyuszika7h> Client> [12:06:08] [MCO Availability Checker #1/ERROR]: Couldn't connect to Realms
12:06 < nyuszika7h> seems legit
12:06 < nyuszika7h> ._.
12:06 < nyuszika7h> why doesn't the launcher pass _JAVA_AWT_WM_NONREPARENTING to MC?
12:07 < nyuszika7h> I'll try wmname
12:07 < nyuszika7h> community/wmname 0.1-2
12:07 < nyuszika7h>     An utility to set the name of your window manager
12:07 < nyuszika7h> "an"
12:07 < nyuszika7h> seems legit
12:09 < nyuszika7h> heh
12:09 < nyuszika7h> nyuszika7h@ymlyna ~ % wmname --help
12:09 < nyuszika7h> nyuszika7h@ymlyna ~ % wmname
12:09 < nyuszika7h> --help
12:09 < nyuszika7h> dav1d: what was the wmname trick?
12:09 < dav1d> wmname LG3D
12:09 < nyuszika7h> thanks
12:09 < dav1d> but it has sideeffects
12:09 < dav1d> clicking on links automatically switches to chromium etc.
12:09 < dav1d> I don't like that
12:09 < nyuszika7h> dav1d: doesn't work :/
12:10 < nyuszika7h> oh
12:10 < dav1d> well, the parenting thing works for me
12:10 < nyuszika7h> toggling floating mode on and off helps
12:12 < nyuszika7h> also
12:12 < nyuszika7h> there really should be an option to log commands in the vanilla server
12:12 < nyuszika7h> and make "Game rule has been updated" messages more meaningful, like CraftBukkit does it
12:13 < nyuszika7h> well
12:13 < nyuszika7h> ok, MC starts fast with just ALSA
12:13 < nyuszika7h> time to install PA
12:13 < nyuszika7h> what I never understood is, why are ALSA volumes so weird without PA?
12:14 < nyuszika7h> you can't have exactly 50% and stuff
12:14 < nyuszika7h> it goes 48, 52, 56
12:15 < nyuszika7h> WTF
12:15 < nyuszika7h> I did "sudo systemctl restart sound.target"
12:16 < nyuszika7h> 'Welcome to emergency mode!"
12:16 < nyuszika7h> back
12:19 < nyuszika7h> ugh, can't get PA to work
12:22 < nyuszika7h> ok, works now
12:31 < nyuszika7h> woo, automatic wireless setup also works now
12:33 < nyuszika7h> weird, I can't reproduce the issue with PA
12:33 < nyuszika7h> 13w41a starts normally
12:34 < nyuszika7h> I wonder why some messages don't have timestamps, btw
14:05 < dav1d> because they are printed rather sent through the logger?
14:09 < nyuszika7h> dav1d: maybe
14:09 < nyuszika7h> s/maybe/I guess/
14:09 < nyuszika7h> also, can't reproduce the PulseAudio issue even with GNOME 3
14:09 < nyuszika7h> it starts up quite fast
14:10 < nyuszika7h> also, the new look of GNOME 3 is so much better than the earlier look
14:10 < nyuszika7h> the default background is slightly more vivid too
14:10 < nyuszika7h> though I still lack a taskbar or similar
14:47 < nyuszika7h> Dinnerbone: for when you're back - I can't reproduce the OpenAL freeze issue
14:48 < nyuszika7h> Linux ymlyna 3.11.4-1-ARCH #1 SMP PREEMPT Sat Oct 5 21:22:51 CEST 2013 x86_64 GNU/Linux
14:48 < nyuszika7h> pulseaudio 4.0-265-g35fe
14:48 < nyuszika7h> I'll post that on Mojira later... however now I have to go.
15:28 < nyuszika7h> friend reproduced bug on:
15:28 < nyuszika7h> Linux laptop 3.7.10-1.16-desktop #1 SMP PREEMPT Fri May 31 20:21:23 UTC 2013
15:28 < nyuszika7h> pulseaudio 3.0
15:34 < nyuszika7h> can't reproduce on:
15:34 < nyuszika7h> Linux dash 3.2.0-4-amd64 #1 SMP Debian 3.2.46-1+deb7u1 x86_64 GNU/Linux
15:34 < nyuszika7h> pulseaudio 2.0
15:35 < nyuszika7h> first uname: Arch Linux, second uname: openSUSE 12.3, third uname: Debian 7.1 (wheezy)
17:59 < AlphaBlend> how long can the player name be in the TAB window?
17:59 < AlphaBlend> i have a feeling the limit was increased a vew versions back
17:59 < Calinou> don't know but it depends on max players (more max players = more columns)
17:59 < Calinou> else it'll look weird
17:59 < AlphaBlend> no, the actual length of each name
18:00 < AlphaBlend> if it was over 15 in the past it would crash the client with name length too long
18:00 < AlphaBlend> that's why i'm saying i feel as if this was changed recently
18:00 < AlphaBlend> as i went on a server that had custom name lengths over 15
18:02 < Thinkofdeath> AlphaBlend: In the snapshot its 16
18:03 < AlphaBlend> hm
19:14 < redstonehelper> if alphablend comes back someone should remind him the scoreboard prefixes and suffixes are displayed as well
19:15 <+sadimusi> die anybody check how many players are sent with the server list ping?
19:20 < Dinnerbone> Up to 12 random players on the server
20:01 < TkTech> Dinnerbone: What was the reasoning behind 12 random players?
20:01 < Dinnerbone> Arbitrary limit, but more than a dozen doesn't make too much sense
20:01 < Dinnerbone> It's only useful for smaller servers
20:03 < TkTech> Just thinking. Plenty of servers don't have query enabled so there's no way to get the full player list.
20:04 < Dinnerbone> The server status packet is not intended to replace query right now
20:05 < Dinnerbone> For larger servers it would just be a waste to send so much data every ping
20:06 < Dinnerbone> I may be adding some data to the status request to ask for larger queries though, so we'll see
20:11 < TkTech> Yeah, a simple optional byte tools can set (or can be used in the client for a "Server Details" frame) to get all of them.
20:12 < TkTech> Chunk the response and give it lower IO priority for the servers with hundreds of players.
21:28 <+clonejo> Minecraft's varints are always unsigned, aren't they?
22:37 < Not-002> [netherrack] thinkofdeath pushed 1 commit to master [+5/-1/±7] http://git.io/Ng4Obg
22:37 < Not-002> [netherrack] thinkofdeath ae1d49a - all: Update to 13w41a. Some things are still broken
22:47 < Not-002> [netherrack] thinkofdeath pushed 1 commit to master [+0/-0/±4] http://git.io/hPwmcg
22:47 < Not-002> [netherrack] thinkofdeath 6dfbc52 - protocol: Corrected UseBed -> PlayerDigging
22:56 < ellisvlad> Hello World!
22:59 < ellisvlad> I don't suppose anyone knows anything about the huge server protocol changes in the snapshot xD
23:00 < Thinkofdeath> ellisvlad: http://wiki.vg/Pre-release_protocol
23:00 < Thinkofdeath> (Not updated to 13w41b yet)
23:04 < ellisvlad> ooh, thank you :P
23:09 < ellisvlad> Fun weekend ahead of me... attempting to add the new protocol to my C++ Minecraft server ;D
23:12 < Drainedsoul> if your C++ server is well-designed it shouldn't be difficult
23:12 < Drainedsoul> if it's difficult, might be time to pay off design debt.
23:19 < ellisvlad> :P It's not going to be hard, just I see myself missing something obvious and getting really mad in a few hours...
23:27 < Not-002> [bravo] MostAwesomeDude pushed 3 commits to master [+0/-0/±3] http://git.io/bYN-Bw
23:27 < Not-002> [bravo] Corbin Simpson 5f24385 - beta/protocol: Use hex numbers in logs.
23:27 < Not-002> [bravo] Corbin Simpson c724d24 - blocks: Add spawn egg. Refs #402.
23:27 < Not-002> [bravo] Corbin Simpson 402ac3a - blocks: Add nether star and beacon. Refs #413.
23:28 < Not-002> [bravo] MostAwesomeDude pushed 1 commit to master [+0/-0/±1] http://git.io/Nl91Dw
23:28 < Not-002> [bravo] Corbin Simpson 8206e48 - policy/recipes/blueprints: Add recipe for crafting beacons. Refs #413.
23:30 < Drainedsoul> ellisvlad: what C++ are you using?
23:36 < ellisvlad> Drainedsoul: What do you mean? :P C++11, linux?
23:36 < Drainedsoul> yeah, C++84, 98, 11, 11 w/some 14, etc.
23:47 < ellisvlad> mainly 98, with a bit of 11 to make chunk mapping a little easier :P
23:47 < Drainedsoul> :[
23:47 < Drainedsoul> saddest thing I've heard all day
23:48 < Grum> <ellisvlad> I don't suppose anyone knows anything about the huge server protocol changes in the snapshot xD <-- i do ;)
23:49 < dav1d> haha
23:49 < dav1d> would be horrible if you wouldn't
23:49 < ellisvlad2> ahaha
23:49 < Grum> not really, jens knows nothing about it for example
23:51 < ellisvlad2> I imagine you've been asked this a lot... But, why are all the strings becoming JSON formatted? :P
23:52 < Thinkofdeath> ellisvlad2: They are removing § and replacing with json
23:54 < Not-002> [netherrack] thinkofdeath pushed 1 commit to master [+0/-0/±6] http://git.io/qA4X6g
23:54 < Not-002> [netherrack] thinkofdeath 4804a47 - protocol: Updated to 13w41b
23:56 < benbaptist> yeah, JSON seems more logical and cleaner than using JSON.
23:56 < benbaptist> erm
23:56 < benbaptist> yeah, JSON seems more logical and cleaner than using the dollar symbol for seperating fields*
23:57 < ellisvlad2> I prefer to use escape characters ;) my server software only converts to JSON right before teh packet is sent :P
23:58 < benbaptist> Ah, heh
23:58 < benbaptist> I do that as well with my server software
23:58 < benbaptist> I should probably consider updating it to the snapshots, or, at least, 1.6.4
23:58 < Drainedsoul> why would your server software use special strings internally
23:58 < Drainedsoul> isn't that what data structures are for?
--- Day changed sam. oct. 12 2013
00:00 < benbaptist> woah, the java virtual machine is suddenly crashing every time I open the game.
00:00 < ellisvlad2> :P but I mean in my little string builder class, any text formatting etc. is still done with escape codes :P
00:01 < Drainedsoul> ...isn't that what data structures are for
00:01 < ellisvlad2> xD I suppose, I still like to avoid JSON until the final stages of dealing with an outbound packet though :P
00:03 < ellisvlad2> I'm sure a lot of the code in this project would make many programmers cry :P
00:04 < Drainedsoul> yeah I do that to, but I don't use escape codes internally
00:04 < Drainedsoul> https://github.com/RobertLeahy/MCPP/blob/master/include/chat/chat.hpp#L17
00:04 < Drainedsoul> internally chat messages etc. are just a stream of tokens
00:04 < ellisvlad2> *finds a particularly horribly written piece of code*...
00:05 < benbaptist> heh
00:06 < benbaptist> In Python, I mostly just use parameter='hello'-type variables when passing onto a function when working with any packet, including ones with JSON
00:06 < benbaptist> the JSON is done within the packet's function
00:06 < ellisvlad2> https://gist.github.com/ellisvlad/69368c76c6733f5e4432
00:06 < benbaptist> https://github.com/benbaptist/pymine2/blob/master/packets.py
00:06 < ellisvlad2> It will make you cry!
00:07  * benbaptist cries
00:08 < ellisvlad2> :P
00:10 < Thinkofdeath> http://wiki.vg/Pre-release_protocol Updated 13w41b
00:11 < Drainedsoul> sprintf, what language is this
00:11 <+sadimusi> benbaptist: how do you create a packet? just create the dict manually?
00:12 < benbaptist> sadimusi, You pass parameters to the function like you would normally, although with key variables, you can define them out of order. e.g. login_request(entity_id=5, difficulty=2, level_type='magical')
00:12 < benbaptist> and I don't have to specify every parameter. some can be kept at their defaults
00:12 <+sadimusi> benbaptist: so you define all packets twice?
00:13 < dav1d> clonejo: btw I missed your query, sorry about that, I'll be there tomorrow
00:13 < dav1d> internet issues ate my highlights
00:13 < benbaptist> sadimusi, No, I don't understand what you mean
00:13 <+sadimusi> ellisvlad2: I should totally register a new minecraft account called "';drop table users;--"
00:13 < benbaptist> LOL
00:14 < ellisvlad2> xD
00:14 < benbaptist> the new snapshot's protocol version is 0?
00:14 <+sadimusi> benbaptist: where is this login_request function defined? all I can find is the parsing method which defines every packet
00:14 <+sadimusi> yes
00:14 < ellisvlad2> I run injection checks on the player's username when they join the server xD
00:14 < benbaptist> wpah. weord/
00:15 < benbaptist> sadimusi, https://github.com/benbaptist/pymine2/blob/master/packets.py line 57
00:15 < Drainedsoul> or you could just use prepared statements
00:15 <+sadimusi> benbaptist: oh, I didn't really read the method names
00:16 <+sadimusi> benbaptist: I thought those were all types
00:16 <+sadimusi> benbaptist: so you do define them twice. once for parsing and once for sending
00:16 < barneygale> you probably want a packet object with .read() and .write()?
00:18 <+sadimusi> benbaptist: maybe take a look at this https://github.com/sadimusi/mc4p/blob/master/mc4p/messages.py#L54
00:18 <+sadimusi> benbaptist: you probably don't want to copy the way packets are constructed from mc4p, but the definitions are pretty good
00:20 < benbaptist> yeah
00:20 < Thinkofdeath> Grum: Any idea what causes this: http://hopper.minecraft.net/crashes/minecraft/MCX-318554/ (Seems to happen sometimes when connecting to my server but not vanilia)
00:21 < benbaptist> That looks pretty cool
00:46 < Grum> Thinkofdeath: you edited the server
00:49 < Grum> Thinkofdeath: 13w41b - Removed 0x08, 0x09, 0x0A, Removed unused fields from Player Position and Look and Join Game  <-- we did far more than that
00:50 <+sadimusi> Grum: did you really remap all the packet ids after removing those?
00:50 < Grum> obviously
00:50 < Grum> and that will continue to happen
00:50 <+sadimusi> that would have been my next question
00:50 < Grum> god order will even change when we rename them
00:50 < Grum> (and i sort the classes :D)
00:50 <+sadimusi> this will make it really difficult to memorize packet ids :/
00:51 <+sadimusi> guess we'll have to refer to them by name in the future
00:51 < dx> well, he wanted packet ids to be varints for a reason
00:51 < Grum> you cannot even name them by their number
00:51 < dx> (i believe that reason is because he hates remembering packet ids)
00:51 <+Amaranth> He expects to have more than 255 of them?
00:51 < Grum> unless you want to conveniently forget the other 4 'packet 0's :D
00:52 <+sadimusi> true
00:52 < Grum> i kinda do not want to have more than 30 :p
00:52 < dx> from what i gathered grum wanted to avoid reusing packet ids
00:52 < Grum> no i said we could if we cared to
00:53 < Grum> i do not care ot
00:53 < Grum> i am actually going to remove the ids
00:53 < Grum> and have it auto-assign
00:53 < dx> poor ids
00:53 < dx> we'll miss them
00:53  * Amaranth tries to think of what packet sends a single bool
00:53 < Drainedsoul> Player?
00:53 <+Amaranth> Err, single byte
00:53 <+sadimusi> that is just one giant fuck you to everyone building any kind of protocol library
00:53 < Grum> sadimusi: why? we can do it
00:53 < dx> yes indeed, you can say "fuck you"
00:54 <+Amaranth> He can build a protocol library :P
00:54 < Grum> we have a protocol library O.o
00:54 <+sadimusi> yes, but yours automatically updates :P
00:54 <+Amaranth> The numbers really aren't that useful
00:54 < Grum> oh, i am considering making all packets default scoped =)
00:54 <+Amaranth> Default scoped... what?
00:54 < Grum> yeah erm
00:54 < Grum> whats it called
00:55 < Grum> not public :P
00:55 < Grum> oh package .... errr fuck lost the name
00:55 < Dinnerbone> Default. But that's still as silly as you suggested it earlier.
00:55 <+Amaranth> Package private
00:55 < dx> i'm so lost with all these changes that i have no idea if they make sense anymore, or if grum and dinnerbro just went insane out of rage for the old protocol
00:55 < Grum> 'no modifier class access level' -- package private
00:55 <+Amaranth> Making things package private makes CraftBukkit a lot harder :P
00:55 <+Amaranth> Have to import things from mc-dev with zero changes just to make the compiler shut up
00:55 < Grum> its packets
00:56 < Grum> no-one does shit with packets
00:56 <+Amaranth> So not a lot harder
00:56 < Grum> and better, no-one SHOULD Do shit with packets >:)
00:56 <+Amaranth> CraftBukkit imports like 5 of them
00:56 < dx> a lot of people do shit with packets from bukkit plugins, but nobody wants to admit it
00:56 < Grum> dx: hehe yeah enjoy doing all that through reflection
00:56 <+Amaranth> dx: No one wants to _support_ it, big difference
00:56 < dx> fun
00:57 <+sadimusi> Grum: pretty much every protocol implementation out there is built on the assumption that the packet ids stay the same between updates
00:57 < Grum> sadimusi: so they are all wrong
00:57 < Grum> i mean, they also assumed they were unique
00:57 < Grum> they were, until they arent
00:57 <+sadimusi> adding protocol state isn't a huge problem
00:57 < Grum> you think this is painful? hehe imagine what happens when player names are not unique anymore (or rather, not the leading identifier for a person)
00:57 < barneygale> yeah, that's fine...
00:58 < barneygale> also, the eventual move to uuids won't really affect /protocol libraries/ will it?
00:58 < Dinnerbone> Not likely
00:58 < barneygale> remapping packet IDs between versions is a rung above that on the ladder
00:58 < Grum> yes it will .. obviously
00:58 < Grum> but not massive
00:59 < Dinnerbone> Nothing talks in names over protocol
00:59 <+sadimusi> Grum:  course it's possible to switch everything to packet names, but you're just forcing everything to do a lot of work for virtually no benefit
00:59 < Grum> i mean 1.7 is the bigger change :P
00:59 < dx> oh man
00:59 < Dinnerbone> But you'll get small impact in terms of extra knowledge about player profiles
00:59 < dx> i'm sorry we whined so much about json chat
00:59 < Grum> sadimusi: we're not forcing you to make libraries
00:59 < dx> really really sorry
00:59 < Grum> dx: oh what? lol ...
00:59 < Grum> BUT IT SUCKS HOW COULD HTEY EVER USE JSON
00:59 < Grum> OMG SO BIG SO SUCK
00:59 <+AndrewPH> such suck
01:00 <+sadimusi> I would have expected NBT :)
01:00 <+Amaranth> I actually argued that they should have used NBT
01:00 <+Amaranth> Because doing half of one and half of the other is worse than doing all NBT
01:00 < Dinnerbone> I hear NBT is an awesome format that we and everybody loves
01:00 < Grum> nbt sucks so badly its ....... sigh
01:00 < dx> here we go again?
01:01 < barneygale> I quite enjoyed implementing nbt decoding, but as a format it sucks :P
01:01  * Dinnerbone waddles off
01:01 <+Amaranth> Storing data in NBT is probably more bytes than doing so in JSON
01:01 < Drainedsoul> what kind of data?
01:02 <+Amaranth> Let's just say "in general"
01:02 <+Amaranth> You could make test cases that show it either way
01:02 < Drainedsoul> Storing numbers -- for example -- in JSON is pretty inefficient once you get above 4 digits.
01:02 < Grum> bad/nonexisting api, not tested, just sad
01:02 < dx> Amaranth: but it doesn't even have a binary string data type
01:02 <+Amaranth> And then any overhead on bandwidth JSON would add is fixed with gzip
01:02 < dx> have fun storing chunks in json :P
01:03 <+Amaranth> dx: Nah for that you use msgpack of course :D
01:03 <+Amaranth> Or "bson", that abortion of an encoding
01:03 < Grum> not sure howe we got into chunks
01:04 < Grum> also storing chunks is trivial
01:04 < Grum> its byteblobs. ....
01:04 < Grum> the entities+data is different however
01:04 < Grum> and then mostly data
01:04 <+Amaranth> Actually msgpack wouldn't be a terrible idea if you really are looking to get rid of NBT
01:05 <+Amaranth> It fits your criteria. It's fast, small, can store binary data, and you don't have to write it yourself
01:06 < Drainedsoul> writing metaprograms to write a packet parser
01:06 < Drainedsoul> so much fun
01:08 < Grum> i dont like msgpack .... for the fact it has 'nil' :(
01:09 < dx> definitely a deal breaker
01:14 < Grum> hehe hardly
01:15 < barneygale> Grum: how are you going to assign packet IDs, if you're not doing it by hand anymore? arbitrary (e.g. hash of class)?
01:17 < Grum> id++
01:18 < barneygale> for some ordering of classes that might change between versions?
01:19 < dx> changing between versions doesn't seem to be an issue
01:19 < Grum> like we do now, we register server and client bound packets by hand, we just do ++ on each registration, done? 
01:19 < Grum> i am not sure why changing protocol between versions matters
01:19 < dx> Grum: it matters with the old way of thinking about packet ids
01:19 < Grum> it literally doesn't work otherwise
01:19 < Grum> well, stop the old way of thinking O.o
01:20 < dx> i'm just pointing it out lol
01:20 < Grum> me too lol
01:20 < dx> lol
01:21 < dx> if i keep laughing out loud so often people around me will ask me what's so funny and i won't know how to explain it
01:23 < Grum> just tell them: "You should have been there"
01:44 < TkTech> sadimusi: Happen to enjoy Flask?
01:44 <+sadimusi> TkTech: yes, why?
01:45 <+sadimusi> (I wrote both minecraftskins.net and my bachelor's thesis with flask)
01:54 < dx> flask is <3
01:55 <+sadimusi> oh, and http://face.minecraftservers.org/TkTech runs on flask as well
01:55 <+sadimusi> so does http://stats.minecraftservers.org
01:55 <+sadimusi> so you could say I like flask very much
01:55 < dx> tktech's face runs on flask, the rest of it is php
01:55 <+sadimusi> the main site is php
01:55 <+sadimusi> sadly
01:56 < dx> i guess i wasn't so far off
01:56 <+AndrewPH> flask is gr8 :)
01:56 < dx> sadimusi: got any experience with "plone"?
01:56 <+sadimusi> never heard of it
01:57 < dx> python cms that uses a framework that is probably one decade old
01:58 < dx> ...welp, 15 years old
01:58 < dx> i'm going to have to mess with plone soon and i'd like to know how far it is from the niceness of flask and others
01:59 <+sadimusi> just telling from a few random doc pages I wouldn't use that thing voluntarily
02:00 <+AndrewPH> I've got a flask thing running via uwsgi behind nginx, and it can handle 1.5k concurrent connections without an issue, I got pretty amazed.
02:00 <+AndrewPH> ('course that's on a static template page, but with mysql it hits 250, my current connection limit, and doesn't sweat. mysql doesn't like it though lol)
02:00 <+sadimusi> the face service runs on nginx as well, but I haven't yet seen it under real user load
02:01 < dx> AndrewPH: nice!
02:01 <+AndrewPH> loader.io
02:01 <+sadimusi> looks cool
02:01 <+AndrewPH> ^not my app, I use it for load testing stuff
02:01 < dx> cloud based DoS
02:01 <+AndrewPH> hehe yeah
02:02 <+sadimusi> AndrewPH: I assumed as much
02:02 <+AndrewPH> but you have to confirm your site as your own
02:02 <+AndrewPH> so there's that
02:02 < dx> cool then
02:03 <+sadimusi> AndrewPH: without an option to use random URLs this service is pretty useless
02:03 <+sadimusi> for my purposes
02:03 <+AndrewPH> sadimusi: provide a way for the server to treat one url as random urls?
02:03 < dx> a random redirection url?
02:04 <+AndrewPH> yeah but all internal
02:04 <+AndrewPH> like no redirects are sent to the browser
02:04 <+sadimusi> but that means work for me :P
02:04 <+AndrewPH> true
02:04 < dx> lol
02:04 <+sadimusi> maybe I'll try that later
02:04 <+sadimusi> right now I have to get the main site running properly
02:04 < dx> well that kind of "redirect" is trivial with wsgi
02:06 <+sadimusi> the faces already look pretty great :) http://d.pr/i/qAI1
02:16 < TkTech> sadimusi: Trying a one-night project, going to bug you to try it tomorrow/tonight :)
10:50 < Thinkofdeath> Grum: Well back to looking for changes then
10:55 < Grum> Thinkofdeath: most of them are subtle things where we read stuff properly (signed vs unsigned), we renumbered some of those constants send in the code
10:56 < Grum> some packets only had one constant, so they just turned into a packet of that type
10:57 < Grum> soon we'll be working on the whole x,y,z conendrum
10:57 < Grum> we'll probably for block-coords pack it into a long
10:58 < Grum> for entity-coordinates we'll use that and 3xbyte for 1/256th of a block as resolution
10:58 < Grum> (which at 64+24 is still cheaper and more accurate than 3 floats)
10:59 < Thinkofdeath> Nice
11:00 < Not-002> [netherrack] thinkofdeath pushed 1 commit to master [+0/-0/±2] http://git.io/U6oCBQ
11:00 < Not-002> [netherrack] thinkofdeath 3fb1748 - protocol: minor fix
11:00 < Not-002> [netherrack] thinkofdeath pushed 2 commits to master [+0/-0/±2] http://git.io/E1SNIw
11:00 < Not-002> [netherrack] Cubixmeister 19ceb3e - protocol: Major optimizations in cfb8. Avoids 2 copies for every streamed byte.
11:00 < Not-002> [netherrack] thinkofdeath ed20de4 - Merge pull request #68 from Cubixmeister/master protocol: Major optimizations in cfb8.
11:54 <+md_5> Grum can we do something about 1.7 server replying to 1.6 ping :p
11:54 <+md_5> fix it!
11:55 <+md_5> add a magic header that is the ascii "minecraft" string to the start of each connection
11:57 < ellisvlad> 1.7 server replying to 1.6 ping is fine, I think. You could have a different MotD displayed, like: "We are on Minecraft 1.7! Reconnect in the updated cilent!"
11:58 <+md_5> ellisvlad no, a) vanilla doesnt, b) you cannot reliably detect a 1.6 ping from 1.7
11:59 < ellisvlad> I mean it could be added to vanilla, or left as an option for custom servers. I noticed 1.6 ping uses a different Packet ID, doesn't it?
12:00 < ellisvlad> 1.6 does give the protocol version in 0xFE C->S
12:04 < ellisvlad> Also, just a little thing I love about the kick packet S->C when pinged... Changing the string for the version number :P http://puu.sh/4O4ny.jpg
12:06 < ellisvlad> Ahh, exploring all my old screenshots... Good memories :P
12:35 < Grum> md_5: what? 1.7 server replies to 1.6 pings
13:26 < Thinkofdeath> Grum: Ok so I might of missed a few changes :P http://wiki.vg/Pre-release_protocol#13w41b
13:30 < Grum> Thinkofdeath: told ya! nice job :D
13:31 < Thinkofdeath> Grum: You still have more Entity IDs to change to VarInts :)
13:32 < Dinnerbone> We have only done an initial review on about a third of the packets
13:32 < Grum> took ~3-4hrs :p
13:32 < Dinnerbone> There's lots more to do
13:32 < Thinkofdeath> Ah ok then
13:32 < Grum> lots more to come on mondays ;)
13:32 < Grum> oh that note, i'm going to think of how to store x,y,z in a long!
13:33 < Grum> especially unsure about the whole unpacking and keeping the sign properly
13:35 < Grum> Thinkofdeath: you might want to make a note @ Protocol version; that eventhough it is both 0 its 100% backwards incompatible for the two snapshots
13:35 < Grum> (because of the port varint->short change)
13:36 < Thinkofdeath> Grum: Done
20:51 < Drainedsoul> so arrays in the protocol still aren't prefixed with a consistent type?
20:56 < Thinkofdeath> Drainedsoul: Not yet
20:56 < Thinkofdeath> Drainedsoul: <DB> There's lots more to do  <- So it may still happen
20:57 < Cubix> Have anyone thought about replacing CFB with more efficient mode?
21:07 <+md_5> <Grum> md_5: what? 1.7 server replies to 1.6 pings
21:07 <+md_5> did not for me
21:07 <+md_5> conclusion: buggy
21:07 < Drainedsoul> is there any other conclusion wrt Minecraft?
21:09 < Thinkofdeath> md_5: Works fine here (Server:13w41b Client:1.6.4)
21:11 <+md_5> Thinkofdeath I was using 41a
21:11 <+md_5> so mebbe
21:18 < Grum> yeah 41a != ok
21:18 < Grum> b works
21:19 <+md_5> Grum how did you do it? check 0xfe01fa ?
21:20 < Grum> worse
21:32 <+Amaranth> Timing?
21:33 < Grum> nah we read the full packet and check every single little piece of it
21:33 < Grum> if anything doesnt work out we revert to the normal protocol
22:35 < Not-002> [netherrack] thinkofdeath pushed 2 commits to master [+0/-0/±15] http://git.io/NrxvSA
22:35 < Not-002> [netherrack] thinkofdeath 8061749 - protocol: packets now have ids automatically assigned to them
22:35 < Not-002> [netherrack] thinkofdeath 8df8614 - protocol: Updated packets to correctly reflect 13w41b changes. Fixed VarInt encoding and added VarInt length support for arrays
22:36 < Drainedsoul> thinkofdeath: "[...] added VarInt length support for arrays" ?
22:37 < Thinkofdeath> Drainedsoul: http://wiki.vg/Pre-release_protocol#Statistics
22:37 < Drainedsoul> thinkofdeath: ah okay
--- Day changed dim. oct. 13 2013
00:28 < frostyfrog> Hmmm, out of curiosity, is it possible to play LAN with a single account? (Doesn't appear to be possible anymore to me)
00:29 <+sadimusi> you just have to spoof the username
00:29 <+sadimusi> or are you talking about the snapshot?
00:29 < frostyfrog> nope, not the snapshots.
00:30 < frostyfrog> My paren't don't really want to give my 5 year old brother an email address, that why I was asking. (Probably should've joined #minecraft and asked there though >_>)
00:30 <+sadimusi> just use a throwaway address if that's the issue
00:31 <+sadimusi> even though it would certainly be possible to spoof the name, we won't help you pirate the game
00:33 < frostyfrog> I always see the issue of using a throwaway email as... I can't change it later when he actually can get an email (as far as I understand anyways)
00:34 <+sadimusi> you never need to receive any emails besides the initial activation
00:34 <+sadimusi> so just use a long and random address and forget about it
00:34 < frostyfrog> but you have to log in with the email, correct?
00:34 <+sadimusi> right
00:34 <+sadimusi> so maybe not long and random
00:35 <+sadimusi> why not just use your parent's address? there probably the ones paying for it anyway
00:36 < frostyfrog> Yeah, but again. the issue there comes down to transfering the game to a different email address later.
00:36 <+sadimusi> why can't you just keep using your parent's address? do they need accounts for themselves as well?
00:41  * frostyfrog sighs, as they relize that this conversation is going nowhere.
00:41 < frostyfrog> Darn typos :/
00:41 <+sadimusi> it just sounds more like you don't want to pay for it
00:42 <+sadimusi> setting up an email address, disposable or not, really only takes a minute 
00:43 < frostyfrog> More like, we already have 3 copies for different people, but doing it your method would, in essence, require us to by 2 more. One for the "family", owned by the parents, and one for each individual (As they are allowed to get an email)
00:46 <+sadimusi> if you already have 3 copies, where is the problem?
00:46 < frostyfrog> if all 3 of those people are playing together, brother number 4 can't join in.
00:46 <+sadimusi> then just buy one more copy
00:47 <+sadimusi> you will be able to change the email address later btw
00:49 < frostyfrog> Oh hey look, it is possible :) Why didn't you say so before?
00:49 <+sadimusi> because I didn't think it way
00:49 <+sadimusi> *was
00:49 < frostyfrog> Sweet! Thank you sadimusi =)
02:19 < Not-002> [bravo] brutal-chaos pushed 1 commit to master [+0/-0/±1] http://git.io/dUGNUA
02:19 < Not-002> [bravo] Justin Noah 81ad3f9 - readme: dropped Windows support
12:06 < Drainedsoul> why is player UUID a string in the new "Spawn Player" packet?
12:09 < SinZ> uuid is slowly replacing player names
12:10 < SinZ> so playername renaming will work more
12:14 < Drainedsoul> but isn't a UUID just a 128-bit integer?
12:47 < ellisvlad> Quick question about the Ping response in the new protocol..
12:47 < ellisvlad> What are the last four bytes?
12:47 < ellisvlad> http://puu.sh/4OVvO.png
12:52 < ellisvlad> Or in fact the whole last line xD
12:57 < Thinkofdeath> ellisvlad: http://wiki.vg/Pre-release_protocol#Ping
12:58 < ellisvlad> ...wow xD
12:59 < ellisvlad> I swear I do actually read the wiki pages first :P
12:59 < Thinkofdeath> :)
13:00 < ellisvlad> ...that would explaine why when I used the time from my sample ping packet the client showed a ping of a few minutes xD
13:00 < Thinkofdeath> lol
13:08 < ellisvlad> Woo! http://puu.sh/4OW8H.png
13:21 < Thinkofdeath> ellisvlad: Got the icon working yet?
13:50 < Grum> Drainedsoul: depends on implementation
20:10 < LaxWasHere> Haaai
20:11 < pdelvo> hey
20:30 < iBotPeaches> notifico :(
21:48 < Not-002> [netherrack] thinkofdeath pushed 5 commits to master [+1/-1/±18] http://git.io/kgr2WQ
21:48 < Not-002> [netherrack] thinkofdeath 96bb21f - protocol: minor fix
21:49 < Not-002> [netherrack] thinkofdeath 53d6668 - entity: slightly cleaned up entity handling
21:49 < Not-002> [netherrack] thinkofdeath ad27365 - entity: fix ticking
21:49 < Not-002> [netherrack] thinkofdeath adcca4f - all: Change event handling
21:49 < Not-002> [netherrack] thinkofdeath d7c84c4 - world: support time and world unloading. Also fixed some data races
21:49 < Eviltechie> another not bot
21:49 < Eviltechie> :/
21:51 < dx> Eviltechie: if you look at the name, it's supposed to handle up to 999 instances!
21:51 < Eviltechie> that's dandy
21:52 < dx> but right now it's more like 999 channels per instance
21:52 < Eviltechie> I guess I'll have to /ignore Not-[.*?] then
21:52 < dx> yeah
21:52 < dx> errr
21:52 < dx> your regexp is weird
21:52 < Eviltechie> it might be wrong
21:53 < dx> just do Not-[0-9]+
21:54 < Eviltechie> meh
21:54 < dx> don't meh me, yours is broken
21:55 < Eviltechie> meh meh meh meh
22:18 < pdelvo> Not-\d{3}
22:34 < Eviltechie> pdelvo: what if it's in base 16?
22:34 < pdelvo> what if its not? :D
22:49 < Drainedsoul> I don't see how you can be on IRC and get irritated by a bot that spits a few lines out
22:50 < Drainedsoul> amidst the sea of people leaving and joining, that's what bothers you?
22:50 < Eviltechie> When people talk, the channel shows up in white. And then I get hopeful that something interesting is going on.
22:50 < Eviltechie> parts/joins are dark blue, and I can safely ignore those
22:51 < Drainedsoul> well something is going on.  Someone pushed a commit :D
--- Day changed lun. oct. 14 2013
04:39 < TkTech> Eviltechie: Your client so archaic it doesn't support /ignore -regex Not-(.*)
06:14 < Eviltechie> TkTech: Like irssi archaic?
06:23 < TkTech> ./IGNORE -regexp -pattern "Not-..." * ALL
06:23 < TkTech> ThatWasEasy™.
06:24 < Eviltechie> What if the bot suddenly starts doing something besides commit messages
06:24 < Eviltechie> Wouldn't want to miss that
06:27 <+AndrewPH> it won't
06:28 < Drainedsoul> thinkofdeath: You around?
09:24 < Not-002> [MCPP] RobertLeahy pushed 3 commits to master [+4/-0/±1] http://git.io/HZcZKw
09:24 < Not-002> [MCPP] RobertLeahy 5aea36f - Variant
09:24 < Not-002> [MCPP] RobertLeahy eca18d0 - JSON
09:24 < Not-002> [MCPP] RobertLeahy 16b7b00 - Makefile Improvements
10:28 < Not-002> [MCPP] RobertLeahy pushed 1 commit to 1_7_protocol [+0/-0/±3] http://git.io/10dwwA
10:28 < Not-002> [MCPP] RobertLeahy 43eb462 - Start of 1.7 Protocol
12:20 < Thinkofdeath> Drainedsoul: Yep?
12:30 < Drainedsoul> Thinkofdeath: Your gists with JSON in them have too many commas
12:31 < Thinkofdeath> Which ones?
12:31 < Drainedsoul> https://gist.github.com/thinkofdeath/e882ce057ed83bac0a1c there's a comma on line 83
12:32 < Drainedsoul> https://gist.github.com/thinkofdeath/6927216 lines 4, 11, 14
12:33 < Drainedsoul> when pitch etc. are sent as bytes, in steps of 2pi/256, is that sent as an unsigned byte, where the angle is between 0 and 2pi radians, or is it sent as a signed byte, between -pi and pi radians? :|
12:34 < Thinkofdeath> Signed
12:35 < Thinkofdeath> (Fixed the json)
12:35 < Drainedsoul> thanks
12:35 < Drainedsoul> (cool)
15:11 < dividuum> where is this channel publicly logged?
17:00 < Eviltechie> mcblockit has a freenode mask thingy?
18:00 <+Amaranthus> Eviltechie: They hand them out like candy these days
18:42 < dividuum> hm. why is LoginSuccess sent before EncryptionRequest when the server is on onlinemode=false?
18:43 < dividuum> oh. maybe it isn't. strange
18:44 < dividuum> ok. didn't switch into the correct state. sorry for the noise :-)
19:37 < Drainedsoul> so with 0x0E, play, S->C, it either ends with one signed 32-bit integer (which is zero), or a signed 32-bit integer followed by three signed 16-bit integers, just so I'm clear?
19:41 < Thinkofdeath> Drainedsoul: https://gist.github.com/thinkofdeath/75c3e5364745b57589f8 
19:53 < Drainedsoul> thinkofdeath: Thanks
20:40 < Drainedsoul> do metadata dictionaries still use UTF-16 internally?
20:44 < Thinkofdeath> Drainedsoul: Looks like they are UTF-8 too
20:45 < Drainedsoul> oh sweet
20:49 < Not-002> [netherrack] thinkofdeath pushed 1 commit to master [+0/-0/±1] http://git.io/vKDM_Q
20:49 < Not-002> [netherrack] thinkofdeath 90be52e - protocol: update metadata to use utf-8 strings
20:55 < Not-002> [netherrack] thinkofdeath pushed 1 commit to master [+0/-0/±2] http://git.io/IOvyUQ
20:55 < Not-002> [netherrack] thinkofdeath 8e5b0b3 - blocks: LockedChest -> StainedGlass
22:09 <+md_5> Thinkofdeath is the B protocol fully updated?
22:09 < Thinkofdeath> md_5: should be
22:10 < Thinkofdeath> http://wiki.vg/Pre-release_protocol#13w41b
--- Day changed mar. oct. 15 2013
00:03 < Drainedsoul> does anyone have any canned example of bad NBT?
00:03 < Drainedsoul> *examples
00:58 < dividuum> i really like those unique packet id for searching on the wiki %)
01:00 < dividuum> hm. does Confirm Transaction really use unsigned byte and signed byte for the different directions?
01:17 < Drainedsoul> does it matter?
01:17 < Drainedsoul> if you use one consistently in your implementation, it shouldn't be relevant because the bit representation will be identical
01:20 <+clonejo> unless he uses < or > :P
01:30 < Drainedsoul> well it's an ID, I wouldn't think that you'd numerically compare IDs
01:50 <+Amaranth> That reminds me, I wonder when they'll get rid of integer ids for enchantments
01:51 <+Amaranth> And move potions to the item metadata format instead of data values
01:51 <+Amaranth> Since they're getting rid of magic numbers
01:57 < dx> it would make sense to keep magic numbers for enchantments though
01:57  * dx cough
01:58 < dx> is that item metadata format going to survive, or will it be replaced by json as everything else?
02:00 < Drainedsoul> if IDs are magic numbers, what will you replace them with that's less magic
02:01 < Drainedsoul> how is "wood plank" less magic than 5
02:01 < Drainedsoul> other than the fact that "wood plank" is 10x longer.
02:05 < Eviltechie> why is 5 a wood plank?
02:05 < Eviltechie> Why not 10, or elevendy billion
02:06 < Drainedsoul> why is the exact arrangement of bits "wood plank" a wood plank?
02:06 < Drainedsoul> the protocol isn't written for human consumption, machines can consume fixed-width integers much easier than strings.
02:07 < Eviltechie> I'm not sure where you got the idea it would be sending strings
02:07 < iBotPeaches> change them all to strings prefixed with "minecraft_item_metadata_"
02:07 < Eviltechie> :/
03:38 < Not-002> [MCPP] RobertLeahy pushed 5 commits to 1_7_protocol [+0/-4/±8] http://git.io/GSdYfQ
03:38 < Not-002> [MCPP] RobertLeahy 59cc7a3 - Variant Documentation
03:38 < Not-002> [MCPP] RobertLeahy 9b7acf1 - Remove Metadata
03:38 < Not-002> [MCPP] RobertLeahy de115f1 - Compression Fix
03:38 < Not-002> [MCPP] RobertLeahy 3ffec25 - NBT Overhaul
03:38 < Not-002> [MCPP] RobertLeahy 7180177 - Packets
03:42 < dividuum> the entity id of the client bound Animation packet (0x0b) has to be a varint
05:20 < Drainedsoul> http://wiki.vg/Slot_Data the last example, libz does not seem to like the last four bytes of that.
06:05 < iBotPeaches> Drainedsoul: lol spells CAFE BABE
06:06 < zml> lol, somebody wanted to start a classfile there
06:52 < Not-002> [MCPP] RobertLeahy pushed 4 commits to 1_7_protocol [+0/-0/±5] http://git.io/tZzadA
06:52 < Not-002> [MCPP] RobertLeahy 2bf6f0c - JSON
06:52 < Not-002> [MCPP] RobertLeahy 9797680 - NBT
06:53 < Not-002> [MCPP] RobertLeahy 8557cc5 - Compression Fix
06:53 < Not-002> [MCPP] RobertLeahy de134a7 - Packets - NBT and Slot
07:25 <+md_5> Thinkofdeath string arrays are varint and byte arrays are short?
10:03 < Thinkofdeath> md_5: Looks like it
10:16 < Thinkofdeath> md_5: ej.class is where they are handled
14:51 < Jckf> Does anyone here have a nice PHP class to interface with the new authserver?
15:56  * frostyfrog might not, does a new one need to be coded? Has the auth server changed recently?
16:03 < Jckf> "New" since the "new" launcher
16:03 < Jckf> Going to use it for SSO on Minecraft.no, since we have a lot of tool with separate user databases today
16:03 < Jckf> *tools
16:54 <+sadimusi> Jckf: so your users have to enter their mojang account credentials?
17:02 < TkTech> sadimusi: If Mojang refuses to add OpenID/OAuth what are else can you do?
17:02 < TkTech> It was painful getting Notch just to add the skins URL.
17:02 <+sadimusi> the usual fake server you have to connect to
17:03 < TkTech> Which is limited in what you can do. Authorization vs authority.
17:04 < TkTech> For the skindex, I wanted you to be able to click a button and bam, skin changed.
17:04 <+sadimusi> that's almost possible
17:05 <+sadimusi> but in Jckf's use case authorization is the only concern if I got this right
17:07 <+sadimusi> TkTech: https://minecraft.net/skin/remote.jsp?url=http://www.minecraftskins.net/static/skin/harveydent.png <- this does exactly what you want, just with the authentication on the other end
17:07 < TkTech> sadimusi: I know. I had him add that URL.
17:09 <+sadimusi> of course I would welcome OpenID as well, but even without it it's still unacceptable to just ask for the password
17:22 < TkTech> sadimusi: (Which I believe he only did because I said I'd start storing passwords.)
17:23 <+sadimusi> TkTech: maybe the same would work for openid
17:23 < TkTech> This was back when the most used sites were minecraft.net -> forum -> wiki -> <wide gap> -> skindex -> <everyone else>
17:23 < TkTech> I would have been storing around a million passwords a month assuming every unique logged in.
17:24 < TkTech> Don't think any one site has that much "push" these days, other than the wiki/forum.
17:26 <+sadimusi> maybe not a single one, but a few combined might have
17:26 <+sadimusi> especially server lists are having problems with fake votes
17:27 <+sadimusi> if the big ones started requiring mojang accounts it would certainly trigger some kind of reaction
17:29 < TkTech> That's true, but it's an empty ultimatum. These days storing passwords would make you an instant cracking target.
17:29 < TkTech> That's a lot of liability that I don't think most lists would go for.
18:03 < edk> TkTech: sadimusi Jckf  Kyle made a thing for that http://mcidentify.com/user/session
18:03 < edk> er
18:04 < edk> i'm not sure if that's the correct url, it redirects there and i didn't notice. *shrug*
18:04 < edk> http://demo.mcidentify.com/test
18:05 < Jckf> I do not intend to store passwords. All I need is a valid session, since that indicates the user is real
18:05 < edk> The thing is, anyone who takes your word for it is insane.
18:06 < edk> the trustworthy (and safe) way to not store people's passwords is by not asking for them
18:07 <+sadimusi> Jckf: it's already hard enough to convince people not to give their password to random websites. legitimate websites asking for them only makes it worse
18:15 < Jckf> Would it help if I said the tools aren't for end-user, but administration?
18:15 < Jckf> Most of our users will never be asked for their Minecraft account
18:16 <+sadimusi> that makes it a bit better
18:16 <+sadimusi> but why do you need mojang accounts for the admins?
18:23 < Jckf> Makes it simpler for me to code and manage
18:31 < Kyle> Jckf: by all means, use my site, but it's a similar issue to the password thing
18:31 < Kyle> trust trust and more trust
18:31 < dividuum> is there a way to man-in-the-middle minecraft connections without having the users credentials?
18:31 < Jckf> dividuum: I haven't read up on how the launcher does this, but I appears to me to only require the user's password once. From there on it keeps a token that it validates?
18:32 < dividuum> that's the idea. i had some problems with that
18:32 <+sadimusi> dividuum: there's no way, at least not known to us
18:32 <+sadimusi> dividuum: (I'm counting the access token as credentials)
18:32 < dividuum> https://mojang.atlassian.net/browse/MCL-1513 (sorry for bringing this up again) :}
18:33 <+sadimusi> you could try proxying the auth server and compare your program to the RI
18:33 < dividuum> i guess i'll have to reread the proxy.py in mc4p then. i somehow remembers that there was no authentication needed
18:35 < dividuum> for my current project i have to forward the connection on behalf of the user
18:35 < dividuum> and i really don't want the credentials. i'm currently trying find ways of doing that. or maybe getting close to it
18:36 <+sadimusi> on which site do you run your program
18:36 <+sadimusi> ?
18:36 < dividuum> wait. i can show you (i think. let me see)
18:38 < dividuum> point your minecraft 1.6.4 to miners-movies.com
18:39 < dividuum> i'll let you figure out the way this is supposed to work :)
18:39 < dividuum> the portal to your right allow you to forward your connection to another server for recording purposes
18:40 <+sadimusi> so you just forward the connection between your own servers?
18:40 < dividuum> for that i need a way to authenticate as the user. the only way to do this at the moment is asking for the username/password
18:40 < dividuum> no. i can forward to any other server
18:40 < dividuum> otherwise i'd disable online-mode
18:41 < Jckf> dividuum: Can't you just run your back-end servers in offline mode and have the inaccessible to the world?
18:41 < dividuum> so it's not bungeecord
18:41 < dividuum> the user can forward to any server
18:41 < Jckf> Any? The address is user profided?
18:41 < Jckf> *provided
18:42 < dividuum> yes
18:42 < dividuum> what you can do with that then is this: dividuum.made.miners-movies.com (point your 1.6.4 minecraft there)
18:43 < dividuum> you can record yourself playing on the server and later watch the footage in game
18:43 < Jckf> Oh. You're recording the packets?
18:43 < dividuum> yes
18:43 < Jckf> Haha. Cool
18:45 < dividuum> it would be awesome to have another authentication flow somehow
18:46 < dividuum> i thought about providing a client side mod that responds to signed messages and returns the session id
18:47 < dividuum> but that feels ugly :-)
18:47 <+sadimusi> if you use a client mod at least do it the other way around
18:48 < dividuum> yeah. then i don't have to store anything. that would be nice. now i just have to figure out how to write those mods...
21:29 < Not-002> [netherrack] thinkofdeath pushed 1 commit to master [+0/-0/±1] http://git.io/4DHrjg
21:29 < Not-002> [netherrack] thinkofdeath 15ddddc - protocol: Animation's EntityID type corrected to VarInt
22:06 < Not-002> [fCraft] fragmer * r2232 2 files : Fixed linewrapper occasionally inserting redundant leading color codes, forcing an extra line to be printed (thanks AndrewPH).
22:17 < Not-002> [FemtoCraft] fragmer * r131 2 files : Updated LineWrapper, now based on fCraft 0.900_r2232_dev
22:22 < Not-002> [netherrack] thinkofdeath pushed 3 commits to master [+1/-0/±4] http://git.io/SANEBw
22:22 < Not-002> [netherrack] thinkofdeath 75061e5 - /: Removed unused code
22:22 < Not-002> [netherrack] thinkofdeath 725e002 - /: Support changing the new ping
22:22 < Not-002> [netherrack] thinkofdeath c73c607 - protocol: clean up readVarInt
22:24 < Thinkofdeath> http://wiki.vg/Pre-release_protocol#Multi_Block_Change <- If Chunk X (or Z) is negative wont the VarInt always be 5 bytes?
22:24 < Not-002> [fCraft] fragmer * r2233 2 files : Fixed linewrapper occasionally inserting redundant leading color codes, forcing an extra line to be printed (thanks AndrewPH) (backport from 0.90x)
22:49 <+clonejo> Thinkofdeath: doesn't MC's implementation of varints differ between signed and unsigned types? (http://en.wikipedia.org/wiki/LEB128#Signed_LEB128)
22:50 < Thinkofdeath> clonejo: Nope always unsigned as far as I can tell
22:51 <+clonejo> okay
23:24 < Not-002> [fCraft] fragmer * r2234 2 files : 0.643 final
23:24 < Not-002> [fCraft] fragmer * r2235 2 files : 0.643 tag
--- Day changed mer. oct. 16 2013
03:42 < Not-002> [MCPP] RobertLeahy pushed 3 commits to 1_7_protocol [+0/-0/±8] http://git.io/t-nG-w
03:42 < Not-002> [MCPP] RobertLeahy 34e3ed4 - No Static Linking
03:42 < Not-002> [MCPP] RobertLeahy 2114837 - GCC 4.8.1
03:42 < Not-002> [MCPP] RobertLeahy 7946b3b - Packet Serialization Performance Improvement
03:54 < iBotPeaches> gcc 4.8???
03:54 < iBotPeaches> how did I miss this
03:57 < Drainedsoul> it's been around a while o_o
03:57 < Drainedsoul> I've been using GCC 4.8 for almost a year.
03:57 < Drainedsoul> just downloaded MinGW GCC 4.8.1 today
05:26 < Not-002> [MCPP] RobertLeahy pushed 2 commits to 1_7_protocol [+0/-0/±2] http://git.io/wocMAQ
05:26 < Not-002> [MCPP] RobertLeahy 05ccd02 - Packet Documentation
05:26 < Not-002> [MCPP] RobertLeahy e93dc30 - Serialize Fix
06:23 < Drainedsoul> Thinkofdeath: So how is a negative chunk encoded?
09:24 < Thinkofdeath> Drainedsoul: The int is treated as unsigned (-1 as 0xFFFFFFFF for example)
09:29 < SinZ> varints do a zigzag encoding for signed numbers
09:29 < SinZ> unless MC reinvented varints
09:30 < Thinkofdeath> SinZ: zigzag encoding isn't used in MC
09:31 < Thinkofdeath> https://gist.github.com/thinkofdeath/e975ddee04e9c87faf22
09:31 < SinZ> zigzag seems like a good way to do it for varints, so you don't waste bytes
13:22 <+clonejo> SinZ: you don't need zigzag for small signed varints, you can just add or remove leading 1 bytes for negative numbers
18:13 < Cubix> Does anyone know nice library for anvil world editing?
19:03 < TkTech> Might help if you specify a language.
19:04 < Cubix> it doesn't really matter, i just needed to spawn world for few things
19:04 < Cubix> pymclevel from mcedit is nice ;)
19:04 < Cubix> scan world*
23:36 < Drainedsoul> SinZ: "good way" + Mojang product, lol
23:38 < SinZ> Drainedsoul: heh
23:38 < Drainedsoul> I'll have to update my varint decoder/encoder, since I wrote it the right way
23:38 < Drainedsoul> with zig zag
23:38 < Drainedsoul> shucks
--- Day changed jeu. oct. 17 2013
01:15 < Drainedsoul> cppreference.com's verification questions as the output of CPP code, cool
01:15 < Drainedsoul> *ask
06:38 < Grum> <Drainedsoul> I'll have to update my varint decoder/encoder, since I wrote it the right way <-- so which 'way' is that?
06:38 < Drainedsoul> zig-zag encoded
06:39 < Grum> we're staying compatible with what google uses in their protobuf spec
06:39 < Grum> for obvious raisins
06:40 < Grum> i mean, i know you love incompatibility and all, sometimes its good not to care for the things you do not use to remain compatible with something else >.>
06:40 < Drainedsoul> https://developers.google.com/protocol-buffers/docs/encoding#types Google specifies ZigZagged types
06:41 < Grum> yeah, so we just used 'int32'
06:41 < Grum> i don't see the problem
06:43 < Drainedsoul> I just don't understand the rationale behind using VarInts and then expending the same number of bytes -- or one extra byte -- to encode negative values.
06:44 < Grum> i dont understand the rational of increasing the size of all the things we send because we do not use it for negative numbers anywhere?
06:46 < Drainedsoul> Thinkofdeath seems under the impression that 0x22, S->C, Play uses VarInts for the X/Z coordinates of the chunks in a multi block change.  I, personally, have not tested or verified this
06:47 < Grum> meh there you lot go away when your numbers
06:48 < Grum> and he's right, so it seems
06:48 < Grum> we have other plans for the chunk-coordinates however
06:51 < Drainedsoul> are arrays going to get a standard type as a length prefix? D:
06:51 < Drainedsoul> *:D
06:53 < Grum> ?
06:55 < Drainedsoul> at least as of my current understanding, collections of items in the protocol are prefixed with their length in -- variously -- bytes, shorts, and ints
07:05 < Grum> we have quite some irregularities to get rid off
07:05 < Grum> but we havent actually done any work on that
07:06 < Grum> we did a short look over what we had and decided we needed to ponder some things a bit more
07:06 < Grum> i prefer to make 'chunkpos' a 'readable type', same for blockxyz, entity.xyz etc
07:07 < Grum> we have about 10 ways to indicate 'xyz' in the code
07:07 < Grum> s/ code/ protocol/
07:10 < Drainedsoul> that'd probably be for the best.  Glad that you're holding off on changes until you're sure about them.
07:10 < Grum> well we have some concepts right now
07:10 < Grum> but other shit came up hehe
07:10 < Grum> x,y,z is storable in an long
07:10 < Grum> downside; resolution of y becomes 12bits
07:11 < Grum> as it is legal to have positions under 0 (not for blocks, but for entities, think of hanging of a ladder), we just have 11 bits of resolution
07:11 < Grum> putting a virtual roof on the world at 2047
07:12 < Grum> beyond that we're sending entity coordinates over the wire in many formats
07:13 < Grum> and one of the main issues we're having in general is the compounding of float/double inaccuracies over time
07:13 < Drainedsoul> yeah I always thought it was odd that floating point and fixed point were used for positions
07:14 < Grum> you know the bug where an item falls, jumps up, falls again etc?
07:14 < Drainedsoul> I've seen it once or twice
07:14 < Grum> that is because on the server its *on* the ledge, on the client its off
07:14 < Grum> because we crop the resolution while sending it over the wire to 1/8000
07:14 < Grum> now, you might wonder, where does the 8000 come from, no idea
07:15 < Drainedsoul> haha
07:15 < Grum> i'm guessing because it was easier to type than 8192
07:15 < Grum> (its send in a short iirc)
07:15 < Grum> -8000 to 8000 'fits' nicely
07:15 < Grum> anyhow!
07:15 < Drainedsoul> well, at least if you truncate the range to that, you don't have weird problems like x<0 && abs(x)==x
07:15 < Grum> that made me wonder, we would't have any errors if we'd use an integer type
07:16 < Grum> and then we figured, what would be a sane resolution
07:16 < Grum> initially we thought 1/256th of a block is ok!
07:16 < Grum> because then an entityposition would be long+bytebytebyte
07:16 < Grum> (the long containing the block xyz)
07:16 < Grum> and then we realized that might be saner to be able to represent a full axis in a single integer
07:17 < Grum> and as we need 26 bits for x/z, +8 ... obviously doesnt fit in an int =)
07:17 < Grum> so right now our idea is at 26+6, giving an 'inblock resolution' of 1/64
07:17 < Grum> which should be more than plenty
07:17 < Drainedsoul> if you have the option of ZigZag you could save bytes in most cases, and not lose resolution by sending X/Y/Z as VarInts
07:18 < Grum> yes but you do not save anything significant
07:18 < Grum> to represent 26 bits you need at least 27 bits when zigzagging
07:18 < Grum> so that is 4 bytes
07:19 < Grum> with the growing point to 3 bytes at 8k
07:19 < Grum> and you barely ever will use 4 bytes
07:20 < Grum> as that is 1-32million
07:20 < Drainedsoul> yeah.  Doesn't the MC world break down thanks to the limit of floats long before +/- 2*10^9?
07:20 < Grum> no that is because of other stupidness
07:20 < Grum> rather than doing a gltranslate when we draw the vertices of the chunk to the 'high number'
07:20 < Drainedsoul> oh, someone told me it was because of single-precision floats, and that seemed plausible
07:20 < Grum> we incur the wrath of float-errors
07:21 < Grum> and add the chunk x,y,z to each vertexes x,y,z
07:21 < Grum> which obviously ....... is stupid :P
07:21 < Drainedsoul> yeah that's the issue with floats.  They have all sorts of weird gotchas
07:21 < Grum> you should *NEVER EVER* draw a vertex with a x/z > 16*chunk-render-distance
07:22 < Grum> and using gltranslate it'll never have a bigger value than 16 ;)
07:22 < Grum> ogl internally works with floats, no matter what you do
07:22 < Grum> you can give it ints, it translates it to floats at the first chance
07:23 < Drainedsoul> do you mean single-precision or just floating point math in general?
07:23 < Drainedsoul> never worked with OGL
07:23 < Grum> internally ogl just uses floats
07:23 < Grum> so minecaft would lend itself really well for a pure integer-based-system
07:23 < Grum> for the world at least
07:23 < Grum> so snap everything to a 1/64 grid
07:23 < Grum> and send pure ints to the videocard
07:24 < Drainedsoul> yeah most of Minecraft is very discrete
07:24 < Grum> anyhow, that is not really the way to solve the issues
07:24 < Grum> we can solve some of the 'tearing' in the world by using indexed buffers
07:25 < Grum> (so we give all the locations of the vertices and then give the list of indices to draw the actual faces)
07:25 < Grum> rather than repeating the vertice x,y,z all the time, and then getting problems because you 'offset them' :p
07:26 < Drainedsoul> what -- if I may ask -- put you guys on such a technical crusade?
07:27 < Grum> its needed?
07:27 < Grum> this has been going on in the background for a long long time
07:27 < Drainedsoul> ah okay.  I can't argue with that.  Just that from the outside looking in MC development looks very feature-driven
07:27 < Grum> this change will be one that wont happen soon
07:27 < Grum> maybe for jens
07:27 < Grum> not for me
07:28 < Drainedsoul> well that's good.  Most people sweep it under the rug, but some people notice that MC is crumbling under the weight of technical debt, and appreciate efforts to pay it off :P
07:30 < Grum> i dont actually care for features
07:30 < Grum> i just see a huge pile of ... things that need fixing
07:30 < dx> i like this guy
07:30 < Drainedsoul> that's a perspective I can definitely appreciate.
07:33 < Grum> i like technical debt
07:33 < Grum> but its not trivial to get rid of :)
07:33 < Drainedsoul> it definitely isn't.  And it's hard to convince some people that it does need getting rid of.  That's the problem I run into a lot.
07:36 <+AndrewPH> pls rewrite minecraft in asm in 3 hours it's not that hard it's just writing code
07:48 < Grum> Drainedsoul: there is multiple ways to do it
07:49 < Grum> we're picking the hard one
07:49 < Grum> but the one with the guaranteed result
07:49 < Grum> you can do the 'big bang' implementation
07:49 < Grum> basically put us in a room for 2 years and we'll have a brand new spanking piece of code that probably doesnt do what we want
07:49 < Grum> or do it on 'live' code, like we're doing now
07:52 < Drainedsoul> yeah total rewrites are for pretty dire situations.  You throw the baby out with the bathwater as it were.
07:53 < Grum> it would have been totally justified
07:53 < Grum> minecraft is not 'one of those games' however
07:54 < Drainedsoul> I wouldn't know anything about how justified it is or not.  I've never looked at the decompiled source etc..  I wouldn't want to have to work on it, though, from what I've heard :P
07:55 < Grum> working on it is a love/hate relationship
07:57 < Drainedsoul> better than hate/hate
07:58 < Grum> hehe
07:58 < Grum> i love fixing broken shit like this :)
07:58 < Drainedsoul> I'm glad.  I just get really angry.  We had "broken shit" made up by a contractor at work earlier this year, I was so furious to have to work on it.
07:58 < Grum> oh me too ;)
07:58 < Grum> and then i fix it ;)
07:58 < Drainedsoul> I guess that's where the "hate" component comes from.
07:59 < Grum> its not always fun fixing
07:59 < Grum> and the fix is not always fun
07:59 < Grum> or good, or even better than how it was
08:00 < Drainedsoul> yeah a lot of fixing is making something awful into something you're proud of, and a lot of it is making something awful into something just as awful, but which actually works
08:01 < Grum> problem is most of this actually works
08:01 < Grum> but its awful
08:01 < Grum> or shouldnt work ;)
08:01 < Grum> so in order make it 'sane' you have to break it
08:01 < Grum> and then bend your nice code to 'do what it roughly did before'
08:01 < Drainedsoul> yeah this project back at the beginning of the year, there was a lot of that.  Total insanity that "worked", but it had to be broken and then fixed.
08:02 < Drainedsoul> when dealing with non-technical people, the definition of "works" is bent in ways that I don't appreciate.
08:03 < Grum> like with this soundsystem, its not really broken in a way that it is 'broken'
08:03 < Grum> but MAN its an abomination
08:03 < Grum> and you cannot do shit easily
08:03 < Grum> everything is hardcoded
08:04 < Grum> the only 'flexible' part is that it plays sound: 'foo.bar' and then you have: foo/bar1.ogg foo/bar2.ogg foo/bar3.ogg and it picks a random one
08:04 < Drainedsoul> hardcoding is a pet peeve of mine.  I'm cursed with having to work on PHP/classic ASP, and hardcoding values seems to be a cultural thing in those communities.
08:05 < Grum> so we're extracting the whole thing into 'hardcoding' onto 'playing a sound event by name'
08:05 < Grum> so it would be: game.block.door.open
08:05 < Grum> and then we can 'register' sounds to that event
08:05 < Grum> with weights
08:05 < Grum> (if you wanted to)
08:07 < Drainedsoul> Registration of events, handlers, etc. is quickly becoming my favourite pattern
08:07 < Grum> and all the soundevents <-> soundfiles is done in json
08:07 < Grum> in resourcepacks
08:07 < Grum> in a stacking way
08:08 < Drainedsoul> you guys should've used XML :P
08:08 < Grum> fuck you :P
08:08 < Grum> nothing worse than xml, oh wait, using xml