<<Prev Next>> Scroll to Bottom
Stuff goes here
L1[00:01:24] ⇦ Quits: Brokkoli (~Brokkoli@p5B23CE53.dip0.t-ipconnect.de) (Quit: Die Sprache der Politik ist daf�r gemacht, dass L�gen wahr klingen und das T�ten angemessen wirkt. (George Orwell))
L2[00:13:30] ⇨ Joins: mallrat208 (~mallrat20@107.145.144.41)
L3[00:32:06] ⇦ Quits: mallrat208 (~mallrat20@107.145.144.41) (Quit: Leaving)
L4[01:10:30] ⇦ Quits: Neal (~Neal@47.146.41.184) (Ping timeout: 186 seconds)
L5[01:28:33] ⇦ Quits: Dimtree (~dimtree@75-110-143-121.nbrncmtk01.res.dyn.suddenlink.net) (Quit: Peace)
L6[01:35:52] ⇨ Joins: Dimtree (~dimtree@75-110-143-121.nbrncmtk01.res.dyn.suddenlink.net)
L7[01:48:47] ⇦ Quits: Dimtree (~dimtree@75-110-143-121.nbrncmtk01.res.dyn.suddenlink.net) (Quit: Peace)
L8[01:51:46] ⇦ Quits: Larry1123 (Larry1123@irc.larry1123.net) (Ping timeout: 198 seconds)
L9[01:55:35] ⇨ Joins: Dimtree (~dimtree@75-110-143-121.nbrncmtk01.res.dyn.suddenlink.net)
L10[02:00:03] <MCPBot_Reborn> [TEST CSV] Pushing snapshot_20171111 mappings to Forge Maven.
L11[02:00:07] <MCPBot_Reborn> [TEST CSV] Maven upload successful for mcp_snapshot-20171111-1.12.zip (mappings = "snapshot_20171111" in build.gradle).
L12[02:00:17] <MCPBot_Reborn> Semi-live (every 10 min), Snapshot (daily ~3:00 EST), and Stable (committed) MCPBot mapping exports can be found here: http://export.mcpbot.bspk.rs/
L13[02:18:13] ⇨ Joins: Larry1123 (Larry1123@irc.larry1123.net)
L14[02:30:09] ⇦ Quits: Larry1123 (Larry1123@irc.larry1123.net) (Ping timeout: 183 seconds)
L15[02:35:57] ⇦ Quits: illy (~illy@cpe-24-166-23-169.cinci.res.rr.com) (Quit: Leaving)
L16[02:39:48] ⇨ Joins: Larry1123 (Larry1123@irc.larry1123.net)
L17[02:47:34] ⇦ Quits: Larry1123 (Larry1123@irc.larry1123.net) (Ping timeout: 186 seconds)
L18[02:57:41] ⇨ Joins: Larry1123 (Larry1123@irc.larry1123.net)
L19[03:20:06] ⇦ Quits: Larry1123 (Larry1123@irc.larry1123.net) (Ping timeout: 186 seconds)
L20[03:22:20] ⇨ Joins: Nedelosk (~Nedelosk@ip-109-90-75-57.hsi11.unitymediagroup.de)
L21[03:27:52] ⇨ Joins: Larry1123 (Larry1123@irc.larry1123.net)
L22[03:55:34] ⇨ Joins: Noppes (~Noppes@ip56530f2e.direct-adsl.nl)
L23[04:03:55] ⇨ Joins: Javaschreiber (~Thunderbi@88-209-32-73.nga.highspeed-baumann.de)
L24[04:07:21] *** PaleOff is now known as PaleoCrafter
L25[04:55:16] ⇦ Quits: McJty (~jorrit@ptr-9197ufq6mer20biruua.18120a2.ip6.access.telenet.be) (Quit: Leaving)
L26[05:01:57] *** PaleoCrafter is now known as PaleOff
L27[05:04:58] ⇦ Quits: Javaschreiber (~Thunderbi@88-209-32-73.nga.highspeed-baumann.de) (Ping timeout: 198 seconds)
L28[05:05:34] ⇦ Quits: RichardG_ (~richardg8@201.37.246.64) (Ping timeout: 198 seconds)
L29[05:19:06] ⇨ Joins: RichardG (~richardg8@201.37.246.64)
L30[05:19:06] MineBot sets mode: +v on RichardG
L31[05:25:03] ⇨ Joins: mallrat208 (~mallrat20@107.145.144.41)
L32[05:26:50] ⇨ Joins: ollieread (~ollieread@exia.ollieread.com)
L33[05:32:11] ⇨ Joins: Javaschreiber (~Thunderbi@88-209-32-73.nga.highspeed-baumann.de)
L34[05:45:42] ⇨ Joins: brandon3055 (~Brandon@pa49-199-233-70.pa.vic.optusnet.com.au)
L35[05:53:47] *** PaleOff is now known as PaleoCrafter
L36[06:27:02] ⇨ Joins: SanAndreasP (~SanAndrea@mue-88-130-48-235.dsl.tropolys.de)
L37[06:33:04] ⇨ Joins: ben_mkiv (~ben_mkiv@p4FED5D77.dip0.t-ipconnect.de)
L38[07:22:41] ⇦ Quits: masa (~masa@86-60-221-191-dyn-dsl.ssp.fi) (Quit: server maintenance)
L39[07:24:54] ⇦ Quits: brandon3055 (~Brandon@pa49-199-233-70.pa.vic.optusnet.com.au) (Read error: -0x1: UNKNOWN ERROR CODE (0001))
L40[07:27:04] ⇨ Joins: brandon3055 (~Brandon@pa49-199-233-70.pa.vic.optusnet.com.au)
L41[08:11:00] ⇨ Joins: ScottehBoeh (~ScottehBo@ip-48-84-52-196.southampton.uk.amsterdamresidential.com)
L42[08:11:17] ⇦ Quits: ScottehBoeh (~ScottehBo@ip-48-84-52-196.southampton.uk.amsterdamresidential.com) (Client Quit)
L43[08:16:25] ⇨ Joins: KGS (~KGS@h-158-174-9-50.NA.cust.bahnhof.se)
L44[08:18:47] ⇦ Quits: justJanne (~justJanne@2001:bc8:33e7:200::1) (Ping timeout: 383 seconds)
L45[08:20:11] ⇨ Joins: justJanne (~justJanne@2001:bc8:33e7:200::1)
L46[08:21:46] ⇦ Quits: Javaschreiber (~Thunderbi@88-209-32-73.nga.highspeed-baumann.de) (Ping timeout: 198 seconds)
L47[08:24:25] ⇨ Joins: Brokkoli (~Brokkoli@p5b23ce53.dip0.t-ipconnect.de)
L48[08:26:44] ⇨ Joins: Javaschreiber (~Thunderbi@88-209-32-73.nga.highspeed-baumann.de)
L49[08:30:23] ⇨ Joins: ScottehBoeh (~ScottehBo@ip-48-84-52-196.southampton.uk.amsterdamresidential.com)
L50[08:30:30] <ScottehBoeh> Hello
L51[08:30:44] <ScottehBoeh> How can I prevent the textures of this puddle entity from glitching with eachother?
L52[08:30:45] <ScottehBoeh> https://i.imgur.com/6HciQXj.png
L53[08:31:04] <ghz|afk> either don't put them on the same exact Y position
L54[08:31:27] <ScottehBoeh> I was thinking that I could slightly nudge the Y position using their X or Z position values
L55[08:31:27] <ghz|afk> or you would have to draw them manually messing with the depth testing settings
L56[08:31:43] <ScottehBoeh> (since the X and Z are basically always unique for each puddle)
L57[08:31:48] <Ordinastie> or use glOffset
L58[08:32:01] <Ordinastie> see text on signs
L59[08:32:08] <ScottehBoeh> kk
L60[08:40:20] ⇨ Joins: h404bi (~h404bi@119.129.119.76)
L61[08:40:32] ⇦ Quits: Javaschreiber (~Thunderbi@88-209-32-73.nga.highspeed-baumann.de) (Ping timeout: 383 seconds)
L62[08:41:35] <ScottehBoeh> How do I remove the shadow underneath entities, again?
L63[08:53:06] <PaleoCrafter> just set the shadow size to something <= 0, it seems?
L64[09:09:33] ⇨ Joins: Javaschreiber (~Thunderbi@88-209-32-73.nga.highspeed-baumann.de)
L65[09:13:02] <ben_mkiv> i have trouble debugging an NPE
L66[09:13:03] <ben_mkiv> at net.minecraftforge.fml.common.registry.EntityRegistry.doModEntityRegistration(EntityRegistry.java:185)
L67[09:13:03] <ben_mkiv> at net.minecraftforge.fml.common.registry.EntityRegistry.registerModEntity(EntityRegistry.java:170)
L68[09:13:08] <ben_mkiv> but those line numbers arent correct
L69[09:13:27] <ben_mkiv> updated mcp and forge, did a gradlew clean, but the issue remains
L70[09:13:35] <ben_mkiv> mcpmappings^
L71[09:24:42] ⇨ Joins: quadraxis (~quadraxis@cpc77293-basf12-2-0-cust699.12-3.cable.virginm.net)
L72[09:35:18] <PaleoCrafter> how do you know they're incorrect?
L73[09:37:23] <ben_mkiv> line 170 is empty for example
L74[09:37:27] <ghz|afk> the lines will be correct if you use the right mappings
L75[09:37:39] <ghz|afk> and the right forge version
L76[09:37:44] <ben_mkiv> snapshot_20171106
L77[09:37:46] <PaleoCrafter> considering those are Forge classes, mappings shouldn't affect anything
L78[09:37:47] <ben_mkiv> for 1.12.2
L79[09:38:01] <ghz|afk> yeah that's not what forge itself is compiled with
L80[09:38:06] <ghz|afk> wait right
L81[09:38:11] <ghz|afk> nevermind ;P
L82[09:38:55] <ghz|afk> in my dev env
L83[09:38:58] <ghz|afk> the first line points to
L84[09:38:58] <ghz|afk> FMLLog.log.trace("Automatically registered mod {} entity {} as {}", mc.getModId(), entityName, entry.getRegistryName());
L85[09:39:21] <ghz|afk> are you passing in a registry name? ;P
L86[09:39:38] <ben_mkiv> EntityRegistry.registerModEntity(new ResourceLocation(UrbanMechs.MODID, name), entity, UrbanMechs.MODID + "." + name.toLowerCase(), 25, instanceOfMyClass, 80, 3, true, primaryColor, secondaryColor);
L87[09:39:40] <ben_mkiv> thats my call
L88[09:39:42] <PaleoCrafter> more likely that it can't find a container
L89[09:39:50] <ghz|afk> in that case
L90[09:39:53] <ghz|afk> where do you call it from?
L91[09:39:58] ⇨ Joins: malte0811 (~malte@185.134.128.197)
L92[09:40:05] <ben_mkiv> preInit()
L93[09:40:08] <ben_mkiv> event
L94[09:40:38] <ghz|afk> wait, "instanceOfMyClass"
L95[09:40:46] <ghz|afk> that is not your @Mod class, is it?
L96[09:40:52] <ben_mkiv> it is
L97[09:40:56] <ghz|afk> weird name, then
L98[09:40:57] <ghz|afk> ;P
L99[09:41:03] <ben_mkiv> its instance in the code ;)
L100[09:41:52] <ghz|afk> the only difference from what I do, so far as I can tell
L101[09:41:55] <ghz|afk> is that I use
L102[09:42:00] <ghz|afk> @SubscribeEvent public static void registerEntities(RegistryEvent.Register<EntityEntry> event)
L103[09:42:13] <ghz|afk> instead of preinit
L104[09:43:25] <ben_mkiv> which actually work for me, too
L105[09:43:27] <ben_mkiv> thanks much :)
L106[09:46:19] <ben_mkiv> weird, as i've looked at this https://github.com/Shinoow/AbyssalCraft/blob/ec0e12e416168457576ddb26b6c70bc630df2a62/src/main/java/com/shinoow/abyssalcraft/init/EntityHandler.java#L42
L107[09:53:57] <ben_mkiv> so theres no error, but also no spawnegg for the entity, did i expect too much? xD
L108[09:58:38] ⇨ Joins: masa (~masa@86.60.225.191)
L109[10:03:59] <ben_mkiv> because it gets never called....
L110[10:04:34] <ghz|afk> did you put @Mod.EventBusSubscriber on the class? :P
L111[10:05:38] <ben_mkiv> public void registerEntities(RegistryEvent.Register<EntityEntry> event)
L112[10:05:42] <ben_mkiv> this works, but not as static
L113[10:05:48] <ben_mkiv> but then i get the same old error
L114[10:05:56] <ghz|afk> static works with @Mod.EventbusSubscriber
L115[10:06:06] <ghz|afk> non-static works with MinecraftForge.EVENT_BUS.register(theclass)
L116[10:06:18] <ghz|afk> anyhow
L117[10:06:20] <ben_mkiv> MinecraftForge.EVENT_BUS.register(this);
L118[10:06:22] <ghz|afk> if the same thing happens
L119[10:06:24] <ben_mkiv> thats what i call in preinit
L120[10:06:26] <ghz|afk> then that wasn't the issue
L121[10:06:31] <ghz|afk> yeah
L122[10:06:34] <ghz|afk> if you make the events static
L123[10:06:48] <ghz|afk> and write @Mod.EventBusSubscriber on the class
L124[10:07:06] <ghz|afk> like
L125[10:07:07] <ghz|afk> https://github.com/gigaherz/PackingTape/blob/master/src/main/java/gigaherz/packingtape/ModPackingTape.java#L21
L126[10:07:12] <ghz|afk> https://github.com/gigaherz/PackingTape/blob/master/src/main/java/gigaherz/packingtape/ModPackingTape.java#L42
L127[10:07:47] <ben_mkiv> thanks
L128[10:08:38] <ben_mkiv> ok, the static events work now, but as you said its not related to the problem
L129[10:08:59] <ben_mkiv> i'm not sure about that integer modid i'm passing
L130[10:09:00] <ben_mkiv> the 25
L131[10:09:06] <ben_mkiv> just took that from abysallcraft code :>
L132[10:09:13] <ben_mkiv> +s-l
L133[10:09:22] <ghz|afk> wait what
L134[10:09:25] <ghz|afk> integer modid?
L135[10:09:32] <ghz|afk> oh
L136[10:09:34] <ghz|afk> the entity id
L137[10:09:38] <ghz|afk> I just start from 0
L138[10:09:41] <ghz|afk> and increase up
L139[10:09:56] <ben_mkiv> each mod has its own counter?
L140[10:10:08] <ghz|afk> like this: https://github.com/gigaherz/ElementsOfPower/blob/master/src/main/java/gigaherz/elementsofpower/ElementsOfPower.java#L293
L141[10:10:09] <ghz|afk> yes
L142[10:10:14] <ghz|afk> the number is internal for the mod
L143[10:10:22] <ghz|afk> sorry I start from 1 ;P
L144[10:10:42] <ben_mkiv> so 25 should work, too and isnt the problem
L145[10:10:49] <ben_mkiv> but now ik at least what to use it for xD
L146[10:12:04] <ben_mkiv> so i guess my actual entity class might be the problem
L147[10:14:16] <ghz|afk> you never pasted the full stacktrace
L148[10:15:13] <ben_mkiv> https://pastebin.com/hDhBRYpt
L149[10:15:41] <ben_mkiv> registerEntityWithEgg just wraps EntityRegistry.registerModEntity()
L150[10:16:20] <ghz|afk> put a breakpoint in that line
L151[10:16:20] <ghz|afk> FMLLog.log.trace("Automatically registered mod {} entity {} as {}", mc.getModId(), entityName, entry.getRegistryName());
L152[10:16:23] <ghz|afk> see what is null
L153[10:16:28] <ghz|afk> or
L154[10:16:47] <ghz|afk> enable exception breakpoint for nullpointer with the class filter set to that class
L155[10:17:15] <ben_mkiv> do i have to run minecraft from intellij to get the breakpoint infos?
L156[10:17:20] <ben_mkiv> or does it print them to the logs?
L157[10:20:44] <ghz|afk> yes you would have to run from intellij
L158[10:20:53] <ghz|afk> or run as debug and attach
L159[10:21:05] <ben_mkiv> yea, i'm just looking into how to set this up
L160[10:21:10] <ghz|afk> (there's commandline flags to enable debug access)
L161[10:21:13] <ben_mkiv> thanks for now :)
L162[10:25:39] <TechnicianLP> > might there be problems if the size of a texture that gets bound manually does not have a size of a power of 2?
L163[10:26:32] <ghz|afk> I recall using 48x48 textures just fine
L164[10:26:44] <ghz|afk> wait that was in the atlas
L165[10:27:19] <ghz|afk> atlas textures should be square, whatever the size
L166[10:27:29] <TechnicianLP> i remember those sizes generating artifacts on the atlas ...
L167[10:27:31] <ghz|afk> and non-atlas textures should be power of two, but can be rectangular
L168[10:28:32] <quadraxis> size should also be a multiple of 16, for atlas textures at least
L169[10:29:12] <TechnicianLP> well im talking about a nonatlas texture (specifically a gui-texture)
L170[10:29:25] <TechnicianLP> rectangular it is then ...
L171[10:30:04] <ghz|afk> textures CAN be non-power of two, in principle
L172[10:30:09] <ghz|afk> but it's generally not recommended
L173[10:30:14] <ghz|afk> since they can be slower
L174[10:33:00] <ghz|afk> hmf
L175[10:33:05] ⇦ Quits: airbreather (~airbreath@d149-67-99-43.nap.wideopenwest.com) (Ping timeout: 200 seconds)
L176[10:33:09] <ghz|afk> did journeymap have some way to show which chunks are slime chunks?
L177[10:44:36] ⇨ Joins: airbreather (~airbreath@d149-67-99-43.nap.wideopenwest.com)
L178[11:04:06] ⇦ Quits: PitchBright (~PitchBrig@CPE00fc8d8a3ce3-CM00fc8d8a3ce0.cpe.net.cable.rogers.com) (Quit: brb)
L179[11:13:05] <barteks2x> Can anyone with a lot of ram (8G or more) test something for me?
L180[11:13:37] <PaleoCrafter> uhm, sure
L181[11:14:01] <PaleoCrafter> got about 20 gigs to spare right now :P
L182[11:15:38] <barteks2x> it's a very simple test mod should I send the jar or can you write it?
L183[11:15:46] ⇨ Joins: Xaphiosis (~x@220.240.73.109)
L184[11:16:04] <PaleoCrafter> I think I can manage with the code xD
L185[11:16:05] <barteks2x> it's basically sending a single 1MB packet to the client per tick
L186[11:16:18] <barteks2x> on a dedicated server
L187[11:16:29] <barteks2x> and the server needs 2 or more GB allocated
L188[11:17:10] <PaleoCrafter> actually, I wouldn't mind the JAR :D
L189[11:17:32] <PaleoCrafter> don't feel like setting up packet sending etc. right now since I haven't got a project with it handy
L190[11:17:40] <barteks2x> so compiling
L191[11:18:17] <Ordinastie> is your code somewhere ?
L192[11:18:37] <barteks2x> the test code? No. I have that bug in cubic chunks and trying to see if it's something in my code or forge issuye
L193[11:18:48] <barteks2x> but can't really run both client and server at once
L194[11:19:05] <barteks2x> because it needs dedicated server
L195[11:19:27] <Xaphiosis> I have a very strange situation... I have a world generator that scans for the topmost solid block, checks if the block above it is water, and changes that block to a specific one in the mod (debugging)
L196[11:19:38] <Xaphiosis> however, not all blocks that are underwater are changed!
L197[11:20:17] <barteks2x> https://www.dropbox.com/s/qdsh39p4lthfwoc/modid-1.0.jar?dl=1
L198[11:20:28] <Xaphiosis> around beaches, gravel, sand, it just ignores me. any ideas what could be causing that?
L199[11:20:41] <barteks2x> show code
L200[11:21:12] <barteks2x> also this is my test code https://hastebin.com/arajecuyeg.java
L201[11:23:09] <Xaphiosis> barteks2x: was "show code" to me?
L202[11:23:13] <barteks2x> yes
L203[11:23:16] <Xaphiosis> https://github.com/Xaphiosis/CoralReef/blob/master/src/main/java/primetoxinz/coralreef/WorldGenReef.java
L204[11:23:47] <Xaphiosis> it's that file, but with all superfluous checks removed, so it only checks isOceanFloor
L205[11:24:09] <Xaphiosis> I am trying to fix it, I did not write that part btw
L206[11:24:38] <barteks2x> where is it called?
L207[11:25:32] <Ordinastie> memory derp here, how do you register an event handler again ?
L208[11:25:33] <Xaphiosis> https://github.com/Xaphiosis/CoralReef/blob/9bdfdbdb1a733a27a1908ba9c6358fd24b6b722e/src/main/java/primetoxinz/coralreef/GeneratorReef.java#L49
L209[11:26:00] <Ordinastie> ah, I remember now ><
L210[11:26:06] <Xaphiosis> I also stripped down that file to ignore all biome / noise checks and just try place the block
L211[11:27:14] <Xaphiosis> so as far as I can tell, either at the time the generator runs, those blocks are not under water, not solid blocks, the way the check for "is this water" is wrong, or the scan for a solid block downwards from chunk.getTopFilledSegment() is wrong
L212[11:27:39] <barteks2x> I think the issue may be that generator from the chunk next to that replaces the blocks afte ryour onw runs
L213[11:27:48] <Ordinastie> wait, I can't test your code, I can't launch server in dev :/
L214[11:27:52] <barteks2x> I see only 1, not very efficient and not completely foolproof way to solve it
L215[11:28:25] <Xaphiosis> setting the generator weight to a higher level does not help? I thought it would be ok if mine ran last, so I set it to 1000
L216[11:28:41] <barteks2x> yes, but it's generator FROM A DIFFERENT CHUNk replacing the blocks
L217[11:28:49] <barteks2x> unless it's all th eblocks that are missed
L218[11:28:51] <Xaphiosis> oooooooooh
L219[11:29:06] <barteks2x> just do the replacement between chunkX/Z*16+1 and chunkX/Z*16+16+15
L220[11:29:22] <Xaphiosis> right, spatial locality, running each generator happens on a per-chunk level
L221[11:29:39] <barteks2x> it's not going to be efficient and will break for generators that cause cacading world gen, but most of the tiome it will work
L222[11:30:47] <Ordinastie> ok fixed it
L223[11:30:55] <Xaphiosis> hmm. won't that dump a bunch of warnings about extra chunks getting loaded? or is the overlap order as you suggested safe?
L224[11:31:06] <Xaphiosis> (x+,z+) direction
L225[11:31:13] <barteks2x> the suggested overlap is just on the edge of being safe
L226[11:31:25] <barteks2x> 1 block more in either direction and it won't be safe
L227[11:31:35] <Xaphiosis> hmm. maybe you can explain another misconception to me by the way of this
L228[11:32:13] <Xaphiosis> when I write block code in other domains, a chunk location scan would start at chunkX*n + offset
L229[11:32:29] <Xaphiosis> and at offset=15 you'd roll over to next chunk
L230[11:32:41] <Xaphiosis> but in minecraft, the scan starts at chunkX*n + 8 + offset
L231[11:32:43] <Xaphiosis> how come?
L232[11:32:50] <barteks2x> read this https://www.reddit.com/r/feedthebeast/comments/5x0twz/investigating_extreme_worldgen_lag/
L233[11:33:10] <barteks2x> specifically, "the cause" section
L234[11:33:40] ⇦ Quits: Nedelosk (~Nedelosk@ip-109-90-75-57.hsi11.unitymediagroup.de) (Read error: Connection reset by peer)
L235[11:33:47] <Ordinastie> barteks2x, I suppose that's not what you expected to see : https://puu.sh/yjLzG.png
L236[11:34:03] <barteks2x> Ordinastie, try connecting again
L237[11:34:19] <Ordinastie> it worked
L238[11:34:21] <barteks2x> this is a rare forge/vanilla bug that sometims happens for no reason
L239[11:34:26] <barteks2x> (the group error)
L240[11:34:34] <Ordinastie> I've never seen it
L241[11:34:37] <barteks2x> now keep the server runnign and wait
L242[11:34:42] <Xaphiosis> barteks2x: that is very interesting, thank you
L243[11:34:44] <barteks2x> while on theserver
L244[11:34:54] <Ordinastie> what is supposed to happen ?
L245[11:35:09] <barteks2x> server has 3G ram and client 1G?
L246[11:35:12] <barteks2x> *2G
L247[11:35:25] <barteks2x> then it should at some point kick you out of the server with strange out of memory error
L248[11:35:32] <barteks2x> if it doesn't, it's an issue in my mod
L249[11:35:47] <Ordinastie> I think they both have 4G
L250[11:35:54] <barteks2x> should be ok
L251[11:35:57] <barteks2x> wait about 15min
L252[11:36:01] <Ordinastie> server doesn't rise above 200mb
L253[11:36:18] <barteks2x> in task manager or actual java heap?
L254[11:36:28] <Ordinastie> client GC at 1.3GB to 300MB
L255[11:36:41] <barteks2x> the client memory usage won't grow, it's native memory
L256[11:36:55] <barteks2x> which is capped at 1G by default
L257[11:37:30] <Ordinastie> guess I'll watch a video mean while
L258[11:38:01] <barteks2x> does the server memory usage grow?
L259[11:38:12] <Ordinastie> nope
L260[11:38:22] <barteks2x> is it with or without GUI?
L261[11:38:30] <Ordinastie> switches between 191 et 200mb
L262[11:38:32] <barteks2x> It did grow for me
L263[11:38:33] <Ordinastie> the GUI one
L264[11:38:49] <barteks2x> and you are actually sending 1MB packet per tick?
L265[11:38:58] <Ordinastie> let me check your code runs first :p
L266[11:39:32] <barteks2x> hm... actually 1000000 bytes may not be enough to trigger it if it';s packet splitting causing it
L267[11:39:40] <barteks2x> try 1024*1024 instead of 1000000
L268[11:39:48] <Ordinastie> maybe it's not sendind... ><
L269[11:40:15] <Ordinastie> fuck the server tick is static ><
L270[11:40:29] <barteks2x> and?
L271[11:40:36] <Ordinastie> I registered the instance
L272[11:40:51] <barteks2x> I had Mod.EventBusSubscriber
L273[11:41:08] <Ordinastie> yeah, I cherry picked what I needed, didn't take the mod class
L274[11:41:21] <Xaphiosis> barteks2x: I changed it to for (int x = 1; x < 16 + 16; x++) as you suggested, i.e. 1 to 16+15, and removed the 8 offset elsewhere... stuck on "building terrain" now
L275[11:41:45] <barteks2x> did you not forget to remove the +8 then?
L276[11:41:45] <Xaphiosis> so clearly that causes what the reddit post was describing
L277[11:42:17] <barteks2x> in this line BlockPos pos = getTop(world, (chunkX * CHUNK_SIZE) + x + 8, (chunkZ * CHUNK_SIZE) + z + 8);
L278[11:42:42] <Ordinastie> now the packet is received
L279[11:43:03] <Xaphiosis> nope, it's now BlockPos pos = getTop(world, (chunkX * CHUNK_SIZE) + x, (chunkZ * CHUNK_SIZE) + z);
L280[11:43:11] <Ordinastie> and I got kicked too
L281[11:43:25] <barteks2x> Ordinastie, error?
L282[11:43:32] <barteks2x> Xaphiosis, try 16 + 16 - 1
L283[11:43:32] <Ordinastie> same as yours
L284[11:43:39] <barteks2x> the out of memory one?
L285[11:44:01] <barteks2x> now try with 1024*512 size
L286[11:44:11] <Ordinastie> https://puu.sh/yjLVy.png
L287[11:44:12] <barteks2x> just to know if it's packet splitting or not
L288[11:44:32] <barteks2x> yup, that's the error. It's either netty or forge bug
L289[11:44:48] <Xaphiosis> barteks2x: sorry, still stuck
L290[11:45:09] <barteks2x> Xaphiosis, show the code as you currently have it
L291[11:45:20] <Xaphiosis> I can push to a branch the exact thing I have now ... yup, will do, hang on
L292[11:45:47] <Ordinastie> I can'"t even reconnect, I have to relaunch the client
L293[11:46:38] <barteks2x> I know
L294[11:46:46] <barteks2x> that's why I know for asure it's actual memory leak
L295[11:46:52] <barteks2x> and not "just not cleaning up fast enough"
L296[11:47:18] <Ordinastie> I got the group error again...
L297[11:47:29] <Xaphiosis> barteks2x: https://github.com/Xaphiosis/CoralReef/tree/debug_water/src/main/java/primetoxinz/coralreef
L298[11:47:55] <Xaphiosis> the files are as before WorldGenReef and GeneratorReef
L299[11:48:25] <Xaphiosis> I fixed quite a few issues with this mod so far, even managed the port to 1.12, but can't figure this one out at all :(
L300[11:48:26] <barteks2x> Xaphiosis, reefFromNoise will crash
L301[11:48:48] <Ordinastie> still kicked with 1024*512
L302[11:48:59] <Xaphiosis> oh, hang on, let me disable the noise lookup
L303[11:49:08] <barteks2x> Ordinastie, can you halve it until you are not kicked?
L304[11:49:25] <barteks2x> I would like to see if there is some cutoff or does it happen with any packets
L305[11:49:47] <barteks2x> sucks when you can't test it because not enough ram :(
L306[11:50:51] <Ordinastie> and the group error again
L307[11:51:01] <barteks2x> probably sending some packet too early
L308[11:51:11] <Ordinastie> probably
L309[11:52:27] <Xaphiosis> barteks2x: ok, it runs again, ignoring any noise tests
L310[11:52:38] <Xaphiosis> but, [04:51:51] [Server thread/WARN] [FML]: Coralreef loaded a new chunk (9, -6 Dimension: 0) during chunk population, causing cascading worldgen lag. Please report this to the mod's issue tracker. This log can be disabled in the Forge config.
L311[11:52:49] <Ordinastie> 256 doesn't seem to kick
L312[11:53:02] <Xaphiosis> and also, the blocks are still not converted, they're the same as before
L313[11:53:51] <Xaphiosis> I'm looking at a bunch of sand and dirt that's two blocks under water, not changed over to the new block
L314[11:54:19] <barteks2x> Xaphiosis, that's probably the cascading part
L315[11:54:24] <barteks2x> Ordinastie, and 512 does?
L316[11:54:30] ⇦ Quits: RichardG (~richardg8@201.37.246.64) (Read error: Connection reset by peer)
L317[11:54:31] <Ordinastie> it did yes
L318[11:54:41] <barteks2x> well, can you bisect it?
L319[11:55:08] <Ordinastie> oh, nvm, it just kicked me
L320[11:55:17] <Xaphiosis> sigh. anything else I can do? does this same stuff happen to oregen etc, people just don't notice?
L321[11:56:32] ⇨ Joins: McJty (~jorrit@ptr-9197ufqjxhrc0jf4y8b.18120a2.ip6.access.telenet.be)
L322[11:56:38] <LexMobile> Vanilla only has 2 spots that cascade, large trees and some large lakes. So no, you shouldn't do it.
L323[11:56:39] <Ordinastie> damn, loading client in dev is so slow for me :/
L324[11:56:50] <barteks2x> Xaphiosis, I do have an idea but it's going to be hacky
L325[11:57:03] <Xaphiosis> hacky is fine, so long as I get to understand what I'm doing in the end
L326[11:57:07] <barteks2x> you would register your own IWorldEventListener
L327[11:57:16] <barteks2x> and during population, find all the changed blocks
L328[11:57:30] <barteks2x> not exactly fast but would probably work
L329[11:57:42] <LexMobile> what is he trying to do?
L330[11:58:02] <barteks2x> tryign to replace some kidn of blocks underwater during population
L331[11:58:23] <Xaphiosis> all blocks that are under water. eventually I only want to work on a subset of them
L332[11:58:46] <Xaphiosis> I found the problem when my perlin noise results were not matching what appeared in the world
L333[11:58:53] <LexMobile> And why would that need to listen for anything?
L334[11:59:25] <barteks2x> because even after replacing the blocks, next populator may replace them again from different chunk
L335[12:00:01] <Xaphiosis> happens near beaches a lot, am unfamiliar with how beach worldgen works at all
L336[12:00:24] ⇦ Quits: ScottehBoeh (~ScottehBo@ip-48-84-52-196.southampton.uk.amsterdamresidential.com) (Quit: Leaving)
L337[12:00:35] <barteks2x> because beaches generate things like sand a lot, and it can cross the boundary of the direct 16x16 population area
L338[12:00:41] <LexMobile> They shouldn't. Decorators donly work on the 4 chunks in the corner, or more specifically the top left and tries to stay out of the other three.
L339[12:00:57] <barteks2x> so if he replaces the blocks just in the 16x16 area it won't work
L340[12:01:04] <barteks2x> he would have to replace it in the whole 32x32 area
L341[12:01:31] <barteks2x> I suggested 30x30 area to avoid block updates triggering loading chunks, and it still didnt work and caused cascading worldgen somehow
L342[12:02:58] <Xaphiosis> I'm going to try skip the water check altogether and replace all solid blocks at chunk.getTopFilledSegment(), just in case it's the water check
L343[12:03:37] <Xaphiosis> yeah, not the water check. happens quite a lot on land too
L344[12:04:21] <barteks2x> also Xaphiosis you get get a good starting posityion as the top block with getHeight or getPrecipitationHeight
L345[12:04:41] <barteks2x> instead of starting from getTopFilledSegment
L346[12:04:57] <Xaphiosis> ok, will add that to the mix
L347[12:05:35] <Xaphiosis> but clearly something isn't making sense to me, because some giant mushrooms' tops are replaced, and some not
L348[12:06:01] <LexMobile> Interesting, Modded genertors run after any vanilla ones do...
L349[12:06:03] <Ordinastie> 128 kick again
L350[12:06:30] <LexMobile> Show a SS of what you're seeing Xap?
L351[12:06:33] <barteks2x> Ordinastie, so thsi means it will probably always happen. interesting how it doesn't break almost all mods that use netowking
L352[12:06:56] <barteks2x> unless it does and noone has noticed
L353[12:07:52] <LexMobile> ?
L354[12:07:55] <barteks2x> and right, the only way to have yuour generator run before vanilla is to listen for Pre event
L355[12:08:16] <PaleoCrafter> ah, Lex, see Twitch or should I PM you? :D
L356[12:08:27] <barteks2x> lex, if you are asking about the networtking issue - there is a strange netty memory leak that happens only on dedicated server
L357[12:08:39] <Xaphiosis> hang on, screenshot incoming
L358[12:08:58] <barteks2x> after enough packets are sent, this happens https://puu.sh/yjLVy.png
L359[12:09:39] <barteks2x> (this happens clientside)
L360[12:11:56] <LexMobile> humm the wrapper for custom payload stuff is rather crappy.
L361[12:12:03] <LexMobile> It may never be freeing the buffers
L362[12:12:14] <Xaphiosis> lex: https://imgur.com/a/RoliQ (uploading, forgive my australian internet)
L363[12:12:14] <LexMobile> But pretty sure I checked it... then again this is fucking netty hell.
L364[12:12:22] <barteks2x> I also tried to use the builtin netty leak detection, set to paranoid. Nothing
L365[12:12:47] <Xaphiosis> aand uploaded
L366[12:12:49] <LexMobile> So, which blocks are wrong?
L367[12:13:05] <Xaphiosis> all the ones that are not weird looking at the topmost y
L368[12:13:14] <barteks2x> and I tried to find the actual leaked buffers in heap dump but it's hard to find anythign in the LWJGL noise
L369[12:14:02] <barteks2x> and it actually makes my mod require client restart every 15 or so minutes of playing
L370[12:14:13] <Xaphiosis> so everywhere you see soil sand etc that's not under something else, excepting leaves and anything non-solid
L371[12:14:20] ⇨ Joins: Nedelosk (~Nedelosk@ip-109-90-75-57.hsi11.unitymediagroup.de)
L372[12:14:45] <LexMobile> Since it seems to happen pretty universally except for that hill in the back.. it would make me think you have a Y issue in your generator
L373[12:14:48] <barteks2x> Xaphiosis, that actually loooks like fault resulting fro using getTopFilledSegment
L374[12:14:51] ⇨ Joins: RichardG (~richardg8@201.37.246.64)
L375[12:14:51] MineBot sets mode: +v on RichardG
L376[12:15:00] <barteks2x> getTopFilledSegment probably returns Y of the one below actuyally
L377[12:15:34] <barteks2x> I'm stupid for not asking for screenshot and assuming it's sometihing else...
L378[12:15:55] <barteks2x> well, Y of the bottom of the top filled segment
L379[12:16:00] <LexMobile> yup
L380[12:17:08] <Xaphiosis> hmm
L381[12:17:13] <barteks2x> shoudl I now report the netty issue on github?
L382[12:17:24] <LexMobile> Got a mod to reproduce it
L383[12:17:29] <LexMobile> And not a 10,000 feature mod?
L384[12:17:34] <barteks2x> https://hastebin.com/arajecuyeg.java
L385[12:17:43] <barteks2x> this, with just about any amount of bytes sent in the packet
L386[12:18:01] <Ordinastie> currently testing with 512 * 64
L387[12:18:04] <Ordinastie> but not kicked yet
L388[12:18:12] <LexMobile> Then sure. We'll see
L389[12:18:14] <barteks2x> I started with 1M because I thought it's packet splitting being the issue
L390[12:18:29] <Xaphiosis> barteks2x: but that's our starting point, we then scan downwards looking for a solid block, and use that for our blockpos where we change the block. where is that going wrong then?
L391[12:18:40] <LexMobile> It may be I do some hacky stuff witht eh buffers there, but I *think* I free the extra ones...
L392[12:18:49] <barteks2x> Xaphiosis, because it starts below the surface sometimes
L393[12:19:12] <barteks2x> because `getTopFilledSegment` returns the *bottom* of that segment
L394[12:19:18] ⇦ Quits: Delaxarnyazer (~Delaxarny@2a02:a44e:91ce:1:50a3:ad04:9869:2e0f) (Ping timeout: 186 seconds)
L395[12:19:26] <LexMobile> you want world.getTopSolidBlock or whatever the name is
L396[12:19:37] <barteks2x> or just getHeioght or getPrecipitationHeight
L397[12:19:47] <barteks2x> which is instant because it returns value from an array
L398[12:19:52] <LexMobile> Mine skips all fluids IIRC.
L399[12:20:10] <LexMobile> But ya, if you guys can keep lookin into it but reporting ti and getting others incolved would be nice.
L400[12:20:11] <barteks2x> yes but it's slower
L401[12:20:21] <Xaphiosis> getTopSolidOrLiquidBlock?
L402[12:20:22] <LexMobile> I *was* going to do shit today... So I need to take off soon.
L403[12:20:48] <barteks2x> Xaphiosis, contrary top the name that method acvtually skips liquids iirc
L404[12:21:03] <Xaphiosis> haha
L405[12:21:26] ⇨ Joins: Delaxarnyazer (~Delaxarny@2a02:a44e:91ce:1:50a3:ad04:9869:2e0f)
L406[12:24:47] <Ordinastie> finally got kicked with 64 too
L407[12:24:53] <Ordinastie> not gonna test 32 though ><
L408[12:26:16] <barteks2x> it's safe to assume it will happen at least down to 256 bytes
L409[12:28:22] <Xaphiosis> barteks2x: FIXED! after hours and hours, thank you so much!
L410[12:28:59] <barteks2x> Xaphiosis, now make sure to also test it without the 30x30 scan area, it may work well enough
L411[12:29:07] <Xaphiosis> it now starts at BlockPos bp = world.getTopSolidOrLiquidBlock(new BlockPos(x, 0, z)); and walks down through any leaves down to the surface
L412[12:29:12] <Xaphiosis> yes, I already reverted that
L413[12:29:38] <Xaphiosis> for (int x = 8; x < 16 + 8; x++) ... BlockPos pos = getTop(world, (chunkX * CHUNK_SIZE) + x, (chunkZ * CHUNK_SIZE) + z);
L414[12:30:08] <Xaphiosis> 16x16 scan area as described in the reddit post you linked, which I think I now understand
L415[12:37:12] <Ordinastie> ok, that's weird
L416[12:37:23] <Ordinastie> I wanted to check something but not I can't connect to the server
L417[12:37:23] <barteks2x> what is weird?
L418[12:37:26] ⇨ Joins: Neal (~Neal@47.146.41.184)
L419[12:37:26] <Ordinastie> I get that : https://puu.sh/yjNIs.png
L420[12:37:53] <Ordinastie> and client side : Nov 11, 2017 7:35:57 PM io.netty.util.ResourceLeakDetector reportUntracedLeak
L421[12:37:53] <Ordinastie> SEVERE: LEAK: ByteBuf.release() was not called before it's garbage-collected. Enable advanced leak reporting to find out where the leak occurred. To enable advanced leak reporting, specify the JVM option '-Dio.netty.leakDetection.level=advanced' or call ResourceLeakDetector.setLevel() See http://netty.io/wiki/reference-counted-objects.html for more information.
L422[12:38:09] <barteks2x> O.O it worked for you
L423[12:38:16] <Ordinastie> I didn't set anything
L424[12:38:21] <barteks2x> I enabled pedantic leak reportign and it found NOTHING
L425[12:38:35] <barteks2x> at l;east clientside...
L426[12:38:39] <barteks2x> I didn't enable it serversize
L427[12:38:43] <Ordinastie> I can't connect to dedi anymore
L428[12:39:02] ⇦ Quits: Rokiyo (~Rokiyo@101.167.173.217) (Ping timeout: 186 seconds)
L429[12:39:06] <Ordinastie> and the server gobs up memory like crazy until client times out
L430[12:39:30] ⇨ Joins: Rokiyo (~Rokiyo@101.167.173.217)
L431[12:39:47] <Ordinastie> it's because I set the packet size to 1024*1024
L432[12:40:08] <barteks2x> oh so it's just consuming all the server memory before the client runs out of memory
L433[12:40:46] <barteks2x> also already reported here now https://github.com/MinecraftForge/MinecraftForge/issues/4509
L434[12:41:47] <Ordinastie> if I do 1024*1000 it doesn't do it, but I got the direct memory error
L435[12:41:49] <barteks2x> I'm really susprised no users are complainign yet
L436[12:42:04] <barteks2x> so it's packet splitting that causes the serverside leak
L437[12:42:34] ⇦ Quits: McJty (~jorrit@ptr-9197ufqjxhrc0jf4y8b.18120a2.ip6.access.telenet.be) (Quit: Leaving)
L438[12:43:17] <Ordinastie> leak was reported by client
L439[12:43:37] <barteks2x> netty detects them very rarely,m it could detect it now just by chance
L440[12:43:48] <barteks2x> it detects probably one in thousands or so
L441[12:44:10] <barteks2x> so if you see one, it's a sign something is very wrong
L442[12:44:35] ⇨ Joins: RichardG_ (~richardg8@201.37.246.64)
L443[12:44:35] MineBot sets mode: +v on RichardG_
L444[12:44:43] <barteks2x> I'm almost sure it didn't happen in earlier versions
L445[12:44:47] <barteks2x> so I will try to bisect it
L446[12:45:00] ⇦ Quits: RichardG (~richardg8@201.37.246.64) (Read error: Connection reset by peer)
L447[12:45:42] <barteks2x> (I have a cubic chunks server that I can reproduce it on)
L448[12:50:08] <Xaphiosis> ok, nearly 6am, time to try sleeping again. once again huge thanks barteks2x (and lex)!
L449[12:50:30] <barteks2x> happened with the oldest 1.12.2 version :(
L450[12:50:51] ⇦ Quits: Xaphiosis (~x@220.240.73.109) (Quit: Leaving.)
L451[12:51:07] <barteks2x> wait... I didn't change forge version, just installed it
L452[12:51:08] <barteks2x> stupid me
L453[12:52:48] <quadraxis> looking at FMLProxyPacket.toS3FPackets(), should payload.release() be called here in the event that the packet contents are split?
L454[12:54:06] <barteks2x> maybe but that wouldn't explain the issue when they are not split
L455[12:54:17] ⇦ Quits: Delaxarnyazer (~Delaxarny@2a02:a44e:91ce:1:50a3:ad04:9869:2e0f) (Ping timeout: 183 seconds)
L456[12:54:30] <barteks2x> and can't easily look at the code now
L457[12:56:00] ⇨ Joins: Delaxarnyazer (~Delaxarny@ip56572345.direct-adsl.nl)
L458[12:56:54] <barteks2x> Now I;m sure the issue is also in gorge 2486
L459[12:57:02] <barteks2x> *forge
L460[12:59:09] <barteks2x> also any way to remove the coremod version check with commandline argument?
L461[13:17:52] ⇨ Joins: Ipsis (~Ipsis@82-69-71-184.dsl.in-addr.zen.co.uk)
L462[13:21:43] <barteks2x> f*** any way to join 1.12.2 server on 1.12 and 1.12.1?
L463[13:22:19] ⇦ Quits: h404bi (~h404bi@119.129.119.76) (Ping timeout: 383 seconds)
L464[13:27:12] <kashike> barteks2x: no
L465[13:27:50] <barteks2x> It's not like I need it working perfectly... I just don't want to ask the owner of the test server to downgrade
L466[13:28:23] <kashike> the protocol changed
L467[13:28:50] <barteks2x> isn't there a way around it? Or did it change enough that joining the world wouldn't work?
L468[13:28:58] <kashike> http://wiki.vg/Protocol_History#1.12.2
L469[13:29:07] <kashike> 17w31a = 1.12.1 snapshot
L470[13:29:40] <barteks2x> would it at least partially work if I made 1.12.1 pretend it's 1.12.2?
L471[13:30:08] <kashike> you'd need to change the keep alive packets too
L472[13:30:23] <barteks2x> I will just ask him to downgrade...
L473[13:30:36] <kashike> good idea :P
L474[13:31:29] ⇦ Quits: Ipsis (~Ipsis@82-69-71-184.dsl.in-addr.zen.co.uk) (Ping timeout: 183 seconds)
L475[13:33:11] ⇨ Joins: airbreather_ (~airbreath@d149-67-99-43.nap.wideopenwest.com)
L476[13:33:27] <barteks2x> I just hope the time he won't screw up installing forge...
L477[13:35:03] ⇦ Quits: airbreather (~airbreath@d149-67-99-43.nap.wideopenwest.com) (Ping timeout: 186 seconds)
L478[13:40:49] <barteks2x> Caused by: java.lang.ClassNotFoundException: org.apache.logging.log4j.Level -> he did again
L479[13:48:25] ⇨ Joins: illy (~illy@cpe-24-166-23-169.cinci.res.rr.com)
L480[13:50:58] ⇦ Parts: malte0811 (~malte@185.134.128.197) ())
L481[14:22:24] <barteks2x> forge should show a more useful error message when someone deletes the vanill ajar from server installation...
L482[14:23:30] <barteks2x> he deleted it every time thinking it was a leftover jar from installing forge...
L483[14:33:31] <barteks2x> looks like it didn't work even in 1.12 forge 2381
L484[14:51:00] <Ordinastie> !gm func_73660_a
L485[14:52:00] <Ordinastie> fuck, can TE tick after their block has changed ?
L486[15:02:00] <barteks2x> you can scledule something to run the tick after
L487[15:02:18] <barteks2x> but it probably won't be saved if the chunk is unloaded before it happens
L488[15:02:36] <Ordinastie> no, I mean, it happens
L489[15:02:53] <Ordinastie> which fucks up anything you would do using the state
L490[15:03:39] <barteks2x> I don't think you should assume the state is actually your block
L491[15:04:38] <Ordinastie> I don't even, but probably didn't do it right : https://github.com/Ordinastie/MalisisDoors/blob/1.12/src/main/java/net/malisis/doors/tileentity/DoorTileEntity.java#L147
L492[15:04:55] <Ordinastie> but tbh, I should be able to assume it
L493[15:05:07] <Ordinastie> TE shouldn't tick if the block has changed :x
L494[15:06:55] <Ordinastie> I could understand the render cache not being quite in sync with the world, but the TE shouldn't have the issue
L495[15:07:43] <barteks2x> Ive seen the code for setBlockState and it does remove the TE from chunk...
L496[15:08:00] <Ordinastie> and yet : https://pastebin.com/V4kU3uAk
L497[15:08:57] <barteks2x> did you get the error or some user?
L498[15:09:09] <Ordinastie> got reported on the forum
L499[15:09:20] <barteks2x> coudl very well just be corrupted world
L500[15:09:38] <barteks2x> or some mod setting bl;ock directly using ExtendedBlockStorage
L501[15:10:16] <barteks2x> unless he specifically said it happens when placing block in a specific place
L502[15:10:44] <barteks2x> also, the memory lkeak doesn't happen on 1.11.2
L503[15:11:11] <barteks2x> which means it's probably a netty bug...
L504[15:12:01] <barteks2x> or another "surely it won't break anything" change
L505[15:20:08] ⇦ Quits: Javaschreiber (~Thunderbi@88-209-32-73.nga.highspeed-baumann.de) (Quit: Javaschreiber)
L506[15:27:10] <barteks2x> What is the difference between tasks setup and setupForge in forge dev environment?
L507[15:35:46] <ghz|afk> one of them depends on the other
L508[15:35:49] <ghz|afk> dunno what it does extra
L509[15:36:58] <barteks2x> also netty is a stupid liar... it said "To enable advanced leak reporting, specify the JVM option '-Dio.netty.leakDetection.level=advanced'" and thats also what documentation says, but the code actually says it should be io.netty.leakDetectionLevel
L510[15:37:10] <PaleoCrafter> setup depends on setupForge, it seems
L511[15:37:42] <kashike> PaleoCrafter: other way around
L512[15:37:47] <kashike> https://github.com/MinecraftForge/MinecraftForge/blob/1.12.x/build.gradle#L396
L513[15:38:34] <PaleoCrafter> yeah, just realized. gotta learn to read dependency in gradlew tasks --all xD
L514[15:39:38] <PaleoCrafter> considering it only additionally depends on ciWriteBuildNumber which only really does something on CIs (duh), they're basically the same
L515[15:40:05] <kashike> https://github.com/MinecraftForge/MinecraftForge/blob/1.12.x/build.gradle#L228
L516[15:40:09] <kashike> not only on ci :p
L517[15:40:27] <PaleoCrafter> you know what I mean ._.
L518[15:40:30] <barteks2x> now that I know that I should use -Dio.netty.leakDetectionLevel=paranoid and not -Dio.netty.leakDetection.level=paranoid debugging the leak will be managable
L519[15:41:25] <kashike> barteks2x: io.netty.leakDetection.level is valid
L520[15:41:31] <kashike> io.netty.leakDetectionLevel is the OLD property
L521[15:41:34] <barteks2x> the code says otherwise
L522[15:41:36] <kashike> private static final String PROP_LEVEL_OLD = "io.netty.leakDetectionLevel";
L523[15:41:38] <kashike> private static final String PROP_LEVEL = "io.netty.leakDetection.level";
L524[15:41:46] <barteks2x> in ResourceLeakDetercor private static final String PROP_LEVEL = "io.netty.leakDetectionLevel";
L525[15:41:59] <kashike> you're looking at old code then
L526[15:42:00] <barteks2x> this is literally what I have in forge dev environment
L527[15:42:12] <kashike> You don't have netty 4.1.9 then, which is what 1.12 uses
L528[15:42:31] <barteks2x> oh... right
L529[15:42:41] <barteks2x> I had to reload the class
L530[15:42:45] <barteks2x> after setupForge
L531[15:44:04] <barteks2x> well, then I'm almost sure it will show nothing
L532[15:44:07] <barteks2x> as it did before
L533[15:44:33] <kashike> how can I test locally?
L534[15:45:15] <barteks2x> opening browser, wait
L535[15:46:03] <barteks2x> link in this issue https://github.com/MinecraftForge/MinecraftForge/issues/4509
L536[15:46:16] <kashike> examplemod?
L537[15:46:19] <barteks2x> yes
L538[15:46:41] <barteks2x> It has to be tested on actual dedicated server
L539[15:46:50] <barteks2x> there is no issue on integrated server
L540[15:48:52] <barteks2x> this really looks like something people shoudlscream and complain about
L541[15:49:55] <barteks2x> WTF is that? Error:(178, 74) java: incompatible types: net.minecraft.world.gen.ChunkProviderFlat cannot be converted to net.minecraft.world.gen.IChunkGenerator
L542[15:50:30] <quadraxis> ChunkGeneratorFlat
L543[15:50:42] <quadraxis> you seem to be missing a class rename
L544[15:50:50] <barteks2x> hm... maybe I didn't do a clean
L545[15:52:06] <LexMobile> Integrated packets arnt really serialized so they won’t get this issue
L546[16:01:31] <barteks2x> but it does call toBytes/fromBytes. But that would explaion why vanilla using BlockPos.toLong would work on singleplayer
L547[16:02:17] <barteks2x> I can only imagine how many mods would be server incompatible if IMessage wasn't serialized on singleplayer
L548[16:04:08] <kashike> barteks2x: leak detector works here
L549[16:04:26] <barteks2x> didn't work for me. At least not from launcher
L550[16:08:22] <kashike> barteks2x: https://gist.github.com/kashike/dbc429e05171b69776f3a1c714e34f70
L551[16:08:26] <kashike> cause is most likely related to https://github.com/MinecraftForge/MinecraftForge/blob/1.12.x/src/main/java/net/minecraftforge/fml/common/network/FMLIndexedMessageToMessageCodec.java#L101
L552[16:09:44] <barteks2x> well that should help a lot
L553[16:14:14] ⇦ Quits: iPixeli (~iPixeli@5.80.20.223) (Ping timeout: 183 seconds)
L554[16:16:19] ⇨ Joins: iPixeli (~iPixeli@5.80.50.120)
L555[16:16:19] MineBot sets mode: +v on iPixeli
L556[16:19:24] *** PaleoCrafter is now known as PaleOff
L557[16:48:13] ⇦ Quits: ben_mkiv (~ben_mkiv@p4FED5D77.dip0.t-ipconnect.de) (Ping timeout: 383 seconds)
L558[16:52:08] <barteks2x> I don't knoe how it ever worked...
L559[16:52:34] <barteks2x> wait, ChannelRegistrationHandler should free it
L560[17:01:38] <barteks2x> of course... FMLProxyPacket is not ReferenceCounted
L561[17:02:04] <barteks2x> ByteBuf is
L562[17:02:45] <quadraxis> https://github.com/MinecraftForge/MinecraftForge/pull/4510
L563[17:02:54] <quadraxis> I'm a little ahead of you on that
L564[17:06:32] <barteks2x> I think I was just about to figure it out
L565[17:06:56] <barteks2x> why are both needed?
L566[17:07:26] <quadraxis> I did try to have FMLProxyPacket be ReferenceCounted, but I couldn't get it to work right
L567[17:08:16] <barteks2x> I was very close here, I already found that CPacketCustomPayload#processPacket calls release, but it's not called for processing forge packets probably
L568[17:08:21] <barteks2x> (wa sjust testing that)
L569[17:09:11] <barteks2x> also the packet goes through both FMLIndexedMessageToMessageCodec and ChannelRegistrationHandler so why release in both?
L570[17:09:31] <barteks2x> the duplicate has the same reference count
L571[17:09:59] <barteks2x> well, at least the duplicated one isn't shown as leaking
L572[17:10:45] <barteks2x> and right, it probably will fix my issue
L573[17:11:27] <billy> what would cause this code to throw class not found exception on EntityPlayerSP at line 5? I've told it client side, so i'm a bit confused.. https://hastebin.com/kuhiqasuyu.java
L574[17:12:00] <ghz|afk> just because it's client-side, doesn't mean it doesn't exist on the server
L575[17:12:46] <ghz|afk> best choice in that case is to add a method on your proxy
L576[17:12:48] <quadraxis> the packets you see going through both FMLIndexedMessageToMessageCodec and ChannelRegistrationHandler are because the ChannelRegistrationHandler deosn't handle them
L577[17:12:50] <ghz|afk> that only the clientproxy implements
L578[17:12:59] <quadraxis> so it forwards it on
L579[17:13:22] <quadraxis> it needs to clean up after the packets it does handle directly
L580[17:13:34] <ghz|afk> billy: example: https://github.com/gigaherz/ElementsOfPower/blob/master/src/main/java/gigaherz/elementsofpower/client/ClientProxy.java#L139
L581[17:13:42] <barteks2x> oh right. I see
L582[17:13:44] <ghz|afk> https://github.com/gigaherz/ElementsOfPower/blob/master/src/main/java/gigaherz/elementsofpower/network/SpellcastSync.java#L78
L583[17:14:39] <illy> scala has ruined me I keep on trying to type def in my java classes >.>
L584[17:15:29] <LexMobile> Shouldnt payload going out of scope free it?
L585[17:15:40] <barteks2x> it's not C++ stack allocation
L586[17:15:54] <barteks2x> and it's actual native memory here
L587[17:16:08] <barteks2x> (direct buffer)
L588[17:16:31] <billy> not sure I understand, ghz|afk. i have other packets going to the client that use the same setup and none of them throw the exception
L589[17:16:53] <billy> https://github.com/BillyGalbreath/Pl3xForge/blob/master/src/main/java/net/pl3x/forge/network/PacketHandler.java
L590[17:17:23] <ghz|afk> yes that means you are lucky enough that nothing has loaded the handler class in the server
L591[17:17:59] <billy> why would the handler get loaded in the server if I told simplenetwork its handled client side?
L592[17:18:19] <barteks2x> nrcause you are referencing the class somewhere in the code probably
L593[17:18:41] <billy> I assure you, I am not.
L594[17:18:58] <barteks2x> this line does INSTANCE.registerMessage(OpenInventoryPacket.Handler.class, OpenInventoryPacket.class, 1, Side.CLIENT);
L595[17:19:05] <ghz|afk> my point was that you should be going through the proxy as a matter of principle
L596[17:19:08] <ghz|afk> even if it happens to work
L597[17:19:29] <billy> the line right about it does that too, barteks2x. works fine.
L598[17:19:34] <billy> *above
L599[17:19:42] <ghz|afk> maybe it's a matter of referencing Gui
L600[17:19:59] <ghz|afk> or something else that causes some side-effect that causes the class to load
L601[17:20:01] <barteks2x> The line above is public static void init() {
L602[17:20:31] <billy> look at my last link (to my github with the full code)
L603[17:22:45] <barteks2x> could be that using the entity class in the lambda causes it. Who knows
L604[17:22:53] <barteks2x> it should go through proxy anyway
L605[17:23:09] <barteks2x> if it doesn;'t, even seemingly irrelevant changes may break it
L606[17:23:29] <kashike> the proxy system exists for a reason
L607[17:25:01] <quadraxis> lambda probably breaks it
L608[17:26:12] <billy> i know it exists for a reason.. i use it.. just havent here because every tut ive seen on SimpleNetworkWrapper shows this way. I'll route it through proxy even though the Side is already specified in the registration. seems redundant, though
L609[17:26:52] <barteks2x> the proxy is for the JVM to not attempt to load client only classes on the server
L610[17:26:58] <barteks2x> not for the packet to not be registered
L611[17:27:45] <barteks2x> the Side only specifies how the packet and the handler are going to be used, there is no way for it to control classloading
L612[17:28:27] <ghz|afk> yep the side you specify on packet registration is only used to decide if it should register the packet handler on the logical client table
L613[17:28:30] <ghz|afk> or the logical server table
L614[17:28:43] <ghz|afk> it has nothing to do with preventing classloading errors
L615[17:29:50] <barteks2x> How long does it take for new forge build to be available?
L616[17:30:23] <barteks2x> (I assume it is automated)
L617[17:30:51] <barteks2x> nvm, it's there
L618[17:33:52] <ghz|afk> no idea waht the "latency" is
L619[17:33:57] <ghz|afk> but not too long
L620[17:35:57] <barteks2x> 1 minute more than seems reasonable for me :D
L621[17:36:55] <ghz|afk> 30min would be reasonable also ;P
L622[17:38:38] <barteks2x> maybe for C++
L623[17:38:55] <barteks2x> or an IDE
L624[17:39:19] <ghz|afk> I mean as an user
L625[17:39:31] <ghz|afk> as a maintainers for forge, yeah 30min would be reason for concern
L626[17:39:31] <ghz|afk> ;p
L627[17:43:04] ⇦ Quits: Noppes (~Noppes@ip56530f2e.direct-adsl.nl) (Read error: Connection reset by peer)
L628[17:46:25] <barteks2x> so my leak is not fixed :(
L629[17:46:41] <quadraxis> well one thing i found
L630[17:46:46] <quadraxis> vanilla is dumb
L631[17:47:02] <barteks2x> that's not new information
L632[17:47:16] <quadraxis> and SPacketCustomPayload doesn't have the call to release() that CPacketCustomPayload does
L633[17:47:37] <barteks2x> This is server->client packets that do it
L634[17:47:46] <barteks2x> I don't send any client->server packets
L635[17:48:10] <quadraxis> can you see if there are any problems with <1MB packets only?
L636[17:48:44] <barteks2x> uh... ok, so I will have to recompile special version of my little test mod
L637[17:50:54] <quadraxis> I'll go do some tests with the packet splitting test mod you made as well
L638[17:52:18] <barteks2x> well, the test may not trigger packet splitting (i'm not exactly sure)
L639[17:53:08] <quadraxis> I meant the one already in the forge repos
L640[17:53:38] <barteks2x> oh the actual test mod?
L641[17:54:41] <quadraxis> yea
L642[17:54:57] <quadraxis> also the SPacket etc is server-> client
L643[17:55:01] <quadraxis> so it could be an issue
L644[17:55:59] <barteks2x> wait, iird forge client packets are server->client, so why would SPacket be server->client?
L645[17:57:02] <barteks2x> packet side really should specify CLIENT_TO_SERVER or SERVER_TO_CLIENT...
L646[17:57:41] <ghz|afk> hmm
L647[17:57:47] <ghz|afk> I thought S* were the server classes
L648[17:57:48] <quadraxis> SPacket is Packet<blah HandlerServer>
L649[17:57:50] <ghz|afk> and C* were the client classes
L650[17:57:56] <ghz|afk> regardless of which direction they go in
L651[17:58:09] <quadraxis> and of course i mean HandlerClient
L652[17:58:30] <ghz|afk> although it seems I got it wrong :P
L653[17:58:50] <barteks2x> I just always find where it's used and discard the letter
L654[18:00:40] <barteks2x> no issues so far with 256k packets
L655[18:05:59] <barteks2x> 256k packets seem to work fine so it's packet splitting leaking them
L656[18:08:03] <quadraxis> I'm working on that atm
L657[18:17:41] ⇦ Quits: Nedelosk (~Nedelosk@ip-109-90-75-57.hsi11.unitymediagroup.de) (Read error: Connection reset by peer)
L658[18:36:51] ⇦ Quits: SanAndreasP (~SanAndrea@mue-88-130-48-235.dsl.tropolys.de) (Quit: Leaving)
L659[18:43:18] ⇦ Quits: Redfoxmoon (~Red@177.92-221-236.customer.lyse.net) (Ping timeout: 186 seconds)
L660[18:45:21] ⇨ Joins: Redfoxmoon (~Red@177.92-221-236.customer.lyse.net)
L661[18:48:43] ⇦ Quits: illy (~illy@cpe-24-166-23-169.cinci.res.rr.com) (Remote host closed the connection)
L662[18:48:59] ⇨ Joins: illy (~illy@cpe-24-166-23-169.cinci.res.rr.com)
L663[19:31:04] ⇨ Joins: Cast0077 (~Cast0077@24-151-68-108.dhcp.nwtn.ct.charter.com)
L664[21:06:19] ⇦ Quits: Spottedleaf (~Spottedle@node-1w7jr9qqos9g2tc1p0ioryubz.ipv6.telus.net) (Ping timeout: 383 seconds)
L665[21:11:11] ⇨ Joins: Spottedleaf (~Spottedle@node-1w7jr9qqos9g0ifecvv8gvk8h.ipv6.telus.net)
L666[21:40:54] ⇦ Quits: KGS (~KGS@h-158-174-9-50.NA.cust.bahnhof.se) (Ping timeout: 186 seconds)
L667[21:45:59] ⇦ Quits: Cast0077 (~Cast0077@24-151-68-108.dhcp.nwtn.ct.charter.com) (Quit: Poof)
L668[22:00:06] ⇦ Quits: Dark (~MrDark@2607:fcc8:d48b:eb00:ac78:9864:f7e2:7cad) (Ping timeout: 186 seconds)
L669[22:07:05] *** illy is now known as illy|Zzz
L670[22:09:48] ⇨ Joins: McJty (~jorrit@ptr-9197ufnp00ovcehdu1k.18120a2.ip6.access.telenet.be)
L671[22:14:31] ⇦ Quits: illy|Zzz (~illy@cpe-24-166-23-169.cinci.res.rr.com) (Ping timeout: 200 seconds)
L672[22:17:18] ⇨ Joins: h404bi (~h404bi@119.129.117.245)
L673[22:25:10] ⇦ Quits: Lathanael|Away (~Lathanael@p54960B82.dip0.t-ipconnect.de) (Ping timeout: 186 seconds)
L674[22:31:16] ⇨ Joins: Lathanael|Away (~Lathanael@p5496053E.dip0.t-ipconnect.de)
L675[23:39:08] ⇦ Quits: Brokkoli (~Brokkoli@p5b23ce53.dip0.t-ipconnect.de) (Quit: Die Sprache der Politik ist daf�r gemacht, dass L�gen wahr klingen und das T�ten angemessen wirkt. (George Orwell))
<<Prev Next>> Scroll to Top