Time Nick Message 00:38 paramat hmmmm, line 107 of mg_decoration.cpp suggests i can use noise params instead of a fixed fill ratio, how is this done and what format? here's my failed attempt https://gist.github.com/paramat/05d14c973db2dfc080cb 01:14 hmmmm hi paramat, it looks good to me but you need to keep in mind that the rainforest biome is its own distribution along with the noise 01:14 hmmmm so where the decorations actually get placed is the intersection between the areas of that biome and those noise params 01:14 paramat okay 01:14 hmmmm you should change default:junglegrass to something that would stand out better such as default:bricks to verify if it indeed works 01:15 hmmmm try removing the biomes parameter if you can't seem to find any placed decorations 01:20 paramat thanks i'll experiment further, hopefully just my own silly mistake =) 06:56 Zeno` has anyone else tried #1771? 06:56 ShadowBot https://github.com/minetest/minetest/issues/1771 -- Increase step smoothing to fit 1:1 stairs (works well on slabs too) by Calinou 06:57 Zeno` I compared with and without the patch and the new (patched value) seems much more natural to me 06:57 Zeno` while still retaining an acceptable amount of "bobbing" 07:25 RealBadAngel +1 for 1771 07:25 Zeno` Ok, I'll merge now ;) 07:25 Zeno` it looks heaps better IMO 07:26 Zeno` unless you want to merge it 07:26 Zeno` lol :) 07:26 RealBadAngel nah, i do have a mess here with my own patches 07:27 RealBadAngel i think i will strip down the extruded meshes patch just for plantlike 07:27 Zeno` so do I... I deleted some of my still pending branches when cleaning up my repo :-o 07:28 RealBadAngel i cant just delete my own, i do have some interesting stuff stored there 07:28 Zeno` I didn't delete them locally (thankfully) just on github 07:28 RealBadAngel there are water and lava shaders (waiting for better days) ;) 07:29 RealBadAngel have you actually saw the lava shaders? 07:29 Zeno` I saw a youtube of them (awesome) 07:31 RealBadAngel i will merge them once we have clean mapblock mesh 07:32 RealBadAngel but we have to fix all the drawtypes before that 07:33 Zeno` you'll fix all those tonight, right? :D 07:35 RealBadAngel hehe i doubt it 07:35 Zeno` :) 07:35 RealBadAngel today i will just make the plantlike rdy 07:36 Zeno` While you're here RBA: #1872 07:36 ShadowBot https://github.com/minetest/minetest/issues/1872 -- Fixes for Android after recent commits. by KodexKy 07:36 Zeno` it already has sapier's go ahead, but a 3rd opinion cannot go astray 07:36 Zeno` I've asked KodexKy to squash and will commit once he does so 07:37 RealBadAngel im ok with it 07:37 Zeno` The fixes are needed a) because of refactor_main; and b) the global thingy that SN added 07:37 Zeno` cool 07:37 RealBadAngel you dont have to wait for original commiter to squash them 07:38 Zeno` yeah I was in a "being polite" mood hehehe 07:38 RealBadAngel just make a note "please squash your commits in the future" and just do it 07:38 Zeno` Plus it gave me an opportunity to explain why I thought those two should be squashed 07:39 Zeno` If he doesn't squash I'll just do it, but it was a way of opening up communication and it's not super urgent to merge 07:39 Zeno` Unless you use Android of course :) 07:41 Zeno` Have you built for Android? I dunno if it's just because it was the first time I've done it (I did it today) but it was kind of tricky for me 07:42 RealBadAngel i was testing some builds, but havent built for myself, were using sapier's builds 07:43 RealBadAngel i do have 2 devices to run android builds, Samsung Galaxy tab 3 and x86 kitkat 07:43 Zeno` It made me wonder how many people using Android actually use "simple single player" mode and whether or not a "light client" would be worth it 07:44 Zeno` i.e. a client version without the server stuff 07:44 RealBadAngel imho most mobile users wont need single player 07:45 Zeno` yeah, so it might be worth it. I'll look at the feasibility tomorrow I think 07:45 RealBadAngel mobile devices imho are too weak to run server and the client at the same time 07:46 RealBadAngel and there are already some servers out there to let them connect 07:46 Zeno` exactly 07:47 RealBadAngel making the client lighter and faster can only effect in improving gameplay and in the longer run raising number of servers 07:48 RealBadAngel they wont see the reason in doing so if the client will be slow or unplayable 07:49 RealBadAngel also one day i would like to return to the idea of connected servers 07:50 Zeno` I'd like that as well, but there are some tricky parts 07:50 RealBadAngel so players could use in-game teleporters, portals or something like that just to walk from one server to another 07:50 RealBadAngel it shouldnt be very different like mc's loading nether and back 07:51 RealBadAngel will just take some time 07:51 RealBadAngel and restrictions can be defined by target server 07:51 RealBadAngel like player can transfer inventory/ cant tranfer it 07:53 RealBadAngel in the worst case player will travel with unknown nodes, at his own risk 07:54 RealBadAngel server wont crash because of it, it can already handle unkowns 07:54 Zeno` yeah 07:54 Zeno` but are node IDs guarenteed to be the same server to server (I doubt they are) 07:54 RealBadAngel that wont need the id's just names 07:55 Zeno` yeah, that'd work then 07:56 RealBadAngel but even, just for starters, version that wouldnt allow inv transfers would be nice 07:57 RealBadAngel so we could have kinda nexus servers with many portals leading to differnt worlds 07:57 Zeno` but in the long run, wouldn't it be awesome if you could join a world that is rich and coal, and then go to another world that is coal poor and then trade it (for example) 07:57 RealBadAngel exactly 07:57 Zeno` that'd add a whole new dimension of gameplay 07:57 RealBadAngel this is WoM 07:58 RealBadAngel ;) 07:58 RealBadAngel World of Minetest 07:58 RealBadAngel and something voxels games does not have 08:00 RealBadAngel also look at this from another point of view, you have 40 servers running with a few players on each one 08:00 RealBadAngel now let them freely walk from one to another. you have game with many players online 08:01 Zeno` yep 08:01 Zeno` it's something I've thought a lot about as well 08:01 Zeno` it'd require opt-in and cooperation between server operators but I think that that would happen 08:02 Zeno` e.g. I'd have no problem if I could interact my server with lots of others 08:02 RealBadAngel just skip back to main menu/join tab and reuse player data to automatically connect to other server (with name:port) given by previous server 08:02 Zeno` could have industrial worlds, natural worlds, fantasy worlds each offering different resources etc 08:03 RealBadAngel failback is to reconnect to old one 08:03 Zeno` I'd even upgrade my VPS to enable it :D 08:03 Zeno` although I doubt I'd have to 08:03 RealBadAngel that doesnt need so much changes 08:03 RealBadAngel all the code is here 08:04 RealBadAngel just skip the menu and connect to something other 08:05 RealBadAngel only difference is that you have to origin (in case of fail) and destination server's data 08:05 RealBadAngel and make client connect again 08:05 Kodexky Hi 08:06 RealBadAngel rest is just cosmetic (some info for a player just) 08:06 Zeno` yep 08:06 Zeno` hi Kodexky :) 08:07 Zeno` I was just talking about merging your latest PR 08:08 Zeno` glad to finally meet you as well 08:08 Kodexky Yes, i was doing my last check of everything before going to sleep and saw it. 08:08 Kodexky Its a pleasure btw. 08:09 Zeno` Well, I compiled minetest for Android for the first time today 08:09 Zeno` And I didn't enjoy it much at all lol 08:09 Zeno` So I for one am glad you're around 08:09 Kodexky hehe first time its just painful. It took me 2 days. 08:10 Zeno` took me about 6 hours so that's probably the same 08:10 Zeno` and about 7 hours yesterday to install the SDK and NDK :-o 08:11 Kodexky Eh, i was paranoid enough to believe i was doing something terribly wrong. 08:11 Zeno` So was I. I couldn't get the emulator to work on x86_64 either (although I worked it out eventually) 08:12 Zeno` merged 08:12 Zeno` thanks for squashing. As I said, I could have done it but I wanted to take the opportunity to explain why I thought that in this case squasing would be good 08:13 Kodexky puff, i tried the emulator to skip downloading to my tab. tried once, wont do it again ever!!. Hope you get your tab soon. 08:13 Zeno` I got it yesterday :) 08:13 Kodexky i was about to ask. to squash or not to squash? 08:14 Zeno` well, squashing is not always necessary. But I think in this case the 2 commits were so closely related and small that they warrented squashing 08:14 Zeno` e.g. I merged something the other day where there were two commits, but they sort of related to different things (more than this current one) 08:14 Kodexky ok, so its case by case. 08:14 Zeno` mostly 08:15 Zeno` and probably pretty subjective as well :) 08:15 Zeno` so it's hard, but I think keep doing it the way you're doing 08:15 Zeno` someone will ask if they think they need to be squash 08:16 Zeno` let me find an example where I didn't squash 08:16 Zeno` https://github.com/minetest/minetest/pull/1850 08:17 Zeno` Those two commits are obviously not related, so squashing would have been bad 08:17 RealBadAngel general rule is one commit per PR 08:17 Kodexky Ok, i see. I'll try to avoid multiple commits when possible. 08:18 Zeno` yeah 08:18 Kodexky btw 3 errors in 4000 lines is way better than my brightest day. 08:18 RealBadAngel so if you want commit that fixes a, and a second that makes b 08:18 Zeno` I could argue successfully that your two commits were distinct as well (because they fix bugs caused by different changes) but separate PRs are best (or squashed ones if they can be related to each other enough) 08:19 RealBadAngel so make two PRs for them 08:19 Zeno` Kodexky, thanks :) 08:19 RealBadAngel if all your commits are in area of a, squash them 08:21 Kodexky Ok. Im still getting used to git. Until now i've only coded for myself or as a lone dev. 08:22 Kodexky Btw, about what RBA and you where talking... 08:22 Kodexky Please, don't cut down mobile clients. 08:22 RealBadAngel huh? 08:23 Kodexky I mean... 08:23 Kodexky Dont remove server capabilities. 08:23 Zeno` Kodexky, it would be a compile time flag 08:23 RealBadAngel while it can be separated, it will be easy to make two builds 08:24 RealBadAngel one with server and one light with client just 08:24 Zeno` e.g. something like -DSLIM_CLIENT 08:24 Zeno` or something like that 08:24 Zeno` removing it entirely would be silly :) 08:24 Zeno` The official builds, I'd imagine, would be "full" 08:25 Kodexky Ok, i get it now. Sorry im a little overworked today. 08:25 Zeno` This is all at the "idle chit chat throwing ideas about stage" anyway :) 08:25 Zeno` It's ok 08:25 Zeno` RBA overworks me all the time so I am constantly tired 08:25 * Zeno` hides 08:26 Kodexky haha 08:26 RealBadAngel lol 08:29 Kodexky On my personal client i have restored classic mainmenu, i like to test mod there when on the move, i even carry a copy of my home server when possible, i panicked after half reading about slimming. 08:29 RealBadAngel bbl, work time 08:29 Zeno` oh yeah, don't worry about the slimming. If it even happens it'll be optional heh 08:30 Kodexky wonderful 08:30 Zeno` Well... 6:30PM here... I'd better cook some dinner 08:31 Zeno` nice to meet you Kodexky and thanks for the Android fixes 08:32 Zeno` Kodexky, something to think about... 08:32 Zeno` on PC, it's faster to run a local server and then connect the client to that local server (rather than the simple singleplayer mode). Why? I have no idea 08:33 Zeno` Hypothetically it should be identical but it's not 08:34 Kodexky Its faster indeed. Ended doing so myself. 08:36 Kodexky I was not on my machine and using windows. Had to do it that way to cope with the frequent crashes. 08:37 Kodexky well, anyway too tired to think straight. Nice to meet you and RBA as well. Bon appetit, im off to morpheus land. 08:43 Zeno` cya :) 09:40 rubenwardy Hello! Can this be merged? https://github.com/minetest/minetest/pull/1843 10:03 Zeno` rubenwardy, it can 10:03 Zeno` if a dev other than myself agrees 10:04 Zeno` This is the second case of a PR where the argument against merging it is that it takes away control from server owners/game creators 10:05 Zeno` I think it's a weak argument 10:05 Zeno` I haven't met a server owner yet who gives a sh*t 10:06 sol_invictus there are a lot of pull requests blocked by weak arguments 10:07 Zeno` sol_invictus, I still need approval from a second core dev 10:07 sol_invictus Zeno`: I do understand 10:08 sol_invictus just pointing out there is no debate here (: 10:08 Zeno` :) 10:09 Zeno` Most of the people running servers, who I have spoken to, are also players 10:09 Zeno` so this would be acceptable 10:10 sol_invictus a vivid example of a weak argument: https://github.com/minetest/minetest/pull/1689#event-197321674 10:10 Zeno` I can of course put myself in the place of imaginary server operators 10:10 Zeno` But why should I do that? 10:11 sol_invictus I think all those arguments about cheating is complete bulls**t 10:11 sol_invictus you just remove checklocalprivilege and you got everything you want 10:11 Zeno` Well, I think the same arguments saying "this should be a server operators decision" are also bullshit 10:11 sol_invictus no need to be an egghead c++ coder 10:12 Zeno` It's bullshit because I have not seen a game (or a server) where the operator/game dev has made that decision 10:13 sol_invictus aha 10:14 Zeno` look at #1867 10:14 ShadowBot https://github.com/minetest/minetest/issues/1867 -- added enable_build_where_you_stand option by Sokomine 10:14 Zeno` every single player (and most of those players also have a server) want it 10:14 Zeno` saying that "well there might be a server operator who wants to control that" is, IMO, dumb 10:14 Zeno` saying it's not dumb... blocking it because of that is dumb 10:15 Zeno` I mean PilzAdam does have a point, but is it meaningful? 10:15 sol_invictus it's dumb because a server operator can't really control what's happening client-side 10:16 Zeno` Take my server. I want people to be able to place a block where they stand. I'm not going to give them noclip to do that 10:16 Zeno` And I really don't care if they place a block where they stand or not (and I doubt any server operators do) 10:17 Amaz If you change the code, (semi-hacked client) you can do it anyway. 10:18 sol_invictus that's m point, Amaz 10:18 Zeno` of course you can 10:18 Amaz So server operators don't really have that much control. 10:18 Zeno` If I, as a server operator, wanted to force pink clouds as a game decisions... well that's *my* problem 10:19 Zeno` not the engine's problem 10:20 Zeno` I can see where some things that players want and what hosting people want can be in conflict... but these examples are NOT (IMO) examples of them 10:20 Zeno` and I want to merge these requests because I agree with the majority rather than the minority 10:24 Zeno` I will ignore arguments that mention "the game designer or a server operator" unless a game designer or server operator speaks up :) 10:24 Zeno` If they don't then the safe assumption is that they don't care 10:25 Zeno` If players speak up, I'm sure those running servers can as well 10:26 Zeno` without me trying to imagine what they want or don't want 10:29 Megaf [4431056.629056] Out of memory in UB 102: OOM killed process 31900 (minetestserver) score 0 vm:1187744kB, rss:505876kB, swap:518020kB 10:29 Megaf Please fix your memory leaks! 10:31 Megaf Maybe those billions of warnings are cause some kind of memory leak 10:31 Megaf I wouldnt have memory leaks before that 10:31 jin_xi ...so turn off strict module or what its called 11:08 Zeno` can it be turned off? 11:10 jin_xi comment out line 20 in builtin/init.lua 11:10 jin_xi ? 11:10 Zeno` that's not turning it off :P 11:10 Zeno` well it is, but it's not a setting 11:21 jin_xi should only explicitly be turned on for server setup/mod developping 11:45 Zeno` make a pR :P 11:47 Zeno` #1843 ... comments please before I merge? 11:47 ShadowBot https://github.com/minetest/minetest/issues/1843 -- Add setting to customise stack max by rubenwardy 11:47 Zeno` sfan5 ^^^ 11:50 Zeno` celeron55 ^^^ 11:52 Zeno` Merging #1744 as it's been agreed upon (although not as an ideal solution) 11:52 ShadowBot https://github.com/minetest/minetest/issues/1744 -- Only set the camera update keymap when using a debug build by rubenwardy 12:00 Zeno` merging #1874 (sapier + zeno) 12:00 ShadowBot https://github.com/minetest/minetest/issues/1874 -- Fix some undeclared global variables by CraigyDavi 12:00 Zeno` in 10 minutes :) 12:04 Zeno` those fixes are required due to SN "strict" patch thingy 12:42 Zeno` I think I fell asleep! 13:09 Zeno` Wayward_Onezie! 13:11 Zeno` oop, sorry. Wrong window 13:34 kahrl ...why didn't I think of that earlier 13:35 kahrl I found an embarrassingly simple fix for the "thin black line" artifacts on the wield item that doesn't enable anisotropic filtering 13:35 Zeno` fix it then :) 13:35 kahrl (so it fixes that as well as the blurriness problem) 13:36 kahrl replace material.setFlag(video::EMF_ANISOTROPIC_FILTER, true); with material.setFlag(video::EMF_USE_MIP_MAPS, false); 13:37 kahrl I'll merge it in half an hour so we can see what that one will break 13:37 kahrl :P 13:42 kahrl I made it respect the anisotropic_filter setting as well, here is the full commit: https://gist.github.com/kahrl/e7e4d3f233bc2272fb73 13:43 kahrl brb 14:15 PilzAdam https://github.com/minetest/minetest/pull/1867#issuecomment-64518337 we should make a fundamental decision whether we want to optimize the game for players that are _currently_ playing the game, or whether we want to optimize it for more long-term goals (like turning Minetest into a generic engine) 14:16 PilzAdam this would make it easier for everyone to handle such pull requests 14:17 Zeno` Hi PilzAdam 14:17 Zeno` I am glad you're here ;) 14:17 Zeno` Do you know why BlockMen added that "feature" in the first place? 14:18 PilzAdam which "feature"? 14:18 Zeno` And, yes, possibly there needs to be a fundamental decision made, but I'm only interested in the history right now 14:19 Zeno` The "feature" that stops people placing a node where they are standing 14:19 PilzAdam AFAIK it was added because it's more realistic? 14:20 Zeno` In real life, if I am standing right here, I can pour a block of concrete on myself 14:21 PilzAdam but you can't place a block of stone where your feet are 14:21 Zeno` But my point is that the original decision was a fundamental decision as well 14:21 Zeno` And from what I can find there was not a huge issue making this original decision 14:22 Zeno` (I'm just trying to understand; I'm not singling you out... well, I am in a way, but not maliciously) 14:23 PilzAdam I don't see how it was a fundamental decision 14:23 Zeno` So how is it a fundamental decision in making the behaviour optional? 14:24 PilzAdam (it already is optional) 14:24 PilzAdam this is not limited to only this issue 14:25 Zeno` I know 14:25 Zeno` how is it already optional? 14:25 PilzAdam if the server grants you noclip, then you can turn it off 14:25 Zeno` but noclip gives way more than this feature 14:26 PilzAdam not if you dont have fly 14:26 Amaz It allows you to see underground. 14:27 Zeno` ok, let's step back a bit 14:27 Zeno` why is this feature bad? 14:27 PilzAdam Amaz, oh, right 14:27 Zeno` until 6 months ago it was the default behaviour 14:28 PilzAdam Zeno`, because it's counter-intuitive (https://github.com/minetest/minetest/issues/85) 14:28 Zeno` and I do not know ANY server operator or game creator who has a problem with this 14:29 Zeno` well, flying is not exactly intuitive either 14:29 kilbith place_where_you_stand also allow more easily the building of the high, ugly 1x1 towers on servers... 14:29 Amaz If a games wants players not to use it, it can use setting_set or whatever. 14:29 Zeno` kilbith, rubbish. It's still very easy to do that 14:29 Amaz kilbith, it's disabled by default anyway. 14:29 PilzAdam Amaz, Lua games can't change the client configuration 14:30 kilbith actually, it's not that easy as before. 14:30 Zeno` kilbith, it is 14:30 Amaz Ah. I was thinking in singleplayer. 14:30 Zeno` kilbith, join ANY server and make a 1x1 pillar currently 14:31 Zeno` heck join mine and I'll give you unlimited cobble to make one. It's not hard 14:32 Zeno` PilzAdam, as I said in my comment I understand your position 14:32 Zeno` And in many ways agree with it 14:32 Zeno` But... 14:32 pitriss Zeno`: I support to revert to previous state or make it player configurable..:) 14:32 Zeno` The players want this 14:33 Zeno` No server operators have commented on it, so I have to assume they don't care 14:33 pitriss that was worst commit in last year.. 14:33 pitriss if i want realistic building i can go outside and play with real bricks:) 14:34 PilzAdam Zeno`, the question now is, do we follow what the current players want from us, or do we follow what would be better for long-term things (e.g. for future game creators, that want control over this) 14:34 PilzAdam Zeno`, Megaf is a server operator 14:34 celeron55 client-side options are quite cheap and can be removed or replaced with server-side ones without many issues 14:34 Zeno` PilzAdam, we, for now, follow what the players want 14:35 celeron55 i don't really see a reason why that couldn't be a client-side option for now 14:35 Zeno` if a server person wants this added then they can argue their point 14:35 PilzAdam celeron55, are we still in a state where we add things "for now"? 14:35 Zeno` PilzAdam, and MegaF wants this commit 14:35 celeron55 PilzAdam: when iterating on software, you always add things "for now" and remove them later if needed 14:36 Zeno` We should not make up imaginary scenarios 14:36 celeron55 we just should keep track of what is there "for now" and what is meant to be persistent 14:36 celeron55 it might be somewhat unclear 14:40 Zeno` PilzAdam, I think the original commit that disabled this behaviour was not thought through as thoroughly as this commit that adds it back as an option 14:40 celeron55 PilzAdam: also, i don't see this as being important to a game; why do you think it is? 14:41 Zeno` And I think you have guessed now that if I get one core dev to support me merging it I will. I will also, of course, consider a future pull request that reverts that behaviour if there is a good reason to do so. 14:43 PilzAdam celeron55, it may be an unwanted option to hide from players or mobs 14:43 Zeno` PilzAdam, these are may may may mays. There is no evidence that server operators care about that at all 14:44 PilzAdam Zeno`, server operators are not game developers 14:44 Zeno` PilzAdam, point me to a game developer that cares about this :( 14:44 PilzAdam <- 14:45 celeron55 well in that case i don't know, try to solve it somehow 14:45 PilzAdam its a rather minor detail, but I think it's "weird" to be able to walk inside nodes if you placed them inside yourself 14:45 celeron55 clearly the current situation is too inconvenient though (according to what people report) 14:45 Zeno` in that case I say we merge and THEN work it out 14:45 Zeno` imo the behaviour should not have been changed in the first place 14:46 Zeno` and it possibly can be worked out, but the PR is probably good for now 14:46 Zeno` at least it's set off by default 14:47 PilzAdam I never said I oppose to merging it (I think I'm too inactive currently to do that) 14:47 celeron55 like, the thing here is that a game should be able to disallow it, but most games don't care and thus in those cases it should be the user's choice 14:47 Zeno` PilzAdam, will you agree to me merging it then? 14:47 hydalgo ThatGraemeGuy: thank you for the answers! 14:47 PilzAdam no, because I think it should be done better 14:47 PilzAdam I guess I'm "neutral" then? 14:47 Zeno` And then after that we can work on a better solution 14:49 Zeno` Look, the way I see it is that most servers don't care and the players want this (configurable) 14:50 Zeno` This PR will revert to the previous behaviour for those who want it and are smart enough to do it 14:50 Zeno` it seems a small concession to make for the interim 14:51 Zeno` Then when a properly considered change is devises it can be another pR 14:52 Zeno` devised* 14:53 Zeno` At this moment in the evolution of the game I'd rather keep the players happy than game developers. I understand the importance of this issue but atm there is no serious alternative 14:55 Zeno` PilzAdam, I really would like your understanding before I merge this :( 14:55 Zeno` it's not the end of the world 14:56 PenguinDad How about using a server side setting to "block" the client side setting? 14:56 PilzAdam I do understand your point 14:56 Zeno` PenguinDad, that's possible but that has to be another PR 14:57 Zeno` this current PR is merely reverting the behaviour to how it always was but requiring the user to choose the "always was" behaviour 14:59 Zeno` and I really really doubt casual users are going to bother modifying their .conf 14:59 shadowzone I don't mind modifying my .conf but it does get a bit difficult 14:59 shadowzone especially considering I'm only using 1 monitor. 14:59 Zeno` see, even shadowzone can't be fscked doing it 15:00 shadowzone And I'm a pro or what? 15:00 Zeno` err... yeah you're a pro! 15:00 shadowzone Okay. 15:00 shadowzone I guess 15:01 Zeno` C'mon PilzAdam, be reasonable :( 15:02 PilzAdam what do you want from me? 15:03 Zeno` just say "ok merge it" and we'll work out a better solution later :p 15:04 PilzAdam can't you find another dev to do that? 15:04 celeron55 merge it 15:04 Zeno` ok :) 15:04 Zeno` thanks 15:04 Zeno` PilzAdam, we will work on this though ok? 15:05 PilzAdam I don't know how often I will be here 15:05 PilzAdam so be prepared to do it alone 15:05 Zeno` that's ok... it's on my TODO list 15:36 Ritchie hello, i don't like that commit from issue #85 too 15:36 ShadowBot https://github.com/minetest/minetest/issues/85 -- Placing a block on the ground where the player stands 15:37 VanessaE zeno merged a fix for that with commit 9f927900 15:38 VanessaE update to HEAD and set, enable_build_where_you_stand = true in minetest.conf 15:57 PilzAdam https://launchpadlibrarian.net/191262376/buildlog_ubuntu-precise-amd64.minetest_201411261546-0~3107~ubuntu12.04.1_FAILEDTOBUILD.txt.gz 15:58 PilzAdam ^ kahrl 16:09 PilzAdam my local build fails too, using Irrlicht 1.7.2 18:00 RealBadAngel hi 18:00 RealBadAngel whats up? 18:06 kahrl PilzAdam: oh wtf 18:06 kahrl why does everything have to be so hard? 18:07 kahrl is there a way to accomplish the same thing in irrlicht 1.7.2? 18:09 Krock what's bad about using 1.8.1? 18:09 rubenwardy Lack of support in package systems. 18:09 rubenwardy Although that might have changed 18:09 * Krock is not affected :) 18:13 Calinou https://packages.debian.org/search?keywords=libirrlicht-dev 18:13 Calinou https://packages.ubuntu.com/search?keywords=libirrlicht-dev 18:14 Megaf Hi all 18:14 Megaf Zeno`: I have some information on the memory leak 18:17 Megaf Zeno`: PilzAdam: kahrl: It seems like minetest server is always consuming more and more and more RAM, without ever releasing it. 18:17 Megaf http://paste.debian.net/plain/133688 18:21 kahrl PilzAdam: I fixed the build issue for now 18:23 PilzAdam yeah, works now 18:23 PilzAdam but I see minor graphical glitches :-/ 18:24 kahrl I know 18:24 kahrl I hope we can find a way to fix them 18:24 kahrl if not, eventually the distros will move on to 1.8 ;) 18:26 PilzAdam that would be 2017 for mine 18:29 Megaf You dont really care about hardware with limited resources, do you? Why bother fixing something if computer have lots of ram and will not crash because of a stupid bug... 18:30 Megaf Its easier to say that its crashing because the little amount of ram...# 18:30 Megaf like sapier seing that it was crashing because it was arm 18:33 RealBadAngel Megaf, sometimes its just true 18:34 RealBadAngel folks like to turn everything on and then complain 18:34 twoelk yeah, I want everything and I want it now 18:35 RealBadAngel and we should provide i7 for each mt user, sure :P 18:35 RealBadAngel any more wishes? 18:35 Krock i9 pls 18:36 RealBadAngel mt claus has i7s this yr only, so sorry pal ;) 18:36 Krock eww too bad. Then I'll stay on my Athlon 4000+ :P 18:37 RealBadAngel please do notice im developing all that high end effects on core 2 duo, 1,8ghz 18:37 * twoelk takes note that every piece of trivial software and evry tiny game is not whispering anymore but screaming: go! go and buy a new computer 18:37 RealBadAngel and i have good frame rate 18:38 RealBadAngel otherwise i wont be even trying to make a pull with a eye candy 18:39 RealBadAngel one thing i have noticed lately: i switched from HD 4670 to GTX 550 Ti 18:39 RealBadAngel and guess what 18:39 RealBadAngel no frame rate change, at all 18:40 twoelk so the probleme lives somewhere else than better hardware? 18:40 RealBadAngel we got lately fps ups, real ones 18:40 RealBadAngel all thx to optimization of the code 18:40 RealBadAngel and redoing things 18:41 Calinou 550 Ti is not that much better compared to 4670 18:41 Calinou also, game is CPU limited, heavily 18:41 Calinou if I don't limit FPS, then my CPU is capped, not my GPU 18:41 Calinou my GPU has lots of room when I have 200 FPS… 18:41 RealBadAngel calinou at least 3 times faster 18:41 Calinou oh 18:41 RealBadAngel 4670 was budget card 3 yrs ago 18:41 RealBadAngel and i paid 150$ for it then 18:42 RealBadAngel now i would have to pay the same to get used 550ti 18:43 kahrl PilzAdam: are the glitches still there if you enable anisotropic filtering? 18:44 PilzAdam kahrl, I have aniso filtering enabled... 18:45 kahrl hmm, strange 18:45 kahrl what exactly do the glitches look like? 18:46 PilzAdam thin lines 18:46 kahrl that's the kind of thing that anisotropic filtering fixes (if it doesn't cause other problem), but I know what the problem is 18:46 kahrl other problems* 18:47 kahrl RealBadAngel: do you think anisotropic filtering can be enabled on the wieldmesh regardless of resolution? (unlike bi/trilinear filtering) 18:47 PilzAdam do I need to configure anything else except the checkbox in the menu? 18:47 kahrl PilzAdam: no 18:47 kahrl but it isn't applied because of wieldmesh.cpp:283 18:49 puhfa hi yall. we have a mod (no idea which one) that causes get_perlin_map to crash the server (bad_alloc), and i think the mod is calling register_on_generated with the parameters in a wrong order (max, min instead of min, max). so the question is, is the unit of get_perlin_maps size parameter nodes or blocks? 18:49 puhfa https://github.com/ethic-rush/snow/blob/master/src/mapgen_v6.lua#L94 <- i think this results in <= 0 in our case 18:50 kahrl RealBadAngel: in other words, do you agree with https://gist.github.com/kahrl/52e88e9a78863be2ea1c? 18:50 puhfa (sry i thought you were done) 18:50 kahrl puhfa: no problem 18:51 kahrl puhfa: is that related to https://github.com/minetest/minetest/issues/1858? 18:52 puhfa probably. i tracked the c++ source down and the 3-param constructor of Noise was the first one where memory was allocated 18:53 RealBadAngel kahrl, yes 18:53 puhfa we are trying to fix it by ensuring max > min in register_on_generated, but havent tested it yet 18:53 kahrl RealBadAngel: ok, I'll merge it 18:54 RealBadAngel btw, when switched to gtx550ti i have no "holes in meshes" problem at all 18:54 RealBadAngel even for 16x in cache 18:55 hmmmm puhfa, the units are in nodes. the range is inclusive 18:56 Calinou <+RealBadAngel> now i would have to pay the same to get used 550ti 18:56 Calinou what?!?! 18:56 puhfa thanks 18:56 Calinou an used 570 is ~ 70 € 18:56 Calinou where do you buy your cards? 18:56 kahrl PilzAdam: pushed that fix, does it work for you? 18:56 hmmmm mixing up your minp and maxp would result in a negative number, which would be an extremely large unsigned number. so yes, that would cause the issue you're talking about 18:56 PilzAdam kahrl, yes :-) 18:57 kahrl awesome :) 18:57 puhfa would love to find the mod that calls it with such parameters. is there any easy way to find out which one it is? 18:57 hmmmm hmmm... I think I know how to 'fix' the Noise ctor std::bad_allocs... simply throw an exception saying "Your noise parameters are either incorrect or too high for this particular machine to handle!" 18:57 hmmmm puhfa, I think a lua stack backtrace would do the trick 18:58 puhfa nope, https://gist.github.com/crazyR14/8c5bd478997b0fb7df04 18:58 hmmmm but the thing is the exception would need to be handled as a LuaError 18:58 puhfa oh... 18:58 puhfa it actually does 18:58 puhfa says here, snow 18:59 hmmmm hmm 19:00 kahrl #1875 seems obvious enough, should I merge it? 19:00 ShadowBot https://github.com/minetest/minetest/issues/1875 -- Dont corrupt stepheight when setting other properties by CiaranG 19:01 hmmmm kahrl, LGTM 19:03 kahrl done 20:48 Sokomine thanks for merging #1867 i'm sure it'll make a lot of players happy. those who care more about realism than easy steering still can leave the setting off. and if there's really a game one day that wants to enforce that players can't build where they stand, adding a server-controlled priv for that ought to be no problem 20:48 ShadowBot https://github.com/minetest/minetest/issues/1867 -- added enable_build_where_you_stand option by Sokomine 20:48 * Amaz is very happy! 23:46 ShadowNinja hmmmm: It doesn't have to be a LuaError, that exception class doesn't do anything special anymore. Anything derived from BaseException should send it's what() on as a Lua error if it's been called from Lua.