Minetest logo

IRC log for #minetest-dev, 2017-05-07

| Channels | #minetest-dev index | Today | | Google Search | Plaintext

All times shown according to UTC.

Time Nick Message
00:04 paramat fixer's bug confirmed, cannot walk through any 2-node high gap, not just doors
00:08 paramat yes setting collisionbox in MTG fixes the bug
00:09 paramat "for players (0, -1, 0) is ground level" so box has to be set as player:set_properties({collisionbox = {-0.3, -1.0, -0.3, 0.3, 0.75, 0.3},})
00:09 paramat that's the offset issue
00:09 paramat will commit to MTG to fix
00:10 VanessaE um
00:10 VanessaE shouldn't there be some default size set by the engine instead?
00:10 VanessaE sounds like this will break pretty much every subgame in existendce.
00:10 VanessaE existence*
00:14 paramat hmm
00:15 paramat yeah
00:16 paramat ok will try to fix in engine
00:26 paramat found a player box default in the engine which is slightly wrong, will edit
00:41 paramat works
01:04 paramat bugfix #5715 any support? if not will merge in 1-2hrs
01:04 ShadowBot https://github.com/minetest/minetest/issues/5715 -- Player properties: Set correct default collisionbox by paramat
01:04 paramat high priority
01:10 paramat hang on will update docs
01:29 paramat done
03:02 paramat now how about #5704 ? seems obvious to do but it's probably not trivial, maybe best i wait
03:02 ShadowBot https://github.com/minetest/minetest/issues/5704 -- Shaders: Remove unused water surface shader by paramat
03:08 benrob0329 sofar: i can get mt to crash by spamming esc + rightclick with new chests
03:10 benrob0329 debug: https://pastebin.com/6b5cvnqs
03:10 paramat please open an issue?
03:10 benrob0329 ok
03:12 benrob0329 game#1727
03:13 ShadowBot https://github.com/minetest​/minetest_game/issues/1727 -- Chests: Crash on Esc Spam
03:34 paramat i'll merge #5715 in an hour
03:34 ShadowBot https://github.com/minetest/minetest/issues/5715 -- Player properties: Set correct default collisionbox by paramat
03:38 paramat and #5714 (trivial)
03:38 ShadowBot https://github.com/minetest/minetest/issues/5714 -- Replace occurrence of luaL_reg in l_camera.cpp by t0ny2
03:40 paramat weird but good to see PR and issue numbers falling
03:42 Hijiri ground level = player position?
03:42 Hijiri I'm not sure it makes sense to call it ground level since the ground won't be there if an entity's collisionbox has a negative miny
03:46 paramat hmm ..
03:47 GreenDimond joined #minetest-dev
03:48 GreenDimond Found out you can leave a chest in it's open state if you press and hold 'ESC' and rightclick. Don't think it's a bug or anything but it's kinda funny.
03:48 paramat 'foot level'
03:48 paramat feet level
03:48 paramat yes that's been reported :]
03:48 GreenDimond Really? I thought he said it crashed...
03:49 paramat separate report
03:49 GreenDimond Oh? *looking...
03:49 paramat maybe the crash fix will fix both behaviours
03:49 paramat reported on IRC but doubt there's an issue
03:49 GreenDimond Aw man. I kinda like it. Works for decorative purposes :P
03:50 paramat ok i'll improve 5715 docs before merge
03:50 GreenDimond LOL and using that glitch you can make the chest appear as if it opened into the block above it
03:54 GreenDimond http://pasteboard.co/3g7yZK19Y.png :]
03:55 GreenDimond Also interesting http://pasteboard.co/3g8Omzapj.png
03:59 benrob0329 lol, that is to be expected
03:59 benrob0329 MC solves that by making chests slightly smaller
04:00 GreenDimond Expected or not, it's funny :P
04:03 GreenDimond Creates unintentional options for builders kindof http://pasteboard.co/3ghky1C91.png :P
04:06 GreenDimond "MC solves that by making chests slightly smaller" Look closely at the new chest and it *is* actually slightly smaller than the selection box...
04:07 sofar to prevent z fighting
04:07 sofar but mc's chest is a full 2/16 smaller
04:07 GreenDimond ah that makes sense
04:08 GreenDimond Yep
04:08 GreenDimond Also saves the chest state on exit which is nice :P
04:08 GreenDimond since it is a separate node i guess
04:10 GreenDimond If I change the black part of the chest texture to transparent would I see the inside of the outer texture?
04:14 sofar no
04:15 sofar you would see "through" the chest and end up looking at the node below it
04:17 GreenDimond sofar http://pasteboard.co/3gw8NUQHO.png :)
04:18 sofar huh, error in the texture of the rear part
04:18 sofar how did that happen
04:18 sofar looks interesting actually
04:19 sofar I forgot culling is off for the chest
04:19 sofar so yeah that works
04:19 GreenDimond however if I remove the wood block it does the z fighting ting
04:19 GreenDimond *thing
04:19 GreenDimond actually it might be y
04:19 sofar it's just called z fighting no matter the direction
04:19 GreenDimond ah oki
04:21 GreenDimond hrm. if you watch the chest as it opens, it doesnt change the back texture, it squishes it down...
04:21 GreenDimond none of the pixels actually disapear
04:22 sofar oh! that's a UV map issue then
04:22 sofar time to whip out blender
04:22 GreenDimond lol :)
04:22 GreenDimond The chest also shrinks xD
04:23 GreenDimond Well, if you look at it from the correct angle
04:23 sofar yes! it's a UV map bug!
04:23 sofar ding ding ding
04:23 sofar alright, I'll submit a PR to fix
04:23 * GreenDimond found his first bug in Minetest.
04:23 GreenDimond *feels good*
04:23 GreenDimond btw same for locked chest
04:25 GreenDimond oh is there an issue or PR for the problem with panes/bars not connecting to blocks as corners?
04:26 GreenDimond *some blocks
04:27 GreenDimond As in cobble, baked clay (from mod), etc..
04:28 sofar bbl in a bit
04:28 GreenDimond well nvm on the cobble
04:29 GreenDimond gonna go now cyas
04:44 paramat joined #minetest-dev
05:04 sofar ~tell GreenDimond fix pending for the UV map as well
05:04 ShadowBot sofar: O.K.
05:07 paramat will merge #5714 #5715 in a moment
05:07 ShadowBot https://github.com/minetest/minetest/issues/5714 -- Replace occurrence of luaL_reg in l_camera.cpp by t0ny2
05:07 ShadowBot https://github.com/minetest/minetest/issues/5715 -- Player properties: Set correct default collisionbox by paramat
05:08 benrob0329 I guess I rebuilt at a bad time then :P
05:11 GreenDimond joined #minetest-dev
05:11 GreenDimond aw :)
05:13 paramat merging
05:17 paramat done
05:34 GreenDimond joined #minetest-dev
05:35 GreenDimond *sigh* sofar, I want to test, but I can barely get my mods on GitHub. Could you inform me how to merge the PR to my fork? :/
05:35 sofar several ways
05:36 sofar I like this one:
05:36 GreenDimond easiest way?
05:36 GreenDimond or whichever :P
05:36 GreenDimond as long as you can tell me xD
05:36 sofar curl -I https://github.com/minetest/m​inetest_game/pull/1728.patch | git am
05:36 sofar run that in git/minetest_game
05:37 GreenDimond I guess I need curl
05:38 GreenDimond patch is empty..
05:38 GreenDimond must have done it wrong
05:38 GreenDimond wait a sec
05:38 GreenDimond I dont have a git/minetest_game
05:39 GreenDimond i have a minetest_game/.git
05:46 GreenDimond meh :/
05:46 GreenDimond sofar, n00b question, where is the git/minetest_game? I can't find it...
05:46 sofar go in to minetest_game
05:46 sofar that
05:46 sofar that's your minetest_game git tree
05:47 sofar that's what I meant
05:47 GreenDimond ah ic
05:47 GreenDimond meh. still says it is empty
05:47 GreenDimond "Patch is empty. Was it split wrong?"
05:48 GreenDimond I am in /home/user/minetest-daily/m​inetest/games/minetest_game btw
05:50 GreenDimond and now it says "(23) Failed writing body"
05:50 GreenDimond what am I doing wrong o_0
05:53 GreenDimond maybe I am in the wrong minetest_game, sofar?
05:53 sofar oh
05:53 sofar typo
05:54 sofar change -I to -L
05:54 GreenDimond Did something different but still failed
05:55 GreenDimond fatal: previous rebase directory .git/rebase-apply still exists but mbox given. blablabla stuff with % curl: (23) Failed writing body (461 != 1370)
05:55 Fritigern joined #minetest-dev
05:56 sofar git am --abort
05:56 sofar then retry
05:56 GreenDimond huh. worked.
05:57 GreenDimond sofar: UV mapping fixed. i can still get it to stay open though
05:58 GreenDimond takes some doing, but I can
05:58 sofar that was always intended
05:58 sofar fixing that is not that interesting
05:58 GreenDimond oh.
05:58 GreenDimond derp.
05:58 sofar I could care less
05:59 GreenDimond ^tru
05:59 sofar some players may like the visual, and it resets on server reboot properly
05:59 sofar I could work a little harder on player logout
05:59 GreenDimond the other part to test is spam esc and rightclick?
05:59 Hunterz joined #minetest-dev
05:59 sofar yes
06:00 GreenDimond by 'spam' he means press&hold right?
06:00 GreenDimond what I have been doing the past hour :P
06:01 GreenDimond sofar, it never crashed when I did that in the first place, so I guess the fix worked? xD I dunno
06:02 sofar no, just click it really fast
06:02 sofar you have to do it within 0.2 seconds
06:03 sofar two right clicks
06:03 sofar lol
06:03 GreenDimond I can get like 5 in
06:03 GreenDimond if I can do >1 clicks before the formspec shows up, methinks it worked
06:05 GreenDimond looks lood sofar. Im off now. gotta get some sleeps :)
06:05 GreenDimond *good
06:06 GreenDimond thats what happens when you type when falling asleep
06:06 Zeno` sofar, sofar, sofar
06:06 Zeno` it's "I could not care less", or "I couldn't care less"
06:06 Zeno` *sigh*
06:06 Zeno` :P
06:07 sofar I could care more
06:07 GreenDimond but everyone says "I could care less" :P
06:07 GreenDimond I could just not care at all xD
06:07 sofar I could never become a native english speaker
06:08 GreenDimond anywhos, g'nite.
06:08 Zeno` I could care less what everyone else thinks it should be
06:08 sofar hehe
06:08 sofar I could stop watching Pocoyo with my daughter
06:14 benrob0329 sofar: even native speakers say it, so your good :-)
06:46 Krock joined #minetest-dev
06:59 iZacZip_m joined #minetest-dev
07:17 burli joined #minetest-dev
07:27 nerzhul joined #minetest-dev
08:21 DS-minetest joined #minetest-dev
09:05 nerzhul pusher another LINT on l_util and l_settings.h to fix LINT
09:06 QwertyDragon joined #minetest-dev
09:08 nerzhul i think there is a detection rpoblem on PR or they are too old and mergeable and travis doesn't lint :(
09:17 lisac joined #minetest-dev
09:23 nerzhul Krock, Zeno` sofar can you look at #5716 ? It's a fairly patch permitting to solve #5709
09:23 ShadowBot https://github.com/minetest/minetest/issues/5716 -- Player attrs: permits to remove an attribute by setting value to nil by nerzhul
09:23 ShadowBot https://github.com/minetest/minetest/issues/5709 -- Remove player attribute
09:25 nerzhul fairly tiny*
09:30 Krock two comments from my side
09:30 nerzhul i see, which indent ?
09:30 Krock otherwise ok
09:30 nerzhul it's like set_inventory_formspec
09:30 Krock in both lines, see documentation above
09:31 nerzhul above there is value ssection
09:31 nerzhul but look at set_inventory_formspec just after, it's the same indent :)
09:31 Krock the description of  set_inventory_formspec  is indented with four spaces more
09:31 nerzhul oh i see tab vs spaces
09:32 Krock arguments/descriptions always are indented further
09:32 nerzhul it's now fixed :)
09:32 Krock good now.
09:32 nerzhul nice
09:32 nerzhul i will merge it in ~15min it's not a huge change :)
09:34 Krock ack
10:25 nerzhul merged, CI was slow
11:03 kilbith joined #minetest-dev
11:32 Fixer joined #minetest-dev
11:58 DS-minetest joined #minetest-dev
12:49 calculon joined #minetest-dev
13:05 Fixer collision bug still not fixed when connecting to multiplayer servers https://github.com/minetest/minetest/issues/5713
13:08 octacian joined #minetest-dev
13:08 octacian joined #minetest-dev
13:37 nerzhul Fixer, it's a server side fix, servers should update
13:38 Fixer oh boy
13:56 Krock2 joined #minetest-dev
13:59 Krock joined #minetest-dev
14:18 iZacZip_m joined #minetest-dev
14:45 Krock merging game#1726 in 15 mins
14:45 ShadowBot https://github.com/minetest​/minetest_game/issues/1726 -- Add option to disable lavacooling by Ezhh
15:10 Taoki joined #minetest-dev
15:33 red-001 #5719
15:33 ShadowBot https://github.com/minetest/minetest/issues/5719 -- Rework escape/pause menu by red-001
15:39 iZacZip_m joined #minetest-dev
16:00 nerzhul joined #minetest-dev
16:21 Karazhan joined #minetest-dev
16:25 Krock oops. totally forgot the pull. merging now
16:42 Fixer top kek: https://i.imgur.com/aQXXofz.png (some guy died and I had insane fps laaag and it crashed)
16:43 Krock eeks.
16:43 Krock this win 7 style ruins the w2k one
16:50 Fixer more info
16:51 Fixer that handsomepig blowed up around 99 TNT and I crashed at very same moment
16:52 Krock if you had an error dump + the debug database, it would be possible to track down the issue
16:52 Fixer unfortunately i don't
16:53 cx384 joined #minetest-dev
16:59 Fixer he plans to explode another megabomb, i'm ready with debug build
17:01 pandaro joined #minetest-dev
17:02 pandaro left #minetest-dev
17:07 nerzhul for the TNT problem i have a working thing in CPP, it's a because node update run on a single server step (thank you lua :p)
17:08 nerzhul i spread node updates in a queue in CPP on AppleTree and node update is spread on multiple step permitting to nuke 10000 TNT if you want :p
17:11 Fixer that explosion occured on multiplayer server, my client fps got to 0 because of entities (and I was in other part of map), i've seen his death messages and crashed at very same moment during lowest fps
17:12 Krock nerzhul, don't we have limited mapblock updates for that?
17:12 Fixer coincidence? don't think so (tm)
17:18 nerzhul Krock, if you run a while (please do all node updates) do (loop) end
17:18 nerzhul :)
17:18 nerzhul all is in example :)
17:59 Beerholder_ joined #minetest-dev
18:06 octacian joined #minetest-dev
19:01 torgdor joined #minetest-dev
19:09 behaleba- joined #minetest-dev
19:14 Amaz joined #minetest-dev
19:31 Taoki joined #minetest-dev
20:13 YuGiOhJCJ joined #minetest-dev
20:35 paramat joined #minetest-dev
20:35 paramat i think i can fix #5718 will try later
20:35 ShadowBot https://github.com/minetest/minetest/issues/5718 -- "Climbing" out of water seems randomly broken
20:36 paramat also since player collisionbox, step height and footstep bool are now taken from player properties i need to check the defaults and also set these in MTG
20:38 red-001 updated #5719
20:38 ShadowBot https://github.com/minetest/minetest/issues/5719 -- Rework escape/pause menu by red-001
20:40 red-001 has #5715 fixed not being able to climb stairs?
20:40 ShadowBot https://github.com/minetest/minetest/issues/5715 -- Player properties: Set correct default collisionbox by paramat
20:51 paramat not sure
20:51 paramat will test and try to fix later
20:52 paramat could be step height stuff as in https://github.com/minetest/minetest​/issues/5718#issuecomment-299728646
20:56 Beerholder_ Oh my what did I get myself into with this web browser in the Minetest core engine <:-((( Getting stuck on making it available in the C++ Lua API -_-
20:57 Beerholder_ The browser should only ever be started on the client as it spawns a Chrome web browser. Should never be started on the server
20:58 red-001 cef?
20:58 Beerholder_ But it should be scriptable on the server via a server mod, and that is where I get stuck. I can't register my functions as the server does not know anything about the chrome browser
20:58 Beerholder_ Yup, cef ....
20:59 rubenwardy joined #minetest-dev
20:59 red-001 add it to l_client.cpp/.h?
20:59 red-001 that's only used by the client
20:59 red-001 or add a new file e.g. l_browser.cpp
21:00 red-001 probaly better to add it to a new file
21:01 Beerholder_ Yup I have l_webpage.cpp/h, is better indeed. Problem is that if I do it client only, whenever I have a mod on the server (e.g. registering the node, doing a formspec where you can enter the URL and then change the web page shown on a node), I try to do local webpage = core:get_web_page("webpagename")
21:01 Beerholder_ That obviously returns nil -_-
21:03 Beerholder_ No actuall, the get_web_page method is not even found as I do not register it in the server_scripting
21:03 red-001 add a new packet?
21:03 red-001 to_client_goto_site
21:03 Beerholder_ Mmmmmh, ok ...
21:03 Beerholder_ It did cross my mind as to what was possible with packets ...
21:05 Beerholder_ In any case what is done on the server is different as to what is done on the client I guess ... So I will probably have two different l_browser.cpp/h implementations depending on client and server?
21:05 Beerholder_ As in e.g. l_browser_client and l_browser_server.cpp/h
21:06 rubenwardy abstraction would be better
21:06 rubenwardy ie: the lua part works exactly the same, but they call different code in the end
21:06 rubenwardy by lua part I mean l_browser
21:06 Beerholder_ You have a point there yes
21:06 Beerholder_ Just define interface and the implementation is different on server and client
21:07 Beerholder_ Then in the Lua part I can keep everything the same
21:09 red-001 you could use IGameDef for that
21:11 Beerholder_ Ah, ok, so there is a server and client implementation of IGameDef? I can have a look at that, or at least as an example
21:12 Beerholder_ It might still be cleaner to keep it a separate interface though so that I do not interfere too much with the core engine ...
21:13 Beerholder_ I already have some ifdefs in some places where I'd rather not have them XD
21:15 Beerholder_ Come to think of it, I might want to get rid of those by interfacing as well and doing a Null/ stub version of the browser
21:15 Beerholder_ #ifdef _CEF_ENABLED do the actual implementation, else do the dummy with methods that do nothing
21:16 Beerholder_ Then I could switch which source files to use in the makefile
21:16 Beerholder_ Rather than having ifdefs in the code
21:17 Beerholder_ Ok, I will give this some more thought
21:17 Beerholder_ Thanks for the pointers!
21:18 rubenwardy pointers would be helpful if you're doing polymorphism
21:19 paramat #5704 ? simple and obvious
21:19 ShadowBot https://github.com/minetest/minetest/issues/5704 -- Shaders: Remove unused water surface shader by paramat
21:24 paramat will merge #5704 in 3hrs if no objections
21:24 ShadowBot https://github.com/minetest/minetest/issues/5704 -- Shaders: Remove unused water surface shader by paramat
21:27 Beerholder_ void pointers, my favorite kind of pointers ;)
21:54 celeron55 what the...
21:54 celeron55 putting chrome inside minetest sounds like the definition of insanity
21:55 celeron55 but i guess whatever floats your boat
21:56 kilbith we need that in order to use e.g. Scratch inside Minetest to control a robot, in schools
21:56 kilbith that's what Microsoft made recently
21:57 celeron55 that sounds even more insane
21:58 kilbith no, that sounds one of the rare chance for Minetest to succeed commercially
21:58 celeron55 CEF is like 50MB, right?
22:00 celeron55 and controlled by google
22:01 kilbith doesn't seem so
22:01 kilbith https://en.wikipedia.org/wiki​/Chromium_Embedded_Framework
22:01 kilbith but anyway, its license is very permissive
22:04 celeron55 succeeding commercially doesn't need an embedded web browser, it needs someone to ask for money for something
22:04 celeron55 and it's not my goal anyway so it's moot point
22:05 kilbith asking for money requires to have something appealing to sell
22:05 kilbith it's not your goal, but there are commercial forks
22:05 kilbith it's always better when upstream cooperates
22:05 paramat a web browser in MT would very probably not be approved, maybe it will not be a PR though
22:06 celeron55 i'm assuming this is a fork of some kind, which is fine to me
22:06 celeron55 that's what i meant by "whatever floats your boat"
22:06 kilbith as I said, it's always better when upstream is willing to maintain that shit with us
22:06 kilbith (us = forkers)
22:15 Fixer_ joined #minetest-dev
22:16 celeron55 calling it shit sounds like it's some kind of crap nobody wants to touch unless they get paid a lot
22:16 Ritchie_ joined #minetest-dev
22:17 kilbith if it's the only way to put it down your throat, we are willing to pay
22:18 kilbith Beerholder is one of our collaborator btw
22:18 Beerholder_ CEF3 is BSD license
22:19 Beerholder_ And yes it is a fork
22:19 kaeza joined #minetest-dev
22:19 Beerholder_ Whether it will at some time be part of Minetest remains to be seen, but whatever I do will be available on github :)
22:19 celeron55 i'll figure out my opinion about what to do with it once i see the implementation; but if you REALLY want to get it upstream, modularity is key; don't spam existing code with ifdefs
22:20 celeron55 and official builds probably don't want to include it
22:20 Beerholder_ Of course I don't want to spam existing code with ifdefs and the CMake already has a flag to disable it
22:20 Fritigern_ joined #minetest-dev
22:21 rubenwardy plus modularity will make it easier to rebase on top of upstream changes too
22:21 Beerholder_ It started as a proof of concept, and there is some interest which made me pursue it more
22:22 Beerholder_ But in the end if I can make it to be designed in such a way that it fits design principles of Minetest that has the preference of course
22:22 Beerholder_ Hence me lurking around here :)
22:22 celeron55 in any case technically it's like putting a house on top of a car
22:22 celeron55 the other way would make more sense
22:23 celeron55 i don't know what kind of UX you're aiming for though
22:23 celeron55 a public plan would probably benefit everyone
22:27 celeron55 if you want to open scratch from minetest, and not minetest from scratch, i guess then you have no choice
22:28 rubenwardy I had this on my to do list
22:28 rubenwardy (which is very long)
22:28 rubenwardy well, ideas list
22:28 rubenwardy although I was thinking of making the program as a separate thing to launch and communicate with Minetest
22:29 rubenwardy would make implementation and UI much easier
22:29 celeron55 *if you want to open scratch inside minetest, and not minetest inside scratch, i guess then you have no choice
22:29 celeron55 cef sounds like popup windows are not allowed
22:30 celeron55 let's just wait for the plan to be publicized so we can just read it instead of asking random questions
22:35 celeron55 i should note that after 7 years of running minetest with basically no money involved, handling any kind of monetary compensation from a company for something being added and maintained upstream would probably be a disaster
22:36 celeron55 so your only option is to spend your money in making the code so god damn good that people want to pet it like a cat
22:39 nore just saying http://www.catb.org/~esr/jarg​on/html/Z/Zawinskis-Law.html :)
22:40 celeron55 web browsers have basically been operating systems since 10 years ago
22:40 VanessaE nore: already achieved. :)
22:40 Beerholder_ Good thing CEF runs in a separate process and the interfacing with it is minimal
22:41 Beerholder_ Which, btw, makes it replaceable with anything else
22:41 rubenwardy it's certainly a very interesting project
22:41 Beerholder_ Whether this will all happen in the end remains, of course, to be seen
22:42 Beerholder_ But like I said, this started as a PoC and I did not think much about it except for others to have a look at
22:42 rubenwardy and it's a shame that the engine isn't more modular on the C++ side, to allow C++ modules/plugins
22:42 Beerholder_ Seems there is interest, for educational purposes, so I am working on it still, for free by the way and all code available on github
22:42 fireglow[m] joined #minetest-dev
22:43 Beerholder_ But to be honest, the only trick is to just mem copy one buffer to a texture buffer ;)
22:44 VanessaE rubenwardy: the problem there is requiring users to compile mods
22:44 rubenwardy VanessaE, better than requiring users to recompile the entire engine
22:45 VanessaE minetest would need to integrate use of some C++ JIT compiler or something
22:45 rubenwardy and also makes it easier for patches to rebase if they're modular
22:45 rubenwardy ofc there's a limit to how modular you can make things
22:45 rubenwardy and you risk going into factory land
22:45 celeron55 in order to have a module, you have to have an interface that is never changed
22:46 celeron55 or rarely
22:47 celeron55 i'm not sure how hairy that kind of interfaces would get in MT internals; depends of course whether you want API or ABI stability
22:47 celeron55 stable ABI is probably practically impossible
22:48 rubenwardy yeah, it would get complex fast
22:48 celeron55 well... always is, just because it's C++
22:49 celeron55 a C++ API between any parts of minetest requires a lot of datatypes starting from stuff like MapNode, MapBlock, like a thousand irrlicht types, tens of weird constants and defines and whatnot
22:50 rubenwardy Bouncing ideas, but if you made most of the module communication through messages, rather than function calls, it would be a little easier. Although inefficient due to decoding
22:51 celeron55 then you can't do most things
22:51 rubenwardy Disclaimer: I've not done ABI before
22:51 rubenwardy nor am I saying Minetest should do this, it's quite late in the game
22:51 celeron55 and you end up needing awful amounts of serialization and whatever
22:51 rubenwardy exactly
22:51 rubenwardy ruins the point
23:02 MrIbby joined #minetest-dev
23:53 paramat joined #minetest-dev

| Channels | #minetest-dev index | Today | | Google Search | Plaintext