Minetest logo

IRC log for #minetest-dev, 2015-07-25

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

All times shown according to UTC.

Time Nick Message
00:09 est31 Firefox has a "chrome-parity" label for bugs, to show that the bug/feature request is fulfilled by chrome
00:09 est31 what do you think about a "minecraft-parity" label?
00:10 est31 meh, not such a good idea
00:11 est31 bad because becoming minecraft is not our goal
00:11 TBC_x well... imo minecraft can at least show what parts of minetest needs improvement
00:12 TBC_x in terms of gameplay
00:12 TBC_x but I would totally love spatial location in mt
00:13 TBC_x that would break 31000 world size limit
00:13 est31 wuts that
00:14 TBC_x spatial location is a relative position to the objects inside a scene
00:14 TBC_x IE objects don't know about their global position
00:15 TBC_x but rather relative to their cell they're bound to
00:16 TBC_x would allow to run each part of the world in its own thread
00:16 est31 wtf
00:16 est31 thats totally unrelated
00:16 est31 yes, it should be done
00:17 TBC_x and also multiworld support
00:17 TBC_x if anyone desires
00:17 TBC_x when an infinite world is not enough infinite
00:17 est31 we'll never get it infinite
00:18 TBC_x idc, the player must have such feel
00:18 est31 because of the same argument as computers aren't truly turing complete
00:18 TBC_x alright, the world must feel really infinite
00:21 TBC_x this would probably break some wireless mods however
00:21 est31 it would totally break modding
00:22 est31 and anything you want to do on the border
00:22 est31 want to grow a tree?
00:22 est31 want to run cables across?
00:22 est31 just want to let a mob walk over?
00:23 TBC_x the spatial location should not be partitioned against world space
00:23 est31 ?
00:24 TBC_x but rather have a thread claim a block as its origin and simulate everything relative to its position
00:26 est31 still you need some sort of handover
00:26 TBC_x mapblock conflicts could be solved by yielding that thread's woldspace into another
00:26 est31 imagine a player walking
00:26 est31 in one direction for a long time
00:27 est31 overlapping, that should work
00:27 est31 interesting idea, indeed
00:27 TBC_x yeah
00:28 TBC_x if we could deploy server-side multiprocess proxies, we could solve protocol incompatibility and in addition have any cool things spatial localion can bring
00:28 TBC_x such as seamless server transition
00:29 est31 I've thought about multiprocess myself already too, like this overlapping feature, but spatial indexing is a cool feature
00:29 est31 multiprocess and multiserver
00:29 TBC_x yes
00:30 TBC_x needs dynamic mod loading
00:30 est31 ?
00:30 TBC_x if the servers will have different mods
00:31 est31 meh, no
00:31 est31 it wont just work
00:31 est31 we declare some blocks to be HANDLED_BY_OTHER_SERVER
00:31 est31 then the mod just cant modify it
00:31 est31 players can walk over,
00:31 est31 and saos can be handled over too
00:31 TBC_x I made a draft of servers communicating to a process called mod manager that would map node and item IDs to have least conflicts, then sending needed files to clients
00:32 TBC_x if dynamic mod loading wouldn't work
00:32 TBC_x you get sum of all mods on connected servers
00:32 TBC_x at least the media and stuff
00:33 est31 media and stuff should be loaded dynamically, by the client
00:33 TBC_x needs server-side initialized caching too
00:33 est31 ?
00:34 TBC_x well... at least priorities
00:34 TBC_x so the near nodes from mods have higher priority
00:34 est31 ??
00:34 TBC_x nvm
00:36 TBC_x would be good to put a link somewhere to this discussion
00:36 TBC_x because it is imho the way to go
00:36 est31 make a wiki page
00:37 TBC_x forum account enough?
00:39 est31 no its separate account systems
00:40 est31 you need to ask an admin for an account
00:40 est31 dunno if some of them is here
00:41 TBC_x I have even put those Ideas on a paper
00:41 TBC_x except mention of the spatial location part
00:45 TBC_x well... with proxies, we could even plug in a MT 3.x client
00:45 TBC_x ridicilous
00:46 est31 proxies make even mt <-> mc communication possible
00:46 TBC_x yes
00:46 TBC_x just wanted to mention that
00:47 TBC_x if we achieve that point, we may even get minetest on wikipedia :)
00:48 est31 no
00:48 est31 wikipedia only thinks in sources
00:49 est31 and controversy
00:49 TBC_x exactly
00:49 est31 but for actual notability to exist, they need actual entries, best in new york times class articles
00:49 est31 and not just a mention, a whole article
00:49 est31 s/class articles/class newspapers/
00:50 TBC_x have you seen c55's buildat?
00:51 est31 yes
00:51 est31 but I had no closer look at it
00:52 TBC_x well... I really like mods in native code
00:53 TBC_x that enables to use any language the developers please
00:53 est31 what about webassembly
00:53 est31 it gives us compilers for free
00:53 est31 and people developing compilers
00:54 est31 it features a top sandbox
00:56 TBC_x I believe that webassembly is just a hack around poor design choice
00:56 TBC_x just as html5 is
00:58 est31 its something called "aged technology"
00:59 TBC_x it is called Hyper -->TEXT<-- Markup Language
01:00 est31 whats bad about text
01:00 TBC_x nothing
01:00 TBC_x binary blobs are bad in text
01:00 est31 so whats bad about html5 then, there are no binary blobs there?
01:01 TBC_x it moved closer to binary blobs
01:01 est31 html5 moved away from binary blobs
01:02 est31 before you required plugins to do lotsa stuff
01:02 TBC_x yeah
01:02 TBC_x that was an ugly hack too
01:03 est31 so whats the problem
01:03 est31 its less blobs now.
01:03 est31 and video and images have to be blobs, there is no way around that
01:04 TBC_x well... regular video allows you to get played with a player of your choice
01:05 TBC_x hmm
01:05 TBC_x Do you know Plan9?
01:06 est31 its an os no?
01:06 TBC_x yes
01:06 TBC_x very, very interesting
01:06 TBC_x every single resource is a file
01:07 TBC_x well... it's not that far
01:07 TBC_x but almost every single resource is a file
01:07 TBC_x VNC? you just export local display
01:08 TBC_x VPN? you just export local network interface over file server
01:10 TBC_x need to run an expensive operation? just connect to a server that has just the CPU power you need
01:12 TBC_x http://plan9.bell-labs.com/sys/doc/auth.html
01:15 est31 wtf what encoding is this
01:17 TBC_x not windows-1252
01:18 TBC_x ?xml version="1.0" encoding="utf-8"?
01:18 est31 either the html or the transport protocol has to specify the encoding
01:18 est31 have*
01:18 est31 interesting
01:19 est31 either a bug with firefox, or a bug with the page
01:19 est31 where is the w3 validator
01:19 TBC_x its a page bug
01:19 TBC_x probably
01:20 est31 lol
01:21 est31 even w3's page itself has a bug
01:23 est31 seems this is firefox fault
01:28 TBC_x it is in View->Text Encoding->Unicode
01:28 TBC_x in firefox
01:32 est31 writing <?xml seems to be invalid by the spec, but isn't recognized by the validator
01:33 TBC_x well.. you can change document encoding in firefox as you wish
01:33 TBC_x i had that menu hidden with now outdated pentadactyl
01:54 est31 ?
01:55 est31 what about five fingers?
01:55 est31 only know duodenum
03:15 BrandonReese joined #minetest-dev
03:55 est31 RealBadAngel, you can push it then.
03:55 est31 #2953
03:55 ShadowBot https://github.com/minetest/minetest/issues/2953 -- Remove drivers dropdown in the settings tab by kilbith
04:18 est31 hmmmm, what do you think, can we push the logging pr?
04:19 hmmmm what logging pr
04:19 est31 #2628
04:19 ShadowBot https://github.com/minetest/minetest/issues/2628 -- Clean up logging by ShadowNinja
04:19 hmmmm oh that one..
04:19 hmmmm i havent' looked at it in a while
04:19 hmmmm whatever happened to the threading PR?
04:20 hmmmm it needs like 2 or 3 more things fixed and another rebase and then it's good
04:21 hmmmm the logging PR has merge conflicts too anyway
04:23 est31 btw is it good practice to have #defines depend on something, but not include the dependency where they are defined, requiring the "user" to include them?
04:23 hmmmm i wouldn't really say it matters that much
04:23 est31 ok
04:52 RealBadAngel will 2953 disable dstream?
04:52 est31 no
04:52 est31 it will rename it
04:52 est31 to rawstream
04:54 RealBadAngel ok
04:54 RealBadAngel reading the comments i was afraid it was about to remove my fav debugging feature ;)
05:31 jhcole joined #minetest-dev
05:35 leat3 joined #minetest-dev
05:45 leat3 joined #minetest-dev
05:45 cornernote anyone know, whats the largest formspec that will work on all devices ?
05:46 est31 there is none specified
05:46 cornernote anyone used trial and error to find out ?
05:47 cornernote looking at you RBA
05:47 cornernote you did a fair bit of formspec in technic and inventory mods, and also others
05:49 cornernote 17,13 fits
05:49 cornernote but was reported on android phone to be offscreen
05:50 cornernote im going to go with 15,10
06:10 Hunterz joined #minetest-dev
06:30 nrzkt joined #minetest-dev
07:17 leat3 joined #minetest-dev
07:25 Krock joined #minetest-dev
07:33 RealBadAngel cornernote, UI got lately lite mode
07:33 RealBadAngel to fit android devices
07:44 kilbith joined #minetest-dev
07:46 nore joined #minetest-dev
07:57 leat3 joined #minetest-dev
07:59 nrzkt will merge #2953 in ~5min, it was approved by two core devs
07:59 ShadowBot https://github.com/minetest/minetest/issues/2953 -- Remove drivers dropdown in the settings tab by kilbith
08:00 Yepoleb_ joined #minetest-dev
08:35 Krock Too bad.. having a list of supported renders was a nice thing
09:02 FR^2 joined #minetest-dev
09:04 ElectronLibre joined #minetest-dev
09:26 FR^2 left #minetest-dev
09:41 jin_xi joined #minetest-dev
09:46 est31 stupid damn ncurses!
09:47 est31 src/chat.h:103:7: error: expected identifier before numeric constant
09:47 est31 void scroll(s32 rows);
09:47 est31 spot the error
09:47 est31 src/chat.h:103:7: error: expected ‘,’ or ‘...’ before numeric constant
09:47 est31 so? found it??
09:47 est31 https://www.mkssoftware.com/​docs/man3/curs_scroll.3.asp
09:47 est31 "Note that scrl() and scroll() may be macros."
09:47 est31 thanks, macros
09:49 nrzkt :p
09:52 kilbith nrzkt: can you reopen #2411 please ?
09:52 ShadowBot https://github.com/minetest/minetest/issues/2411 -- Colored chat for FreeType builds by nerzhul
09:52 kilbith has been accidentaly closed when you deleted the 0.5 branch
09:52 Miner_59 joined #minetest-dev
09:54 nrzkt i cannot because branch doesn't exists anymore
09:54 kilbith ah bad :(
09:54 VanessaE nrzkt: https://github.com/nerzhul/minetest/commit/6c​f922524deeddc0e0343c7015b2e9c5ae6e2a5f.patch
09:55 VanessaE you can restart it from that point (good luck rebasing it I guess, though :) )
09:55 nrzkt i can, but i'm working on another thing.
09:56 kilbith or maybe est31 since he's involved with the chat
09:57 est31 kilbith, no, sorry, I'm doing something else too
09:57 nrzkt est31: when you get time, can you review #2956 please. I merge particle spawner functions because they do same thing but we have two functions for changing only peer_id...
09:57 ShadowBot https://github.com/minetest/minetest/issues/2956 -- Cleanup server addparticle(spawner) by merge two identical functions. by nerzhul
10:08 casimir joined #minetest-dev
10:11 Calinou joined #minetest-dev
10:28 Amaz joined #minetest-dev
10:29 Miner_59 Hi, Issue #1551 was closed by hmmm but the PR to fix the issue is still open and was never added to source code. I tested and can confirm the bug is still there. You can jump on nodes which have disable_jump as a group when the block is a nodebox which is not a full block high.
10:29 ShadowBot https://github.com/minetest/minetest/issues/1551 -- disable_jump group doesn't work on nodeboxes
10:31 est31 reopened it Miner_59
10:31 nrzkt will merge #2956 in ~10 mins
10:31 ShadowBot https://github.com/minetest/minetest/issues/2956 -- Cleanup server addparticle(spawner) by merge two identical functions. by nerzhul
10:34 Miner_59 ok, thanks.
10:38 Miner_59 And PR #2241 by gregorycu fixes it for me too.
10:38 ShadowBot https://github.com/minetest/minetest/issues/2241 -- Fix for disable_jump group doesn't work on nodeboxes by gregorycu
10:41 kilbith est31, nrzkt : https://github.com/minetest/minetest/pull/2957
10:46 Krock #2948 comments?
10:46 ShadowBot https://github.com/minetest/minetest/issues/2948 -- Fix MSVC number conversion warning by SmallJoker
10:47 est31 Krock, is that the only conversion warning, or is it one of one million?
10:47 nrzkt pushed
10:48 Krock est31, an extreme long warning
10:48 Krock VC suggests to use conversion methods etc. however, this warning is about 15 lines long
10:48 est31 are there other such extreme long warnings?
10:48 Krock nope, it's the only one
10:48 est31 ok
11:09 blaze joined #minetest-dev
11:48 H-H-H joined #minetest-dev
12:53 Megaf joined #minetest-dev
13:22 kilbith_ joined #minetest-dev
14:09 leat3 joined #minetest-dev
14:21 TBC_x hi, is there any better solution to this? http://sprunge.us/Jjib
14:21 TBC_x whoops
14:21 TBC_x w/c
14:26 est31 looks like rust, doesnt it?
14:26 TBC_x yes
14:27 est31 I don't know any rust, but I like its goals.
14:27 est31 want to learn it at some point
14:29 est31 teaching hermatite the minetest protocol?
14:30 TBC_x well... I wanna try to link mt against connection.cpp rewritten to rust
14:30 TBC_x best way to learn imho
14:32 est31 I don't think that this will ever be merged, but development in this direction is good IMO. rust gives many guarantees c++ doesn't
14:33 blaise joined #minetest-dev
14:34 TBC_x yeah, I doubt it get merged
14:41 est31 kilbith_, wanna test something?
14:46 kilbith_ est31: yes ?
14:48 est31 I've written #2955 without testing it, and the issue reporter disappeared
14:48 ShadowBot https://github.com/minetest/minetest/issues/2955 -- Fix recipes getter regression by est31
14:48 est31 it should be easily testable, the reporter gives lua code
14:56 kilbith ok
14:56 kilbith meanwhile i have +2 on 2957
14:56 est31 ok will push i
14:56 est31 t
14:57 est31 later
15:01 hmmmm joined #minetest-dev
15:09 Taoki[mobile] joined #minetest-dev
15:10 Taoki[mobile] joined #minetest-dev
15:16 Ardonel joined #minetest-dev
15:21 kilbith est31: your patch works, will give a proof on the PR
15:29 est31 nice, thanks
15:29 est31 hmmmm, can you have a look at #2955 ?
15:29 ShadowBot https://github.com/minetest/minetest/issues/2955 -- Fix recipes getter regression by est31
15:30 hmmmm what's with all the static keywords
15:30 est31 well, its not put into the header
15:30 hmmmm yeah sure.  looks fine.
15:31 est31 ok, pushing kilbiths pr (has 2 +1) and mine (has 1 +1) then
15:35 kilbith and what about #2948
15:35 ShadowBot https://github.com/minetest/minetest/issues/2948 -- Fix MSVC number conversion warning by SmallJoker
15:37 est31 ok merging that too
15:40 kilbith you have a date for freezing in mind, est31 ?
15:42 est31 August 2.
15:42 kilbith ok sounds good
15:43 est31 Right now I'm sitting on an ncurses terminal, want to get it ready for 0.4.13
15:44 est31 Just hope I get it reviewed faster than areastore
15:44 est31 which should optimally also get into 0.4.13
15:45 VanessaE est31: speaking of, check out the current server stats page
15:45 VanessaE should be enough data there for you to tell if it's working as expected.
15:45 VanessaE (I think so, but you're the expert :) )
15:47 est31 I think even changes in the >10% range, which this is inside are hard to identify in CPU load stats
15:47 est31 CPU varies too much for that
15:48 est31 I have my benchmarks from singleplayer, there I can confirm the speedups
15:48 VanessaE ok
15:58 Miner_59 Can you try to add PRs #2886 and #2241 before the freeze? Use this mod to test: https://github.com/Jeija/minetest-mod-jumping , node for testing bouncing "jumping:trampoline5", and node with disable_jump "jumping:cushion"
15:58 ShadowBot https://github.com/minetest/minetest/issues/2886 -- Fix "bouncy" blocks by Miner59
15:58 ShadowBot https://github.com/minetest/minetest/issues/2241 -- Fix for disable_jump group doesn't work on nodeboxes by gregorycu
15:59 est31 2886 has the basic problem that it removes functionality, no?
15:59 est31 2241 seems ok though.
15:59 leat3 joined #minetest-dev
15:59 Miner_59 no, I will write in german channel ok
15:59 est31 ok
16:28 est31 Tesseract, https://github.com/minetest/minete​st/pull/2898#discussion_r35372179
16:28 est31 so you agree to the name, but only want the comment changed=
16:28 est31 ?
16:29 est31 I can change it to "areas"
16:29 est31 "area" is bad, because "lists" is in plural too
16:39 amadin joined #minetest-dev
16:39 amadin How to set bones walkable, "walkable = true," in "minetest.register_node("bones:bones", {" not help me?
16:42 Calinou they are already walkable by default. Moreover, walkable = true is the default!
16:42 Calinou ask such questions in #minetest by the way
16:43 blaise joined #minetest-dev
16:49 amadin in my mod they are not walkable
16:50 blaise joined #minetest-dev
17:28 Miner_59_ joined #minetest-dev
17:29 leat3 joined #minetest-dev
17:32 est31 hmmmm, miner59 argues you misunderstood #2886 and says that because you've tested with calinou's version.
17:32 ShadowBot https://github.com/minetest/minetest/issues/2886 -- Fix "bouncy" blocks by Miner59
17:33 est31 but somehow I doubt that
17:33 enesbil joined #minetest-dev
17:33 est31 because you seldomly test things, you read code :)
17:56 est31 man this ncurses thingy is fucking fun, I wonder why nobody before did it
17:56 est31 well you missed a hell of a time
18:16 hmmmm hrm
18:16 hmmmm cache_entry_t
18:16 hmmmm aren't identifiers with _t suffixes reserved?
18:17 hmmmm est31, if we want to be POSIX compliant you can't name your type cache_entry_t
18:17 est31 no problem
18:17 hmmmm http://pubs.opengroup.org/onlinepubs​/9699919799/functions/V2_chap02.html
18:17 est31 I was just copying some random stuff from the internet, and was happy as it compiled
18:18 est31 what bout CacheEntryT
18:18 hmmmm is the T really necessary
18:19 est31 its parameter type vodoo
18:19 est31 or how the <T> stuff is called
18:20 hmmmm that's a typename you defined on your own
18:20 est31 Its no real class, just a "helper type"
18:20 hmmmm oh, actually it looks like a typedef
18:20 hmmmm and it's defined inside of your own class
18:20 hmmmm so that'd be LRUCache::cache_entry_t fully qualified
18:21 est31 yes, its private.
18:22 est31 so, I strip the _t and make cache_entry and cache_type CamelCase, is it ok then?
18:22 hmmmm you can do whatever you want really
18:22 hmmmm i thought it was globally defined
18:26 hmmmm how is Area different from VoxelArea?
18:29 hmmmm oh i see, it has an ID as well
18:29 est31 it has also data
18:32 hmmmm unit tests are great
18:32 hmmmm they give a quick summary on how to use whatever interface being tested
18:48 Darcidride joined #minetest-dev
18:53 BrandonReese joined #minetest-dev
19:54 OldCoder joined #minetest-dev
20:12 Guest65346 joined #minetest-dev
20:16 luizrpgluiz joined #minetest-dev
20:20 luizrpgluiz at some future it will be possible to link the worlds minetest through portals, worlds I say is the folder words
20:24 luizrpgluiz would be really cool, even very demanding computer
20:25 luizrpgluiz -.
20:25 luizrpgluiz -:D
20:35 hmmmm i think tbc_x was interested in working on something similar to that
20:39 luizrpgluiz it would be great to servers with the worlds made with the default because it would be a problem trying to connect interconnected worlds with different subgames
20:43 TBC_x if you look up the log from today's morning
20:43 TBC_x well
20:44 TBC_x 19 hours ago
20:49 luizrpgluiz but it is difficult to program? program multi words?
20:55 luizrpgluiz because interconnect different subgames would much conflict
21:12 msantana joined #minetest-dev
21:22 msantana joined #minetest-dev
21:27 CraigyDavi joined #minetest-dev
22:19 est31 ok! the ncurses chat console works now more or less
22:19 est31 has a few bugs though
22:21 TBC_x hmm... I think that sender/receiver is bad place to manage connection
22:21 TBC_x sender/receiver threads
22:22 TBC_x too much shared state
22:24 luizrpgluiz ncurses? chat command to minetest? I thought I would not have it in the game
22:25 est31 luizrpgluiz, wanna try it?
22:25 est31 its alpha state
22:27 luizrpgluiz it would be very useful :D
22:28 luizrpgluiz so I could give commands out of the game without logging into the game and make the commands :D
22:29 luizrpgluiz despite being in alpha, but I wanted to try, but I hope it works on windows and linux
22:29 luizrpgluiz :)
22:29 est31 dont think it works on windows
22:29 est31 I'm not sure
22:30 ElectronLibre left #minetest-dev
22:30 GunshipPenguin joined #minetest-dev
22:31 luizrpgluiz est31:but what is the system that you programmed it?
22:31 est31 ncurses
22:31 GunshipPenguin left #minetest-dev
22:31 est31 its mostly a linux library
22:32 est31 it will probably work if you have mingw on windows
22:32 est31 dunno however if windows builds are with mingw, or bare, or whatever.
22:35 est31 https://github.com/est31/mine​test/commits/ncurses-console
22:36 est31 luizrpgluiz, ^ thats the code, try it, say what you find good, what you miss etc
22:40 kahrl for windows there's also pdcurses
22:42 luizrpgluiz est31: hehehe, I do not even know where I start, because I'm still learning the minetest mods language, however I never mechi with high programming languages
22:42 luizrpgluiz :D
22:42 kahrl I think you have to include curses.h instead of ncurses.h though
22:42 est31 dunno whether I use special curses features
22:42 luizrpgluiz ops mechi = using
22:43 est31 lemme see what happens if I strip the n
22:44 est31 wow, it built
22:44 est31 so porting to pdcurses shouldnt be a big deal
22:44 est31 I however cant do it due to lack of a windows box
22:44 est31 perhaps I should adjust the cmake variable USE_NCURSES and ENABLE_NCURSES
22:50 luizrpgluiz ?
22:51 TBC_x does the console connect to the server or is its integral part?
22:55 est31 TBC_x, its in its own thread, but within the server's process
22:56 est31 if you start in terminal mode, closing the terminal also ends the server
22:57 TBC_x would be cool if you could connect on demand
22:57 TBC_x but probably blacklist the network
23:09 est31 TBC_x, well, you can route it through screen
23:10 est31 you are the plan9 guy :)
23:10 TBC_x in plan9 you feed a file with commands
23:10 TBC_x or directly write to minetest/chat
23:11 est31 here you can run it on screen, and connect to it if you want
23:11 TBC_x so cat /dev/disk > minetest/chat may be interesting
23:11 TBC_x :)
23:11 est31 you can even connect multiple instances of screen
23:14 TBC_x I should stop mocking your efforts, do your best
23:33 est31 hmmmm, why does UASSERTEQ need an extra type?
23:42 est31 I mean thats horrible
23:43 est31 imagine if somebody changes the return type of a function, but forgets to change it for the uasserteq
23:44 est31 ofc its "just" unit tests, but rotting unittests are bad

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