Minetest logo

IRC log for #minetest-dev, 2015-01-15

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

All times shown according to UTC.

Time Nick Message
00:07 cg72 joined #minetest-dev
00:07 cg72 is there any way to have a HTLM or web link in the main menu?
00:12 cg72 BUZZZZZZ!!!!!!!
00:44 crazyR joined #minetest-dev
01:14 TriBlade9 joined #minetest-dev
01:17 TriBlade9 I've written a basic Java UDP client that can sort of connect, at least get the first packets from the minetest server.
01:18 TriBlade9 The idea is to have a intermediary client runnin on the server machine, which forwards the data to a TCP socket, which can then be used by a web-based administration panel. (Since JavaScript can only use TCP sockets)
01:19 cg72 all i want is a button that brings up a webpage
01:20 cg72 in your broswer
01:22 TriBlade9 cg72, Yeah, first you need TCP sockets somehow
01:22 TriBlade9 Trying to use Java to read C++ is just a pain though ;_;
01:22 cg72 just to open a external broswer like firefox
01:23 acerspyro joined #minetest-dev
01:24 TriBlade9 cg72, Wait, what sort of thing are we talking about?
01:24 TriBlade9 I'm talking about an administration interface that lets you communicate/chat/run commands with the server
01:25 cg72 i want i html like link or button in the main menue of minetest to just send your default broswer to a web page
01:26 TriBlade9 Oh okay, nevermind then.
01:26 cg72 lol but a skins server would be nice if you could ;)
01:27 acerspyro We've talked about that already
01:28 cg72 i have a trick for that on my MineTime servers :)
01:28 TriBlade9 Meh
01:28 TriBlade9 I'm already over my head with this idea
01:30 TriBlade9 Oh cg72?
01:30 cg72 lol same here but my c++ is embeded not normal os stuff lol
01:30 cg72 yes TriBlade9
01:30 TriBlade9 What's your trick?
01:31 cg72 my webserver some php and a mod ( buggy atm but kinda works
01:33 TriBlade9 Hmm
01:33 TriBlade9 How'd you pull off the connection packets?
01:34 cg72 i use a prog already on most unix systems
01:34 cg72 its my middle man. not up to minetest security standards but meh
01:38 TriBlade9 Hmm
01:38 TriBlade9 I can't really understand the packet formate well cg72, I'm still trying to implement TOSERVER_INIT
01:45 Wayward_One joined #minetest-dev
01:55 TriBlade9 Anyone have any idea what "SER_FMT_VER_HIGHEST_READ" is in the TOSERVER_INIT section?
02:31 TriBlade9 Jeez, I'm getting to the point where I'd prefer Node.js for this
02:31 TriBlade9 Java uses signed bytes, which makes everything harder. -_-
02:33 acerspyro joined #minetest-dev
02:33 Player_2 joined #minetest-dev
02:47 TriBlade9 Implemented basic test connection in Node.jhs
02:47 TriBlade9 Still trying to get the player connection working
02:59 TriBlade9 Gute Lard, this is tricky.
03:05 Wayward_One joined #minetest-dev
03:15 cg72 left #minetest-dev
03:18 younishd joined #minetest-dev
03:19 TriBlade9 Woot, got the example uptime system implemented using Node.js
03:19 compunerd joined #minetest-dev
03:19 TriBlade9 Now gotta figure out how all the other messages work. ;_;
03:33 Zeno` joined #minetest-dev
03:34 TriBlade9 Starting to get the hang of this... Has packet builders now. Working connect/disconnect system... kinda.
03:35 Wayward_One joined #minetest-dev
03:59 younishd joined #minetest-dev
04:09 chchjesus joined #minetest-dev
04:28 selat joined #minetest-dev
05:01 prozacgod joined #minetest-dev
06:16 paramat joined #minetest-dev
06:30 paramat left #minetest-dev
06:40 Hunterz joined #minetest-dev
06:44 crazyR joined #minetest-dev
07:10 twenty319 joined #minetest-dev
07:26 Kalabasa joined #minetest-dev
07:30 ImQ009 joined #minetest-dev
07:33 leat joined #minetest-dev
08:18 twenty319 joined #minetest-dev
08:25 twenty319 joined #minetest-dev
08:32 twenty319 joined #minetest-dev
08:33 deltib joined #minetest-dev
08:37 kilbith joined #minetest-dev
09:00 cib0 joined #minetest-dev
09:14 NrzMT629253 joined #minetest-dev
09:14 NrzMT629253 Hi all
09:17 kilbith hi nerzhul
09:17 Kalabasa joined #minetest-dev
09:36 NrzMT629253 left #minetest-dev
09:37 NrzMT joined #minetest-dev
09:37 NrzMT ShadowNinja, i merge your improvements, that's great
09:42 twenty319 joined #minetest-dev
10:18 Amaz joined #minetest-dev
10:34 cib0 joined #minetest-dev
10:38 SudoAptGetPlay joined #minetest-dev
10:39 Amaz left #minetest-dev
11:04 gregorycu joined #minetest-dev
11:04 proller joined #minetest-dev
11:05 Zeno` joined #minetest-dev
11:07 T4im joined #minetest-dev
11:10 Zeno` #2116
11:10 ShadowBot https://github.com/minetest/minetest/issues/2116 -- Optimize vertex calculations (maths), perf improvement by nerzhul
11:11 Zeno` can those commits be squashed?
11:12 Zeno` but before that.... pos.y -= sin (posYbuf) * WATER_WAVE_HEIGHT + sin (posYbuf / 7.0) * WATER_WAVE_HEIGHT;  should be pos.y -= sin(posYbuf) * WATER_WAVE_HEIGHT + sin(posYbuf / 7.0) * WATER_WAVE_HEIGHT;
11:12 Zeno` there is no space after a function name and the opening parenthesis
11:13 Zeno` s/is/should be
11:13 gregorycu ...
11:13 gregorycu sin(posYbuf / 7.0)
11:13 gregorycu What does that even mean?
11:13 Zeno` looks like the wave height at that current position
11:14 gregorycu / 7.0 ?
11:14 gregorycu What is that?
11:14 Zeno` magic
11:14 Zeno` good point ;)
11:15 gregorycu Anytime I see rational numbers inside sin, I get concerned
11:15 Zeno` I don't think any division by 7 can be rational
11:15 Zeno` 7 is prime
11:15 gregorycu No, but 7 is a rational number
11:16 Zeno` hmm, yes
11:16 gregorycu I'm not being very clear
11:16 Zeno` neither am I
11:16 Zeno` but there needs to be a comment
11:16 gregorycu Also, dividing
11:17 gregorycu It just... I don't know
11:17 Zeno` yep, so a comment is needed
11:17 Zeno` there are too many magic values in minetest already
11:18 gregorycu At least divide it by some multiple of PI, that's all I'm saying :)
11:18 Zeno` oh,, PI
11:18 Zeno` even more magic :)
11:19 Zeno` what about phi?
11:19 gregorycu Nah, it's better
11:19 Zeno` I like phi
11:19 gregorycu sin(PI) == sin(0)
11:19 Zeno` phi fi foe fum
11:20 Zeno` https://github.com/minetest/minetest/pull/1977
11:20 Zeno` is that still not merged?
11:21 nrzkt hi Zero
11:21 nrzkt zeno*
11:21 Zeno` hi there
11:21 nrzkt sorry for to be late, i'll look at this
11:23 jin_xi joined #minetest-dev
11:23 gregorycu 20% of light related calculations is spent inserting into a std::map
11:24 gregorycu 20%
11:26 gregorycu Sorry, a set
11:30 kilbith Zeno`, 1977 has been merged but not closed yet : https://github.com/minetest/minetest/commi​t/800d912b6f93b4e9e6efdc89b9e7a720807f2b1a
11:32 Zeno` thanks
11:32 Zeno` closed
11:32 rubenwardy joined #minetest-dev
12:02 Kalabasa joined #minetest-dev
12:05 FR^2 joined #minetest-dev
12:12 MinetestForFun joined #minetest-dev
12:21 Zeno` hmm, I cannot apply #2116
12:21 ShadowBot https://github.com/minetest/minetest/issues/2116 -- Optimize vertex calculations (maths), perf improvement by nerzhul
12:24 Zeno` weird
12:32 blaze joined #minetest-dev
12:33 gregorycu joined #minetest-dev
12:33 gregorycu Zeno`: I'm going to investigate if a pooled_allocator can drastically speed up the places where we use std::set
12:40 Zeno` ok #2116 seems fine to me
12:40 ShadowBot https://github.com/minetest/minetest/issues/2116 -- Optimize vertex calculations (maths), perf improvement by nerzhul
12:40 Zeno` any objections to it being merged?
12:40 nrzkt ty :)
12:42 Zeno` will merge in 20 minutes if there are no objections
12:48 cib0 joined #minetest-dev
12:56 jin_xi hm. someone tell me if this looks right: a dump of some registered decorations, registered with no name: http://paste.ubuntu.com/9755619/
12:57 jin_xi documentation says they all should have their id as keys, but only [0] shows up. is this because of the way dump works or is there a discrepancy?
13:20 Amaz joined #minetest-dev
13:24 Zeno` will merge #2131 once travis is finished mucking around
13:24 ShadowBot https://github.com/minetest/minetest/issues/2131 -- Small tweaking (alignment - client tab) by kilbith
13:35 Taoki joined #minetest-dev
13:40 nrzkt what do you think about project structure ?
13:41 Zeno` that it sucks?
13:41 nrzkt all sources in one directory, except somes, it's not a proper way no ?
13:41 nrzkt why don't do a separation, client sources in one subdir, server sources in one subdir and common sources in one subdir
13:41 nrzkt and more, because in CMAKE we have separated compilations, one dir per compilation
14:00 proller joined #minetest-dev
14:00 impact joined #minetest-dev
14:01 shadowzone joined #minetest-dev
14:05 nrzkt #2119 rebase done, i tested it, it works perfect
14:05 ShadowBot https://github.com/minetest/minetest/issues/2119 -- Rewrite client/server packet handling by nerzhul
14:05 nrzkt please test it and merge
14:06 nrzkt if needed i can pull master to solve the conflict mentionned by github
14:06 Zeno` you didn't add the memcpy() improvement?
14:06 nrzkt hmm, forgot it
14:06 nrzkt lot's of comments :p
14:06 Zeno` heh, yes
14:06 nrzkt i add it, test it and merge it if it's valid
14:07 nrzkt before, i merge master to avoid conflicts
14:07 Zeno` memcpy is (on just about every CPU and OS) much, much faster than a loop
14:07 Zeno` ok, I will test
14:09 nrzkt github removed comments because commits disappear, can you give me the comment ? i'm not very familiar with memcpy
14:09 Zeno` wtf
14:09 Zeno` this does not automerge either :/
14:09 nrzkt yes, strange thing, i have the problem before
14:10 nrzkt wait i do the patch and rebase
14:10 nrzkt hmmm can't rebase on merge commit, but it's on the pr
14:11 Zeno` can someone look at why this doesn't merge?
14:12 Zeno` unless whitespace errors are the reason...
14:12 Zeno` hmm
14:24 nrzkt #2119 PR is now complete
14:24 ShadowBot https://github.com/minetest/minetest/issues/2119 -- Rewrite client/server packet handling by nerzhul
14:42 proller joined #minetest-dev
14:47 TriBlade9 joined #minetest-dev
15:06 hmmmm joined #minetest-dev
15:10 gregorycu Is there any reason why we don't use boost?
15:15 nrzkt http://dev.minetest.net/Cod​e_style_guidelines#Use_STL.2C_avoid_Irrlicht_containers.2C_and_no.2C_Boo​st_will_not_even_be_considered.2C_so_forget_it
15:20 gregorycu I assume it's because of this: "adding new dependencies is considered serious business."
15:20 Wayward_One I think there's a problem with this commit, because now when i enable bumpmapping, everything turns red, and the plantlike nodes look weirdhttps://github.com/minetest/minetest/commi​t/148fffb0f23fa437c67639ff3cc69177fb71d76a (#2116)
15:20 gregorycu Goodnight all
15:21 Wayward_One err, #2116
15:21 ShadowBot https://github.com/minetest/minetest/issues/2116 -- Optimize vertex calculations (maths), perf improvement by nerzhul
15:23 hmmmm boost isn't used becauser it's a mess
15:23 hmmmm look at this shit http://www.boost.org/doc/libs/1_55_0b1/li​bs/geometry/doc/html/geometry/design.html
15:24 hmmmm so if you want to triple the compile time while making everything impossible to read, go ahead
15:24 cib0 lolboost
15:26 hmmmm i think a better question is "what can we gain from boost"
15:26 Zeno` gain from boost?
15:26 Zeno` nothing but trouble
15:27 hmmmm Freeminer is a fork of minetest that adds a lot of crap, and yes they added boost too.  maybe you should work on freeminer instead if you think we need boost
15:31 Zeno` Wayward_One, reverted
15:33 Wayward_One ok, cool :)
15:35 nrzkt Wayward_One, that's strange.... somebody else can verify my calculs ?
15:37 nrzkt the only thing i see is color
15:38 T4im alphachannel was messed up too, thus the weird plants
15:38 nrzkt i haven't the problem in local... don't know where does this problem comes from
15:38 nrzkt left #minetest-dev
15:39 nrzkt joined #minetest-dev
15:39 kilbith i confirm : https://lut.im/XW54nky3/7BHE3Amc
15:39 nrzkt what's the hell...
15:39 crazyR_ joined #minetest-dev
15:40 nrzkt what is the difference here: color = 0.05*base.rgb + diffuse*base.rgb + 0.2*specular*base.rgb; => color = (0.05 + diffuse + 0.2 * specular) * base.rgb
15:40 nrzkt it's the only thing which can cause the problem
15:43 T4im hmm you miss a semicolon
15:43 T4im :)
15:43 nrzkt oh
15:43 nrzkt you are right
15:43 nrzkt ...
15:43 T4im two even
15:43 nrzkt yes it's a copy paste
15:43 nrzkt i fix it now
15:45 nrzkt sorry...
15:45 kilbith be very careful before stating a finished PR ;)
15:45 kilbith but to err is human
15:46 Zeno` To be fair, I should have caught that
15:46 nrzkt in fact my client doesn't seem to see this was missing... whereas i do a make clean before relaunch him
15:46 nrzkt travis cannot compile shaders ? it will be useful to track those problems
15:47 T4im well isn't that executed on the graphics card? maybe some gpu's can handle that
15:47 T4im others don't
15:47 nrzkt intel GPU like it xD
15:47 T4im ati failing here
15:47 T4im :D
15:50 T4im btw did you consider using mesa's glsl optimizer?
15:50 nrzkt me ?
15:50 T4im I'd expect it to catch something like that too
15:50 T4im https://github.com/aras-p/glsl-optimizer
15:51 nrzkt i somes can the the PR now, with semicolon fix and tell me if that's better because Intel GPU cheats ? :p
15:51 T4im well the gpu's do optimize their glsl execution.. of course they will do so differently well
15:52 T4im mesa's extracted one would be one way to make that independend, but of course that means that it will run slower on some gpu's faster on others than before
15:52 T4im iirc then this was more benefitial for mobile gpu's
15:54 T4im ah yes, they linked the blogarticle about the background http://aras-p.info/blog/2010/09/29/glsl-optimizer/
16:04 Calinou joined #minetest-dev
16:07 T4im nrzkt, Zeno`: nerzhul/math_improvements works for me now
16:10 nrzkt perfect ! sorry for the inconvenience with ';'
16:11 ImQ009 joined #minetest-dev
16:17 kaeza joined #minetest-dev
16:17 Hunterz joined #minetest-dev
16:30 Wayward_One joined #minetest-dev
16:37 ElectronLibre joined #minetest-dev
16:56 rubenwardy joined #minetest-dev
17:03 hmmmm hey, does anybody know why on android, blocks in sqlite need to be read before they are deleted?'
17:14 Krock joined #minetest-dev
17:14 roniz joined #minetest-dev
17:15 ShadowNinja hmmmm: So that they're deleted before they're written.  Existence is checked first IIRC.  Might be able to skip the check and ignore inexistence errors though.
17:16 PilzAdam joined #minetest-dev
17:16 hmmmm i was asking if the latter part of what you said is okay..
17:18 cib0 joined #minetest-dev
17:18 ShadowNinja I don't know, you'd have to look up docs or just try it.
17:18 ShadowNinja There might be a DELETE IF EXISTS statement.
17:22 acerspyro joined #minetest-dev
17:23 Krock #2095 is still present in current HEAD
17:23 ShadowBot https://github.com/minetest/minetest/issues/2095 -- Display (sub)game icons again by SmallJoker
17:29 leat joined #minetest-dev
17:30 johnnyjoy joined #minetest-dev
17:30 Krock looks like irrlicht does not like \\ in paths
17:31 Krock I replaced all double-backslashes and solved the icon display problem
17:31 Krock *replaced with normal slash
17:31 Krock talking about this issue: http://pastebin.com/iSp94uG3
17:33 kahrl I think there is a deeper issue that needs to be fixed
17:33 kahrl not just the symptoms hidden
17:34 kahrl everything in GUIFormSpecMenu unescapes strings way too often
17:34 kahrl (for example, try displaying a \ in a label)
17:35 kahrl why wasn't that a problem until recently though?
17:35 hmmmm LevelDB doesn't error on block load failure... is this behavior desired?
17:35 Krock no idea. I think there was a commit which broke the system
17:36 Krock I could say you my latest working build. using bisect is a horror on my system
17:38 kahrl actually, I think 3c3887bb19 is the problem
17:38 kahrl try reverting what it did to unescape_string
17:38 kahrl bbiab
17:39 Krock first bad build: 150108 latest good build: 150103
17:40 Krock if 3c388 is in between, then it's the one
17:44 Krock yeah. it's in between
17:46 Krock https://github.com/minetest/minetest/commit/3c38​87bb1#diff-2a505d2691feba087023af2ada625da5L465
17:47 acerspyro http://imgur.com/ http://i.imgur.com/BwvKHmU.png
17:47 acerspyro oops, first link failed
17:47 acerspyro http://imgur.com/E4pGYeJ
17:48 Krock acerspyro, more 3d-like signs?
17:49 acerspyro No
17:49 acerspyro Menu buttons
17:49 Krock we aren't minecraft devs
17:49 acerspyro I am aware
17:49 leat joined #minetest-dev
17:49 acerspyro But the current buttons are ass
17:49 Krock ShadowNinja, https://github.com/minetest/minetest/commit/3c38​87bb1#diff-2a505d2691feba087023af2ada625da5L465 is skins all backshashes
17:49 Krock *skips
17:49 sapier joined #minetest-dev
17:53 Krock updated #2095, untested but should work fine
17:53 ShadowBot https://github.com/minetest/minetest/issues/2095 -- Display (sub)game icons again by SmallJoker
17:55 Krock joined #minetest-dev
18:08 kaeza https://github.com/kaeza/minetest/commit/​e31096daeed880216a3762756f5cbe4c66c02f5a
18:08 kaeza <jordan4ibanez> This is so weird, why am I getting this error? 13:01:21: WARNING: Undeclared global variable "t" accessed at .../jordan/minetest/bin/../bui​ltin/common/serialize.lua:143
18:10 kaeza hmmmm, ShadowNinja ? ^
18:10 hmmmm looks good to me
18:11 hmmmm i for one love the strict module.  if only lua could have static typing now :(
18:15 kaeza test suites could be useful too
18:16 kaeza that code path is only taken if an unsupported type is found
18:16 kaeza (duh!)
18:17 hmmmm this seems to happen constantly with scripting languages
18:17 hmmmm undeclared variable gets used in an error case
18:17 hmmmm nobody knows about it until the error happens
18:22 nrzkt joined #minetest-dev
18:34 cib0 joined #minetest-dev
18:36 hmmmm question:  what the hell is the point of the MapBlock reference counting?  you're not referencing a mapblock without being envlocked, so I don't understand
18:39 sapier ShadowNinja: is there any comforable way to find a suitable icon in that android action bar iconpack?
18:40 Calinou can you screenshot where it would be used, sapier?
18:41 Calinou we can make them ourselves if neede
18:41 Calinou +d
18:41 sapier ShadowNinja suggested to use one of those icons for the autohiding settings bar
18:42 hmmmm also shouldn't MapSectors be deleted once they're empty?
18:42 sapier I made a button myself but that's just a spacer my skills at button design are quite limited. I'd prefere not to merge my spacers as last time I did this noone ever did create a bette icon ;-)
18:43 sapier btw can someone explain to me why we move down a ladder on sneak?
18:44 Calinou because it's the most obvious key
18:44 Calinou Shift by default
18:44 Calinou like in first person shooters
18:45 DFeniks joined #minetest-dev
18:45 sapier yes but there that button is used for crunch ... and there move down would be logic
18:58 sapier still any suggestions for the settings bar button?
19:01 hmmmm settings bar button?
19:01 hmmmm oh you mean android interface
19:02 shadowzone joined #minetest-dev
19:03 sapier yes I made my initial code more generic this way we could add multiple autohiding buttonbars too
19:04 sapier e.g. combine inventory, drop and settings in different bars
19:16 selat joined #minetest-dev
19:17 hmmmm oh man
19:17 hmmmm there's endless minetest crap to work on
19:17 hmmmm to improve the liquid flowing situation, I propose:
19:18 hmmmm - detecting mass amounts of liquid flow concentrated in a certain area and speeding it up with a voxelmanipulator
19:18 hmmmm - not using a UniqueQueue.  these are really slow data containers and it's rather hot
19:18 hmmmm what's wrong with a prereserved std::vector
19:20 hmmmm preallocated... reserved... whatever
19:29 sapier hmmm havent UniqueQueues been changed lately
19:29 sapier ?
19:30 Kodexky joined #minetest-dev
19:33 MinetestForFun joined #minetest-dev
19:42 est31 joined #minetest-dev
19:44 Kodexky Hi sapier. For the autojump thing. Would be acceptable to expose BS?
19:44 casimir joined #minetest-dev
19:45 sapier no
19:45 sapier BS is a internal value that might change
19:45 Kodexky ok
19:45 Calinou currently 10
19:45 Calinou save files store position in real units, not node units…
19:46 Calinou so changing BS would break at least this
19:46 sapier yes and it's actually more node size
19:46 sapier guess that's a historic bug
19:47 sapier but we shouldn't expose BS
19:47 Kodexky stepheight then?
19:48 sapier btw I once had a pull request making stepheight a server sent property I just did close it because someone wanted me to fixe some other issues in it ... and it was just a unsuccessfull try to fix another issue
19:50 sapier if someone wants to fix the remaining errors in there https://github.com/sapier/minetest/​tree/configurable_player_stepheight
19:51 Kodexky i'd like to try it.
19:54 sapier look for the pull request there are some comments telling what about this wasn't liked
19:55 Kodexky once i get back at my pc. i'll close my pr. Didn't expect it to be merged, just wanted to try anyway, sorry.
19:57 sapier You can leave it open too it's just my opinion maybe everyone else does have a different one
19:59 shadowzone left #minetest-dev
20:08 SudoAptGetPlay joined #minetest-dev
20:09 Nihao joined #minetest-dev
20:15 Nihao left #minetest-dev
20:15 Kodexky Ok, well then, i'll let it get more comments. I'll work with your commit meanwhile. Thanks. (I can swear i never saw it go by, got to stop working while exhausted.)
20:17 hmmmm humm
20:18 hmmmm m_database_delete was never finalized in android
20:18 hmmmm this was probably intentional, just interesting
20:18 sapier where?
20:18 hmmmm no you don't need to
20:18 hmmmm you only delete before reading, you never need a finalized delete statement
20:19 hmmmm so if all operations pass in theory it should work
20:19 sapier well I added a workaround for sqlite on android I did never fully understand why it fails.
20:19 hmmmm the only problem with this setup is
20:19 sapier as of api it should've worked
20:19 hmmmm what if the write fails
20:19 hmmmm but the delete does not
20:19 hmmmm i realize there's hardly any practical chance of that happening but it's a theoretical problem
20:20 hmmmm databases are supposed to be transactional in nature, i.e. ACID
20:20 sapier if there's any chance for it to fail we should handle it
20:21 hmmmm is it possible to combine delete and insert statements into one statement?
20:21 sapier not handling an error was exactly why the android sqlite error wasn't realized till about a day prior release
20:21 sapier I don't think we do it that way
20:21 hmmmm well it'd be 'handled' as in simply regenerated the next time the load returns nothing, you'd also get a save error
20:22 ElectronLibre left #minetest-dev
20:23 sapier that's what happened on android ;-)
20:23 hmmmm oh right
20:23 hmmmm that failure mode that appears at face to be not so problematic is actually horribly problematic
20:24 hmmmm but I don't think we ever did find the *root* cause of that fiasco
20:24 sapier yes on android it caused mud flowing to float whole map
20:24 sapier on android I did
20:24 hmmmm and the "fix" I made to voxelmanipulator to prevent that happening caused trees to get chopped off :(
20:25 sapier I just don't understand why sqlite behaves different there
20:25 hmmmm we need to work on map.cpp a lot.
20:25 sapier on android REPLACE doesn't work
20:25 hmmmm just because it existed for a long time doesn't mean it's perfect... a lot of those interfaces, frankly, suck
20:25 sapier possible
20:26 sapier yet leveldb can't even be compiled on mips ;-)
20:26 hmmmm =/
20:26 sapier ok it couldn't I haven't checked it recently
20:27 kilbith PilzAdam: waiting for your suggestions on game#400
20:34 ShadowBot joined #minetest-dev
20:40 ImQ009 joined #minetest-dev
20:44 Megal joined #minetest-dev
20:44 Wayward_One joined #minetest-dev
20:48 hmmmm lol
20:48 hmmmm you can't do x = y = 5 in Lua?
20:49 sapier I'd not bet on it
20:49 hmmmm oh noes :(
20:50 hmmmm well i guess this is correct behavior
20:50 hmmmm if you delete a block underneath a structure, it'll regenerate the whole chunk, but the structure will block direct sunlight, so dirt_with_grass won't be placed
20:52 sapier Any comments on prefixing android only commits by Android: ?
20:52 hmmmm i think it's helpful
20:53 hmmmm maybe it could be like (Android) component: Fix thingie here
20:54 sapier Hmm but "component" doesn't make sense if we don't do this on regular commits too
20:54 hmmmm I do
20:54 hmmmm the ones I don't do it on are more general commits affecting more than one component, usually
20:55 sapier I see, hmm I don't really like using brackets at the beginning, what about "Android, compoent:"
20:56 hmmmm i'm not picky
21:00 Wayward_One joined #minetest-dev
21:07 exio4 joined #minetest-dev
21:12 impact joined #minetest-dev
21:15 sapier nope leveldb still doesn't support mips
21:15 shadowzone joined #minetest-dev
21:21 ShadowNinja kaeza: serialize() fix pushed.
21:23 hmmmm https://github.com/kwolekr/minetest/commit​/99de5237c11cf81c72992351b69b758be68f42bc
21:24 ShadowNinja hmmmm: I fixed the finalizing bug in my SQLite3 PR.
21:25 ShadowNinja (And possibly a few other bugs that I don't remember)
21:25 hmmmm oh no, merge conflicts..
21:26 sapier shouldn't you return the retval of sector->deleteBlock in deleteBlock
21:26 sfan5 hmmmm: https://github.com/kwolekr/minetest/commit​/99de5237c11cf81c72992351b69b758be68f42bc#​diff-c6d551e431c1fb79f5cf714da9b6c604R416 the comment here is wrong
21:26 sfan5 s/comment/text/
21:26 Wayward_One joined #minetest-dev
21:27 hmmmm i thought deleteblock is void
21:27 hmmmm oh whoops @sfan5
21:27 sapier you did add it returning bool ;-)
21:27 hmmmm i mean sector::deleteblock
21:27 PilzAdam hmmmm, the mapblock design is usually hidden in the API
21:28 sapier hmm let me check :-)
21:28 PilzAdam i.e. mods don't / can't care in what format we store the map
21:28 hmmmm it's meant for chat commands
21:28 hmmmm gotta be able to repair messed up map
21:28 sfan5 hmmmm: minor nitpick: https://github.com/kwolekr/minetest/commit​/99de5237c11cf81c72992351b69b758be68f42bc#​diff-1305560bd8befb32862f0feeefabd02eR1883 other functions seem to use pos1 instead of p1
21:29 hmmmm ahh ok
21:29 hmmmm didn't notice that
21:33 Miner_48er joined #minetest-dev
21:42 hmmmm sfan5:  is that a good way to get a position to begin with?
21:42 hmmmm should the person fly into the mapblock they want to delete or aim at it
21:42 sfan5 i think the current approach makes the most sense
21:49 roniz joined #minetest-dev
21:49 hmmmm yeah, MapSector::deleteBlock() has a void return
21:49 hmmmm I don't see SN's commit in there that adds the delete finalize
21:50 hmmmm no more concerns I don't think
21:50 sapier sounds like a design error if database delete can fail and we loose that information till lua api ... but that's obviously not topic of this commit
21:51 hmmmm sapier: no, MapSector is not Map
21:51 hmmmm maybe I should've named my thing eraseBlock() instead so people don't confuse it for the MapSector method
21:52 hmmmm but nobody ever modifies Map so
21:52 sapier once you provide it ppl will use it
21:52 hmmmm well think of it this way
21:52 hmmmm a MapSector post-0.3.x is inherently all in memory
21:53 hmmmm whereas a Map is on disk and in memory
21:53 hmmmm so if you delete a block from a Map, what should you expect?
21:58 sapier I assume you know that code way better then me so if you think it's right this is most likely true
21:58 hmmmm i'm just trying to make up a justification
21:58 acerspyro joined #minetest-dev
21:58 hmmmm our naming scheme sucks
21:59 acerspyro naming scheme?
21:59 acerspyro Forniture mod?
21:59 hmmmm naming scheme.
21:59 sapier naming scheme?
22:03 ShadowNinja hmmmm: minetest.string_to_pos may be helpful for that command...
22:03 hmmmm possibly
22:03 hmmmm i was just following suit with the rest of the command there which use regex
22:03 hmmmm commands
22:04 hmmmm what do you think about rounding minetest.pos_to_string() to 1 decimal place by default?
22:04 ShadowNinja local pos1, pos2 = unpack(param:split(" ")) pos1, pos2 = minetest.string_to_pos(pos1), minetest.string_to_pos(pos2)
22:04 sapier I'd use at least 2
22:05 ShadowNinja hmmmm: No, you could add a rounding param though.
22:05 sapier ShadowNinja's suggestion is even better
22:05 hmmmm yeah i was just going to say that
22:06 hmmmm i figured nobody really wants to store like 500000000000000000000 decimal places that are nothing but floating point error anyway
22:06 ShadowNinja Conterting to/from strings should result in the same value (with a very slight difference to account for rounding errors and the like).
22:06 sapier but someone already complained about lua's precision isn't enough to keep exact positions ;-)
22:07 hmmmm lol
22:07 sapier it's relevant if you wanna pass positions via formspecs for example
22:08 ShadowNinja sapier: Did you tell them about libgmp?  ;-)
22:09 hmmmm lol, he was being sarcastic
22:09 hmmmm oh you had the ;-) too
22:16 hmmmm why is string_to_pos in builtin/game but pos_to_string is in builtin/common
22:19 cib0 joined #minetest-dev
22:29 hmmmm deactivateFarObjects: id=22 m_static_exists = true but static data doesn't actually exist in (18,0,-2)
22:29 hmmmm hrmm
22:29 hmmmm how did four objects get into those blocks?
22:30 sapier if you find the reason you're gonna fix a long mysterious issue
22:30 proller joined #minetest-dev
22:30 crazyR joined #minetest-dev
22:30 hmmmm well i know the reason
22:30 hmmmm i just caused it by deleting a range of mapblocks
22:31 sapier ah :-)
22:31 sapier that could be an issue
22:31 hmmmm so I assume that in the average case when you see this error, the mapblocks are loaded for sure
22:32 sapier if you delete mapblocks you'll always have entities beeing stored in there active
22:32 hmmmm 'cause that would be a rather obvious reason if you didn't cover that base already
22:32 sapier at least you always could have
22:33 hmmmm agh :(  thanks a lot ShadowNinja, I still need to use regexes to extract the positions
22:33 hmmmm I wonder if I'll break anything by making parentheses optional
22:34 hmmmm %(? %)?
22:34 ShadowNinja hmmmm: Why?
22:34 hmmmm erm wait a minute
22:34 hmmmm it already is
22:35 hmmmm this is not working for some reason..
22:35 TriBlade9 joined #minetest-dev
22:36 hmmmm ahhh... oops
22:37 hmmmm well then :) I guess it's not as easy as unpack(string.split(" "))
22:37 hmmmm you can have spaces in between the positions
22:37 Wayward_One joined #minetest-dev
22:38 hmmmm how do I split (4, 1, 4) (1, 2, 3)
22:39 hmmmm blah:split(") (") works but the parentheses are supposed to be optional.. besides, i have to readd the parentheses after the split
22:41 nrzkt good night all
22:41 nrzkt i need a final review on PR #2119
22:41 ShadowBot https://github.com/minetest/minetest/issues/2119 -- Rewrite client/server packet handling by nerzhul
23:06 loggingbot_ joined #minetest-dev
23:07 deltib joined #minetest-dev
23:07 Eivel joined #minetest-dev
23:09 shadowzone joined #minetest-dev
23:11 Brains joined #minetest-dev
23:11 HLuaBot joined #minetest-dev
23:11 Kray__ joined #minetest-dev
23:11 dzho joined #minetest-dev
23:11 7GHAAMDMO joined #minetest-dev
23:11 book`_ joined #minetest-dev
23:11 leat joined #minetest-dev
23:11 Topic for #minetest-dev is now Minetest core development and maintenance. Chit-chat goes to #minetest. Consider this instead of /msg celeron55. http://irc.minetest.ru/minetest-dev/ http://dev.minetest.net/
23:13 celeron55 joined #minetest-dev
23:13 VargaD joined #minetest-dev
23:13 daswort joined #minetest-dev
23:13 pitriss joined #minetest-dev
23:13 Taoki_1 joined #minetest-dev
23:13 JZTech101 joined #minetest-dev
23:13 OldCoder joined #minetest-dev
23:13 crazyR joined #minetest-dev
23:13 hintss joined #minetest-dev
23:13 zat joined #minetest-dev
23:13 CraigyDavi joined #minetest-dev
23:13 ShadowNinja joined #minetest-dev
23:13 pixelcrumbs|away joined #minetest-dev
23:13 disableclouds joined #minetest-dev
23:13 T4im joined #minetest-dev
23:13 n4x joined #minetest-dev
23:13 ottodachshund joined #minetest-dev
23:13 Ritchie joined #minetest-dev
23:13 ShadowBot joined #minetest-dev
23:13 Guest23455 joined #minetest-dev
23:13 Wayward_One joined #minetest-dev
23:13 acerspyro_ joined #minetest-dev
23:13 Du_Draig joined #minetest-dev
23:13 SmugLeaf joined #minetest-dev
23:13 Anchakor1 joined #minetest-dev
23:13 puhfa joined #minetest-dev
23:13 diemartin joined #minetest-dev
23:13 TriBlade9 joined #minetest-dev
23:13 roniz joined #minetest-dev
23:13 Megal joined #minetest-dev
23:13 MinetestForFun joined #minetest-dev
23:13 johnnyjoy joined #minetest-dev
23:13 kilbith joined #minetest-dev
23:13 monty joined #minetest-dev
23:13 Robby joined #minetest-dev
23:13 psedlak joined #minetest-dev
23:13 Sokomine joined #minetest-dev
23:13 Aaron1011 joined #minetest-dev
23:13 blaise joined #minetest-dev
23:13 eeew joined #minetest-dev
23:13 sfan5 joined #minetest-dev
23:13 AnotherBrick joined #minetest-dev
23:13 VanessaE joined #minetest-dev
23:13 chrisf joined #minetest-dev
23:13 Aaron1011_ joined #minetest-dev
23:13 peacock486 joined #minetest-dev
23:13 jin_xi joined #minetest-dev
23:14 hmmmm joined #minetest-dev
23:14 kahrl joined #minetest-dev
23:14 younishd joined #minetest-dev
23:14 shadowzone joined #minetest-dev
23:15 prozacgod joined #minetest-dev
23:15 shadowzone https://github.com/minetest/minetest/pull/2135
23:20 Ritchie joined #minetest-dev
23:24 est31 joined #minetest-dev
23:26 Wayward_One joined #minetest-dev
23:31 kahrl shadowzone: I think it gives the wrong kind of impression if a new player clicks on "Help" and the first thing (after the controls) is how to cheat
23:32 kahrl the concept is good though
23:33 shadowzone It's not cheating
23:33 kahrl so maybe talk about basic gameplay instead; OTOH the gameplay of voxel games might be common knowledge by now
23:33 kahrl giving yourself unlimited items is not cheating?
23:33 shadowzone In singleplayer it's fine
23:34 shadowzone And without the "give" privs they cannot do that.
23:34 kahrl it's fine but it's still cheating
23:35 crazyR joined #minetest-dev
23:38 kahrl perhaps the "cheaty" privs should be granted by default in creative mode singleplayer
23:38 hmmmm wow that network cleanup is huge
23:39 hmmmm soo, who wants to review it
23:40 kahrl *tumbleweed*
23:40 VanessaE haha
23:41 shadowzone I'm gonna sit here and be quiet as you review it.
23:42 cg72 joined #minetest-dev
23:44 hmmmm VanessaE, did you see /deleteblocks??
23:44 hmmmm merry christmas
23:44 VanessaE no, I haven't seen it.  been working on mesh stuff
23:44 VanessaE (see pipeworks and random_buildings/cottages)
23:45 hmmmm (=
23:45 hmmmm i am a little concerned about environment.cpp:1860
23:45 hmmmm emergeBlock with create_blank == false can possibly return null, and then it's dereferenced without checking a few lines later
23:45 hmmmm why haven't we seen any crashes yet
23:48 VanessaE what else would it do as opposed to crashing?
23:48 VanessaE I mean, theoretically
23:49 hmmmm explode
23:49 VanessaE heh
23:49 hmmmm and burn
23:49 hmmmm sometimes it crashes AND burns
23:52 exio4 joined #minetest-dev
23:53 hmmmm jesus christ some people are insane
23:54 shadowzone hmmmm, tell me 'bout it.
23:54 shadowzone on #minetest
23:54 hmmmm heh good point =]
23:55 cg72 hmmmm is a fucking cunt who will not see the truth nor is willing to be human towards anyone!  he evidently gets off on butchering the code of a 12yo girl instead of trying to point her in the right direction.
23:55 hmmmm ...
23:55 cg72 FUCK YOU AND FUCK ALL THE EGOTISTICAL DEVS HERE
23:55 hmmmm please take it to #minetest
23:55 cg72 ;)
23:55 cg72 EAT MY PUSSY YOU FUCKING DOUCHE
23:56 * shadowzone facepalms
23:56 VanessaE O_o
23:56 hmmmm do we have a landmine #2 here or something
23:56 hmmmm hahaha
23:56 cg72 Hmmmm your luck i have reasons not to come after yo uirl
23:56 hmmmm alright seriously, leave now
23:57 cg72 i have unlimited ips and nicks i will never leave
23:58 acerspyro lol
23:58 acerspyro wtf is going on
23:58 hmmmm preteen drama, evidently =]
23:58 hmmmm if we had a mailing list these problems wouldn't exist
23:59 acerspyro lol
23:59 acerspyro Mailing list rage

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