Minetest logo

IRC log for #minetest-dev, 2014-10-01

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

All times shown according to UTC.

Time Nick Message
00:11 chchjesus joined #minetest-dev
00:20 luke joined #minetest-dev
00:37 exio4 joined #minetest-dev
00:57 shadowzone joined #minetest-dev
00:57 shadowzone joined #minetest-dev
01:00 exio41 joined #minetest-dev
02:12 ibloat joined #minetest-dev
02:12 ibloat joined #minetest-dev
02:22 Zeno` Can the new mouse behaviour be made an option?
02:29 Zeno` The change to node box highlight was an option and not as "severe" a change as this
02:29 Zeno` s/was/is
02:29 kahrl Zeno`: it will probably changed in the way SN mentioned
02:30 kahrl be changed*
02:30 kahrl so there won't be a need for the option
02:30 Zeno` oh ok
02:35 Miner_48er joined #minetest-dev
03:29 Atlas|1618033 joined #minetest-dev
03:32 Atlas|1618033 Looking for a good place to get started on documentation, both within the minetest code and maybe a few references to get me started.
03:46 VanessaE does anyone know whatever became of sapier?
03:48 Zeno` Well, in the books I read he was a god and sort of disappeared once the humans colonised the planet
03:48 VanessaE hah
03:49 Zeno` The series was called Terra Incognita
03:49 Atlas|1618033 Vacation maybe?
03:50 VanessaE that might be, didn't think of that
03:50 Zeno` here it is: http://books.google.com.au/books?id=hLlvEyNPx​1sC&pg=PT207&lpg=PT207&dq=terra+i​ncognita+series+sapier&source=bl&ots=​BAZ0ykjhjI&sig=jO7J9y34cmzKzPRG4qfAjnQY2i​E&hl=en&sa=X&ei=13krVKbjLJeeoQS4x​ICgCw&ved=0CDoQ6AEwBQ#v=onepage&q=ter​ra%20incognita%20series%20sapier&f=false
04:22 kaeza joined #minetest-dev
04:23 doomtron joined #minetest-dev
04:31 doomtron When waving water is selected, the water block, which is above another block(i.e. wood), doesnt get its texture loaded. how can I change it?
04:41 VanessaE it sounds like you're using one of RealBadAngel's glistening water test forks?
04:41 VanessaE normally the texture is still loaded, shaders or not
04:44 rickmcfarley joined #minetest-dev
04:46 proller joined #minetest-dev
04:54 VanessaE that is, normallyit's loaded if you're using mainline.  with his test code, it isn't (the shader provides some of the texture algorithmically and some of it from a loaded normalmap)
04:57 sol_invictus joined #minetest-dev
04:57 RealBadAngel hi
04:57 rickmcfarley hello
05:00 VanessaE speak of the dev...angel ;)
05:01 RealBadAngel hehe
05:01 RealBadAngel just call me and i come ;)
05:12 RealBadAngel https://github.com/minetest/minetest/pull/1685
05:14 RealBadAngel any constructive objections? i need that code be merged for my next patches
05:15 VanessaE I guess the highlight only shows the dig progress if cracks are disabled?
05:17 VanessaE the code seems okay but I'm no expert.
05:17 VanessaE I have no objection.
05:21 RealBadAngel not really
05:21 VanessaE RealBadAngel: one caveat:  "more than 16 steps for cracks? Longest one i saw is 10."  HDX has a 16-frame crack animation.
05:22 VanessaE (if I remember right)
05:22 RealBadAngel i made the cracks divided into 16 steps no matter the method used
05:22 VanessaE but that's just a note really.
05:22 RealBadAngel at least animation of it
05:22 RealBadAngel textures (frames) are scaled to fit
05:23 RealBadAngel highlighting is using it to fade away the halo
05:23 VanessaE right
05:23 VanessaE not a feature I'll use.
05:24 * VanessaE <-- luddite who prefers the crack animation :)
05:24 RealBadAngel ofc you will
05:24 RealBadAngel ;_
05:24 RealBadAngel erm i meant ;)
05:25 RealBadAngel but 16 is ok
05:25 RealBadAngel thats the max ive chosen
05:25 RealBadAngel and theres really no need for longer crack animations
05:26 RealBadAngel haven has 10 frames, game 5
05:27 RealBadAngel also when we are on the edge of flexibility and speed i will always choose hardcoded speed
05:29 doomtron VanessaE: i'm using default git version
05:30 RealBadAngel the area im working on with slight changes can efect with serious fps gains
05:30 Taoki[mobile] joined #minetest-dev
05:31 RealBadAngel lately i was comparing builds from 3-4 months away (were extracting old patches) to the current stage
05:31 RealBadAngel occasion was good, since now im on galium drivers
05:31 RealBadAngel gain is huge
05:32 RealBadAngel 10-15 fps before (pre caching era) compared to now
05:33 RealBadAngel now i get 45-60 on galium
05:38 Hunterz joined #minetest-dev
05:40 doomtron https://mediacru.sh/9Ctwwh0H7d4y
05:40 VanessaE oh THAT?
05:40 VanessaE that's an old glitch
05:40 VanessaE olllllllllld
05:40 doomtron hmm
05:41 VanessaE btw, protip:  /time 12000  before you take a screenshot :)
05:41 doomtron x)
05:41 VanessaE RealBadAngel: he talks about the top surface of the water under that log is missing, because the surface of the water has been lowered and those faces are being culled.  I guess you planned to fix that soon?
05:42 RealBadAngel yes
05:42 VanessaE ok
05:42 VanessaE doomtron: sufficient? :)
05:42 RealBadAngel im gonna move all the vertices animations to mapblock_mesh(animate)
05:42 RealBadAngel so, out of the shaders
05:43 RealBadAngel and drop this silly water jelly animation
05:43 doomtron more or less
05:43 sol_invictus joined #minetest-dev
05:43 RealBadAngel water surface shaders will take the place
05:44 doomtron I would like to do it myself, just to get into the code. Could you tell me more about that?
05:45 RealBadAngel do what?
05:45 doomtron patch minetest
05:46 RealBadAngel oh boy, you would get lost in the jungle of my branches ;)
05:46 doomtron you a real bad angel
05:46 RealBadAngel ;)
05:47 VanessaE bahaha
05:53 kaeza joined #minetest-dev
06:34 Miner_48er joined #minetest-dev
07:01 NakedFury joined #minetest-dev
07:30 ImQ009 joined #minetest-dev
07:40 Zeno` joined #minetest-dev
08:03 rickmcfarley joined #minetest-dev
08:05 Anchakor joined #minetest-dev
08:22 RealBadAngel https://github.com/minetest/minetest/pull/1685
08:22 RealBadAngel mergeing this in an hour
08:22 RealBadAngel this is an blocker for me for further work
08:40 NakedFury joined #minetest-dev
08:50 RealBadAngel anyway, i am requesting to remove c55 from AV head
08:50 RealBadAngel hes doing nothing
08:50 RealBadAngel hes just a blocker
08:51 RealBadAngel either that or i stop contributing
08:53 iqualfragile joined #minetest-dev
08:55 mberends joined #minetest-dev
08:58 alexxs joined #minetest-dev
09:00 ImQ009 joined #minetest-dev
09:04 Amaz joined #minetest-dev
09:23 ImQ009 joined #minetest-dev
09:26 hax404_ joined #minetest-dev
09:27 psedlak joined #minetest-dev
09:34 mberends joined #minetest-dev
09:38 hax404_ joined #minetest-dev
09:40 casimir joined #minetest-dev
09:42 jin_xi joined #minetest-dev
09:46 proller joined #minetest-dev
09:48 RealBadAngel c55 being head of AV is no go
09:49 RealBadAngel hes not a born coder
09:51 RealBadAngel since quite a while im one and only to know whats up in meshes
09:51 RealBadAngel i do have circa 10 branches open with pulls rdy
09:52 jin_xi Hey RealBadAngel there is something with shaders:
09:52 RealBadAngel this is ridiculous since im working on a few files
09:52 RealBadAngel jin_xi, yes?
09:52 jin_xi night/day cycle used to be smooth over all the blocks
09:52 jin_xi now its block per block again
09:52 jin_xi can this be avoided?
09:52 RealBadAngel never was
09:53 jin_xi sure
09:53 jin_xi ?
09:53 RealBadAngel it is updated when time allows
09:53 RealBadAngel more to do, easier to spot
09:54 RealBadAngel can be done faster tho
09:54 PilzAdam joined #minetest-dev
09:54 RealBadAngel and im on it
09:55 RealBadAngel Hi PilzAdam
09:55 RealBadAngel PilzAdam, when you will start to contribute again?
09:55 PilzAdam hi RealBadAngel, have you seen my message in #minetest?
09:56 RealBadAngel no
09:56 RealBadAngel can u copy it?
09:56 PilzAdam [16:21:22] <PilzAdam> RealBadAngel, why isn't there a smooth day-night-cycle with shaders anymore?
09:58 RealBadAngel im not aware of any glitches
09:58 RealBadAngel can you reproduce the issue?
09:59 jin_xi set time to dawn, look in one direction, then look around in other directions when it got a little darker
10:00 Zeno` actually I have a few issues with shaders :(
10:00 RealBadAngel so go on
10:00 Zeno` if I stand at the bottom of a cliff, for example, it looks like night
10:00 PilzAdam RealBadAngel, basically mapblocks seem to have fixed light levels at sunrise/sunset instead of a smooth transition
10:01 PilzAdam also if you turn arround after sunset (like jin_xi said), you see bright mapblocks for a second that then update their light level
10:01 PilzAdam it's like the old non-shader behavior
10:01 Zeno` but then "flicks" back to day
10:02 proller joined #minetest-dev
10:05 PilzAdam RealBadAngel, http://i.imgur.com/bjei5Bz.png this is how it looks if mapblocks slowly update their light at night
10:06 RealBadAngel PilzAdam, and you blame me for that?
10:07 RealBadAngel engine sucks balls not me :P
10:07 PilzAdam blame? I just asked if you know what caused this regression and possibly fix it
10:07 RealBadAngel i know
10:08 RealBadAngel but every piece of code im posting is put away
10:08 RealBadAngel im tired of full mouth contributors  that do nothing  but stop anothers
10:09 RealBadAngel PilzAdam, do you get it?
10:10 PilzAdam RealBadAngel, I guess the problem with most of your changes is that it doesn't really fit into the goals of other core devs
10:11 RealBadAngel PilzAdam, problem is that most want flexibility and i want speed
10:12 RealBadAngel we are doing game
10:12 RealBadAngel we want it to be fast
10:12 Zeno` I want speed
10:13 VanessaE in a game, speed is more important than flexibility
10:13 RealBadAngel not crawling around to be meeting everybodys wishes
10:13 VanessaE particularly when we already have enough flexibility as it is (generally speaking)
10:13 Zeno` I'd much prefer speed over some fancy abstracted (and let's face it, the src code is not that tidy) "perfect" implementation
10:14 PilzAdam well, I can only speak for myself and I don't care about the "fancy visuals" like bumpmaps and parallax occlusion and what not; it's nice to have, but your additions tend to break previously working things
10:14 RealBadAngel PilzAdam, my changes tend to work previous things faster
10:15 Zeno` come on... I get 20fps on minetest and 200fps on skyrim with max settings... something seems wrong there
10:15 RealBadAngel damn faster
10:15 proller anything wrong
10:15 RealBadAngel Zeno`, ofc faulty meshes handling
10:16 proller not only meshes
10:16 RealBadAngel proller, im on meshes only
10:17 proller but anything around wrong too
10:17 RealBadAngel i can say whats wrong here
10:17 proller i'm too, but impossible to fix all problems
10:17 RealBadAngel ofc not
10:17 proller faster write from scratch
10:17 RealBadAngel lol no
10:18 proller or start from hexa...
10:18 RealBadAngel easier to have  a base
10:18 jin_xi poor proller
10:18 proller why?
10:18 jin_xi forked the wrong shit
10:19 RealBadAngel lol
10:19 RealBadAngel the shit was right
10:19 RealBadAngel approach was wrong
10:19 proller wrong but it playable
10:19 Zeno` well the other thing affects speed is having node ids and meta in a class
10:20 Zeno` they should be separate arrays so that data proximity is preserved
10:20 Zeno` there are cache misses all over the place and this is even before things get to the rendering stage
10:20 jin_xi look, maybe we can be more productive than stating all is wrong. like with saying what is wrong in each case and musing whether there is a way to make it better
10:20 RealBadAngel Zeno`, im moving all the things to startup only
10:21 RealBadAngel now im about to cache facedir and nodebox rotations
10:22 Zeno` it's not impossible to fix, jin_xi
10:22 RealBadAngel propably this way i will exclude the on the fly modifications
10:23 proller first wrong thing is architecture
10:23 RealBadAngel but i will gain serious speed up
10:23 jin_xi Zeno`: thats my point
10:23 Zeno` if a node ID is part of an object that is, say, 400 bytes, then the next id is 400 bytes away (at least)
10:23 proller mt is singleplayer 1.1 threaded game
10:23 Zeno` these can be kept so they're always 1 byte away
10:24 proller with lua tumor
10:24 Zeno` Kind of like how int a[3][3][3] is faster (in general) if you split it into three single dimensional arrays
10:25 jin_xi well, i think its no question if lua is gonna stay in mt proller
10:25 RealBadAngel cool
10:25 PilzAdam bisected the day-night transition: bf91d623c8fc1fede7b600d1b0f4be1bc6b37c44 is the first bad commit
10:25 PilzAdam "Let lighting be done only CPU side. Remove finalColorBlend implementation from shaders." by RealBadAngel
10:25 RealBadAngel abstraction layer is being discussed again
10:26 Zeno` take voxelmanip as an example. On my i7 it takes 60-70ms just to populate the Lua stack. This should me more like < 1ms
10:26 Zeno` be*
10:26 proller voxelmanip is another shit
10:26 RealBadAngel PilzAdam, whats wrong with this?
10:26 PilzAdam it makes a non-smooth day night transition
10:26 Zeno` and that's because the data layout is wrong. simple thing to fix
10:26 RealBadAngel ah oh
10:26 jin_xi RealBadAngel: shaders cant make smooth day night transition before cpu can catch up
10:27 PilzAdam a smooth day-night-transition was one of the reason we added shaders in the first place
10:27 RealBadAngel you have lived with that code for years
10:27 RealBadAngel and now you blame me?
10:27 RealBadAngel read what the pull does
10:27 RealBadAngel and stfu
10:28 RealBadAngel PilzAdam, youre notin more than full mouth
10:28 RealBadAngel sorry to say that, but thats true
10:28 RealBadAngel fix the carts instead
10:30 proller jin_xi, maybe js? slow shit too, but designed for peoples
10:30 RealBadAngel PilzAdam, and the shaders was the thing you seem to hate the most
10:30 RealBadAngel so double stfu
10:31 RealBadAngel im pretty tired of havin code ready but BEGGING for it to be merged
10:32 RealBadAngel it is my hobby to code for mt
10:32 RealBadAngel but begging is not
10:33 RealBadAngel either something is gonna change or i will just simply fork
10:33 proller ^ it was one of reason for forking
10:33 RealBadAngel and leave
10:34 proller join us ;)
10:34 RealBadAngel no, not you
10:34 proller you almost ready
10:35 RealBadAngel dunno
10:36 PilzAdam <RealBadAngel> you have lived with that code for years <- what? it's 2 months old and I haven't played Minetest since then
10:36 PilzAdam or are you saying that removing the smooth day-night-transition was intentional?
10:37 RealBadAngel PilzAdam, im saying i havent touched that
10:37 PilzAdam you did in bf91d623c8fc1fede7b600d1b0f4be1bc6b37c44
10:37 RealBadAngel i can now say i know where that code is
10:38 edlothiol joined #minetest-dev
10:39 RealBadAngel PilzAdam, please fill the issue
10:39 RealBadAngel with all the details
10:39 PilzAdam "fill the issue"?
10:39 RealBadAngel atm i dont know what you are talking about
10:40 PilzAdam ah, ok
10:40 PilzAdam before shaders were added there was a non-smooth day-night-cycle; basically some steps of fixed light values of mapblocks
10:41 RealBadAngel nothing changed
10:41 VanessaE PilzAdam: I am pretty sure that the mapblocks have always done that
10:41 RealBadAngel thats still here
10:41 PilzAdam then shaders were added and the day-night-transition code was moved to them, so the day-night-transition was smooth instead of having fixed light steps
10:42 PilzAdam and in bf91d623c8fc1fede7b600d1b0f4be1bc6b37c44 you now removed the code from the shaders, so the transition is non-smooth again
10:42 RealBadAngel theyre smooth enugh
10:42 Amaz
10:42 PilzAdam [12:05:29] <PilzAdam> RealBadAngel, http://i.imgur.com/bjei5Bz.png this is how it looks if mapblocks slowly update their light at night
10:43 PilzAdam ^ this is not smooth enough
10:43 RealBadAngel but no point to feel sorry bout loosing that piece of code
10:43 RealBadAngel we are at the doorstep to have real lights
10:43 Amaz :D
10:44 RealBadAngel im learning irrlicht all the time
10:44 RealBadAngel we were doing many things faulty
10:44 PilzAdam RealBadAngel, you talked about speed earlier; it is actually faster to execute the code in the shaders, since you don't have to rebuild all meshes at sunset/sunrise
10:44 RealBadAngel including lights
10:45 PilzAdam and rebuilding the meshes is slow and has visual glitches (as you can see in the screenshot linked above)
10:45 RealBadAngel PilzAdam, rebuilding meshes is the point of getting them ready
10:45 RealBadAngel at startup
10:46 BlockMen joined #minetest-dev
10:46 RealBadAngel ie grab the mesh, apply rotations, calculate bounding boxes and store them
10:47 RealBadAngel on mapblock mesh refresh just fetch the correct mesh
10:49 RealBadAngel and what you can see on screenshot is lag, not light code
10:49 RealBadAngel and sucking balls multi threading
10:50 jin_xi RealBadAngel: the point is that shaders can avoid this issue with a little code duplication, so why not keep doing it
10:50 BlockMen 10:43 <PilzAdam> ^ this is not smooth enough <-- /me agrees
10:51 PilzAdam https://github.com/minetest/minetest/issues/1705
10:51 Zeno` I don't think that screenshot has anything to do with RBA's code
10:51 PilzAdam I suggest to mark this as blocker since it's a regression to previous stable release
10:51 RealBadAngel what about folks that cannot use shaders?
10:52 jin_xi they are stuck with the old way
10:52 PilzAdam RealBadAngel, they always have non-smooth transition anyway
10:52 RealBadAngel also, i dont like having sticker on my forehead beacause im just around :P
10:53 RealBadAngel jin_xi, shaders are not a solution to the faulty lighting code
10:54 RealBadAngel purpose of the GPU code is to make gfx shine
10:54 jin_xi but there is no need to rub it in your face
10:54 RealBadAngel not to fix cpu faults
10:55 RealBadAngel PilzAdam, and this code will never return there
10:55 RealBadAngel no way for such nasty workarounds
10:56 BlockMen so better keep shit than having workarounds?
10:56 jin_xi huh? is it so nasty?
10:56 Amaz Just one question, will that be fixed by different code at some point?
10:56 PilzAdam RealBadAngel, seriously, wtf are you talking?
10:56 Amaz s/Just one question, will/Will
10:57 PilzAdam light calculation is faster and looks better when done in GPU
11:02 RealBadAngel PilzAdam, what have u smoked today?
11:03 RealBadAngel since when light it gpu side here?
11:04 RealBadAngel never was
11:04 RealBadAngel and the shader code was never close to it
11:04 jin_xi its just about the final color blend, nothing more
11:05 RealBadAngel ofc
11:05 RealBadAngel but no matter where
11:05 RealBadAngel cpu side is faster
11:06 RealBadAngel putting it into shaders was like shooting own feet
11:06 Zeno` lighting is faster on CPU?
11:06 RealBadAngel damn
11:06 RealBadAngel theres no lighting
11:07 RealBadAngel we do have just dumb code
11:07 RealBadAngel stop calling it lighting
11:07 Megaf I think light and shadow calculations should be moved from server to  client
11:07 Zeno` I'm not sure what else to call it :)
11:07 RealBadAngel best way to handle that is to trash it
11:08 Megaf server will just say to client where the light sources are and the client would render everything
11:08 RealBadAngel see?
11:08 Megaf and in case of plantations, client would calculate where theres light and shadow and then it would tell the server
11:08 RealBadAngel calculate shadows on the servers side....
11:08 Megaf RealBadAngel: we already have bugs related to shadows on server side
11:08 RealBadAngel sun god is smiling in his grave....
11:09 Megaf sometime we get those dark areas for no reason
11:09 Zeno` why would shadows or light need to be server side (except for an approximation to account for light-reliant ABMs?)
11:09 RealBadAngel we dont have shadows at all
11:09 Megaf ok, so how do you call those dark areas under trees and buildinds and on buggy areas?
11:10 RealBadAngel bugs of so called lighting systems
11:10 RealBadAngel but no shadows
11:10 Amaz Are you planning on redoing the lighting system?
11:11 RealBadAngel im planning to redo mesh system
11:11 RealBadAngel but it seems i will have to fork mt to do so
11:11 Amaz :(
11:11 RealBadAngel im tired enough
11:12 RealBadAngel nobody is payin me for my job
11:12 Megaf RealBadAngel: do we really need this mesh thing?
11:13 Megaf and nobody is paying neither of us
11:13 RealBadAngel and nobody should crying on me what i did wrrong
11:13 RealBadAngel im the one who cares
11:13 RealBadAngel i can fix things based on feedback
11:13 proller not one
11:14 Amaz RealBadAngel, your commits are the best commits of the past month, if not further back than that.
11:15 Amaz I'm not joking either.
11:18 RealBadAngel seriously? we dont have commits that brings something new
11:19 RealBadAngel those who did were taken by freeminer
11:19 RealBadAngel most shame on us we let circuits let go
11:24 RealBadAngel we dont have even such basic feature as storing data in db
11:25 RealBadAngel mt development is in dead end
11:26 RealBadAngel nobody knows what other dev is doing, neither understand his code
11:26 RealBadAngel all the voices are about whitespaces or the code style
11:27 RealBadAngel dev branch was supposed to merge experimental code
11:28 RealBadAngel to be polished and used in stable
11:28 Amaz The node highlighting is brilliant.
11:28 RealBadAngel but all think nothin can be merged here until its stable
11:28 RealBadAngel so how?
11:29 RealBadAngel we shall hire testers to test our branches?
11:29 Amaz The devs need to open up to unstable content.
11:30 RealBadAngel nobody can create stable code on first attempt
11:30 Amaz Yep.
11:30 Amaz And it probably won't be tested if it isn't merged.
11:31 RealBadAngel if its merged folks can find holes
11:31 RealBadAngel authot can fix them
11:31 RealBadAngel ie improve the code
11:31 Amaz +100
11:32 RealBadAngel but not just by begging a piece to be merged
11:32 RealBadAngel thats fuckin sick
11:32 proller one more understood ;)
11:32 BlockMen RealBadAngel, you were also blocking things from mergin because you did not like it
11:32 BlockMen and now you complain that same is done with your code
11:33 BlockMen wtf
11:33 RealBadAngel BlockMen, i did so once
11:33 Zeno` Well, that does raise a question... why aren't there branches?
11:33 RealBadAngel for a reason
11:33 Zeno` What reason?
11:34 Zeno` If Alan Cox didn't have his own linux branch Linux would not be the OS it is today
11:34 Zeno` I would test a RealBadAngel branch
11:34 RealBadAngel BlockMen, and i said "veto"just once. in 2 years
11:35 BlockMen Zeno`:  https://github.com/RealBadAngel <-- many RealBadAngel branches
11:35 Amaz What was the veto for?
11:36 Zeno` those are not branches
11:36 BlockMen Zeno`, i ment in his minetest repository
11:37 RealBadAngel most of the code is never published
11:37 RealBadAngel i am being blocked all the time
11:38 RealBadAngel since im working on a few files just, all my code depend on previous changes
11:39 Zeno` BlockMen, they're still not branches. Unless you mean his avatar, but I thing those look more like wings
11:40 RealBadAngel i cannot even push bugfixes to the engine
11:40 RealBadAngel engine is too outdated for that
11:41 BlockMen Zeno`, wtf? https://github.com/RealBadAngel/minetest <<-- click on "branch"-Buttin to see all branches
11:41 BlockMen *button
11:41 RealBadAngel BlockMen, im not publishig everythin
11:42 Zeno` those are private branches
11:42 BlockMen Zeno`, a "RealBadAngel"-Branch on minetest/minetest wouldnt be something else
11:42 Zeno` I disagree, but anyway
11:42 Zeno` I don't want to argue
11:43 BlockMen lel
11:43 RealBadAngel BlockMen, problem is not i do have branches
11:43 RealBadAngel problem is i do have idle code
11:43 PilzAdam [13:13:51] <RealBadAngel> i can fix things based on feedback <- I just gave you feedback about one of your commits and you start ranting about how awful everyone here is and how bad Minetest's code base is
11:43 BlockMen RealBadAngel, i read you complains here
11:44 RealBadAngel i keep coding but it seems for no reason
11:45 RealBadAngel PilzAdam, good, please keep doing so
11:45 RealBadAngel if i know something is wrong, tested, i can fix that
11:46 RealBadAngel but not when code is lyin on the shelve
11:46 RealBadAngel do you get it?
11:46 PilzAdam your first reaction when I told you that you probably broke old behavior was [12:06:30] <RealBadAngel> PilzAdam, and you blame me for that? [12:07:08] <RealBadAngel> engine sucks balls not me :P
11:47 PilzAdam then you attacked me about some completly different thing
11:47 RealBadAngel PilzAdam, sorry for that attidude
11:47 PilzAdam then you started saying how bad Minetest was designed and that your code was the only correct thing to do; even if it breaks things
11:47 RealBadAngel i know
11:47 RealBadAngel but i want to push things
11:48 RealBadAngel i can focus on things i broke or the things i may know whats broken in
11:49 PilzAdam there is basically one main reason why I "blocked" most of your changes: you changes break stuff (which isn't that bad) but you also don't fix your bugs
11:49 PilzAdam like your changes to the dropped items a while ago; I told you that they are too small now and hard to point at/collect
11:50 PilzAdam you didn't fix this at all
11:50 RealBadAngel ehm, i did
11:50 RealBadAngel theyre now nice, rotating  and flashy
11:51 RealBadAngel ofc official repo is too outdated to accept that
11:52 PilzAdam what?
11:53 RealBadAngel highlighting code is takin source mesh
11:53 RealBadAngel and is rotating the selection as source does
11:53 PilzAdam why are you talking about highlighting now?
11:54 RealBadAngel you mentioned items
11:54 RealBadAngel i mentioned changes i did
11:54 PilzAdam the visual size of items is too small after you change that made size depend on item count
11:54 PilzAdam *your
11:54 RealBadAngel halo comes this way
11:55 RealBadAngel it puts the cube/mesh into copy of bigger size
11:55 RealBadAngel doesnt it?
11:56 RealBadAngel with smaller things i can add pulsation
11:57 RealBadAngel grow the things over time wih sin
11:59 RealBadAngel PilzAdam, and yes, that was intentiona
11:59 RealBadAngel piece of cobble has to be smaller than a whole stack
11:59 PilzAdam it's way too hard to see and pick up saplings falling from trees now
12:00 RealBadAngel maybe a fine tune is needed then
12:01 RealBadAngel im not claimin to be an oracle
12:01 RealBadAngel feel free to propose a change
12:02 PilzAdam I did propose a change over 2 months ago
12:03 PilzAdam also I mentioned that inventory_image was meant to also override the texture used for dropped items; your change broke that too
12:03 RealBadAngel sorry, but every time i asked for your opinion you have chosen to not respond
12:03 PilzAdam we even talked about adding a new drawtype for entities, like "wieldhand", but using the inventory_image
12:04 PilzAdam but haven't fixed that
12:04 PilzAdam and then you wonder why people don't accept your new changes
12:05 RealBadAngel they cannot accept something thats not in
12:05 PilzAdam this is not meant to attack you or to put you under pressure; I just want you to understand why I tend to "block" your new changes
12:05 PilzAdam you seem to not care about them after they got merged
12:06 RealBadAngel and sorry to say most of you (except of kahrl maybe) dont know what im doing
12:06 RealBadAngel i would like to push fixes
12:07 RealBadAngel i really would
12:07 RealBadAngel but i am not allowed to do so
12:08 PilzAdam I don't see pull requests to fix the item size issue
12:09 RealBadAngel beacause i still do have another issuses open?
12:09 PilzAdam bugfixes have higher priority then your new / other pull requests
12:10 PilzAdam *than
12:10 RealBadAngel i know what branches are for
12:10 RealBadAngel ok
12:11 RealBadAngel lets talk items size
12:11 PilzAdam do you need a comparison screenshot?
12:11 RealBadAngel is there a proposal?
12:12 RealBadAngel talk me code
12:12 PilzAdam https://github.com/minetest/minetest/commi​t/18fe277d949a82fcc6fbc47c734df514d3ea1f52​#diff-8d585596eb0f55074bb385d0a609f570L18
12:12 PilzAdam here, you change visual size from 0.5 to 0.3
12:13 PilzAdam thats definitely too small for saplings
12:13 RealBadAngel i see
12:14 RealBadAngel but i strongly suggest do not touch it now
12:14 RealBadAngel i am doing caching of meshes atm
12:14 PilzAdam https://github.com/minetest/minetest/commit/18​fe277d949a82fcc6fbc47c734df514d3ea1f52?diff=un​ified#diff-8d585596eb0f55074bb385d0a609f570R46
12:14 RealBadAngel at startup
12:15 PilzAdam and here it's even worse; you make it 0.15 for single item stacks
12:15 RealBadAngel it was at purpose
12:16 RealBadAngel so one cobble is small compared to whole stack
12:16 PenguinDad joined #minetest-dev
12:17 Zeno` I dunno, Pilz. They look ok to me
12:18 PilzAdam RealBadAngel, as I said, at leasts for saplings is too small
12:18 Zeno` Pilz, what is "too small"?
12:19 RealBadAngel lack of glasses
12:19 PilzAdam Zeno`, they are like half the size they previously had
12:20 Zeno` I think that's a far less worse regression than the inventory/craft-grid regression
12:20 RealBadAngel "before" is not meaning the same as "now" neihter "good"
12:21 RealBadAngel and term "used to"  used on dev branch is like a joke
12:24 RealBadAngel open an issue with it, describe what you want, if you will have upvotes, i will think bout it
12:24 RealBadAngel fair?
12:26 diemartin joined #minetest-dev
12:27 RealBadAngel anyway, goin to take a nap now
12:28 PilzAdam https://github.com/minetest/minetest/issues/1706
12:28 RealBadAngel celeron55_, you should rea above
12:30 celeron55_ so you made them smaller and now they are too small?
12:30 celeron55_ they were fine before, why did you change them
12:30 celeron55_ seems like a trivial revert to me
12:30 jin_xi size now shows how many are in a stack, so single ones are small
12:31 celeron55_ that's tupid
12:31 celeron55_ +s
12:31 RealBadAngel imho they represent the number of itemsin the stack
12:31 celeron55_ it's stupid
12:31 celeron55_ if there are many, make it look like many
12:32 RealBadAngel yes, its how it works
12:32 Zeno` well, the broken "Add inventory right click drag and drop" is a trivial revert as well
12:32 celeron55_ big != many
12:33 RealBadAngel single item == small whole stack == big
12:33 Zeno` at least until it's fixed to behave in a more sane manner, which kharl has suggested it will be
12:33 PilzAdam I added a screenshot to the issue
12:33 celeron55_ (i'm not going to argue further; you know my opinion now)
12:36 PilzAdam Zeno`, what is broken about it?
12:37 Zeno` PilzAdam, it behaves in a non-standard manner (compared to every other UI ever created)
12:37 RealBadAngel celeron55_,  youre head: https://github.com/minetest/minetest/pull/1685
12:37 Zeno` if you right-click drag, you do not expect it to drop (or swap!) things the mouse happens to cross over
12:38 Zeno` anyway, wait
12:38 Zeno` https://github.com/minetest/minetest/pull/1704
12:38 Zeno` my patch is not right
12:38 Zeno` but the idea is there
12:38 RealBadAngel celeron55_, youre going to pretend im doin anythin, im doing it wrong ot close your eyes?
12:38 PilzAdam Zeno`, it behaves exactly the way I expected it to work
12:39 Zeno` PilzAdam, it does not behave how I expected it to behave
12:39 PilzAdam and AFAIK it works in MC the same way, so your argument that itÄs different from any other UI is invalid
12:39 Zeno` you right-click and move something and it drops and swaps?
12:39 PilzAdam s/Ä/'/
12:39 Zeno` that doesn't make sense
12:39 Zeno` it doesn't work the same way in MC
12:40 Zeno` in MC you have to hold shift (IIRC) to get that behaviour
12:40 Zeno` and, not totally related, isn't inventory lag already bad enough without adding this?
12:40 PenguinDad Zeno`: you don't have to hold shift for that
12:42 Zeno` PenguinDad, I will use the same analogy as last night. If you right click and drag an icon in your DE you do NOT expect it to drop and/or swap with something in a window just because the mouse move crosses over into that window
12:42 PilzAdam "Drag item stacks around with the right mouse button to place one of that stack in each dragged-over slot, if that slot is not occupied already. "
12:42 Zeno` you expect it to drop when you release
12:42 PilzAdam from MC wiki
12:42 kaeza the only thing different from MC, is that in MC you must click the stack to take it, and then right click and hold
12:42 Zeno` PilzAdam, how to drop half the stack into another slot?
12:43 PilzAdam Zeno`, right click the slot, and the left click on the other slot
12:43 PilzAdam (without holding right click, of course)
12:43 Zeno` not very intuitive
12:44 celeron55_ RealBadAngel: dunno, i think you and i should be more concerned about whether other people like it than whether we like it
12:44 Zeno` holding right button means... hold
12:44 Zeno` not drop
12:44 PilzAdam holding right click to fill multiple slots with a single item is the logical consequence of a single right click to place one item
12:44 Zeno` I disagree
12:45 Zeno` and so does every UI study out there
12:45 PilzAdam Zeno`, so you generally disagree with the idea that holding a stack in your mouse is possible without holding a mouse button?
12:45 Zeno` in general, yes
12:45 PilzAdam ok, then thats a different issue
12:46 shadowzone joined #minetest-dev
12:46 PenguinDad Zeno`: this would be totally unintuitive to me :/
12:46 PilzAdam the RMB holding to place single items fits the current style of the inventory; if you want to change the whole style of the inventory then don't focus on one feature
12:47 Zeno` no, no, the current inventory (before that commit) worked perfectly intuitively
12:47 kaeza aren't there things more important to be concerned about than another "OMG lava is not renewable"-style boycott?
12:47 shadowzone Hi Zeno`.
12:47 Zeno` this is not like that kaeza
12:48 Zeno` this is a usability issue
12:48 ImQ009 joined #minetest-dev
12:48 PenguinDad Zeno`: but it's a very subjective usability issue
12:48 PilzAdam Zeno`, you seem to be the only person who doesn't like the current style
12:49 Zeno` I love the current style. I dislike the new style
12:49 Zeno` and, no, I'm not the only person
12:49 Zeno` VanessaE and ShadowNinja agree that it's not correct
12:49 PilzAdam you like the current style? there was never "hold to move items"
12:50 Zeno` PilzAdam, revert to just before that commit and try it
12:50 Amaz I think you should remember that Minetest is not a DE, Zeno...
12:50 kaeza I've used the "right click source, click target" style and I don't think it hinders the usability in any way, but that said, the new feature is kinda different from MC
12:50 Zeno` it's not a DE, but that part of it is a UI
12:51 Zeno` I never implied that it was a DE btw
12:51 kaeza in MC, you can drag & drop with any button
12:51 Zeno` I was using a DE as an analogy
12:51 PilzAdam Zeno`, you are right; I was not aware that this worked before
12:51 Amaz Sorry.
12:52 Amaz I read UI studies as DE studies.
12:52 kaeza the feature being discussed, in MC, is only activated after grabbing an item, and started by [right-]clicking on an empty space
12:52 kaeza (and dragging)
12:55 Zeno` that would be ok kaeza
12:55 Zeno` because it requires an... "interaction" by the user to say "yes, do this"
12:56 Amaz That would work nicely.
12:56 PilzAdam I agree to what ShadowNinja said here: https://github.com/minetest/minete​st/pull/1704#issuecomment-57382533
12:59 casimir btw. swapping items was an issue even before that commit (for left click).
13:03 kaeza joined #minetest-dev
13:04 Atlas|1618033 joined #minetest-dev
13:53 Amaz joined #minetest-dev
13:54 diemartin joined #minetest-dev
13:57 ignamv joined #minetest-dev
13:59 proller joined #minetest-dev
14:04 BlockMen joined #minetest-dev
14:05 zat joined #minetest-dev
14:07 kaeza joined #minetest-dev
14:12 GrimKriegor joined #minetest-dev
14:13 BlockMen comments on https://github.com/minetest/minetest/pull/1694 ?
14:14 Amaz Merge!
14:16 proller joined #minetest-dev
14:17 ignamv joined #minetest-dev
14:17 ignamv joined #minetest-dev
14:18 hmmmm joined #minetest-dev
14:18 ignamv left #minetest-dev
14:19 ignamv joined #minetest-dev
14:24 proller joined #minetest-dev
14:33 PilzAdam BlockMen, seems good, but someone should definitely rework the settings tab so there is enough place to add all settings
14:34 BlockMen for that we need a larger formspec/mm maybe?
14:35 PilzAdam maybe just a scrollbar
14:37 BlockMen or dynamic. scrollbar on small screens/windowsize larger formspec on larger screens/windowsize
14:40 Atlas|1618033 joined #minetest-dev
14:45 BlockMen sfan5, ok with mergin https://github.com/minetes​t/minetest_game/issues/323 ?
14:50 Atlas-1618033 joined #minetest-dev
14:56 Atlas|1618033 joined #minetest-dev
14:58 iqualfragile another way of eleminating the grid would be to use a texture that is bigger then one block
14:58 iqualfragile preferably a prime number
14:58 iqualfragile but that would have to be optional, as it really changes the look a lot
14:58 iqualfragile would be great for landscapes
15:27 Atlas|1618033 Have you thought about rotating the textures? Use the same texture but randomly flip and rotate it to break up the pattern. It might be less intensive than applying individual textures to each node.
15:28 PilzAdam Atlas|1618033, that would require a way for mods to disable it, since it would look weird for e.g. chests
15:29 sfan5 >that would require a way for mods to disable it, since it would look weird for e.g. chests
15:29 sfan5 what does prevent us from doing that?
15:29 Atlas|1618033 Only apply it to naturally occuring nodes
15:30 PilzAdam sfan5, nothing?
15:31 sfan5 PilzAdam: why are you saying "we can, but ..." then?
15:31 PilzAdam Atlas|1618033, tree nodes occur "naturally"
15:31 PilzAdam sfan5, it was meant as "if someone is going to do that then keep in mind that..."
15:31 sfan5 I see
15:32 Atlas|1618033 PilzAdam: Yeah, true. But we could cut those out as well.
15:33 Atlas|1618033 Water and lava should wait. Don't know what it would do to liquids...
15:33 PilzAdam something like textures = {[1] = {"default_grass.png", randomize=true}}  would work best, I guess
15:35 proller joined #minetest-dev
15:39 Atlas|1618033 My concern would be the overhead involved having to calculate it per node. Perhaps only applying it to grass nodes at first to see what happens.
15:44 Calinou joined #minetest-dev
15:59 Atlas|1618033 joined #minetest-dev
16:02 Megaf_ joined #minetest-dev
16:16 lstep_ joined #minetest-dev
16:18 darkrose joined #minetest-dev
16:18 terminator joined #minetest-dev
16:21 rubenwardy joined #minetest-dev
16:23 MinetestForFun joined #minetest-dev
16:35 sapier joined #minetest-dev
16:45 zat joined #minetest-dev
16:57 rickmcfarley joined #minetest-dev
17:20 shadowzone joined #minetest-dev
17:26 diemartin joined #minetest-dev
18:05 EvergreenTree joined #minetest-dev
18:18 iqualfragile_ joined #minetest-dev
18:24 Hunterz joined #minetest-dev
18:44 rickmcfarley left #minetest-dev
18:51 Miner_48er joined #minetest-dev
18:55 BlockMen joined #minetest-dev
19:06 VanessaE randomly flipping/rotating a texture would just plain look horrible, honestly.  However, being able to randomly choose between several textures would work.
19:07 sfan5 ^
19:07 VanessaE I believe there's a mod that does this
19:08 VanessaE but of course it does so by defining lots of new nodes
19:08 sfan5 which is not exactly a good idea
19:09 VanessaE yeah
19:13 VanessaE there IS that "special tiles" feature that was being worked on (is that in mainline?  I don't remember).  It could easily be exploited to this end.
19:13 VanessaE at least for two textures anyway
19:19 ShadowNinja Comments please?  https://github.com/minetest​/minetest/pulls/ShadowNinja
19:21 Anchakor_ joined #minetest-dev
19:21 sapier same as before, if you add a version api you'll enable mods to do things bases uppon a version number which may or may not tell anything about the really supported features on forks
19:21 VanessaE ah HAH
19:21 VanessaE THERE HE IS
19:21 VanessaE *pounce*
19:21 BlockMen ShadowNinja, https://github.com/minetest/minete​st/pull/1694#issuecomment-57516988
19:23 ShadowNinja sapier: Yes, thet's why I just provide the strings, not the version numbers.
19:23 ShadowNinja that*
19:23 VanessaE sfan5: I think this one was it:  https://forum.minetest.net/v​iewtopic.php?f=11&amp;t=8128
19:23 ShadowNinja BlockMen: Oh, didn't see that.  :-P
19:23 sapier minetest modders are known for "creative" solutions. Having just a different datatype is almost a invitation ;-)
19:24 ShadowNinja BlockMen: BTW, _game#323 looks good.
19:24 BlockMen well, sfan5 did not aswer yet
19:24 ShadowNinja (Or whatever 32* one it was you mentioned, the grass time one)
19:24 ShadowNinja tile*
19:24 BlockMen i asked if hes ok with mergin
19:25 BlockMen ShadowNinja, concering the preload items cb, do we rly need to keep it?
19:25 ShadowNinja BlockMen: Can't you at least consider opinions from the other devs?  :-[
19:25 sfan5 I would've answered if not * BlockMen has quit (Read error: Connection reset by peer)
19:25 sfan5 BlockMen: yes, ok with that
19:25 BlockMen i think its not something to be chaned that often
19:25 ShadowNinja BlockMen: Nah, that's O.K.
19:25 BlockMen ok, so can i merge my pull then?
19:25 ShadowNinja We need a better settings menu though.
19:26 ShadowNinja Auto-generated too, so you just add the setting and it's type to a list.
19:26 ShadowNinja BlockMen: Seems O.K. to me.  Does it look good without a seperate texture?
19:27 BlockMen well, not great
19:27 BlockMen but thats why i added a new drawtype
19:27 BlockMen and not changed the glasslike to do that jib
19:27 BlockMen so all that use that new drawtype are supposed to provide two textures
19:28 ShadowNinja BlockMen: You need to find a way to make it look good with all modes.
19:28 BlockMen ^ thats impossible
19:28 BlockMen and you know that
19:28 ShadowNinja BlockMen: special_times?
19:28 ShadowNinja special_tiles*
19:29 BlockMen to make that work i would also need to change glasslike_framed
19:29 BlockMen and i think then RBA would complain
19:30 Salbei joined #minetest-dev
19:31 ShadowNinja BlockMen: Well, any comments on my pulls ^?
19:32 sapier did anyone care about the font sizes fixes by now?
19:32 indriApollo joined #minetest-dev
19:33 ShadowNinja sapier: I don't like that your patch adds so many font files.  Is there a good reason not to go freetype-only?
19:33 sapier yes, we don't even have it as defaul
19:33 sapier t
19:34 sapier and freetype code is buggy as hell, do you want to maintain it?
19:34 ShadowNinja sapier: So?  If we remove the option for non-freetype builds freetype will be the default.
19:34 sapier not freetype itself the freetype adapter
19:34 ShadowNinja sapier: No, I don't know much about freetype, but it seems like it's a lot better than the image method.
19:34 BlockMen ShadowNinja, for what we need versions API?
19:34 sapier I did look at it to fix it but I don't wanna learn freetype internals
19:35 ShadowNinja BlockMen: Informational purposes.
19:35 BlockMen hmm...
19:36 ShadowNinja BlockMen: Eg, "Minetest version 0.4.10, IRC mod version 0.2" when running /CTCP MinetestServer VERSION on a minetest server with the IRC mod.
19:36 sapier ShadowNinja: in theory it's better but actually our freetype code just renders the freetype font to textures resulting in almost same as those pictures. One major difference, that "rendering code" is not checking anything
19:37 ShadowNinja sapier: Hmmm, and I guess adding proper freetype support to Irrlicht is more than we can handle?
19:37 sapier I'd suggest to switch to freetype once irrlicht does support it natively but not while we're using that ugly adapter
19:38 sapier I'm quite sure I don't know enough about fonts to even fix that adapte
19:38 sapier r
19:38 BlockMen sapier, does it allow using more than one font or just another way to change skin font?
19:39 BlockMen ShadowNinja, so its mainly for IRC mod?
19:39 sapier right now we have two fonts, it's not fully generic
19:39 sapier but changes wouldn't be that big, yet we don't even support different font sizes within minetest ;-)
19:40 BlockMen ik :\
19:40 sapier my patch would support them but without minetest using it ;-)
19:40 sapier and that patch is only half of it, the second part are those formspec fixes from the guy refusing to use github ;-)
19:41 ShadowNinja BlockMen: That's the use case I have in mind.  It could also be used to remove get_server_status and build the string in Lua (hmmm, we already have access to the version anyway, just in a harder-to-access format).
19:43 ShadowNinja BlockMen: My API provides access to the full version hash though, not just the simple string.
19:44 sapier to be honest as most ppl here usually refuse to even support stable version with their mods I doubt those ppl will fix their mods for different versions
19:46 BlockMen ShadowNinja, i thought current get_version() is mm only?
19:48 sapier You're right blockmen, at least I didn't add it to minetest api on implementing it (unless it's been merged by accident)
19:48 sapier e.g. when scriptapi was refactored
19:49 ShadowNinja BlockMen: "mm only"?
19:49 BlockMen mm = main menu
19:49 ShadowNinja BlockMen: Oh, yes, it is.
19:50 BlockMen so the main question is not string vs. complete version info
19:50 BlockMen the question is: should mods get acces to versions
19:50 ShadowNinja BlockMen: The game only has get_server_status, which can be parsed for the version string.
19:51 BlockMen oh, ic
19:51 ShadowNinja (Which I WILL implement in the IRC mod if I can't get a proper API function for it.)
19:57 BlockMen ShadowNinja, why VERSION_EXTRA_STRING ?
19:58 ShadowNinja BlockMen: Because it might be usefull.
19:59 Amaz joined #minetest-dev
19:59 BlockMen i think its an bad idea. mods could exclude single versions/builds
19:59 BlockMen im against adding this
20:00 BlockMen this = version extra
20:01 ShadowNinja BlockMen: Alright, otherwise O.K.?
20:02 BlockMen ShadowNinja, if "minetest_version_hash" does not contain the version extra aswell, yes
20:03 ShadowNinja BlockMen: I think hash does, but you'd have to parse the version-extra out, which mods can already do with get_server_status.
20:04 BlockMen get_server_status returns   minetest_version_simple
20:04 BlockMen so no, they can currently not
20:04 ShadowNinja Trying to make it impossible for mods to exclude versions is pointless and counter-intuitive.  Mods can provide a warning that a specific build is broken in a certain way and that they hould expect a certain bug if you let them.
20:05 BlockMen wait, wait. you told me mods should not be able to override the ids of statbars (because that would be bad), but they should be able to exclude certain builds
20:06 ShadowNinja Eg, say 0.4.10 had a bug that causes flashing textures with that mod.  You can warn users about that and avoid getting spammed with "FIx plx" issues about that.
20:06 BlockMen could you be so nice and find a sane line what mods should be able to and what not
20:07 ShadowNinja Mods should be able to do just about anything that doesn't break security, but certain things are almost always bad and should be discouraged.
20:07 BlockMen lol
20:07 BlockMen so mods can disallow certain builds but the statbar is protected once and for all
20:07 ShadowNinja It's not funny, I'm getting quite tired of this.
20:07 proller joined #minetest-dev
20:07 BlockMen its indeed not funny
20:08 BlockMen get your version by server_status
20:08 ShadowNinja BlockMen: Also, I'm not sure what you're refering to.
20:08 BlockMen thats enough
20:10 ShadowNinja BlockMen: Alright, doing that now.
20:12 ShadowNinja I can't get the PROJECT_NAME this way though, so it will show up as "Minetest" even with other forks.
20:14 ShadowNinja http://sprunge.us/cXYD?diff
20:18 ShadowNinja Works with some tweaks: * Received CTCP-VERSION answer from TestServer: Minetest version 0.4.10-dev, IRC mod version 0.2.0.
20:20 ShadowNinja https://github.com/kaeza/minetest-irc/commi​t/d0ebde388419fcf102d8397d392cf652f313a2fb
20:20 ShadowNinja ^ That was about 20x easier than arguing about implementing a cleaner way to do it.
20:20 ShadowNinja And took even less time.
20:21 ShadowNinja It only takes about 4 lines too, so it's not all that bad.
20:22 ShadowNinja A whole lot better than my method that looked for the executable, opened it, read it into memory, and searched for the VER= in the build_info.
20:28 Atlas|1618033 joined #minetest-dev
20:30 ShadowNinja BlockMen: Next pull?
20:31 ShadowNinja Ok, Taoki: I think you were looking for a way to get the version?  See my method, if probably the most elegant way to get the version currently available, and, by the looks of it, that will ever be available.
20:32 ShadowNinja Oh*
20:32 ShadowNinja s/if/it is/
20:32 Taoki ShadowNinja: Get what version?
20:32 ShadowNinja Taoki: The Minetest core version.
20:33 Taoki I already do with GIT, for years
20:34 ShadowNinja Taoki: Oh, well this works on systemwide installs and builds without git.  It also doesn't require os.execute.  It doesn't give you the hash though.
20:34 sapier I still don't understand why this is needed for regular operation of a mod ;-) yes it's nice to know but for what else then curiosity?
20:34 Taoki What works? :P
20:35 ShadowNinja Taoki: https://github.com/kaeza/minetest-irc/commi​t/d0ebde388419fcf102d8397d392cf652f313a2fb
20:35 ShadowNinja sapier: I'm done arguing about it.
20:36 sapier meaning your just gonna add it without argueing? ;-)
20:37 ShadowNinja sapier: No, meaning I'm going to ignore it for a few more months and use hacky ways to get the version instead.
20:38 ShadowNinja sapier: Feel free to approve one of my other pulls though.  :-)
20:38 sapier for what do you use the version number?
20:39 ShadowNinja sapier: /connect irc.inchra.net; /ctcp ShadowServer version
20:39 sapier https://github.com/minetes​t/minetest/pull/1631/files can someone explain to me why we replace a speaking name by a nonsense name?
20:40 ShadowNinja sapier: Argue with iqualfragile about that.
20:40 sapier ShadowNinja did you try your code?
20:41 ShadowNinja sapier: What code?
20:41 sapier for what I tried "asInt()" doesn't work for that crapy json lib
20:41 sapier nor anything else except "asString()"
20:41 ShadowNinja sapier: Oh, yes, it works, at least on my computer.
20:41 sapier strange
20:42 sapier If those "asType" functions would work lot of that code could be done way more simple
20:49 sapier https://github.com/minetest/minetest/pull/1535 seems to be fine (code wise). Guess we can live with that const char* for some more time
20:49 sapier https://github.com/minetest/minetest/pull/1418 completed?
20:49 sapier if yes merge it
20:50 sapier left #minetest-dev
21:09 hmmmm ooh what's going on
21:10 hmmmm looks like I missed it :(
21:17 proller 1535 - nooooo
21:36 kahrl joined #minetest-dev
21:47 casimir joined #minetest-dev
22:02 ShadowNinja proller: Open thousands of files and write to them every few seconds, plus a final write at shutdown?  You must just want MT to be slower to make FM look better.  Before my changes I saw my disk usage spike to 80-100% for 2-3 seconds on every save interval (which was why I had said interval at about 15 seconds instead of the default of 5 seconds or something like that).
22:07 shadowzone joined #minetest-dev
22:07 shadowzone joined #minetest-dev
22:12 hmmmm heh
22:13 hmmmm no matter what minetest does - short of merging it in - would make minetest as bad as freeminer
22:14 hmmmm my real life friend got hooked onto minetest about a month ago and found freeminer on google and read that it was supposed to be a better version so he tried that out
22:14 proller why  "Open thousands of files" ?
22:14 hmmmm he said that it often crashed, was very buggy, and he much preferred minetest
22:16 proller ShadowNinja, your patch fixes consequences of very bad design by uglifying code
22:17 proller and not solves real problem
22:18 proller hmmmm, what you make for minetest in last year except cool jokes about freeminer ?
22:18 ShadowNinja The few times I've looked at freeminer's commit history I've found that the majority of the commits were fixes.
22:19 hmmmm i won't deny that i haven't done much, but that doesn't change the fact that freeminer is a prime example of what minetest would have become if we let him have his way with the codebase
22:19 hmmmm oh god, I'm so glad freeminer is a separate project
22:20 ShadowNinja Also, anyone: http://strawpoll.me/2687768
22:21 proller and lot of fixes about mt code
22:21 kahrl ShadowNinja: missing #5: I don't care because I /ignored him
22:21 proller now nobody build/test freeminer under windows
22:22 shadowzone o.O
22:22 shadowzone Who is proller?
22:22 proller under linux some problems exists, but fixes ^ in progress
22:23 shadowzone Cause I'm not gonna vote till I find out.
22:23 ShadowNinja kahrl: That goes as "Yes", since you've already done it, just not globally.
22:24 ShadowNinja shadowzone: He's talking in here.  You might have him on /ignore.
22:24 shadowzone I don't.
22:24 shadowzone I've never even heard of him.
22:25 Miner_48er joined #minetest-dev
22:25 exio4 he was(is?) a coredev
22:25 ShadowNinja shadowzone: http://irc.minetest.ru/minet​est-dev/2014-10-01#i_3957766
22:25 ShadowNinja exio4: Was.
22:25 exio4 k
22:26 shadowzone What'd he do?
22:28 ShadowNinja shadowzone: He trolls and is generally unhelpfull and annoying.  Just about all he does is criticise Minetest, it's devs, and the code they write; while telling us about how everything is fixed in Freeminer.
22:28 shadowzone OH.
22:28 proller yes, saving players problem was solved in freeminer ;)
22:28 shadowzone So a pain in the butt?
22:29 proller ans solved from root without uglyfying code
22:29 proller because i understand what i do
22:29 shadowzone I voted.
22:39 proller ShadowNinja, and im criticise only bad or useless code
22:44 Atlas|1618033 If this is the usual goings on in here, I think I can drop my cable...
23:11 proller joined #minetest-dev
23:39 Atlas|1618033 joined #minetest-dev

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