Time  Nick         Message
00:04 kahrl        #1891
00:04 ShadowBot    https://github.com/minetest/minetest/issues/1891 -- Display serverlist flags as icons by kahrl
00:05 kahrl        (thanks to john_minetest for creating the icons)
00:05 hmmmm        the icons could probably be more compact
00:05 hmmmm        i really wish that interface scaled with the screen size
00:05 kahrl        yeah, let's see how that looks
00:06 hmmmm        well, that's another problem for another day
00:06 hmmmm        autoscaling formspecs.  lol.
00:09 kahrl        http://i.imgur.com/EonsYv1.png
00:11 kahrl        btw, the large space in front of clients_max is because there are some servers with clients_max = 1000
00:11 kahrl        wonder if that should be limited in the menu
00:12 paramat      hmmmmm > "what if each biome defined a "size factor".." hmmmmm > "..I don't want to manually adjust the spread, i think the spread should be adjusted by the number of biomes currently active". good idea, i realised players will not need to adjust octaves, persistence or seed of the heat/humid noises, so we can auto-set the spreads of those noises from a new per-mapgen 'biome size' parameter multiplied by sqrt(number
00:12 paramat      of active biomes). so heat/humidity noises do not need to be settable through .conf at all =) i will work on this
02:45 hmmmm        hrmmm
02:46 hmmmm        feeling conflicted on where to place getNoiseParams()
02:47 hmmmm        Settings::getNoiseParams() would be nice... I like NoiseParams blah = g_settings->getNoiseParams("np_blah");  but should I put something that specific inside of settings.cpp?
02:47 hmmmm        currently, getNoiseParams() is just a macro defined in noise.h, not an actual method in Settings.
03:20 OldCoder     Excuse silly question. After 3 years I need to learn how to wield an item.
03:20 OldCoder     How may I wield an item in my inventory? Google shows nothing for such a basic question.
04:44 Zeno`        I cannot build latest commits :)
05:19 Zeno`        one-liner (missing include) so pushed fix
05:22 cg72         Zeno` you saved the day
05:26 cg72         Zeno`>= real_dev
05:41 hmmmm        uh oh
05:41 hmmmm        how did algorithm not get included
05:41 hmmmm        that's probably thanks to my use of std::remove
05:42 hmmmm        =/  "builds for me" != "builds"
05:44 hmmmm        all of those clang warnings are pretty much trivial to fix
07:57 paramat      https://github.com/minetest/minetest/pull/1892 Add node_stone to mgv5 biomes
08:00 paramat      nore if the MTgame team decide to add snow biomes to mgv6 i might be able to do that (after 0.4.11 as i am busy) but it will have to be enabled using a parameter to not break old worlds
10:28 Zeno`        Will merge #1893 as a simple bug fix in about an hour unless anyone objects
10:28 ShadowBot    https://github.com/minetest/minetest/issues/1893 -- Fix node 0,0,0 being highlighted when enable_node_highlighting == false by Zeno-
10:55 kahrl        Could the mainmenu tabs that are not singleplayer be expanded downward so they occupy the space of the missing game button bar?
11:00 Zeno`        Hi kahrl ... <Zeno`> Will merge #1893 as a simple bug fix in about an hour unless anyone objects
11:00 ShadowBot    https://github.com/minetest/minetest/issues/1893 -- Fix node 0,0,0 being highlighted when enable_node_highlighting == false by Zeno-
11:02 Zeno`        I wish I knew what time I wrote that comment
11:02 Zeno`        I think it must have been almost an hour ago. Feels like it anyway
11:03 kahrl        Zeno`: hmm
11:04 kahrl        right now the code handles the case where enable_node_highlighting changes in-game, your patch does not
11:04 kahrl        Zeno`: doesn't your irc client show message times?
11:05 Zeno`        karhl, I checked if it could be enabled and disabled in-game; I'll just use the g_settings() method... I was debating with myself about that anyway
11:05 Zeno`        kahrl, my IRC client probably does. I've never looked at the menu :)
11:05 kahrl        Zeno`: alternatively you could initialize m_highlighted_pos to some out-of-range value in the Client constructor, say 31337,31337,31337
11:07 Zeno`        would that be a better way do you think?
11:08 kahrl        it would mean fewer ifs and thus easier code, so probably yeah
11:08 Zeno`        ok, good idea
11:09 kahrl        where's RBA btw? he meant to fix this :)
11:09 Zeno`        I haven't seen him for a few days
11:15 Zeno`        pushed update
11:18 kahrl        looks good
11:20 Zeno`        ok, I'll merge in a bit
11:21 Zeno`        The reason I didn't do that (out-of-range thing) in the first place was that I wasn't sure if that was actually checked. I am assuming now that it is. I'll have to look at those sections of code more closely one day soon
11:21 Zeno`        I.e. I wasn't sure of side-effects
11:27 kahrl        hmm, maybe it's a little inefficient because it adds update mesh tasks for the mapblock of 31337,31337,31337
11:28 kahrl        does that actually cause MapBlockMesh() to be called?
11:28 kahrl        I hope not
11:30 Zeno`        well actually it probably does :/
11:31 Zeno`        It appears to get added to the mesh m_queue anyway :p
11:31 kahrl        damn
11:31 kahrl        oh
11:31 kahrl        could you try initializing the strangely named m_show_hud to false instead?
11:32 Zeno`        pretty sure I tried that, but will do again because I'm not totally sure now
11:33 Zeno`        hmm, I mustn't have. /me looks at the code again
11:34 Zeno`        initialising to false seems to work, I just want to triple check
11:35 Zeno`        I'm not sure I like that name
11:35 Zeno`        it's used in a very strange place
11:35 kahrl        yeah I'm not sure what RBA smoked when he chose that named, but it must've been good ;)
11:36 kahrl        s/named/name
11:36 Zeno`        toggling hud on/off in game makes it come back ;)
11:36 Zeno`        actually I can probably modify thegame.cpp since it's passed to setHighlighted
11:37 Zeno`        and make it always false if node highlighting is disabled
11:37 Zeno`        there's already a check there where it's called anyway
11:37 Zeno`        seems hacky though :/
11:38 kahrl        why, it seems reasonable
11:38 Zeno`        well, probably only because the param name is show_hud :)
11:39 kahrl        it's already forced to false if the pointed type is not a node
11:39 kahrl        rename it :)
11:40 Zeno`        and it would add 2 mesh tasks that are not normally there when enable_node_highlighting is false. Perhaps where show_hud is toggled might be better. I'll look more closely
11:41 Zeno`        yes, that's the perfect place
11:41 kahrl        I think there is already a bug
11:41 kahrl        in master I mean, that you'll fix with that
11:42 kahrl        ah no, sorry
11:44 kahrl        I thought that with enable_node_highlighting, if you disabled the hud, moved the mouse off the node and enabled the hud again, it would show the old highlight
11:44 kahrl        but it doesn't do that because the code in processPlayerInteraction runs every frame
11:44 kahrl        which makes me wonder what the point of the line in toggleHud is
11:45 Zeno`        it's to stop the highlighting when the hud is hidden
11:45 kahrl        oh, because of the pointed != runData->pointed_old
11:46 kahrl        makes sense now, I find it confusing though as you might be able to tell ;)
11:46 Zeno`        wth am I going to rename it to, though lol
11:47 Zeno`        force_highlighting_off :/
11:48 kahrl        bool highlighted?
11:49 Zeno`        I'm not crazy about either suggestion but I guess they're both better than show_hud
11:49 Zeno`        hide_highlighting
11:49 Zeno`        I'll go with that for want of a better name
11:50 kahrl        why make it negative?
11:50 Zeno`        I am a negative person
11:50 kahrl        :)
11:51 Zeno`        nah, show_highlights
11:51 Zeno`        show_highlighted
11:52 kahrl        yeah
11:59 Zeno`        ok, done
11:59 Zeno`        tested with node highlighting on/off and also toggling hud/on off with highlighting on and off
12:02 kahrl        Zeno`: nice, merge it
12:02 Zeno`        ok. For paranoia's sake I'll wait for Travis
12:02 kahrl        https://github.com/minetest/minetest/blob/master/builtin/fstk/tabview.lua#L31
12:03 kahrl        ^ why does this check tab.size when the rest of the code uses tab.tabsize? and why x,y instead of width,height?
12:03 kahrl        sapier: ?
12:05 kahrl        well tabsize works but it also moves the tab buttons, which is undesirable
14:46 Zeno`        The font manager is not receiving callbacks (e.g. changing font_size in minetest.conf has no effect)
14:46 Zeno`        Strangely enough this mechanism introduced looks almost exactly the same as my proposed solution
14:48 Zeno`        It's possible to change parseConfigLines() to call all of the callbacks but they would have to be all called for each known settings string at the time they're called :/
14:57 Zeno`        Any issues with merging #1853?
14:57 ShadowBot    https://github.com/minetest/minetest/issues/1853 -- Add name of node pointed at to debug by Zeno-
15:00 Zeno`        hmm changing font size does work but not for formspecs
15:02 * Zeno`      pokes kahrl
15:05 Zeno`        Should the tile name remain?
17:44 Calinou      font_size no longer works? I'm stuck with a tiny (yet Freetype) font
17:44 Calinou      increasing GUI scaling doesn't increase all elements, and makes some elements too large
17:46 Krock        Calinou, call sapier
17:48 Calinou      sapier()
17:49 Krock        yes.
17:53 VanessaE     minetest.sapier(fix_fonts, true)
17:55 sol_invictus I already reported this problem yesterday
17:56 sol_invictus http://irc.minetest.ru/minetest-dev/2014-11-30#i_4044526
17:56 Krock        minetest.get_player_by_name(sapier):add_task(fix_fonts)
17:58 sol_invictus I think you either should define fix_fonts functions or pass it as a string
17:58 VanessaE     string-named functions are deprecated :)
17:58 Krock        ^
17:58 Krock        sol_invictus, it's a global variable
17:59 sol_invictus hah, tasks passed around as globals is an interesting concept
18:14 Krock        Calinou! Your turn!
18:16 VanessaE     minetest.sapier(fix_fonts, true)
18:16 VanessaE     Server terminated after calling an instance of font_size:  undefined
18:16 VanessaE     ;)
18:19 sol_invictus good one VanessaE
19:59 Kodexky      Hi. Android build is broken again. fontengine.cpp is missing in Android.mk
20:01 Krock        \o/
20:03 sapier       well it's gonna be fixed prior release ;-)
20:22 kahrl        Kodexky: fixed
20:26 sapier       you're to fast for me kahrl ;-)
20:35 twoelk       could it be that minetest.conf.example is missing a space in line 65 right after the hash?
20:37 CraigyDavi   Yep, and capital letter
20:40 twoelk       my little experimental parser keeps tripping in that line (among other things)
20:42 sapier       hmm but your parser should be able to handle minor style issues like that ;-)
20:42 kahrl        twoelk: the comment indicator is "#", not "# "
20:43 twoelk       I try to split comment from example
20:45 kahrl        well... the .example is really meant for human consumption, not for being parsed
20:46 twoelk       well I though it might work :-/
20:46 kahrl        I'll fix that line but I can't promise your parser will keep working
20:48 twoelk       was fiddling with an out of game configuration gui that takes the possible configuration keys from that file, instead of hardcoding everything
20:48 twoelk       not really working as intended, maybe never will
20:48 sapier       twoelk that's not gonna work
20:49 twoelk       :-(
20:49 sapier       we can't guarantee description will always follow some style, and we don't even want to
20:50 twoelk       I'm ok with the first   -   but the second?
20:52 * twoelk     sulks at being sabotaged before even getting on the road
20:52 sapier       it's not sabotaged but why force ppl to follow some exactly defined style in a comment?
20:53 sapier       sometimes a different style is more readable and readability is goal for that file
20:54 twoelk       oh well at least my feasabilty tests added a missing space lol
21:02 kahrl        sapier: do you think we could/should make the non-singleplayer tabs larger?
21:02 Kodexky      kahrl: :)
21:03 kahrl        the game button bar isn't shown on them so the space could be used productively for the actual tab
21:03 sapier       kahrl the tab headers or the tabs itself?
21:03 kahrl        the tabs themselves, expanding them downward to occupy the space of the game button bar
21:03 sapier       would be a minor change
21:04 sapier       you can specify a size per tab only if none is specified original size is used
21:04 kahrl        I tried setting tabsize on them, that works but it moves the tab header bar
21:04 kahrl        is there a way to avoid that?
21:05 sapier       let me have a look but if I remember correct a tabview tries to be centered. we could add a "fixed" mode but that'd be strange on higher resolutions
21:05 sapier       hmm we'd need to separate x and y centering
21:06 sapier       not yet supported as fstk tries to remain simple ;-)
21:06 kahrl        another minor thing, https://github.com/minetest/minetest/blob/master/builtin/fstk/tabview.lua#L31
21:06 kahrl        this should check for .tabsize, not .size, and .width/.height, not .x/.y
21:07 kahrl        oh and 'table' instead of table
21:07 sapier       ok if I remember correct there's already a open fstk update pull request, I'll change it in there
21:11 sapier       I'll add x and y alignment options too there
21:11 kahrl        thanks :)
21:12 sapier       hmm not sure if this is what you need ;-)
21:12 sapier       because the y position of a x sized tab will still be different from a x+2 sized tab
21:12 kahrl        hmm
21:13 sapier       that's the problem with "we want a fixed size main menu"
21:13 sapier       we'd have to do a very very special handling to get same position
21:14 kahrl        yeah I can't think of a clean way to do it either
21:15 sapier       we'd have to calculate the exact position of the original menu and then foce position the bigger one ... no idea how to do this in a clean way
21:16 kahrl        make the bigger size the default and (if the tab activates the game button bar) 1. position the tab 2. subtract space for the game button bar from the max y 3. add the game button bar?
21:17 sapier       button bar and menu are independent objects
21:17 sapier       I'd prefere not to link them
21:18 kahrl        that's a bit counterintuitive because its visibility depends on which tab is selected
21:18 sapier       fstk ain't a full gui toolkit ;-)
21:19 sapier       there is limited parentship but it's not as complex and full featured as in gtk or qt
21:19 kahrl        I see
21:20 sapier       for what I remember the buttonbar is child of tabview, at same level as tab itself
21:20 sapier       but I'd have to verify this it's some time since I did that
21:23 sapier       is it really that crucial to have it at exactly same position? modstore uses different size too?
21:24 kahrl        yeah
21:24 kahrl        it's quite annoying if you click one tabheader, move mouse to the right or left and the cursor is no longer on the tabheader
21:25 kahrl        *tabheader bar
21:25 kahrl        it's not a problem with sub-dialogs because you don't enter or leave them through the tabheader bar
21:26 sapier       you're right