Time  Nick          Message
00:25 sapier1       https://github.com/minetest/minetest/pull/724 now including world config
00:28 PilzAdam      sapier1, https://gist.github.com/PilzAdam/5609715
00:29 PilzAdam      first loading screen with black background and then a grey screen
00:31 sapier1       I wonder what you're doing ;-) ... I'll try a clean build to figure out
00:39 Exio          i just compiled
00:39 Exio          and got the same error
00:39 PilzAdam      maybe because the settings in minetest.conf are not present?
00:39 sapier1       strange my clean build gives an error to but not same as yours (I can't create a world)
00:40 Exio          what do you have in your minetest.conf?
00:40 sapier1       ok after starting second time I get same error
01:05 sapier1       ok it's working for me on a fresh cloned minetest tree now
01:06 sapier1       no curl included so plz give feedback about this setup
01:19 RealBadAngel  http://imgur.com/a/Eq09o#0
01:19 RealBadAngel  i played a bit more with shaders
01:20 RealBadAngel  theres compare from left to right: normal+some specular, only normal, then no shaders at all
01:25 sapier1       looks interesting ... imho the leftmost blocks (up to chest) look a little bit to plastic
01:31 hmmmm         well they look about just as plasticy without the normal map.. it's not a really good sand texture i guess
01:32 hmmmm         is there no diffuse lighting in minetest?
01:34 iqualfragile  sapier1: they will continue looking like plastic untill real lightningsources are defined in minetest
01:35 sapier1       hmm :-/
01:37 RealBadAngel  vec3 lVec = normalize(vec3(0.0, -0.4, 0.5));
01:37 RealBadAngel  this is my lightsource atm ;)
01:38 iqualfragile  i think that i should be possible to hack something…
01:38 RealBadAngel  like what?
01:38 iqualfragile  a) calculate where the sun/moon should be using the daytime and use that to generate the vector
01:38 hmmmm         i think we should start the move to real lighting
01:39 hmmmm         the imagined problems really aren't that bad
01:39 RealBadAngel  its definitely needed
01:39 hmmmm         hmm.. https://github.com/minetest/minetest/pull/608 this pull request isn't being merged
01:39 hmmmm         https://github.com/minetest/minetest/pull/611 no idea what this even does
01:39 iqualfragile  b) put a litle test into the blockload/unload code wich checks if the node emits light (i think there is a group for it), put them into a list and use these as lightsources
01:40 hmmmm         https://github.com/minetest/minetest/pull/640 this wasn't added yet?  hmm ..
01:41 hmmmm         still wondering about this https://github.com/minetest/minetest/pull/670  celeron doesn't seem to like it but i don't see any problem, never got his final word
01:41 iqualfragile  it seems to me that 611 enables the use of two different animations on one model at the same time
01:41 iqualfragile  (like shaking the head an d walking or whatever)
01:42 hmmmm         ah... forgive me about the whole animation thing, i'm completely unknowledgable about that
01:42 hmmmm         i'd rather leave it to somebody who does know
01:43 iqualfragile  but i think it good that you take a look at the pull requests
01:43 hmmmm         https://github.com/minetest/minetest/pull/729 this was going to be included with my world creation parameter interface, but i won't add it because i know it'll screw up what sapier's working on
01:43 iqualfragile  i think that its important to give those more attention as a successful pullrequest motivates the programmer to do more work
01:44 iqualfragile  successfull& quick
01:44 hmmmm         true... there are lots of reasons to not like outside code though
01:45 RealBadAngel  imho 611 does exactly nothing
01:45 iqualfragile  nope
01:45 RealBadAngel  this commit just copies 3 lines and pastes them in different order
01:45 iqualfragile  it moves two lines around!
01:46 iqualfragile  maybee order is important for animations, i dont know, i am not realy knowledagble either
01:46 hmmmm         i guess outside code isn't as bad as it otherwise would be, because we don't exactly have a set development timeline and nobody's being put on the spotlight if there's a problem
01:48 hmmmm         what do you guys think about this https://github.com/minetest/minetest/pull/675 ?
01:48 hmmmm         pilzadam seems to be taking care of it maybe, or he forgot about it
01:50 RealBadAngel  from what i can see khonk have fixed everythinh
01:50 RealBadAngel  and rain looks good
01:50 hmmmm         with all the memory leaks and warnings fixed, along with those other miscellaneous problems, i'm going back to work on the decoration things, and when i'm done with that, i'll do the world parameter configuration hopefully by the time sapier is done with his thing
02:06 ShadowNinja   Hmmmm: I noticed one warning (repeated a few times) http://pastebin.ubuntu.com/5682388/ Is this fixable?
02:07 hmmmm         i guess so, but i'd rather not
02:07 hmmmm         why is it that every time i think i fix all the warnings for everybody's setup, here's another one
02:08 RealBadAngel  http://imgur.com/a/Eq09o#3
02:08 hmmmm         wow nice
02:08 sapier1       guiConfigureWorld won't exist anymore if formspec menu is added so most likely fixing is wasted work
02:08 RealBadAngel  added another pictures to comapre (another pack)
02:09 hmmmm         i feel like it looks a bit nicer than the other one with normal + specular, what is that
02:11 RealBadAngel  3 is specular+normal, 4 shaders off, 5 only normal
02:44 RealBadAngel  ive changed a few normal maps
02:44 RealBadAngel  to get right effects all kind of surfaces need different made ones
02:45 RealBadAngel  http://imgur.com/a/Eq09o#6
02:49 VanessaE      those are looking goodf
02:49 VanessaE      good*
02:50 VanessaE      still too dark without proper bumpmap-aware lighting though
02:59 RealBadAngel  http://imgur.com/a/Eq09o#7
02:59 RealBadAngel  imho caves are lookin good now
03:01 VanessaE      not bad.
03:03 RealBadAngel  good enough for a start ;)
03:03 RealBadAngel  im off to work, cya laters
03:03 VanessaE      ok, seeya
06:20 celeron55     hmm, so the scriptapi split wasn't merged yet? i'd be super happy if somebody would do that (this version is ready-to-go https://github.com/kahrl/minetest/commits/scriptapi_separation) as i'm quite busy with other stuff
06:20 celeron55     i could merge it but wouldn't have time to fix any problems found later to be caused by it, so it doesn' feel right to merge it
06:22 celeron55     i recommend considering the scriptapi freeze to continue until that is merged
11:31 PilzAdam      sapier, https://github.com/minetest/minetest/pull/696/files#L1R1054 what if there are 2 exact same craft recipes?
11:32 sapier        the first one will be removed ... as far as I know this is already an error case not handled correctly
11:33 PilzAdam      why not remove all?
11:34 sapier        would be another option I don't have any preference which one to implement
11:34 sapier        but I'd recommend not to merge this until scriptapi separation is merged
11:34 kahrl         does it happen often that there are duplicate crafting recipes?
11:35 sapier        there's some duplicated code in this function that can be moved to a separate function to be used by both luaapi fcts
11:35 sapier        imho this shouldn't happen at all kahrl
11:35 sapier        I guess first one to be found will be only one to work either
11:35 sapier        but I haven't checked this
11:35 kahrl         if there aren't, removing all would add 100% to the runtime of remove_craft on average
11:35 kahrl         sapier, yes
11:36 kahrl         well, the first from the back since it checks them in reverse order
11:36 PilzAdam      just checked, it can happen that 2 exactly same recipes are registered
11:36 PilzAdam      so calling remove_craft() once doesnt help at all
11:36 sapier        and what does minetest do in this case?
11:36 sapier        it would work pilzadam as you could check in lua if there's another one
11:37 kahrl         if you registered it n times it would then be registered n-1 times
11:37 sapier        it's a minot change but imho multiple identical recieps are a very rare case
11:38 PilzAdam      I would prefer to be sure that all recipes are removed
11:38 sapier        current behaviour is symetric to register_craft function
11:38 sapier        we could stop register_craft from registering a second time
11:38 sapier        giving an error message in this case
11:39 sapier        are there any legit cases for multiple recieps?
11:39 kahrl         the way it is implemented now that would make register_craft take quadratic time if counting all crafting recipes in all mods
11:40 PilzAdam      just dont break in removeCraft() is the easiest way I guess
11:40 kahrl         I mean O(n) if n recipes are registered, total O(n^2)
11:40 sapier        hmm comment of function is wrong too
11:41 PilzAdam      sapier, rebase and fix these things after scriptapi split is merged
11:43 sapier        yes ... still decision of how it should behave has to be made ;-) remove all or symetric to regsiter remove only one
11:43 kahrl         I wonder if the CraftDefManager should keep indexes to make searching of recipes faster?
11:43 sapier        imho performance isn't that critical for an initialization function
11:43 kahrl         it crafting too slow if many recipes are registered?
11:43 kahrl         is*
11:44 PilzAdam      I wonder if all the crafting stuff can be moved completly to Lua
11:44 PilzAdam      bbl
11:44 sapier        hmm could be done as far as I know list operations aren't that slow within lua
11:44 sapier        at least with jit
11:45 sapier        without ... maybe not a good idea
11:45 kahrl         but when crafting, checking each single crafting recipe in lua would be slow I imagine
11:46 sapier        I think there is lots of optimization possible there but I don't have heared anyone complaining about crafting performance ;-)
11:46 sapier        ok very bad english sentence :-)
11:47 kahrl         they might not realize the bad performance and assume it is just lag
11:47 sapier        do we have any counters in place to validate our assumption?
11:48 kahrl         I don't think so but it would be useful
11:49 sapier        yes :-) but this is a task for someone else :) I've already got plenty of work with scriptapi and mainmenu ;)
11:49 kahrl         a simple TimeTaker in getCraftResult and similar functions would be enough
11:50 sapier        another topic i18n for lua ... sfan5 has some code in place doing things like that, I'd need i18n for mainmenu
11:50 kahrl         is sfan5's code for serverside lua?
11:50 sapier        I'd prefere to add a single solution for mainmenu and scriptapi
11:51 kahrl         you mean single solution on the interface side?
11:51 sapier        it's for scriptapi so it should work serverside too with (some) changes
11:51 kahrl         because I think internally they would have to work very differently
11:51 sapier        I guess it wouldn't be that different
11:52 kahrl         where can I find sfan5's code?
11:52 sapier        it's just replacing text by text ;-) don't know why there should be any difference for mainmenu and scriptapi, but both are servers side of course just different lua interpreters
11:53 kahrl         well on the server side you have many different clients that may want different languages
11:53 sapier        http://forum.minetest.net/viewtopic.php?id=5989
11:54 kahrl         "This is only useful when you play in Singleplayer or when you run e.g. a Spanish Minetest Server"
11:54 sapier        hmm you're right kahrl didn't think about this issue by now
11:55 sapier        ok that's not a problem for mainmenu as this is client side ... maybe a single solution isn't as easy as I thought first time
11:56 kahrl         I still would say the interface for mainmenu and scriptapi should be the same or at least similar
11:56 kahrl         so modders don't have to learn a whole new thing just to make a menu
11:57 sapier        yes ... maybe if client tells it's settings to server server can use this information for sending data e.g. in formspecs
11:57 sfan5         are you talking about i18n for lua?
11:57 sapier        yes
11:58 sapier        e.g. L.getText("text to translate",localization(optional))
11:58 sapier        if localization is not passed client/server setting is used if passed it tries to use the given one
11:58 kahrl         yes that's a start, but then there are things like itemdefs that are registered before any client is connected
11:59 sapier        true this only works for client specific generated content
12:00 sapier        maybe we can't fix everything with a single mechanism
12:01 Zeg9          Why not sending the .po's and using gettext client side ? (warning: didn't read everything up there)
12:02 sapier        interesting idea but what to do if text contains dynamic data?
12:04 Zeg9          Maybe server could somehow tell which parts to translate
12:04 sapier        this is a lot of work to do ;-)
12:04 Zeg9          Like, "_(Current time is: )"..time
12:05 Zeg9          Parts inside _(...) would be translated.
12:05 Zeg9          Of course, parsing this is a bit annoying...
12:05 sapier        you need to add a parser on any sting usage within client ... quite a few changes ;-)
12:05 kahrl         Zeg9: different languages have different grammar rules, this might limit the languages you can translate to
12:06 Zeg9          It is the same thing when you call gettext, afaik.
12:09 kahrl         I think the common solution is to use sprintf. Which has security implications
12:10 sapier        kahrl that's an issue I've never seen a generic solution for
12:11 sapier        different grammar rules ;-)
12:11 Zeg9          Is it possible to have gettext looking in multiple locale directories?
12:12 sapier        no matter what we do we need to fix formspecs
12:13 sapier        there's missing escape support ... I guess I need to add this first
12:33 PilzAdam      anyone against this? https://github.com/minetest/minetest/pull/656
12:34 drizz         not I
12:35 PilzAdam      should saplings grow on dirt_with_snow?
12:35 VanessaE      no.
12:35 drizz         also, who the hell is creating clones on my server
12:35 drizz         http://servers.minetest.net/
12:36 VanessaE      drizz: looks like mine and jpappsvr01.dyndns.org have also been attacked recently
12:37 VanessaE      maybe also that 198.211.125.161 one.
12:37 drizz         heh, someone's been busy :p
12:39 drizz         02:49:29: ACTION[ServerThread]: Server: A mismatched client tried to connect from 71.164.76.159
12:39 drizz         02:49:33: ACTION[ServerThread]: Server: A mismatched client tried to connect from 71.164.76.159
12:39 drizz         02:49:37: ACTION[ServerThread]: Server: A mismatched client tried to connect from 71.164.76.159
12:39 drizz         etc.
12:39 Exio          iptables!
12:40 drizz         any forum moderator here mind looking up the ip?
12:41 PilzAdam      echo "71.164.76.159" >> ipban.txt
12:41 VanessaE      that IP also tried my server.
12:41 VanessaE      21:46:49: INFO[ServerThread]: Server: New connection: "kRs9IkyfXesZke2D8Fg" from 71.164.76.159
12:42 VanessaE      good thing I don't give interact by default
12:44 drizz         I'm surprised it actually manages to do a DoS on the server O.o
13:30 celeron55     kahrl, sapier: i have a branch somewhere (apparently not public) that implements item description translations by allowing lua to pass them as {lang="foo", lang2="bar"}
13:31 celeron55     it works by sending them all to the client, which is wrong though; the right thing would be making the client send the language at initialization and let the server choose one
13:31 Exio          i'd prefer that
13:32 Exio          so the client would be able to change the language "in-runtime" in a long-term future
13:32 sapier        me to and of course I don't want to write a table at any place where text is used
13:34 celeron55     sapier: a small lua library can be used to make it easier
13:35 celeron55     as far as the api of MT itself is concerned, that's how it would see it work
13:35 celeron55     implement eg. a function T() that will work like T("english text") and will return a table with all the translations
13:36 celeron55     Exio: it's not good because there can be tons of items and tons of languages
13:36 sapier        hmm but this only works if you don't need to concatenate basic text and dynamic text
13:38 celeron55     afaik item descriptions don't have dynamic text
13:38 Exio          celeron55: well, that is true
13:38 celeron55     and actually, maybe sending all the descriptions at once would work; they're dwarfed by all the other stuff anyway i guess... but it's really up to if anyone *actually* needs instantaneous language switching
13:38 sapier        traditionaly they don't true, but once you do menu in formspec you need to create a single textline to describe the formspec
13:39 sapier        ok this is true for any formspec not only menu
13:39 celeron55     sapier: true; that applies to formspecs
13:39 celeron55     so they need something more fancy
13:40 sapier        I see two possible options, 1) make formspec plain prior creating formspec string
13:40 sapier        2) completely redefine formspec string to support spacers
13:41 sapier        later one most likely will break compatibility
13:42 celeron55     some translation syntax inside formspec could work...
13:43 sapier        maybe but it'd be quite crazy to do this in a compatible way ... I already needed to add some metadata to add e.g. color support
13:43 sapier        using 4/5 different delimiters doesn't ease formspec parsing too ;-)
13:44 sapier        and I somehow don't want to use xml for formspecs ;-)
13:44 celeron55     well the formspec format was designed basically for implementing a chest and a furnace using a single thing; it's not really a surprise it doesn't scale well to the extent required these days
13:45 sapier        true ... I'm surprised I was able to improove it to an extent where it is possible to almost identical recreate mainmenu
13:45 sapier        almost because I'm still missing treeview support and I'd prefere not to add it :-)
13:53 drizz         mk@pyre ~ % ls ~minetest/worlds/world/players -1 | awk "length==19" | wc -l
13:53 drizz         20861
13:54 drizz         apparently having loads of user files makes the server run poorly, even after a restart
13:54 Zeg9          Nice, now you have to find a way to delete the bad ones. Shouldn't be too hard.
13:55 celeron55     drizz: that is a known bug and it's still waiting for someone to care enough to fix it
13:55 drizz         oic
13:55 drizz         Zeg9: I just did :p
13:56 drizz         they're all 19 characters, ranging from a-z, A-Z and 0-9
13:56 Zeg9          Quite stupid imo, they could have used a random length
13:57 drizz         agreed
13:57 Zeg9          (in the hope the responsible for this isn't in this room...)
13:57 iqualfragile  ^^
14:01 drizz         find . -name "???????????????????" -exec rm {} \;
14:01 drizz         ezpz :p
14:01 iqualfragile  i have once fuzzed a email-subscription field with random input… got the owner on a lot of blacklists as it seems that there are a lot of domains consisting of 6 alphanumeric characters
14:01 sapier        http://imgur.com/a/DZ7HO#0
14:02 PilzAdam      anything against this? https://github.com/minetest/minetest/pull/463
14:03 iqualfragile  sapier: nice
14:03 iqualfragile  but we should dissalow VanessaE to use colours
14:03 Zeg9          sapier: so you finally made the tabs??
14:04 sapier        somehow ;-) ... it's not exactly tabs but abuse the tab header as butonlist
14:04 Zeg9          Oh so they act as buttons, right?
14:04 sapier        similar but not exactly
14:05 Zeg9          okay, and what about using the new elements on servers?
14:05 sapier        a tabheader only has a single name and submits clicked tabnumber as value
14:05 sapier        if this is added you most likely will be able to use every element from within scriptapi too
14:06 Zeg9          Great work :p
14:06 sapier        but I haven't tested right now you should expect minor glitches that need to be fixed
14:07 Zeg9          About #12: is this multiple formspecs?
14:08 sapier        no just new formspec element box ... a plain box that can be drawn everywhere
14:09 sapier        color is configurable
14:10 sapier        "(16:04:30) PilzAdam: anything against this? https://github.com/minetest/minetest/pull/463" ... he can't stop adding things to make rebase more difficult ;-P
14:11 PilzAdam      nobody said something about a menu freeze
14:12 sapier        :-) maybe I should proclaim a menu freeze ? ;-P ... don't know if someone supports this
14:13 Exio          the menustuff where sapier is working on
14:13 Exio          is too damn big, i think he should get the "freeze" for that
14:13 sapier        btw PilzAdam differences between old and new have droppen even more as I have added support for [] in formspec text elements
14:13 Exio          unless anyone wants to rewrite the menu and write other 4000 or 5000 lines
14:15 sapier        most of the changes done to menu the last days already have been included in formspec ... some bugs have been fixed "by accident" ;-)
14:15 sapier        still there's one big problem .. localization
14:15 Zeg9          Just gettext every buttons and tabs, it should do it...
14:16 Zeg9          also checkboxes and labels, everything static
14:16 sapier        zeg9 it's lua ;-P
14:16 Zeg9          I know, do this when parsing the formspec
14:16 Zeg9          For the menu it should be enough I think
14:16 sapier        won't fix all the problems
14:17 sapier        and will have sideeffects to ALL formspecs within game
14:18 sapier        so unless at least some core developers agree with it I won't do a change that big
14:26 sapier        how do I enable gettext support?
14:26 PilzAdam      -DENABLE_GETTEXT=1
14:27 Zeg9          The only problem still there is dynamic content... Especially screenshot #1 "World: <worldname>"
14:29 sapier        this can be fixed by two labels ... ugly but working
14:29 sapier        does anyone know name of gettext development package on debian?
14:29 Zeg9          sapier: what about alignment of text inside labels?
14:30 sapier        no chance with current formspec
14:30 sapier        -DENABLE_GETTEXT=1 is just ignored
14:30 PilzAdam      sapier, on ubuntu its just gettext
14:31 sapier        I guess this is same for debian and its installed... any idea why it's ignored
14:31 PilzAdam      I also have libgettextpo0, but I dont know if its needed
14:31 sapier        me too
14:32 sapier        I wonder if I should get an error message on missing libs if -DENABLE_GETTEST=1 is specified
14:35 sapier        argh ... gettext-base != gettext :-/
14:49 sfan5         may I merge this? https://github.com/minetest/minetest/pull/729
14:50 PilzAdam      sapier, ^
14:51 hmmmm         sfan5, please don't.... sapier is reworking all of that and that merge will just cause more trouble
14:51 hmmmm         a world creation config was pretty much the next thing on my list too
14:51 sfan5         but.. but.. world seeds..
14:52 PilzAdam      anyone against using the header from minetest.net in the menu? (this does not conflict with sapier's changes)
14:54 Exio          sfan5: you can use fixed_map_seed for now
14:54 Exio          :P
14:54 sfan5         Exio: thats not user friendly
14:55 Exio          the idea is merging sapier's menu and then add a webrow.. err, i mean, user-friendly-menu later, isn't it?
14:58 RealBadAngel  you mean "menu start"? ;)
14:58 RealBadAngel  every os shall have it ;)
15:00 PilzAdam      also, should I enable the old menufooter.png?
15:01 PilzAdam      https://www.dropbox.com/s/eiw2a2tnqv9329q/Menu.png
15:02 Calinou       not really needed
15:02 PilzAdam      this would apply to all tabs except Singleplayer
15:02 PilzAdam      so, the header is enough?
15:03 Zeg9          With menufooter there are 2 "minetest" logos
15:04 PilzAdam      https://github.com/PilzAdam/minetest/commit/c67f72890c3f9154dd9124310c61ba55bc2806bd
15:04 PilzAdam      is this ok?
15:11 sapier        no no the idea is adding my menu and after that everyone can customize menu to her/his needs
15:11 PilzAdam      not everyone can do this
15:11 PilzAdam      we need a proper default menu
15:13 sapier        that's requirement for "adding my menu" ... but "proper" won't be perfect as this might be different for everyone ;-)
15:17 Zeg9          Maybe, we could make a setting where you can choose between 3 (more or less) predefined layouts. But that's not the current problem
16:06 sapier        https://github.com/minetest/minetest/pull/724 added gettext i18n to formspecs (thx to zeg9 for the idea)
16:06 sapier        it's disabled for in game formspecs to avoid undesired sideeffects but enabling is quite easy if decision is made
16:07 sapier        wooh I broke the 10k lines changed barrier
16:11 Zeg9          sapier: is it possible yet to change the in-game menu?
16:12 sapier        no I haven't touched in game
16:13 sapier        only main menu
16:13 sapier        but if you meant this menu yes it's fully customizable and there's a config option which menu to load
16:34 celeron55     sapier, kahrl: here's the stuff i played around with (the newest commit does nothing useful): https://github.com/celeron55/minetest/tree/modtranslate
16:36 sapier        celeron for the time beeing I added gettext support to guiFormSpec menu ... this works bur rewuires manual po-file creation as parser obviously can't catch any strings
16:37 sapier        and it's working only as long as client has po files
16:54 PilzAdam      here is some info about the farming mod Im working on currently: https://gist.github.com/PilzAdam/5613541
16:56 sapier        any difference to current farming mod?
16:56 PilzAdam      yea, you dont need water to make bread
16:56 sapier        ohhhhh quite important ;-P
16:56 PilzAdam      also the code is competly rewritten and should be faster
16:57 sapier        believe me it's quite difficult to bake bread without water ... I know what I'm talking about
16:58 PilzAdam      the only thing that I need is the textures for the 8 wheat growing steps; the on RealBadAngel gave me are copied from MC
16:58 sapier        hmm I think I need to add fixed size formspecs to in game mode too
16:58 PilzAdam      and the current farming mod has only 5 or so
16:58 sapier        not for menu but to make some of the elements usefull e.g. tabheader
17:09 FourFire      Hello
17:09 FourFire      I daresay it's about time someone rewrote Minecraft in C++
17:09 FourFire      Good on you guys!
17:10 PilzAdam      FourFire, please no chit-chat here
17:10 FourFire      Does Minetest integrate cubic chunks (or similar) functionality, or is that planned?
17:12 sapier        FourFire -->#minetest
17:13 FourFire      I note that the URL http://celeron.55.lt/~celeron55/minetest/   "has been reported as an attack page and has been blocked based on your security preferences."  according to My version of firefox
17:15 rubenwardy    ooh nice PilzAdam. Any item name changes relevant to me?
17:15 rubenwardy    brb
17:23 kahrl         if you look at the diagnostic page for 55.lt google basically says "we found nothing but let's mark the domain as suspicious for the hell of it"
17:38 VanessaE      iqualfragile: disallow colors?  *insert completely confused look here*
18:05 VanessaE      minor glitch:  don't bob the hand when switching between two empty slots in the hotbar
18:26 sfan5         VanessaE: s/glitch/annoyance/
18:28 VanessaE      that too
18:31 PilzAdam      fixed
18:32 VanessaE      that was easy :D
19:23 iqualfragile  VanessaE: your window decorations look horrible :P
19:23 PilzAdam      ^ true
19:26 VanessaE      window decora....you mean for my window manager?
19:26 VanessaE      :P
19:26 VanessaE      I *like* purple thank you.
19:27 iqualfragile  ironicaly your username is purple in pidgin
19:31 VanessaE      and in xchat, in fact
19:32 PilzAdam      you are blue in xchat
19:32 VanessaE      PilzAdam: as are you
19:33 VanessaE      I am purple, everyone else is blue, nick highlights are yellow
19:37 iqualfragile  -.-
19:44 VanessaE      (why are we discussing this in -dev anyway?)
19:45 iqualfragile  (i dont know)
21:27 ShadowNinja   How do I set the rectangle for GUIStaticText? setMaxSize() forces word wrapping.
21:27 ShadowNinja   (and setMinSize() is ignored)
21:29 ShadowNinja   I am trying to add resizing to this. But I think it looks best using the full width. https://github.com/minetest/minetest/pull/608
21:31 ShadowNinja   Hmmm, I think I got it but still wrapping...
21:32 kahrl         does it still wrap if you do setWordWrap(false)?
21:34 ShadowNinja   kahrl: It probably wouldn't but I want it to if it reaches the end of the screen.
21:35 kahrl         hmm not sure I understand, when do you want it to wrap and when not?
21:36 ShadowNinja   kahrl: I want it to wrap at the end of the screen. It is wrapping at the first space.
21:36 kahrl         huh, I haven't seen it do that
21:38 ShadowNinja   kahrl: This is what I am doing: http://pastebin.ubuntu.com/5685128/
21:40 ShadowNinja   Before it was screensize.X - 10
21:47 kahrl         I think I know why
21:48 kahrl         if word wrap is enabled getTextWidth returns the width of the longest *broken* line
21:48 kahrl         so at maximum the current width of the gui element
21:48 ShadowNinja   ^ I just thought that.
21:48 kahrl         which starts out at zero
21:48 kahrl         well actually max(width of the gui element, width of a single word)
21:49 kahrl         so disable word wrap before calling getTextWidth
21:49 PilzAdam      the background could also be a bit darker
21:50 ShadowNinja   Is there a function that will get the line length without accounting for that or one I can pass getText() to?
21:50 PilzAdam      80 seems good
21:50 ShadowNinja   kahrl: I whould have to disable and re-enable it. Very hacky.
21:51 kahrl         without word wrap getTextWidth does this: return font->getDimension(Text.c_str()).Width
21:52 sapier        irrlicht gui is quite limited some important features are simply missing
21:53 PilzAdam      could it happen that the box gets bigger than the screen?
21:53 kahrl         PilzAdam: yep, there would need to be a min(..., screensize.X - 10)
21:56 ShadowNinja   PilzAdam: Already thought of that.
21:57 ShadowNinja   min not declared in this scope. What file has it?
21:57 PilzAdam      https://github.com/PilzAdam/minetest/commit/05286f2ca766adec82614860be9a3c0cdd69f5ad
21:57 PilzAdam      ^ this seems to work
21:57 kahrl         oops, I mean MYMIN
21:58 ShadowNinja   Right, caps.
21:59 PilzAdam      everyone agrees to push that?
22:00 PilzAdam      (also fixed comment and positioning of ,)
22:03 kahrl         should the alpha be configurable?
22:03 kahrl         btw I tested the commit and it seems to work fine
22:03 PilzAdam      the alpha of other things isnt configureable too
22:04 kahrl         hmm
22:04 kahrl         that's true
22:05 kahrl         still I think there's people who don't care about the chat and who might be annoyed that it obstructs the screen more
22:05 kahrl         but who don't want to disable the chat entirely in case there's something important
22:06 PilzAdam      we need some kind of highlight when a server messages is printed or when someone pings the player
22:08 kahrl         we'd need to move away from IGUIStaticText and draw the chat ourselves (which wouldn't be too tough admittedly)
22:10 PilzAdam      maybe 3 steps when pressing F2: "Chat shown" -> "Chat filtered" -> "Chat hidden"
22:10 PilzAdam      where in filtered mode only server/error messages and messages containing the player name are shown
22:10 PilzAdam      so people can talk to each other in a server with 100 people on
22:11 kahrl         I like that idea
22:13 kahrl         and actual highlights could be added in the console (the backend kind of supports adding highlights to text fragments actually)
22:18 kahrl         suggestion: make the nick inside the <> bold if it's a line you said. otherwise, if the line mentions your name, make it red
22:18 kahrl         actually red might not be a good color on a black background, suggestions?
22:25 ShadowNinja   Anything to add before I push?
22:30 ShadowNinja   https://github.com/ShadowNinja/minetest/tree/chat_background
22:33 PilzAdam      I already have that ready push
22:34 ShadowNinja   PilzAdam: You wrote the same thing?
22:34 PilzAdam      like half an hour ago or so
22:35 ShadowNinja   Well you could have told me...
22:35 PilzAdam      <PilzAdam> https://github.com/PilzAdam/minetest/commit/05286f2ca766adec82614860be9a3c0cdd69f5ad
22:35 PilzAdam      <PilzAdam> everyone agrees to push that?
22:38 ShadowNinja   Oh, only saw the second one. And no, the comment has to be fixed and I didn't know the font was available in that scope.
22:39 ShadowNinja   It also looks like that line is too long.
22:39 PilzAdam      <PilzAdam> (also fixed comment and positioning of ,)
22:40 ShadowNinja   Well that is what my commit has. ;-)
22:44 ShadowNinja   Also, this can be closed. What do you think of the set_node_as idea? https://github.com/minetest/minetest/issues/726
22:51 PilzAdam      why do I get a loop if I run WStrfnd().next() on the result of a WStrfnd.next()?
22:51 PilzAdam      +infinite
22:52 PilzAdam      oh, and that whole thing is in while (!fnd.atend()) (where fnd is of the first wstring)
22:53 kahrl         are you modifying ChatBackend::addMessage?
22:53 PilzAdam      yea
22:53 kahrl         me too :P
22:53 ShadowNinja   khonkhortisan: If I read this correctly it is an irrlichit issue. You should open it there. https://github.com/minetest/minetest/issues/678
22:53 PilzAdam      https://gist.github.com/PilzAdam/5616239
22:54 kahrl         haha I'm also doing the filtered chat thing
22:54 kahrl         sorry
22:54 PilzAdam      I got everything else ready
22:54 PilzAdam      just this stupid thing wont work
22:57 ShadowNinja   PilzAdam: Well first you need to pass it playername...
22:57 PilzAdam      playername given to ChatBackend at when its created
22:58 PilzAdam      I should call it m_playername
22:59 kahrl         use line.find(narrow_to_wide(playername)) != std::wstring::npos
22:59 kahrl         I think that would work
23:00 ShadowNinja   fndname is syntax-hilighted. You might be overriding something that it needs.
23:06 PilzAdam      seems like addMessage is called in an infinte loop...
23:07 PilzAdam      oooops
23:07 * PilzAdam    better doesnt say what he had done wrong
23:07 sapier        hehe don't bother happend to each of us ;-P
23:12 sapier        btw plz test formspec menu it's as complete as possible now so it needs some testing to find the remaining glitches
23:20 PilzAdam      kahrl, are you ok with these 2 commits? https://github.com/PilzAdam/minetest/commits/master
23:21 kahrl         well I just coded the same thing as your second commit so let's compare ;)
23:23 ShadowNinja   Can someone merge IPv6 support? It seems to work and rebased easily last I checked.
23:25 ShadowNinja   As long as you don't include the menu checkbox or have sapier rebase his menu it shouldn't add issues for anything else.
23:26 PilzAdam      ShadowNinja, the problem is that I have no way to test this, so I dont feel very secure when merging it
23:26 kahrl         PilzAdam: https://github.com/kahrl/minetest/commit/a330a9c1dd901cd3e3822dc1d01b519a6324c914
23:27 ShadowNinja   Well it didn't seem to break anything so no harm. If there is an issue you can fix it.
23:27 kahrl         (I borrowed your commit message :D)
23:28 PilzAdam      kahrl, I store the playername in wstring so I dont have to call narrow_to_wide() everytime
23:28 PilzAdam      oh, you have colors
23:28 PilzAdam      I guess you won
23:29 kahrl         heh
23:29 Exio          what game are we playing?
23:29 PilzAdam      c++
23:30 kahrl         the best move is to play
23:32 PilzAdam      hmm.. the colors are only in the chat console
23:33 PilzAdam      any  chance to get the normal chat colored?
23:33 kahrl         I figured it might be too in-your-face in the HUD chat
23:33 kahrl         also laziness ;)
23:34 kahrl         IGUIStaticText only supports one color afaik
23:34 PilzAdam      are you sure you want color.setRed(0); instead of color.setBlue(0); & color.setGreen(0)?
23:34 kahrl         nah those were just test colors
23:35 kahrl         the solution for the normal chat would be to draw it ourselves, but that would mean we'd also have to handle word wrapping ourselves
23:35 PilzAdam      I would be fine with having the colors in the normal chat; its sometimes unreadable without colors
23:36 kahrl         are those colors readable against sky/cave/landscape background?
23:37 PilzAdam      if not then we can make the background darker
23:37 PilzAdam      is it possible to use bold instead of colors in the normal chat?
23:37 khonkhortisan ShadowNinja, I opened a short one
23:38 kahrl         I looked into that (I originally wanted to draw the name bold on CHATLINE_SELF)
23:38 kahrl         it is possible by loading a bold font
23:51 kahrl         I think I'll go ahead and use irrlicht's breakText() to draw the recent chat without using CGUIStaticText
23:52 kahrl         "use" as in write a similar function that doesn't do as much stuff (breakText is over 200 lines long) but is sufficient
23:55 kahrl         I'll do that tomorrow