--- Log opened Sat Feb 20 00:00:22 2016
--- Day changed Sat Feb 20 2016
20160220 00:00:22< vultraz> zookeeper: we haven't activated HW acceleration yet
20160220 00:08:13-!- louis94 [~~louis94@91.178.242.231] has quit [Quit: Konversation terminated!]
20160220 00:20:08-!- gfgtdf [~chatzilla@x50ab6fcd.dyn.telefonica.de] has joined #wesnoth-dev
20160220 00:25:01-!- Appleman1234 [~Appleman1@KD119104018024.au-net.ne.jp] has quit [Ping timeout: 250 seconds]
20160220 00:38:19< gfgtdf> vultraz: i'd say this could be ported to giu2
20160220 00:39:36< vultraz> gfgtdf: ok, i'll make it gui2
20160220 00:39:48< vultraz> gfgtdf: btw when can you undo recruiting/recalling?
20160220 00:39:54< vultraz> gfgtdf: bc i tried to in master and you can't
20160220 00:39:59< gfgtdf> vultraz: yes you can
20160220 00:39:59< vultraz> is it only specific circumstances 
20160220 00:40:06< gfgtdf> vultraz: thats what these files do
20160220 00:40:26< gfgtdf> vultraz: well you can not undo them if they eigher uncovered shroud or called the rng .
20160220 00:40:48< gfgtdf> vultraz: this specialyl includes all units with random traits
20160220 00:41:23< vultraz> is that the default for recruits?
20160220 00:41:33< gfgtdf> vultraz: how do you mean ?
20160220 00:41:55< vultraz> I just started the first scenario of Two Brothers which has no fog or shroud, recruited a unit, but couldn't undo that
20160220 00:42:10< gfgtdf> vultraz: well yes most likeley it has rndom traits
20160220 00:42:51< vultraz> bc if every time you recruit a unit it has random traits, how would you ever undo the recruit unless it was a unit with no traits
20160220 00:44:09< gfgtdf> vultraz: some unittypes like spaerman gain 2 randomly picked traits from 'strong', 'quick', 'intelligent' and 'ressilent'. Other unittypes like ghose have no such traits
20160220 00:44:18< gfgtdf> vultraz: ghosts*
20160220 00:48:30< vultraz> ahh ok
20160220 00:49:40< gfgtdf> vultraz: if you recruit a unit like a spearman you can never unto that
20160220 00:57:56-!- oldlaptop [~quassel@192.183.14.123] has joined #wesnoth-dev
20160220 01:00:09-!- irker938 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20160220 01:00:09< irker938> wesnoth: Charles Dang wesnoth:master 100e46347308 / src/actions/ (undo_recall_action.cpp undo_recruit_action.cpp): Convert two GUI1 popups to GUI2 https://github.com/wesnoth/wesnoth/commit/100e463473081c0bc059024a2deb5a57aa41fcf7
20160220 01:00:25< vultraz> gfgtdf: ^
20160220 01:02:31< gfgtdf> vultraz: looks good
20160220 01:02:41< shadowm> The word is 'either', by the way, not 'eigher'.
20160220 01:14:04< irker938> wesnoth: Charles Dang wesnoth:master da86bcb7a2ad / src/ (13 files in 9 dirs): Cleaned up some GUI1 includes https://github.com/wesnoth/wesnoth/commit/da86bcb7a2ada3301cde23450aef605a1eda635e
20160220 01:15:08< vultraz> yay for testing the removal of each individual include! :P
20160220 01:16:32< vultraz> I tried running python's cppclean but that tool sucks. 
20160220 01:19:00< vultraz> gfgtdf: did you say you were going to work on porting the hotkeys dialog to gui2?
20160220 01:20:41< gfgtdf> vultraz: i tried, but the problem was that it seems nontrivial to implement the 'pres the hotkey you want to bind the hotkey to (esc to abort)' dialog to gui2 since it has an additional levle of indirection to the sdl events, this is specailyl true for joystik events afaik.
20160220 01:21:21< vultraz> hm, yes, that would be a problem..
20160220 01:22:00-!- Appleman1234 [~Appleman1@KD119104007035.au-net.ne.jp] has joined #wesnoth-dev
20160220 01:34:11-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has quit [Quit: Connection closed for inactivity]
20160220 01:50:47-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit []
20160220 02:02:57-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Remote host closed the connection]
20160220 02:27:26< vultraz> I've always wondered why the minimap is so blurry at small map sizes
20160220 02:27:37< vultraz> it's almost like it's scaled down and then back up
20160220 02:53:15-!- gfgtdf [~chatzilla@x50ab6fcd.dyn.telefonica.de] has quit [Quit: ChatZilla 0.9.92 [Firefox 44.0.2/20160210153822]]
20160220 02:59:14-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20160220 03:12:16< irker938> wesnoth: Wedge009 wesnoth:master 964d7bb09f8f / src/wesnoth.cpp: Restore loading of user preferences at start-up (bug #24427) https://github.com/wesnoth/wesnoth/commit/964d7bb09f8f6438a28850061e0d0517a5e63ec3
20160220 03:12:18< irker938> wesnoth: gfgtdf wesnoth:master fe3709dba7b5 / src/wesnoth.cpp: Merge pull request #606 from Wedge009/bug_24427_fix https://github.com/wesnoth/wesnoth/commit/fe3709dba7b5f6ea6bf3f265aa9fed5391ba179a
20160220 03:30:13-!- prkc [~prkc@gateway/vpn/privateinternetaccess/prkc] has quit [Remote host closed the connection]
20160220 04:22:05< vultraz> Aginor: so I'm thinking, once we have ogl support, would the entire gamemap be rendered as one surface? because I'm think if so, then we could just scale that surface down for the minimap instead of rendering each minimap 'tile'
20160220 05:07:57< vultraz> celticminstrel: could you help me understand structs?
20160220 05:08:07< vultraz> i keep reading about them and looking at them but they don't really make sense
20160220 05:08:41< vultraz> are they kinda like classes?
20160220 05:08:58< vultraz> they seem to be able to have constructors and private members 
20160220 05:22:14< shadowm> vultraz: In C++, structs are classes for all intents and purposes, except that their default member visibility is public instead of private.
20160220 05:24:55< shadowm> Also, scaling down an 80x80x72x72 surface ought to be expensive no matter whether it's the CPU or the GPU doing the job.
20160220 05:27:20< shadowm> *member and base class visibility
20160220 05:29:28< shadowm> But in reality people tend to treat structs and classes differently in design. I'll let someone else explain.
20160220 05:29:39< shadowm> *in API design.
20160220 05:31:55< shadowm> And if you want to learn C++ as opposed to just write code for Wesnoth, I'd also suggest you research `union` types.
20160220 05:32:33< shadowm> Yes, that is homework. Teachers don't dispense this stuff just because they want to give you bad grades, you know.
20160220 05:32:54< shadowm> At least I'm charitable enough as to make it optional.
20160220 05:37:02< vultraz> ok, what would be the practical application of such a type
20160220 05:40:18< vultraz> like, to take the example on cppref, why would I choose a union with a string and vector over a struct with the same members
20160220 05:42:07< celticminstrel> They are exactly like classes.
20160220 05:43:09< celticminstrel> Typically, "struct" is used for a class that's primarily just a bag of data - ie, it just has a bunch of public data members and not much else.
20160220 05:43:21< celticminstrel> Putting a string in a union is dangerous.
20160220 05:43:39< celticminstrel> Putting a vector in a union is also dangerous.
20160220 05:43:54< celticminstrel> Putting any class type in a union is dangerous.
20160220 05:45:06< celticminstrel> A union is completely different from a struct or class.
20160220 05:45:17< celticminstrel> It's in the name.
20160220 05:46:19< celticminstrel> (Strings and vectors can be safely put in a union, but you need to do some extra work using stuff like placement new.)
20160220 06:10:50-!- celticminstrel is now known as celmin|sleep
20160220 06:12:47-!- irker938 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20160220 06:53:52-!- Kwandulin [~Miranda@p200300760F0BC5EA197BF73357FE461C.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160220 08:01:44-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20160220 08:01:45-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Read error: Connection reset by peer]
20160220 09:21:20-!- iceiceice [~chris@wesnoth/developer/iceiceice] has joined #wesnoth-dev
20160220 09:22:17< iceiceice> vultraz, the correct answer is that you should never use a union type in modern C++ code
20160220 09:22:30< iceiceice> its only purpose is compatibility with C, and really obscure forms of type punning
20160220 09:22:56< iceiceice> if you need something liek that in C++, you should use a "tagged union" or a variant type
20160220 09:23:40-!- aidanhs [~aidanhs@81.4.110.234] has quit [Ping timeout: 255 seconds]
20160220 09:25:01-!- aidanhs [~aidanhs@81.4.110.234] has joined #wesnoth-dev
20160220 09:37:51< vultraz> iceiceice: so why did shadowm tell me I should learn of it
20160220 09:38:04< iceiceice> you still should i guess,
20160220 09:38:09< iceiceice> if only because SDL will sometimes pass you union types
20160220 09:38:11< iceiceice> since it is a C lib
20160220 09:38:19< iceiceice> SDL event is a union type for instance
20160220 09:38:34< iceiceice> or maybe contains them, i would have to look it up again :)
20160220 09:38:57< vultraz> I should probably get something straight... is C++ superior to C? For some reason I've always had the impression that C is 'old' and 'worse' than the 'new' C++
20160220 09:39:22< iceiceice> C++ is a different language built on C
20160220 09:39:33< iceiceice> whether it is superior is highly contentious
20160220 09:39:47< iceiceice> they are both basically the most widely used languages depending on how you cont
20160220 09:39:49< iceiceice> *count
20160220 09:40:15< iceiceice> C is relatively stripped down and primitive
20160220 09:40:29< iceiceice> C++ adds a bunch of stuff on top of that, but they also wanted to be largely compatible
20160220 09:40:40< iceiceice> the thing is like,
20160220 09:40:46< vultraz> ok, so C is not some obsolete thing 
20160220 09:40:48< iceiceice> most of the infrastructure is written in C
20160220 09:40:52< iceiceice> the entire linux kernel is written in C
20160220 09:40:57< iceiceice> a good chunk of OS X is written in C
20160220 09:41:06< iceiceice> and allt he useful libraries are writtne in C
20160220 09:41:33< iceiceice> if you are making a programming language, and it can't talk to C *easily* then its a major drawback, for many kinds of programming
20160220 09:41:59< iceiceice> there was some quote i remember that Bjarne Stroustrup wrote like, "inside C++ there is a much smaller and simpler language that is trying to get out"
20160220 09:42:19< iceiceice> "but if I had made that language instead of C++, it would have only been some obscure language with a cult following"
20160220 09:43:00< iceiceice> usually, you should be trying to write within the clean, nice idiomatic C++ stuff
20160220 09:43:17< iceiceice> and avoid doing the parts of C that are considered obsoleted
20160220 09:43:24< iceiceice> but sometimes you gotta do what you gotta do
20160220 09:44:06< vultraz> so, for example, I should avoid C-style function pointers?
20160220 09:44:31< iceiceice> you can't always avoid that i guess
20160220 09:44:37< iceiceice> and that's not really obsoleted in C++ either
20160220 09:45:03< iceiceice> you can use the boost::function stuff instead, which is more featureful, but its also pretty different and sometimes slower
20160220 09:45:27< iceiceice> usually in C++ when they obsolete some C thing, its like they make it strictly safer, with no runtime penatly
20160220 09:45:42< iceiceice> for instance like, you should not use printf
20160220 09:45:55< iceiceice> (i think most people would agree with this)
20160220 09:46:46< iceiceice> because its very primitive, you have to use all these obscure format specifiers which sometimes have portability problems, and if you get the specifier wrong you just get undefined behavior
20160220 09:46:48< vultraz> what is printf?
20160220 09:47:14< iceiceice> in C++ you've probably done stuff like
20160220 09:47:19< iceiceice> int my_number = 5;
20160220 09:47:29< iceiceice> std::cout << "The number is " << my_number << std::endl;
20160220 09:47:42< iceiceice> "std::cout" is C++, its not in C
20160220 09:47:48< iceiceice> the normal way you do that in C is something like
20160220 09:48:01< loonycyborg> I think the much smaller language stroustrup was talking about that is inside C++ isn't even procedural 
20160220 09:48:06< vultraz> I usually use cerr
20160220 09:48:09< iceiceice> printf("The number is %d\n", 5);
20160220 09:48:10< vultraz> but continue
20160220 09:48:10< loonycyborg> It's more like haskell :P
20160220 09:48:35< iceiceice> loonycyborg, you might be right, maybe i should dig up the quote
20160220 09:48:38< vultraz> iceiceice: that looks more like lua
20160220 09:48:47< iceiceice> lua derives a lot of stuff from C
20160220 09:48:49< vultraz> string.format, specifically 
20160220 09:48:54< iceiceice> but in lua that is very safe
20160220 09:48:55< iceiceice> like,
20160220 09:49:02< iceiceice> if you passed it a table instead of 5
20160220 09:49:07< iceiceice> or you passed it 5.5
20160220 09:49:12< iceiceice> it would give you an error
20160220 09:49:22< iceiceice> in C it will just crap some bytes on the screen probably, and / or crash
20160220 09:49:41< iceiceice> the specifiers and the types just have to match and if you mess it up its on you
20160220 09:50:30< iceiceice> some people would say that printf is still better then cout, i mean that is also an argument
20160220 09:50:33< iceiceice> because it can be faster
20160220 09:50:49< iceiceice> but then i would say like, you should use a C++11 version of printf
20160220 09:50:58< iceiceice> because you can use variadic templates to make it typesafe
20160220 09:51:35< iceiceice> idk
20160220 09:51:41< iceiceice> theres always something better you can use than printf
20160220 09:51:48< iceiceice> or in c++11 you can often use "std::to_string"
20160220 09:52:00< iceiceice> which basically does the same thing as printf, but it figures out the format specifier for you at compile time
20160220 09:52:26< loonycyborg> It always seemed to me that lua goes too far in implicit type conversion
20160220 09:52:49< loonycyborg> you can do stuff like adding strings and numbers
20160220 09:52:55< loonycyborg> without explicit conversion
20160220 09:53:56< iceiceice> yeah i mean
20160220 09:54:05< iceiceice> i think not many people like that feature
20160220 09:54:14< iceiceice> idk
20160220 09:54:26< iceiceice> at least it doesnt have undefined behavior though or something
20160220 09:54:56< iceiceice> one thing that always amused me also though was like
20160220 09:55:03< iceiceice> the people who say that printf is superior because, it works better with i18n
20160220 09:55:13< iceiceice> because you can take the formatter string and mark it as translatable
20160220 09:55:26< iceiceice> so your translators will translate "I have %d dogs" or something
20160220 09:55:55< iceiceice> while if you have like `std::cout << "I have " << dogs << " dogs"` its kind of a hot mess for the translators
20160220 09:55:58< loonycyborg> %d things wouldn't work too well in some language
20160220 09:56:27< iceiceice> i always thought it was super squirrely though to have the formatter strings be translated
20160220 09:56:33< loonycyborg> like some languages require words to have different form depending on their syntactical role
20160220 09:56:36< iceiceice> because like what if the translators mess up the symbols?
20160220 09:56:40< iceiceice> and you get a crash / ub?
20160220 09:57:07< loonycyborg> for example in russian
20160220 09:57:46< loonycyborg> "у меня есть %d собак" wouldn't work for very low numbers :P
20160220 09:58:01< vultraz> so like, one area I've always found really messy about c++ (just from the look of it) is all this integer stuff. unsigned vs signed and precision values and doubles and longs and Unit8s and Unit16s and Uint32s, etc, etc. Is C also like that?
20160220 09:58:27< iceiceice> loonycyborg, yeah i guess that like
20160220 09:58:34< iceiceice> the gettext folks have thought abot this to death
20160220 09:58:49< iceiceice> you are supposed to provide like n different translations of that
20160220 09:58:52< iceiceice> for the n differnet plural forms
20160220 09:58:53< loonycyborg> they have some solution to stuff like that
20160220 09:59:13< iceiceice> i wonder if there's ever been a hack attack where like
20160220 09:59:24< iceiceice> someone gives you some translated content file for some application
20160220 09:59:42< iceiceice> and it messes up a format specifier in a printf used in translation, and causes buffer overflow and takes over your machine
20160220 09:59:47< vultraz> why do there have to be half a dozen different data types just for some specific byte size
20160220 10:00:37< iceiceice> well
20160220 10:00:43< iceiceice> ok there's two different questions there really
20160220 10:00:53< loonycyborg> it's portability issues
20160220 10:00:54< iceiceice> one is, do you need to have uint8, uint16, uint32, uint63 etc.
20160220 10:00:57< iceiceice> *64
20160220 10:01:03< iceiceice> and the answer is yes, you really do need all of those types
20160220 10:01:04< loonycyborg> there's buttload of architectures
20160220 10:01:22< loonycyborg> and each of them has own properties of arithmetic types
20160220 10:01:31< iceiceice> i think i never fully understood though why the standard doesnt mandate like, size of int, size of float
20160220 10:01:45< loonycyborg> so applications make own type aliases
20160220 10:01:48< iceiceice> i mean realistically the programmer usually makes assumptions about what they are
20160220 10:01:56< loonycyborg> to chose type most advantageous to them on each arch
20160220 10:02:09< iceiceice> yeah but how many programs will really work if int is 2 bytes for sintance
20160220 10:02:17< vultraz> when I code something I usually just say, "it's a number? it must be an int"
20160220 10:02:49< loonycyborg> different types encode different desired properties
20160220 10:02:57< vultraz> but then sometimes I see code examples where you fetch an index and it's stored as unsigned int
20160220 10:03:04< vultraz> sometimes it's a regular int
20160220 10:03:30< loonycyborg> e.g. if you absolutely want int type of 32bits you use uint32_t
20160220 10:03:41< loonycyborg> if you don't care about amounts of bits
20160220 10:03:54< loonycyborg> but want natural word size for current arch
20160220 10:04:01< loonycyborg> then you use plain int
20160220 10:04:01< vultraz> I guess I just haven't written anything where you have to care about bit size
20160220 10:04:21< iceiceice> the thing is like
20160220 10:04:33< iceiceice> if you get overflow on the int its UB
20160220 10:04:51< iceiceice> so you have to be cognizant of the limit usually?
20160220 10:05:06< iceiceice> i mean you are not goint o use "std::numeric_limits<int>::max()" tests everywhere, its so ugly
20160220 10:05:18< iceiceice> vultraz, you should read this probably:
20160220 10:05:18< iceiceice> http://blog.llvm.org/2011/05/what-every-c-programmer-should-know.html
20160220 10:05:30< iceiceice> thats one of the best articles i read about those kinds of issues
20160220 10:05:55< loonycyborg> a lot of applications would never notice whether word type is 16bit, 32bit or 64-bit
20160220 10:06:20< vultraz> how come in languages like lua I can just say "local some_number = 209890" and it's totally ok
20160220 10:06:58< vultraz> is it because it's higher up?
20160220 10:07:12< vultraz> and the lower you get the more you have to worry about this?
20160220 10:07:23< iceiceice> loonycyborg, do you know if rust has a "word size int" type?
20160220 10:07:29< vultraz> I've heard that c++ is a 'middle-tier' language
20160220 10:07:39< loonycyborg> I don't know anything about rust
20160220 10:07:44< iceiceice> i guess it must
20160220 10:07:50< iceiceice> :/
20160220 10:07:58< iceiceice> i am really meaning to try to learn rust one of these days
20160220 10:08:51< iceiceice> vultraz, in lua, the typing is dynamic
20160220 10:08:57< iceiceice> it represents decimal numbers with double precision
20160220 10:09:38< iceiceice> i actually dont know if it represents integers as flat ints
20160220 10:09:53< iceiceice> and intelligently switches the represntation
20160220 10:09:58< iceiceice> or if numbers are always floating point
20160220 10:09:59< loonycyborg> It represents all numbers with floats
20160220 10:10:09< iceiceice> hmm ok
20160220 10:10:11< loonycyborg> internally
20160220 10:10:56< vultraz> iceiceice: hm... are languages like haskell and rust on the lua-ish tier of handling these things?
20160220 10:11:02< vultraz> or are they more like c++
20160220 10:11:10< vultraz> bc tbh I like not having to worry about this crap :P
20160220 10:11:16< iceiceice> haskell is very different from all of thes language imo
20160220 10:11:18< loonycyborg> haskell has very strict type system
20160220 10:11:24< iceiceice> rust is really like, C++ killer-type language
20160220 10:11:30-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20160220 10:11:42< iceiceice> it is definitely in the same domain as C++ and sort of cherry picks many of the good things without bringing in the bad
20160220 10:11:50< iceiceice> but its also not as mature as C++
20160220 10:12:45< iceiceice> rust is also very original though,
20160220 10:12:55< iceiceice> there's a lot of stuff it does to make it easier to write bug-free code
20160220 10:12:56< vultraz> but facebook uses haskell
20160220 10:13:09< iceiceice> like above and beyond what other languages do
20160220 10:13:26< loonycyborg> but haskell's default int types are derived from C and have similiar properties
20160220 10:14:10< iceiceice> i think facebook is a huge company that uses a lot fo different technologies
20160220 10:15:03< loonycyborg> it seems haskell is the first pure-functional language that is borerline usable in production :P
20160220 10:15:13< loonycyborg> *borderline
20160220 10:15:31< zookeeper> vultraz, regardless of performance, you shouldn't expect the scaled-down full map to look good as a minimap.
20160220 10:16:04< vultraz> zookeeper: other games do it, somehow
20160220 10:16:08< vultraz> have nice minimaps, that is
20160220 10:16:30< zookeeper> uh, yes, obviously?
20160220 10:18:10< iceiceice> loonycyborg, the thing with haskell that i remmber was like,
20160220 10:18:23< iceiceice> ok suppose you study it long enough that you can do simple IO using monads and such
20160220 10:18:35< iceiceice> now like, you want to interact wiht C libraries...
20160220 10:18:46< iceiceice> so what, you have to like make a monad for every single library?
20160220 10:18:52< vultraz> zookeeper: but we do not have a nice minimap
20160220 10:18:54< vultraz> at all
20160220 10:18:57< vultraz> that's the issue :)
20160220 10:19:12< vultraz> it has weird patterns that show up if you repeat a terrain 
20160220 10:19:21< vultraz> it has jagged hex edges
20160220 10:19:26< vultraz> and horrible resolution on small maps
20160220 10:19:27< vultraz> etc
20160220 10:19:33< iceiceice> i mean it seems to have the familiar problem, like,
20160220 10:19:46< loonycyborg> monads represent C style control flow
20160220 10:19:53< zookeeper> vultraz, name me a mainline map which exhibits the problem
20160220 10:19:59< iceiceice> if your distance from C is too far then a lot of thing become harder than they would be
20160220 10:20:06< loonycyborg> so it's natural that they end up in wrappers for C libraries
20160220 10:20:32< vultraz> zookeeper: httts1
20160220 10:20:48< iceiceice> i mean it sucks but thats just kind of the infrastructure that we have
20160220 10:21:05< loonycyborg> I think there are needed higher order wrappers
20160220 10:21:10< loonycyborg> purely functional ones
20160220 10:22:07< iceiceice> mm 
20160220 10:22:16< iceiceice> in theory i wrote some haskell programs in a class like 12 years ago
20160220 10:22:20< iceiceice> but lookign at it now...
20160220 10:22:27< iceiceice> i did not realize that haskell had exceptions
20160220 10:22:33< iceiceice> i thought they didnt do that
20160220 10:23:50< zookeeper> vultraz, oh _those_ kind of patterns
20160220 10:23:56< vultraz> yes
20160220 10:24:06< vultraz> the hell are they
20160220 10:25:16< loonycyborg> what haskell IO monads are used for is basically implementing the Command pattern
20160220 10:25:51< loonycyborg> so we should eliminate use of IO monads everywhere we don't need Command pattern
20160220 10:26:19< zookeeper> vultraz, well, we could just force the "toggle minimap terrain coding" thing on... :p
20160220 10:26:31< loonycyborg> they shouldn't be used unless we want to encode particular sequence of actions
20160220 10:29:12< vultraz> zookeeper: i don't really like the color coding that uses
20160220 10:29:31< vultraz> zookeeper: plus there;s still the border hex issue, where the edges appear jagged 
20160220 10:31:00< zookeeper> i see no jaggies. obviously the hexes are drawn as blocks as they should, but they're not jagged.
20160220 10:32:18< vultraz> no i meant
20160220 10:32:33< vultraz> the edges are hex-shaped
20160220 10:32:45< vultraz> and this is especially obvious on small maps
20160220 10:35:04< vultraz> we don't draw border hexes
20160220 10:37:47< zookeeper> "we don't draw border hexes" what does that even mean?
20160220 10:38:32< zookeeper> and what else would the edges be but hex-shaped? they're hexes, they need to be shaped like hexes so you can actually tell something about the terrain by looking at the minimap
20160220 10:40:04< vultraz> i mean we should do like we do in the main map and draw a 'half-hex' to fill in the gaps
20160220 10:40:17< vultraz> so the minimap is always a rectangle
20160220 10:42:27< zookeeper> oh you mean like if you have snow at the map edge, there's small white-black jaggies at the minimap edge
20160220 10:42:29< zookeeper> sure, sure
20160220 10:42:48-!- zombah [~zombah@2a02:28:3:1:214:4fff:fe47:5920] has joined #wesnoth-dev
20160220 10:56:58< vultraz> well i was specifically referring to how it looks at smal resolutions 
20160220 10:57:00< vultraz> er
20160220 10:57:02< vultraz> small map sizes
20160220 10:57:03< vultraz> like this https://www.dropbox.com/s/q1rmhpdvyqpcr0m/minimapbad.PNG?dl=0
20160220 11:05:00< vultraz> also, random note
20160220 11:05:12< vultraz> drawing in the editor is really fast now O_O
20160220 11:05:32< vultraz> sdl2 is great
20160220 11:08:37< vultraz> hm
20160220 11:08:45< vultraz> actually it's faster the smaller the map is
20160220 11:08:47< vultraz> interesting
20160220 11:11:02< vultraz> hm
20160220 11:11:12< vultraz> the colors for the minimap terrain coding could use improvement
20160220 11:11:15< vultraz> snow looks like water
20160220 11:11:22< zookeeper> and i was under the impression that sdl2 made things a bit slower
20160220 11:12:13< vultraz> zookeeper: oh?
20160220 11:24:23-!- mjs-de [~mjs-de@x4db5c8ca.dyn.telefonica.de] has joined #wesnoth-dev
20160220 11:28:26-!- prkc [~prkc@46.166.190.184] has joined #wesnoth-dev
20160220 12:34:18-!- Kwandulin [~Miranda@p200300760F0BC5EA197BF73357FE461C.dip0.t-ipconnect.de] has quit [Ping timeout: 250 seconds]
20160220 12:44:07-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160220 12:45:17-!- aeonchild is now known as dlihcnoea
20160220 13:14:10-!- Kwandulin [~Miranda@p200300760F0BC500197BF73357FE461C.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160220 13:25:12-!- louis94 [~~louis94@91.178.241.211] has joined #wesnoth-dev
20160220 13:56:01-!- dlihcnoea is now known as enchi
20160220 14:26:32-!- Kwandulin [~Miranda@p200300760F0BC500197BF73357FE461C.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160220 14:36:57-!- nkr [Elite14718@gateway/shell/elitebnc/x-vpxpgyvzbfxngwpy] has joined #wesnoth-dev
20160220 15:01:15-!- zombah [~zombah@2a02:28:3:1:214:4fff:fe47:5920] has quit [Quit: Leaving]
20160220 15:25:04-!- Kwandulin [~Miranda@p200300760F0BC5007CA8A41547A38CFC.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160220 15:33:54-!- fr_rrr [~Necrospor@unaffiliated/necrosporus] has joined #wesnoth-dev
20160220 15:37:11-!- Necrosporus_ [~Necrospor@unaffiliated/necrosporus] has quit [Ping timeout: 250 seconds]
20160220 15:41:29-!- louis94 [~~louis94@91.178.241.211] has quit [Ping timeout: 240 seconds]
20160220 15:49:23-!- louis94 [~~louis94@91.178.241.211] has joined #wesnoth-dev
20160220 15:55:14-!- mjs-de [~mjs-de@x4db5c8ca.dyn.telefonica.de] has quit [Remote host closed the connection]
20160220 16:04:31< celmin|sleep> vultraz: zookeeper: i don't really like the color coding that uses
20160220 16:04:32< celmin|sleep> ...so do you want to make those colours customizable too? :|
20160220 16:09:45< zookeeper> of course not. but of course the colors can be changed.
20160220 16:10:25< zookeeper> although i have no idea how those are determined. since they're not in terrain.cfg, i must presume that they're an average of the tile or something.
20160220 16:16:56< celmin|sleep> Aren't they the colours of the terrain icons?
20160220 16:23:58< celmin|sleep> They really aren't in terrain.cfg.
20160220 16:24:12< celmin|sleep> Maybe they could be derived from the icon_image= key though? I dunno.
20160220 16:52:45-!- celmin|sleep is now known as celticminstrel
20160220 17:08:45< vultraz> celticminstrel: in the same file the orbs' colors are in
20160220 17:09:00< vultraz> lines 150 and below
20160220 17:09:13< vultraz> celticminstrel: maybe you could tweak them, since I'm not exactly sure how the range works
20160220 17:09:53< celticminstrel> That file is only 110 lines long.
20160220 17:10:00< zookeeper> which file?
20160220 17:10:16< celticminstrel> The orb's colours are in game_config.cfg
20160220 17:10:41< vultraz> oh
20160220 17:10:51< vultraz> i meant data/core/team_colors.cfg
20160220 17:10:57< vultraz> team-colors
20160220 17:11:28< zookeeper> what are you both on?
20160220 17:11:46< celticminstrel> Oh, those extra colour ranges are used for the terrain background colours?
20160220 17:12:00< zookeeper> ohh there it is
20160220 17:12:02< zookeeper> i missed it, duh
20160220 17:12:06< vultraz> I'm guessing
20160220 17:12:17< vultraz> the code seems to indicate so
20160220 17:12:30< zookeeper> humm. why are they ranges, when only a single color is used?
20160220 17:12:39< vultraz> I don't know
20160220 17:12:49< vultraz> that's why I asked celticminstrel to look into it :)
20160220 17:12:59< celticminstrel> I think I know.
20160220 17:13:06< zookeeper> "to recolor the background of terrain type icons" <- like, what does that even mean
20160220 17:14:31< celticminstrel> The little icons shown on the sidebar to identify the terrain types.
20160220 17:14:39< zookeeper> anyway, considering how those values have never seen any revisions, it's a fair bet that they could be improved
20160220 17:14:50< celticminstrel> Those seem to be white with transparent background though...
20160220 17:15:09< nkr> hey is there a project roadmap?
20160220 17:16:07< zookeeper> celticminstrel, right, yeah. i wonder why they are transparent, and not just... you know, whatever color one wants them to be.
20160220 17:17:24< celticminstrel> Also, data/core/images/terrain/symbols seems to be a mixed hodgepodge of stuff that would probably be better split into two separate directories.
20160220 17:19:21< celticminstrel> Since the colours are ranges, why don't the terrain icons actually use magenta?
20160220 17:19:47< vultraz> again, don't know
20160220 17:19:57< vultraz> possibly we're looking in the totally wrong place
20160220 17:20:48< vultraz> hm
20160220 17:20:50< celticminstrel> Grepping the code probably won't help. If we're right, it probably uses the base terrain's ID to look up the correct colour range.
20160220 17:20:52< vultraz> nope, it's the right place
20160220 17:21:08< celticminstrel> ...so if you made a terrain with id=blue, you'd get a blue background.
20160220 17:21:17< celticminstrel> Oh, wait.
20160220 17:21:19< vultraz> it changes the displayed color if you edit those
20160220 17:21:52< celticminstrel> If you make a completely new terrain that's not based on anything and you assign it an icon, it's shown on a magenta background.
20160220 17:21:59< celticminstrel> Is that background also an image somewhere?
20160220 17:22:34< vultraz> seems the terrain minimap colors only use the first color
20160220 17:23:01< vultraz> celticminstrel: are you talking about the minimap?
20160220 17:23:19< celticminstrel> No, I'm talking about where it shows the selected terrain in the sidebar.
20160220 17:23:41< celticminstrel> It has one or more icons followed by the name of the terrain, I think. And the coordinates.
20160220 17:23:59< vultraz> oh yeah, the icon is magenta which is tced
20160220 17:23:59< celticminstrel> Or maybe coordinates and no name. Something like that.
20160220 17:24:02< vultraz> i believe
20160220 17:24:15< celticminstrel> Is that icon an image somewhere?
20160220 17:24:46< celticminstrel> ...
20160220 17:24:58< celticminstrel> The terrain_group image set is duplicated in two places. Why?
20160220 17:25:04< celticminstrel> ......
20160220 17:25:09< vultraz> what
20160220 17:25:11< celticminstrel> So is the terrain type icons set. o.O
20160220 17:25:15< vultraz> it;s duplicated
20160220 17:25:17< vultraz> O_O
20160220 17:25:27< celticminstrel> (The terrain_group set is what the editor uses.)
20160220 17:25:43< celticminstrel> Yeah, they're in data/core/images/terrain/symbols and also in images/icons/terrain.
20160220 17:26:30< vultraz> yes I just noticed that too
20160220 17:26:35 * celticminstrel wonders whether images/icons/profiles will ever be finished and used.
20160220 17:26:37< vultraz> maybe it's so both umc and core can use them
20160220 17:27:00< vultraz> celticminstrel: well it is finished
20160220 17:27:08< vultraz> technically
20160220 17:27:14< celticminstrel> I don't consider it finished.
20160220 17:27:30< celticminstrel> ...there are alignment icons there which are amusingly simplistic.
20160220 17:27:43< celticminstrel> I'm not sure what I think of them.
20160220 17:27:56< celticminstrel> Anyway, the ones that are composed of text aren't really finished.
20160220 17:29:08< vultraz> oh 
20160220 17:29:10< vultraz> right
20160220 17:29:11< vultraz> ok
20160220 17:29:17< celticminstrel> So that means... ambush, concealment, cures, drain, heals, nightstalk, petrify, plague, poison, regenerate, skirmisher, slow, submerge, teleport.
20160220 17:29:26< celticminstrel> ...what's local_tod for?
20160220 17:29:28< vultraz> ok you're right
20160220 17:29:31< vultraz> no idea
20160220 17:29:52< Ravana_> [time_area] I guess
20160220 17:29:57< celticminstrel> Some of them I'm surprised to be unfinished - slow obviously should be a snail, heals could be an ankh or something.
20160220 17:31:48< celticminstrel> I think it's probably images/buttons/icon-base-{16,32}.png that's used for the terrain icons.
20160220 17:32:09< vultraz> ah yes
20160220 17:32:12< vultraz> that's it
20160220 17:36:31< vultraz> celticminstrel: so will you adjust the colors or should I
20160220 17:37:03< celticminstrel> I don't really care about the colours. I think I recall a forum thread discussing them though.
20160220 17:37:54< vultraz> [03:04:30]	celmin|sleep	vultraz: zookeeper: i don't really like the color coding that uses
20160220 17:37:57< vultraz> :|
20160220 17:38:45< celticminstrel> That was me quoting you telling zookeeper that you didn't like it. Except apparently my copy-paste missed the timestamp.
20160220 17:39:45< vultraz> oh
20160220 17:39:51< vultraz> I thought  it was you saying you didn't like it
20160220 17:39:58< vultraz> well, I'll work on it then
20160220 17:40:16< celticminstrel> Yeah, I can see why you'd misunderstand that.
20160220 18:05:00-!- mjs-de [~mjs-de@x4db5c8ca.dyn.telefonica.de] has joined #wesnoth-dev
20160220 18:15:00-!- louis94 [~~louis94@91.178.241.211] has quit [Read error: Connection reset by peer]
20160220 18:15:10-!- louis94 [~~louis94@91.178.241.211] has joined #wesnoth-dev
20160220 18:23:29-!- louis94 [~~louis94@91.178.241.211] has quit [Ping timeout: 240 seconds]
20160220 18:47:05-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has joined #wesnoth-dev
20160220 18:56:12-!- trewe [~trewe@bl20-23-134.dsl.telepac.pt] has joined #wesnoth-dev
20160220 19:06:25-!- pyndragon [~pydsigner@unaffiliated/pydsigner] has quit [Ping timeout: 255 seconds]
20160220 19:18:54< vultraz> hm
20160220 19:19:02< vultraz> colors not that simple to pick 
20160220 19:28:35-!- Appleman1234 [~Appleman1@KD119104007035.au-net.ne.jp] has quit [Ping timeout: 248 seconds]
20160220 19:44:06< vultraz> I must ponder this further
20160220 19:46:01-!- trewe [~trewe@bl20-23-134.dsl.telepac.pt] has quit [Ping timeout: 255 seconds]
20160220 19:47:15-!- trewe [~trewe@bl20-9-181.dsl.telepac.pt] has joined #wesnoth-dev
20160220 19:52:18-!- trewe [~trewe@bl20-9-181.dsl.telepac.pt] has quit [Ping timeout: 276 seconds]
20160220 21:00:18-!- enchi [enchilado@defocus/yummy/enchilado] has quit [Read error: Connection reset by peer]
20160220 21:14:48-!- Alduin_ [~Alduin@p200300864421BA00E29467FFFE0EE8C4.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160220 21:35:01-!- louis94 [~~louis94@91.178.241.211] has joined #wesnoth-dev
20160220 21:39:15-!- louis94 [~~louis94@91.178.241.211] has quit [Ping timeout: 248 seconds]
20160220 21:42:56-!- Appleman1234 [~Appleman1@KD119104009060.au-net.ne.jp] has joined #wesnoth-dev
20160220 21:49:45-!- Kwandulin [~Miranda@p200300760F0BC5007CA8A41547A38CFC.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160220 22:05:59-!- Alduin_ [~Alduin@p200300864421BA00E29467FFFE0EE8C4.dip0.t-ipconnect.de] has quit [Ping timeout: 240 seconds]
20160220 22:19:16-!- louis94 [~~louis94@91.178.241.211] has joined #wesnoth-dev
20160220 22:49:23< vultraz> celticminstrel: is the font scaling/orb pr done?
20160220 22:51:55< vultraz> shadowm: proposal to remove annoying sound effect on gui2 sliders
20160220 22:54:22< vultraz> celticminstrel: actually, let me test it first
20160220 22:58:51< vultraz> celticminstrel: what's this line for? https://github.com/wesnoth/wesnoth/pull/603/files#diff-08aa73adcb83ac851460e778c528958aR67
20160220 23:06:18< vultraz> celticminstrel: orbs display the right color for me now
20160220 23:06:36-!- gfgtdf [~chatzilla@x50ab6d92.dyn.telefonica.de] has joined #wesnoth-dev
20160220 23:12:00< vultraz> edge of the orb images are weirdly fuzzy, though
20160220 23:15:43-!- louis94 [~~louis94@91.178.241.211] has quit [Ping timeout: 255 seconds]
20160220 23:15:53< vultraz> celticminstrel: Defaults doesn't save the state of the orbs
20160220 23:16:09< vultraz> celticminstrel: ie, use Defaults, exit, it will be back to what it was before you reverted
20160220 23:22:46-!- ancestral_ [~ancestral@97-116-184-84.mpls.qwest.net] has joined #wesnoth-dev
20160220 23:24:59-!- mjs-de [~mjs-de@x4db5c8ca.dyn.telefonica.de] has quit [Remote host closed the connection]
20160220 23:29:39< vultraz> blah
20160220 23:29:48< vultraz> you used incorrect widget dimensions
20160220 23:29:51< vultraz> fixing
20160220 23:33:16< vultraz> perfect
20160220 23:42:30-!- irker314 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20160220 23:42:30< irker314> wesnoth: Ignacio R. Morelle wesnoth:master 7fe5437298fc / src/playsingle_controller.cpp: Fix comment https://github.com/wesnoth/wesnoth/commit/7fe5437298fc78c0dea8ca7e3a73f22bd47f2a1a
20160220 23:42:30< irker314> wesnoth: Ignacio R. Morelle wesnoth:master 1e167aaa667a / src/ (9 files in 6 dirs): Fix cases of made-up Doxygen syntax https://github.com/wesnoth/wesnoth/commit/1e167aaa667a352cc408416e8e3ad823619a9902
20160220 23:44:05< ancestral_> Anyone have success getting TLU working in 1.13?
20160220 23:45:36< vultraz> celticminstrel: ok, review should be done
20160220 23:46:07< vultraz> celticminstrel: edited my last comment, so if you got a notification about not being able to add any new friends, period, ignore that bit and refer to the updated comment
20160220 23:50:23< irker314> wesnoth: Ignacio R. Morelle wesnoth:master c04bb9414968 / src/ (addon/client.hpp addon/manager_ui.cpp addon/manager_ui.hpp wml_exception.hpp): Fix cases of people not updating the documentation when refactoring https://github.com/wesnoth/wesnoth/commit/c04bb9414968f5d481682b60855afcd76a1b53ed
20160220 23:50:31< vultraz> ancestral_: haven't tried
20160220 23:50:34< vultraz> is it broken?
20160220 23:51:17< ancestral_> Yeah, some do/while/for loop thing with macros
20160220 23:51:25< ancestral_> (I downloaded from the 1.12 addon server)
20160220 23:52:09< vultraz> oh, yeah, that. probably the same thing I fixed in one of my AtS PRs
20160220 23:53:17< vultraz> (possibly)
20160220 23:53:27< vultraz> not sure
--- Log closed Sun Feb 21 00:00:57 2016