--- Log opened Fri Feb 26 00:00:53 2016
20160226 00:01:16< vultraz> gfgtdf: I didn't feel like doing in in c++03, since it would have involved a static function argument
20160226 00:01:24< vultraz> so far the functionality is unused anyway
20160226 00:01:39< celticminstrel> vultraz: I'm pretty sure I said you could simply loop through manually.
20160226 00:02:22< celticminstrel> vultraz: I forget, are you using CodeBlocks or MSVC?
20160226 00:02:29< vultraz> former
20160226 00:03:06< celticminstrel> Okay, go to project settings.
20160226 00:03:53< celticminstrel> Or build options.
20160226 00:05:59< celticminstrel> I haven't used CB recently, but assuming these screenshots are up-to-date, you want Project build options -> Compiler settings -> Compiler flags -> Enable all warnings [-Wall]
20160226 00:09:36-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160226 00:12:26< celticminstrel> vultraz: You can probably also commit that change.
20160226 00:18:44-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20160226 00:19:06< celticminstrel> Hi.
20160226 00:30:57-!- travis-ci [~travis-ci@ec2-54-160-143-193.compute-1.amazonaws.com] has joined #wesnoth-dev
20160226 00:30:58< travis-ci> wesnoth/wesnoth#8620 (master - f84fc5c : Charles Dang): The build is still failing.
20160226 00:30:58< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/111888615
20160226 00:30:58-!- travis-ci [~travis-ci@ec2-54-160-143-193.compute-1.amazonaws.com] has left #wesnoth-dev []
20160226 00:35:34-!- gfgtdf [~chatzilla@x50ab9478.dyn.telefonica.de] has quit [Quit: ChatZilla 0.9.92 [Firefox 44.0.2/20160210153822]]
20160226 00:35:57-!- mjs-de [~mjs-de@f048100206.adsl.alicedsl.de] has quit [Remote host closed the connection]
20160226 00:40:41< mattsc> Hi
20160226 00:41:46< mattsc> celticminstrel: I think the Wesnoth AI Framework and RCA AI pages are ready.
20160226 00:43:23-!- gb056 [~circuser-@84.93.54.17] has joined #wesnoth-dev
20160226 00:43:38-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has joined #wesnoth-dev
20160226 00:43:41-!- gb056 [~circuser-@84.93.54.17] has quit [Remote host closed the connection]
20160226 00:44:01-!- gb056 [~gb056@84.93.54.17] has joined #wesnoth-dev
20160226 00:46:44-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160226 00:50:17-!- louis94 [~~louis94@91.178.241.169] has quit [Quit: Konversation terminated!]
20160226 00:51:58-!- gb056 [~gb056@84.93.54.17] has quit [Remote host closed the connection]
20160226 01:11:02-!- iceiceice [~chris@50.245.222.235] has joined #wesnoth-dev
20160226 01:11:02-!- iceiceice [~chris@50.245.222.235] has quit [Changing host]
20160226 01:11:02-!- iceiceice [~chris@wesnoth/developer/iceiceice] has joined #wesnoth-dev
20160226 01:13:51< vultraz> celticminstrel: what would you think of a tgroup::set_group_callback_function(boost::function<void(void)>& callback) function that allow callback to be bound to tdispatcher::front_post_child in add_member
20160226 01:14:36< celticminstrel> Sorry, what does this do again?
20160226 01:15:38< vultraz> well, I'm considering it as a way to bind the same function to all members of the group
20160226 01:16:32-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20160226 01:16:41< celticminstrel> So, a shortcut for looping through the group members and binding a left-click callback to each one?
20160226 01:16:54< vultraz> basically
20160226 01:17:05-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160226 01:17:10< celticminstrel> Hmm.
20160226 01:17:58< celticminstrel> I think the signature is wrong, and I'm not sure if it's sufficiently useful to bother. Is there a specific use-case that makes you want this?
20160226 01:18:16< vultraz> no, not really
20160226 01:40:28< irker372> wesnoth: Charles Dang wesnoth:master e67dd6b3dfd1 / src/gui/dialogs/editor/ (editor_edit_side.cpp editor_edit_side.hpp): teditor_edit_side: simplified radio toggles with groups https://github.com/wesnoth/wesnoth/commit/e67dd6b3dfd1774c1acaf426a150cf153170cdb3
20160226 01:40:44< vultraz> celticminstrel: CB already seems to have -Wall on 
20160226 01:41:20< celticminstrel> vultraz: Maybe look for "treat warnings as errors [-Werror]"?
20160226 01:42:01-!- enchi [enchilado@defocus/yummy/enchilado] has quit [Quit: Lost terminal]
20160226 01:42:12< vultraz> hm...nope
20160226 01:42:15-!- aeonchild [enchilado@defocus/yummy/enchilado] has joined #wesnoth-dev
20160226 01:42:16< vultraz> there is -Wextra
20160226 01:42:27< celticminstrel> -Wextra is largely useless noise.
20160226 01:43:19< celticminstrel> Maybe you could add -Werror? The screenshot I was looking at said something about right-clicking.
20160226 01:43:34< celticminstrel> Alternatively, maybe there's a place to directly add arbitrary options to the command-line.
20160226 01:43:45< celticminstrel> You could also look for something like -Wunused.
20160226 01:44:04< vultraz> well -Werror doesn't do me much good if I don't get the warning in the first place :P
20160226 01:44:18< celticminstrel> Right.
20160226 01:44:28< celticminstrel> I think the warning should fall under -Wunused.
20160226 01:44:52< celticminstrel> It might be split into several -Wunused-something options.
20160226 01:45:22< vultraz> nope, not on the list
20160226 01:46:08< irker372> wesnoth: Charles Dang wesnoth:master e5cf2e8eea01 / src/tests/gui/test_gui2.cpp: Attempt to fix gui2 unit tests (post-e23a4764971b) https://github.com/wesnoth/wesnoth/commit/e5cf2e8eea01fc497937458a4d3ec8437270793e
20160226 01:46:22< celticminstrel> Can you add it to the list?
20160226 01:46:41< vultraz> I think so, yes
20160226 01:46:53< vultraz> there's an 'other compiler options' section
20160226 01:47:09< vultraz> do you want me to add -Wunused and -Werror or just one
20160226 01:47:29< celticminstrel> Do you get any warnings right now when you compile?
20160226 01:48:02< vultraz> usually a large number of 'unused variables' in boost
20160226 01:48:17< vultraz> about 40
20160226 01:48:33< celticminstrel> Hmm. Where is Boost installed?
20160226 01:48:54< iceiceice> vultraz, you should usually setup the compiler to use the "-isystem" line instead of "-I" when including boost
20160226 01:49:03< iceiceice> since it will suppress warnings in boost code
20160226 01:49:07< celticminstrel> I was about to get to that. :P
20160226 01:49:09< iceiceice> :)
20160226 01:49:42< vultraz> huh
20160226 01:49:44< vultraz> ok
20160226 01:49:57< vultraz> well, I'll add -Wunused for now, but not -Werror
20160226 01:50:28< celticminstrel> Does anyone else use CB?
20160226 01:50:30< vultraz> then I'll see if the interface provides an easy way to use -isystem with boost 
20160226 01:50:38< vultraz> not usually :P
20160226 01:50:55< vultraz> I wanted to switch to MSVC but the project build time is UNGODLY slow
20160226 01:53:24< irker372> wesnoth: Charles Dang wesnoth:master 90ae1151c72e / projectfiles/CodeBlocks/wesnoth.cbp: CB: add -Wunused compiler flag https://github.com/wesnoth/wesnoth/commit/90ae1151c72eecb6eb9806ded5d2f9d6952b3b4e
20160226 01:53:53< Aginor> hey mattsc 
20160226 01:54:31< vultraz> celticminstrel: are you going to deal with the orbs dialog or should i?
20160226 01:54:41< celticminstrel> I can do it.
20160226 01:54:44< vultraz> ok
20160226 01:55:21< celticminstrel> I dunno if I said this before, but in theory it should be possible to add another colour just by editing the WML (if such a thing is desired).
20160226 02:12:59< mattsc> Hi Aginor 
20160226 02:15:32< mattsc> Hope you’re feeling better.
20160226 02:16:23< Aginor> hi mattsc 
20160226 02:16:23< Aginor> did you want to talk AIs at some point?
20160226 02:17:02-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has quit [Ping timeout: 250 seconds]
20160226 02:17:28< mattsc> Aginor: Yes, I do; but right now I don’t have the time we probably need for that.
20160226 02:18:12< Aginor> fair enough
20160226 02:18:38< mattsc> Did you see what I wrote about the input from Alarantalara?
20160226 02:18:51< Aginor> I'm home sick at the moment so I actually have some time to spare. not that I have much brains at the moment
20160226 02:18:54< Aginor> no
20160226 02:19:14< mattsc> Let me see when that was…
20160226 02:19:52< mattsc> Starting 20160223 22:39:24 in the logs
20160226 02:21:19 * Aginor reads
20160226 02:23:04< Aginor> thanks mattsc 
20160226 02:23:15 * Aginor tries to think
20160226 02:23:39 * mattsc tries that occasionally as well
20160226 02:23:46< mattsc> usually unsuccessful
20160226 02:23:54< Aginor> it certainly feels that way today
20160226 02:24:33< mattsc> Essentially, the bottom line is that the ExpAI is not quite ready for mainlining, but that we probably will only ever get the testing we need for that if we put it into mainline.
20160226 02:24:47< irker372> wesnoth: Charles Dang wesnoth:master 82b089e4788d / src/gui/widgets/group.hpp: tgroup: add a clear() function https://github.com/wesnoth/wesnoth/commit/82b089e4788df4e65c08de284661ca4b0f257edf
20160226 02:24:49< Aginor> I don't particularily like the idea of using it for some maps
20160226 02:24:50< irker372> wesnoth: Charles Dang wesnoth:master dfe6f6c935cc / src/gui/dialogs/addon/ (filter_options.cpp filter_options.hpp): taddon_filter_options: simplify radio buttons with groups https://github.com/wesnoth/wesnoth/commit/dfe6f6c935cc35cfb18bfd364a9d098b05b5d72c
20160226 02:25:04< mattsc> The questions are: Do we do that?  And do we use it for a subset of maps only?
20160226 02:25:15< mattsc> You just answered the second question before I had asked it.  :)
20160226 02:25:32< mattsc> Must be the time zone difference!
20160226 02:25:46< Aginor> it was the easier question to answer :D
20160226 02:26:50< mattsc> Aginor: I am being called away for family duties, but I’d like to follow up on this later.
20160226 02:27:01< mattsc> Not sure if that means later today or some other day.
20160226 02:27:52< Aginor> mattsc: all good
20160226 02:27:55< Aginor> enjoy
20160226 02:28:00< mattsc> I can stay logged in though and will probably come by briefly from time to time.
20160226 02:28:10< mattsc> I’ll try.  ;)
20160226 02:29:08-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has joined #wesnoth-dev
20160226 02:36:08< Aginor> is add-ons just slow or is our network implementation crud?
20160226 02:42:02< vultraz> latter
20160226 02:42:59< vultraz> hopefully loonycyborg will finish the asio redesign soon
20160226 02:43:19< Aginor> yup
20160226 02:44:12-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has quit [Quit: Connection closed for inactivity]
20160226 02:44:27< vultraz> Aginor: btw, I was thinking, before we try to implement OGL, we should switch to c++11
20160226 02:44:43< Aginor> I don't see a compelling reason to switch
20160226 02:44:52< Aginor> and potentiall reasons not to
20160226 02:45:03< vultraz> oh?
20160226 02:45:44< Aginor> of our supported platforms, including all the linux distros we target, which ones come with old enough compilers that they don't have full support?
20160226 02:46:18< iceiceice> Aginor, i think the project should switch because it will help attract students and such who want to work on the project
20160226 02:46:23< vultraz> I don't know personally, but I believe it was concluded (by shadowm, I think), that support was sufficient 
20160226 02:46:30< iceiceice> traditionally a lot of the developers on wesnoth have been students i guess
20160226 02:46:38< vultraz> the only people who need to upgrade their compilers are gfgtdf and zookeeper - windows users
20160226 02:46:38< Aginor> fair enough
20160226 02:46:41< iceiceice> but today any student that leanrs C++ leanrs C++11
20160226 02:46:52< iceiceice> just my 2 cents though
20160226 02:47:01< Aginor> I suspect most students don't leanr c++ today though
20160226 02:47:11< Aginor> js, python and c# more likely
20160226 02:47:12< iceiceice> thats true, but many of them still do i think
20160226 02:47:13< vultraz> iceiceice: I agree - plus it doesn't help that I've been learning an outdated standard :P
20160226 02:48:27< vultraz> plus it'd be nice to have stuff like initializer lists and lambdas
20160226 02:49:12< iceiceice> Aginor, i think C++ is still basically the most popular language for making game engines
20160226 02:49:24< iceiceice> b/c of portability and performance
20160226 02:49:24< vultraz> plus c++17 will be out in the not too distant future 
20160226 02:49:30< vultraz> and we're not even on c++11 yet?
20160226 02:49:44< iceiceice> and safety in knowing the language will let you do anything your crazy game engine-maker heart desires, even bad things
20160226 02:50:33< iceiceice> yesterday i learned a wierd technique I had never heard of
20160226 02:50:36< iceiceice> called "XOR-linked list"
20160226 02:50:44< Aginor> I don't feel particularily strong on the subject :D
20160226 02:51:01< iceiceice> maybe its better that i didn't hear of it until now, its kind of evil i guess :)
20160226 02:51:07< iceiceice> Aginor, yeah i mean
20160226 02:51:10< iceiceice> its really the consensus of the devs
20160226 02:51:23< iceiceice> there will rpobably be a lot of problems to switch to C++11, we discussed it a few times
20160226 02:51:32< iceiceice> and idk who will try to make it happen
20160226 02:51:49< iceiceice> as long as there's someone who doesnt want to get a new compiler probably it wont
20160226 02:52:19< Aginor> I have full support for it, so I don't particularily care :D
20160226 02:52:31< Aginor> I'm more worried about target platforms, but if that's a non-issue
20160226 02:52:52< iceiceice> vultraz, actually i'm thinking about making a C++14 version of spirit_po
20160226 02:53:00< iceiceice> b/c it will be more efficient i guess
20160226 02:53:23< iceiceice> c++14 is not nearly is big a leap as c++11 but it fixes a few crucial problems
20160226 02:53:39< iceiceice> for one C++14 has polymorphic lambdas
20160226 02:53:56< iceiceice> for two, for some damn reason in C++11 the std::string move constructor is not marked "noexcept"
20160226 02:54:16< iceiceice> so all that stuff they tell you about how "vector goes faster, everything is faster because of moves" doesn't actually work if std::string is involved
20160226 02:54:30< iceiceice> b/c most of the time `std::move_if_noexcept` will detect that it should copy instead
20160226 02:54:50< iceiceice> the std:string thing isn't fixed until c++14
20160226 02:55:13< celticminstrel> I thought the reason we haven't switched to C++11 yet was MSVC.
20160226 02:55:18< iceiceice> also there's a new version of boost spirit that requires c++14 which iw oudl ike to learn
20160226 02:55:28< iceiceice> celticminstrel, msvc 2015 supports C++11 fairly well
20160226 02:55:31< iceiceice> and it is free
20160226 02:55:44< Aginor> iceiceice: fairly well?
20160226 02:55:44< iceiceice> actually it supports c++14 fairly well
20160226 02:56:00< iceiceice> iiuc the only thing they are missing is some obscure thing called "Expression SFINAE"
20160226 02:56:14< celticminstrel> But MSVC 2013 already has fairly decent C++11 support (not complete, but at least it has some of the big ones like lambdas and range-for). So MSVC is really only a problem if people are still using much older versions.
20160226 02:56:36< iceiceice> celticminstrel, iiuc MSVC 2013 had a lot of problems,
20160226 02:56:45< celticminstrel> MSVC 2015 is free? Is that something like MSVC Express or a full version?
20160226 02:56:57< celticminstrel> I've used MSVC 2013, and it worked well enough for my purposes.
20160226 02:56:58< Aginor> don't we support MSVC down to version 9.0 at the moment?
20160226 02:56:58< iceiceice> like in C++11, in some cases the compiler is required to default generate certain special member functions like move constructors 
20160226 02:57:06< iceiceice> but it fails to do it
20160226 02:57:17< iceiceice> maybe its only fixed after an upgrade pack or whatever ms calls it
20160226 02:58:00< iceiceice> celticminstrel, https://www.visualstudio.com/en-us/products/visual-studio-community-vs.aspx
20160226 02:58:50< celticminstrel> Yeah, I recall C++11 having problems with defaulted move constructors.
20160226 02:58:56< celticminstrel> Sorry, MSVC2013 I mean.
20160226 03:00:12< celticminstrel> Is it worth upgrading from MSVC2013 Ultimate to the thing at that link on my Win7?
20160226 03:00:31< Aginor> that reminds me
20160226 03:00:43< Aginor> doesn't VS2015 require win8 or newer?
20160226 03:00:51< iceiceice> celticminstrel, i dont know, i have only poked around in the "community" thing
20160226 03:00:52< celticminstrel> I have no idea!
20160226 03:01:01< iceiceice> i generally dont use ides
20160226 03:01:04< celticminstrel> I know that Office XP is officially unsupported on Win7, but works anyway.
20160226 03:01:17< celticminstrel> Though, that's the opposite direction.
20160226 03:01:55< celticminstrel> I can't see anything on that page about minimum OS requirements.
20160226 03:09:08< Aginor> I have a vague memory of running into it, but I may be wrong
20160226 03:09:58< vultraz> https://www.visualstudio.com/en-us/products/visual-studio-2015-compatibility-vs.aspx
20160226 03:10:42< Aginor> ok
20160226 03:10:47< Aginor> I'm convinsed
20160226 03:12:04-!- travis-ci [~travis-ci@ec2-54-144-242-105.compute-1.amazonaws.com] has joined #wesnoth-dev
20160226 03:12:05< travis-ci> wesnoth/wesnoth#8623 (master - 90ae115 : Charles Dang): The build was fixed.
20160226 03:12:05< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/111909295
20160226 03:12:05-!- travis-ci [~travis-ci@ec2-54-144-242-105.compute-1.amazonaws.com] has left #wesnoth-dev []
20160226 03:19:34< celticminstrel> I think that says that it supports Win7?
20160226 03:19:51< celticminstrel> It's kinda confusing though.
20160226 03:20:08 * celticminstrel finally gets around to proofreading the AI pages.
20160226 03:22:10< Aginor> where's the addon's directory? - I have a number of broken addons now that I have a strong urge to delete
20160226 03:22:41< celticminstrel> I think you can find out by clicking the button in the bottom left corner of the title screen.
20160226 03:23:14< celticminstrel> It's normally placed in a standard settings location for your platform.
20160226 03:23:17< Aginor> thanks celticminstrel 
20160226 03:24:10< vultraz> Aginor: or you can use the Remove Addons dialog
20160226 03:24:47< Aginor> too much hassle
20160226 03:24:56< Aginor> I should raise a bug about my addons experiences too
20160226 03:25:07< mattsc> celticminstrel: no rush with any of this; I’ll just work on them as I have time and let you know whenever I am done with one
20160226 03:25:09< Aginor> I find the usability poor :)
20160226 03:25:26< mattsc> I it takes you a couple of months to get to them, they’re not going to go away  :)
20160226 03:26:33< mattsc> and some of them are pretty massive; AI Framework is probably not as well done as it could be because I was getting really tired of it :P
20160226 03:27:23< vultraz> Aginor: what do you mean
20160226 03:27:54< Aginor> vultraz: it installed the addon-on even though I said no to the dependencies
20160226 03:27:56< celticminstrel> My refactor of the AI config loading means that the values of id= can be used for ai_algorithm-.
20160226 03:28:05< celticminstrel> At least, that was the intent. I haven't done any testing.
20160226 03:28:14< Aginor> I think it would make more sense to abort the install at that point
20160226 03:28:17-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20160226 03:28:30< vultraz> Aginor: we have no way of setting whether an dependency is mandatory or not
20160226 03:28:40< vultraz> Aginor: so depending on the addon, either way makes sense
20160226 03:28:51< vultraz> so right now we just install anyway
20160226 03:29:29< irker372> wesnoth: Andreas Löf wesnoth:guifixes cc760cb33fa4 / src/ (5 files in 2 dirs): Unable undrawing of dialogs used in the title-screen https://github.com/wesnoth/wesnoth/commit/cc760cb33fa49957bc354146bb12027a7b368b25
20160226 03:29:39< celticminstrel> Unable?
20160226 03:29:48< celticminstrel> Do you mean disable?
20160226 03:29:57 * Aginor facepalms
20160226 03:30:17< Aginor> this is why you shouldn't program while sick
20160226 03:30:23< celticminstrel> Heh.
20160226 03:30:39< Aginor> yes, I think I meant disable
20160226 03:30:53< celticminstrel> I was looking through the Lua reference a few hours ago and noticed that it mentions a way to make Lua directly use exceptions instead of longjmp.
20160226 03:30:59< celticminstrel> I wonder if that would be worthwhile.
20160226 03:31:11< iceiceice> celticminstrel, we already do that
20160226 03:31:14< Aginor> anyhow, I'm prepared to argue that guifixes should be merged into main. Does anyone want to take me up on it?
20160226 03:31:17< iceiceice> that's what compiling lua as C++ achieves
20160226 03:31:57< celticminstrel> iceiceice: Really? I got the impression that you have to fiddle with the internals a bit. Something about LUAI_THROW.
20160226 03:32:09< mattsc> celticminstrel: sounds good to me
20160226 03:32:25< celticminstrel> mattsc: Sorry, is that about the Lua thing or something AI-related or what?
20160226 03:32:31< iceiceice> celticminstrel, as i understand, lua detects using preprocessor macros if you are using C++ or not
20160226 03:32:38< celticminstrel> Hmm.
20160226 03:32:43< iceiceice> celticminstrel, if we are using long jmp that woudl be really bad
20160226 03:32:47< celticminstrel> Well, that's nice if it does, I guess.
20160226 03:33:01< mattsc> celticminstrel: sorry, that was about the message that is highlighted for me  (which was your AI refactor id comment)
20160226 03:33:10< celticminstrel> Oh, that, right. Okay.
20160226 03:33:27< celticminstrel> I forget exactly what the problem with longjmp in C++ is.
20160226 03:33:42< iceiceice> celticminstrel, if we are using longjmp in lua, that means that this line here is a memory leak / undefined behavior
20160226 03:33:43< iceiceice> https://github.com/wesnoth/wesnoth/blob/master/src/scripting/game_lua_kernel.cpp#L1053
20160226 03:33:56< iceiceice> and the many more like it
20160226 03:34:12< celticminstrel> Ah, okay.
20160226 03:34:19< mattsc> Aginor: I’d have 20 min right now; but maybe this is not the right time in between all the other discussions
20160226 03:34:28< iceiceice> i guess i never confirmed for a fact that we are doing it, but i hope that whoever set up lua did it right :)
20160226 03:34:37< Aginor> mattsc: possibly not
20160226 03:34:58< Aginor> I'm also starting to feel ready for yet another nap
20160226 03:35:17< celticminstrel> iceiceice: Looks like you're right.
20160226 03:35:32< celticminstrel> There's an #ifdef __cplusplus in ldo.cpp
20160226 03:35:37< iceiceice> ok, thanks for checking :)
20160226 03:35:51< mattsc> Aginor: okay, no worries; I probably won’t have time again tonight though
20160226 03:35:58< mattsc> Is it early afternoon for you at the moment?
20160226 03:36:18< Aginor> it's 16:36 here
20160226 03:36:30< mattsc> late afternoon
20160226 03:36:34< Aginor> mattsc: I'm currently GMT+13
20160226 03:36:39 * celticminstrel is -5
20160226 03:36:40< mattsc> really, you’re only 3 hours off from me?
20160226 03:37:10< Aginor> maybe I'm a day ahead :D
20160226 03:37:20< mattsc> well, yes
20160226 03:37:36< mattsc> but I’m at +8 and it 19:37 here;  so something does not add up
20160226 03:37:46< Aginor> ah, not a day ahead then
20160226 03:38:14< mattsc> oh, plus 13
20160226 03:38:22< mattsc> I’m at -8 actually ...
20160226 03:38:27< mattsc> so the 3h do make sense
20160226 03:38:29< Aginor> that makes more sense
20160226 03:38:42< vultraz> it's currently 2:38 PM for me
20160226 03:38:57< celticminstrel> ...-8 is west coast...
20160226 03:40:50< mattsc> indeed
20160226 03:41:16< mattsc> just as vultraz, I live on an island in the Pacific  ;)
20160226 03:41:28< vultraz> hawaii?
20160226 03:41:31< celticminstrel> Oh fun.
20160226 03:41:39< mattsc> npe, HI is qt -10
20160226 03:41:45< mattsc> s/qt/at
20160226 03:42:15 * Aginor too lives on an island in the Pacific
20160226 03:42:22< mattsc> hehe
20160226 03:42:27< vultraz> mattsc: where, then
20160226 03:42:29< celticminstrel> Presumably on the opposite side.
20160226 03:42:34< mattsc> Vancouver Island
20160226 03:42:43< celticminstrel> That should have been obvious.
20160226 03:43:08< mattsc> It’s barely an Island, but enough that you can’t just drive out of here
20160226 03:45:25< Aginor> celticminstrel, vultraz, should I go and merge guifixes?
20160226 03:45:26< celticminstrel> What's the recommended way of attaching some AI functions to specific units?
20160226 03:45:34< celticminstrel> Filters in the aspects or something?
20160226 03:46:14< mattsc> I’d use Lua CAs with a filter (check out how, for example, the patrol or guardian MAIs do it.
20160226 03:46:25< mattsc> Or you could use BCAs; or FAI unit formuals
20160226 03:46:25< vultraz> Aginor: I'm guessing yes
20160226 03:46:36< mattsc> But I’d go with the first option
20160226 03:46:44< celticminstrel> I saw something about BCAs in legacy docs somewhere. It wasn't very clear.
20160226 03:47:05< mattsc> Yeah, I’ll put that into Creating Custom AIs when I get to it
20160226 03:47:16-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160226 03:47:23< mattsc> That’s the only page that still needs more work than simple clean-up and proof-reading
20160226 03:47:29< Aginor> it does introduce flickering during a redraw, which will happen when a dialog/window is closed or a window event happens
20160226 03:47:59< celticminstrel> I get the impression you're trying a little too hard not to end your sentences with a preposition.
20160226 03:50:56< mattsc> I’m used to writing technical papers and reports.  :)
20160226 03:51:19< Aginor> mattsc: passive voice and all of that? :D
20160226 03:51:35< mattsc> At times, yes.  :)
20160226 03:51:42-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 244 seconds]
20160226 03:52:07< mattsc> But they correct you on informal use of grammar for those, so it probably shows up in my wiki pages.
20160226 03:52:22< mattsc> In other words, I don’t really try, that’s just what I am used to.
20160226 03:52:53< Aginor> ah, yes
20160226 03:53:00< mattsc> Feel free to change things so that they sound better.
20160226 04:06:14-!- iceiceice [~chris@wesnoth/developer/iceiceice] has quit [Quit: Leaving]
20160226 04:13:37< celticminstrel> Honestly, the rule against ending sentences with a preposition was just made up by some 18th(?)-century quack who that English should be more like Latin.
20160226 04:14:23< celticminstrel> Following it religiously can lead to some very awkward-sounding sentences.
20160226 04:15:36< celticminstrel> You seem to also like two spaces after sentences, but I'm not gonna bother changing that since it's collapsed by the HTML rendering.
20160226 04:16:55< celticminstrel> I like the old Lua CA syntax better than the new. I suppose one advantage of the new syntax is that evaluate and execute can share some state.
20160226 04:17:23< celticminstrel> I guess both have their flaws.
20160226 04:17:54< celticminstrel> (Avoiding sentence-ending prepositions usually sounds okay, so I'll only change them if it's really awkward.)
20160226 04:18:57< celticminstrel> (Which so far is only one case.)
20160226 04:21:34< celticminstrel> Hmm... which sounds better? "A huge number of aspects is available", or "A huge number of aspects are available"?
20160226 04:22:03< celticminstrel> The latter sounds better to me, but I think I can see why you'd choose the former.
20160226 04:24:02< celticminstrel> mattsc: Is this really true? "Note that this is the syntax as it shows up in the AI configuration after processing by the engine, but that you cannot simply wrap this into an [ai] tag like this. If you wanted to define it in this way in a scenario, you would have to use the [modify_ai] tag as described here."
20160226 04:24:13< mattsc> celticminstrel: seriously, I don’t mind one way or the other (on any of that, not just the last).  Just change it if you think it sounds funny or whatever
20160226 04:24:17< celticminstrel> Regarding using composite aspect syntax for simple aspects in the [ai] tag.
20160226 04:24:46< mattsc> celticminstrel: I’m not 100% sure, but I tried it and it did not work
20160226 04:24:53< mattsc> so I added that sentence
20160226 04:25:08< celticminstrel> Ah. That seems odd. It should be allowed to work.
20160226 04:25:22< celticminstrel> I don't know if my refactor will fix it.
20160226 04:25:24< mattsc> but then I later found an example where it is done like that, I think — but I have not had time yet to figure out if I did something wrong or what else is going on.
20160226 04:25:35< celticminstrel> Could it be related to version=?
20160226 04:25:52< mattsc> I think it is; I sued version=10710 
20160226 04:25:59< mattsc> and it did not work
20160226 04:26:14< mattsc> I thought I needed it, but maybe that makes it not work
20160226 04:26:21< mattsc> anyways, I’ll check that out tomorrow.
20160226 04:27:02< mattsc> I’m about to take off (it’s still early here, but I’ll be busy with other things for the next couple hours
20160226 04:27:28< mattsc> celticminstrel: AiWML is ready too (except that I have not removed the underscores from links yet)
20160226 04:28:04< mattsc> But I have to admit that I did not go through the descriptions of all the aspects; and I am not planning to do so, at least not anytime soon.
20160226 04:30:15< vultraz> Aginor: what is your opinion on do_gameloop()/
20160226 04:32:39< celticminstrel> vultraz: Do you have a list of dialogues that could use groups?
20160226 04:32:52< vultraz> celticminstrel: I've converted them all already
20160226 04:32:56< vultraz> celticminstrel: except orbs
20160226 04:33:10< celticminstrel> I didn't see you converting advanced graphics options. Maybe I missed it.
20160226 04:33:14< vultraz> oh
20160226 04:33:16< vultraz> right
20160226 04:33:18< vultraz> that
20160226 04:33:19< vultraz> yes
20160226 04:33:23< vultraz> will you do that too?
20160226 04:33:30< celticminstrel> Sure.
20160226 04:36:38-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: So long and thanks for all the fish.]
20160226 04:50:26-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has joined #wesnoth-dev
20160226 05:05:43-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20160226 05:09:06-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has joined #wesnoth-dev
20160226 05:11:57-!- redscorpion [bdd9d9a9@gateway/web/freenode/ip.189.217.217.169] has joined #wesnoth-dev
20160226 05:20:16-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20160226 05:25:20-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has joined #wesnoth-dev
20160226 05:33:47-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has quit [Ping timeout: 250 seconds]
20160226 05:39:29< Aginor> vultraz: what do you mean?
20160226 05:40:59< vultraz> Aginor: it's basically the backbone of the program, one step removed from main(). But I'm wondering if you think it could use come refactoring, especially regarding its handling of titlescreen functionality 
20160226 05:41:13< celticminstrel> Orb dialog simplification done. Now testing.
20160226 05:41:48< celticminstrel> Also updated XCode and MSVC projects with the new file, and implemented remove_member (because why not).
20160226 05:41:59< Aginor> I think we should pick our battles, and is that a battle we want/need now?
20160226 05:42:32< celticminstrel> It works, so let's not fix it. :)
20160226 05:42:50< vultraz> right
20160226 05:42:57< vultraz> I'm just constructing a mental todo list
20160226 05:43:24< vultraz> because I think our general goal is to split the game's functionality into clearly defined areas
20160226 05:43:28< celticminstrel> Are there really so few dialogs that use radiobuttons?
20160226 05:43:46< Aginor> vultraz: separate gui thread and game thread, synchronisation with rendering code
20160226 05:44:10< Aginor> and well established ways to handle it without getting into deadlocks/races
20160226 05:44:52 * celticminstrel greps for "radio" in data/gui, finds six results.
20160226 05:44:59< celticminstrel> Huh.
20160226 05:45:41< Aginor> push the button if you're happy: https://github.com/wesnoth/wesnoth/pull/595 :)
20160226 05:45:47< celticminstrel> (Should've restricted it to data/gui/default/window - one of the results is the radiobutton definition.)
20160226 05:45:55< celticminstrel> That's guifixes?
20160226 05:46:00< Aginor> yes
20160226 05:46:17< celticminstrel> I assume you've tested it?
20160226 05:46:20< Aginor> yes
20160226 05:46:29< Aginor> I keep missing things though :D
20160226 05:46:33 * celticminstrel scans the diff.
20160226 05:46:37< Aginor> so please give it a whirl
20160226 05:46:50< celticminstrel> By "give it a whirl" do you mean test it?
20160226 05:46:55< Aginor> yes
20160226 05:47:11< Aginor> having someone else looking at it with fresh eyes would be a good thing
20160226 05:47:17< celticminstrel> Huh? Does draw_layering extend sdl_handler?
20160226 05:48:02< celticminstrel> Maybe it would be better to look at this one commit at a time instead of all at once?
20160226 05:48:17< Aginor> celticminstrel: some of them contradict themselves slightly
20160226 05:48:19< Aginor> but yes
20160226 05:48:36< Aginor> the draw layering is based around event contexts
20160226 05:48:55< celticminstrel> I'm in the middle of stuff on the main branch, but maybe I'll pull and test this once I'm done with that.
20160226 05:49:03< Aginor> I'll need to extend this further into GUI2 land at some point too, but I'm not too keen on it
20160226 05:49:06< celticminstrel> s/the main branch/my local master/
20160226 05:49:20< Aginor> fair enough
20160226 05:50:01< celticminstrel> Wait wait wait.
20160226 05:50:16< celticminstrel> Does this dynamic layout apply to the theme?
20160226 05:50:25< celticminstrel> I noticed the theme uses entirely static rects.
20160226 05:50:37< celticminstrel> Which cut things off it the font size is increased.
20160226 05:50:42< celticminstrel> ^if
20160226 05:51:43< celticminstrel> If so, maybe ancestral will want to test this out too.
20160226 05:52:10< celticminstrel> I have no idea what some of these changes do.
20160226 05:52:47< celticminstrel> Is it my imagination, or did you make SDL_GPU always active?
20160226 05:54:51< Aginor> I have most certainly not touched any of the SDL_GPU code
20160226 05:55:05< celticminstrel> Well, I noticed some SDL_GPU sections were enabled, though.
20160226 05:55:11< Aginor> I still haven't even looked at compiling it to see if it's viable (I expect not)
20160226 05:55:52< Aginor> I take no responsibility for the font layout thing :D
20160226 05:55:59< celticminstrel> Looks like you did something that I mentioned a few days ago (probably you did it before I mentioned it), which is passing certain events to all handlers instead of just the top of the stack.
20160226 05:56:17< Aginor> yeah, I introduced that a few weeks ago
20160226 05:56:22< Aginor> the global event context
20160226 05:57:04< celticminstrel> Doesn't SDL have a limit on the number of user events it can support?
20160226 05:57:41< celticminstrel> ...did you fix the typo in events.hpp?
20160226 05:57:52< celticminstrel> raise_volativle_draw_all_event
20160226 05:57:52< Aginor> mmm, I think it's around 0x7ffffff-500 or something like that
20160226 05:58:23< Aginor> the previous 8 or whatever it was a limitation imposed by sdl 1.2
20160226 05:58:32< celticminstrel> Ahh.
20160226 05:58:37< celticminstrel> So SDL2 increased the limit.
20160226 05:58:46< celticminstrel> Oh, my build is almost finished.
20160226 05:58:52< Aginor> let me look in the header
20160226 05:59:00-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160226 05:59:20< Aginor> my bad
20160226 05:59:27< Aginor> we have fewer
20160226 05:59:33< celticminstrel> ?
20160226 05:59:45< celticminstrel> Fewer user events?
20160226 05:59:49< Aginor> 63487 user events
20160226 06:00:00< celticminstrel> That's a lot.
20160226 06:00:16< Aginor> but orders of magnitudes smaller than my initial claim
20160226 06:00:59< celticminstrel> I didn't really know what your original claim was, anyway.
20160226 06:01:10< Aginor> billions :D
20160226 06:01:26< celticminstrel> I guess you had around 4 too many F's.
20160226 06:01:54< Aginor> I simply assumed they'd used a whole 32-bit field instead of some silly 16-bit verion
20160226 06:02:13< celticminstrel> Three too many, I think. One less than you probably intended.
20160226 06:02:40< Aginor> yes
20160226 06:03:19-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 252 seconds]
20160226 06:04:36< vultraz> Aginor: do you have anything I can work on?
20160226 06:04:54< vultraz> in general
20160226 06:05:14< Aginor> vultraz: tracking down calls to video.flip(), ensure that they only happen at the end of a draw/draw_all_event. 
20160226 06:05:49< Aginor> vultraz: add a layer to GUI2 that keeps tracks of windows/dialogs and the layers they add, and the position of the components
20160226 06:06:01< vultraz> 36 instances of flip
20160226 06:06:01< Aginor> vultraz: add partial redrawing to GUI1/GUI2
20160226 06:06:13< celticminstrel> Well, the orbs dialog appears to work just the same as before, so I guess that one's done.
20160226 06:06:36< celticminstrel> Oh, the advanced graphics options has two of them.
20160226 06:06:40< Aginor> so that when one window is on top of another, only the affected parts of the window has to be redrawn
20160226 06:06:51< celticminstrel> Two groups, I mean.
20160226 06:08:38< celticminstrel> There is a minor problem in the orb dialog whereby it will crash if the IDs in the WML definition don't follow the correct format, but... it already relied on them being in the correct format (though I don't think it would've crashed before), so whatever.
20160226 06:09:12< Aginor> vultraz: remove those instances of flip and call them from the end of the draw events
20160226 06:09:20< Aginor> it should be possible after guifixes
20160226 06:09:31< Aginor> that'll probably make the interface look a lot less flickery
20160226 06:09:33< celticminstrel> Or, do it on guifixes. :P
20160226 06:09:59< vultraz> can you just merge guifixes?
20160226 06:10:04< vultraz> it's a pita for me to keep switching branches
20160226 06:10:31< Aginor> I can, I want to, I also want to not break stuff and have buy-in from more people before I commit
20160226 06:11:19< Aginor> I do not want to catch a bunch of flak for these changes and hold up any releases further, especially not since I will have trouble being available from Sunday onwards for a week or so
20160226 06:12:09< celticminstrel> I agree that it's a pain to switch branches.
20160226 06:12:25< celticminstrel> Probably this one especially, since gui stuff is included almost everywhere.
20160226 06:12:31 * Aginor is unsympathetic
20160226 06:12:43< vultraz> :P
20160226 06:12:59< Aginor> you did not want me to do this on master
20160226 06:13:14< celticminstrel> The worst branch-switching is when one branch includes changes to the XCode project that are not in the other branch.
20160226 06:13:29< Aginor> I can merge master if you want
20160226 06:14:11< celticminstrel> Wait a bit.
20160226 06:14:29< celticminstrel> I have one XCode project change pending push.
20160226 06:14:57< Aginor> or make your own local branch
20160226 06:16:29< celticminstrel> Uhhh... why the hell does the advanced graphics dialog have a wikidoc comment about lobby sounds options? o.o
20160226 06:17:25< celticminstrel> Advanced graphics might be a bit more complicated than expected. I guess I might as well push now.
20160226 06:17:34-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20160226 06:18:06-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has joined #wesnoth-dev
20160226 06:21:15< celticminstrel> There's really no need to set the new scaling modes while the dialog is still up, is there?
20160226 06:21:49< irker372> wesnoth: Celtic Minstrel wesnoth:master 9a77b81fc7db / projectfiles/ (VC9/wesnoth.vcproj Xcode/Wesnoth.xcodeproj/project.pbxproj): Update XCode and MSVC projects https://github.com/wesnoth/wesnoth/commit/9a77b81fc7dbbcce5bdddfd4b96ab8c45297a090
20160226 06:21:51< irker372> wesnoth: Celtic Minstrel wesnoth:master 5d5f89d31b1f / src/gui/widgets/group.hpp: tgroup: Pre-C++11 implementation of remove_member https://github.com/wesnoth/wesnoth/commit/5d5f89d31b1f6a6200ebee7fc3706667a9265d6f
20160226 06:21:53< irker372> wesnoth: Celtic Minstrel wesnoth:master 63a3312e705d / src/gui/dialogs/ (select_orb_colors.cpp select_orb_colors.hpp): Use widget groups in orb selection dialog https://github.com/wesnoth/wesnoth/commit/63a3312e705d4e1d9ce92b7b7b32d77b1df421ef
20160226 06:22:30< vultraz> celticminstrel: the documentation is probably becuase another file was just copied and the comments not updated
20160226 06:28:01< celticminstrel> vultraz: Any opinion on what I said just before pushing?
20160226 06:28:29< vultraz> celticminstrel: advanced graphics?
20160226 06:28:33< vultraz> yeah, it's a mess...
20160226 06:28:47< vultraz> you should just scrap the existing code and redo it
20160226 06:28:58< celticminstrel> I was thinking of basically rewriting it, yeah.
20160226 06:29:06< celticminstrel> Not entirely scrapping the existing code, but mostly.
20160226 06:30:28-!- Kwandulin [~Miranda@p200300760F0BC585009A1BE2EEA60AA1.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160226 06:30:33< vultraz> iceiceice is a good coder but in this case he didn't really try to make it neat
20160226 06:30:39< celticminstrel> But for the question... it's fine to defer application of the chosen settings to the post_show, right?
20160226 06:31:04-!- redscorpion [bdd9d9a9@gateway/web/freenode/ip.189.217.217.169] has quit [Ping timeout: 252 seconds]
20160226 06:31:08< celticminstrel> Does the Close button trigger a resoult of OK?
20160226 06:31:24< celticminstrel> ^result
20160226 06:32:04< vultraz> well
20160226 06:32:07< vultraz> not really
20160226 06:32:19< vultraz> preferences::set is called every time you click a button
20160226 06:32:27< vultraz> so the result is already set
20160226 06:32:30< vultraz> just likeprefs
20160226 06:32:43< celticminstrel> Yeah, that's what I was saying - I don't need to do that until post_show, right?
20160226 06:32:59< celticminstrel> These settings probably aren't even used by GUI2?
20160226 06:33:10< vultraz> they aren't
20160226 06:33:28< vultraz> changing the button to OK is probably better
20160226 06:36:15< celticminstrel> The ID is OK, but the name is Close, so I guess it'll probably trigger a result of OK.
20160226 06:37:45< vultraz> i mean make it say OK
20160226 06:38:42< celticminstrel> It also has tooltips on the column labels, which don't work. Kinda would be nice to fix that, but it's out of scope for what I'm doing here.
20160226 06:51:57< celticminstrel> For some reason, recently XCode's indexing and code completion for Wesnoth started working again.
20160226 06:52:44< ancestral> vultraz: So…
20160226 06:52:55< celticminstrel> Hmm. Calling image::flush_cache twice in a row shouldn't hurt, right?
20160226 06:52:55< ancestral> I applaud you for the font scaling, because I think it’s a great idea
20160226 06:53:00< ancestral> But I don’t think it works well enough
20160226 06:53:02< celticminstrel> That was me.
20160226 06:53:07< ancestral> within the parameters of the app
20160226 06:53:10< ancestral> Ahhhh
20160226 06:53:23< ancestral> Cool!
20160226 06:54:03< ancestral> I mean, maybe if enough widgets get changed it will react better?
20160226 06:54:30< celticminstrel> If enough widgets get changed?
20160226 06:54:35< ancestral> Well
20160226 06:54:39< ancestral> When the font scales up
20160226 06:54:53< ancestral> Some of the items on the screen get cut off in certain contexts
20160226 06:54:57< ancestral> Like in the game
20160226 06:55:05< ancestral> Buttons too?
20160226 06:55:14< celticminstrel> I don't think buttons get cut off.
20160226 06:55:26< celticminstrel> At least, I didn't see any when I was testing.
20160226 06:55:37< celticminstrel> What I saw getting cut off was mostly the unit sidebar text.
20160226 06:56:26< celticminstrel> BTW, regarding the font problem - were HP/XP/MP the only things in the correct font or the only things in the wrong font?
20160226 06:56:46< ancestral> That’s GUI1
20160226 06:57:12< celticminstrel> So the rest of that is drawn with ttext?
20160226 06:57:24< ancestral> I have an image
20160226 06:57:35< celticminstrel> I remember that image, vaguely.
20160226 06:57:41< celticminstrel> Unless this is a different one.
20160226 06:57:54< Aginor> ancestral: do you feel like testing PR 595?
20160226 06:57:59< ancestral> http://mproud.com/wesnoth/gameplay-fonts.png
20160226 06:58:10< ancestral> Aginor: Sure, let me look at it
20160226 06:58:16< Aginor> ancestral: thanks
20160226 06:58:25< ancestral> celticminstrel: I have problems with the editor at font scaling 150%
20160226 06:58:30< ancestral> missing button images
20160226 06:58:36< ancestral> hex x,y is cut off a little
20160226 06:58:44< ancestral> Wait, maybe that’s my theme
20160226 06:58:49< celticminstrel> I never actually tried the editor.
20160226 06:59:08< ancestral> NOpe
20160226 06:59:29< ancestral> Minor stuff though
20160226 06:59:41< celticminstrel> ...wait, didn't you say TC wasn't broken on your build?
20160226 06:59:57< celticminstrel> Because that screenshot shows no TC whatsoever.
20160226 07:00:01< ancestral> Someone said the icon thing might be fixed
20160226 07:00:10< ancestral> TC was broken for someone else
20160226 07:00:12< celticminstrel> Though I guess it's an older screenshot.
20160226 07:00:25< celticminstrel> That screenshot is from someone else?
20160226 07:00:31< ancestral> The screenshot is mine
20160226 07:00:57< celticminstrel> It shows broken TC - the terrain icons, the orbs, and the units are all magenta.
20160226 07:00:59< ancestral> Interesting
20160226 07:01:24< celticminstrel> If enemies had been visible, I imagine they would've had red ellipses.
20160226 07:01:48< ancestral> It’s not from a recent build
20160226 07:01:56< celticminstrel> Right...
20160226 07:02:18< ancestral> I can see 1.13.2+dev, but recent builds I’ve made are working correctly afaik
20160226 07:02:59< celticminstrel> It's still broken in my build for some reason.
20160226 07:04:49< ancestral> celticminstrel: I bet I could fix the default theme to use vertical height based on the fontsize
20160226 07:05:07< ancestral> which would probably help from clipping
20160226 07:05:15< celticminstrel> That would help, if it works.
20160226 07:05:24< ancestral> However… the default size needs to be bigger.
20160226 07:05:43< celticminstrel> I don't really agree, or at least, not significantly bigger.
20160226 07:05:43-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20160226 07:05:58< celticminstrel> But again, I don't have a retina monitor.
20160226 07:07:53< celticminstrel> And now he's gone. :|
20160226 07:08:21-!- ancestral [~ancestral@97.116.184.84] has joined #wesnoth-dev
20160226 07:08:29< ancestral> Whoops, ran out of power
20160226 07:08:43< celticminstrel> Ah.
20160226 07:08:56< ancestral> It’s not really about retina monitors
20160226 07:09:07< ancestral> Wesnoth is probably not getting true retina support anytime soon
20160226 07:09:14< ancestral> It’s more about larger monitors
20160226 07:09:48< ancestral> I’m proposing upping the size by a few points
20160226 07:09:57< celticminstrel> What's the current font size?
20160226 07:10:31< vultraz> 14
20160226 07:10:38< vultraz> it's a good default size
20160226 07:10:40< vultraz> ancestral: why not?
20160226 07:10:47< ancestral> 14/12/10/9
20160226 07:11:44< ancestral> 18/14/13/11 seems to be working well for me over here
20160226 07:12:18< celticminstrel> 14 feels kinda large to me already.
20160226 07:12:29< celticminstrel> 18 is like, a title size.
20160226 07:13:00< vultraz> 20 is title size
20160226 07:13:08< celticminstrel> Yes.
20160226 07:13:34< celticminstrel> Anything larger than a title size is still a title size. :P
20160226 07:15:03< ancestral> Images coming shortly…
20160226 07:21:00< ancestral> http://mproud.com/wesnoth/wesnoth-mac.jpg
20160226 07:21:25< ancestral> Top is 1.13.x
20160226 07:21:34< ancestral> Bottom is 1.13.x with font workaround
20160226 07:22:16< ancestral> Do you realize how hard it is to read “melee-blade”? and discern the various numbers on the screen?
20160226 07:22:27< ancestral> (In top image)
20160226 07:23:27< celticminstrel> It's not hard at all.
20160226 07:24:08< ancestral> Okay, now move 6 inches away
20160226 07:24:32< celticminstrel> Um, I'm about 3ft away and it's not hard.
20160226 07:24:36< ancestral> We need to remind ourselves
20160226 07:24:41< vultraz> ancestral: tbf, if you look at dota 2 the font isn't really bigger
20160226 07:24:45< vultraz> and that has 12 million players
20160226 07:24:54< ancestral> We know where to find info, and what it says without reading it
20160226 07:25:03< ancestral> vultraz: But the font is thicker
20160226 07:25:13< ancestral> Look, the font workaround is better
20160226 07:25:46< ancestral> But without it, the anti-aliasing is correctly trying to render the fonts at less than 1 pixel thick
20160226 07:26:01< ancestral> That’s what’s causing the fonts to not appear white
20160226 07:26:07< ancestral> In the top image
20160226 07:26:12< ancestral> “Arvith” is bold
20160226 07:26:18< ancestral> It’s readable; it’s pure white
20160226 07:26:36< ancestral> But the other text that’s smaller is not nearly as white
20160226 07:26:39< celticminstrel> I can see TC is fine in that image, too.
20160226 07:26:52< vultraz> ancestral: but with the font workaround, it looks the same as shadowm on linux
20160226 07:26:58< celticminstrel> By the way, the four sizes you listed - what are they for?
20160226 07:27:10< ancestral> Normal, small, tiny and very tiny
20160226 07:27:40< ancestral> I think there’s a dev bias ;-)
20160226 07:27:45< celticminstrel> That's... not very helpful.
20160226 07:27:47< vultraz> what do you mean
20160226 07:27:59< ancestral> We know where to find info, and what it says without reading it
20160226 07:28:26< celticminstrel> The two versions do seem to look different, but only slightly.
20160226 07:28:48< celticminstrel> I don't really think I can say that I know what it says without reading it. I do know where to find it, sure.
20160226 07:28:52< ancestral> The top one, it’s much smaller, the text is harder to read. It’s thinner.
20160226 07:29:25< celticminstrel> "melee-blade" doesn't seem white in either of them (but maybe it's not supposed to be?)
20160226 07:29:59< ancestral> That’s not supposed to be white. It’s supposed to be tan or something
20160226 07:30:01< celticminstrel> Yeah, I guess "10-4 sword" is a bit easier to read in the bottom one, for example.
20160226 07:30:02-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20160226 07:30:29< celticminstrel> Is the only difference sans vs serif?
20160226 07:30:37< ancestral> No serifs here
20160226 07:31:13< ancestral> Deja Vu Sans renders wider. It’s easier to read, certainly.
20160226 07:31:15< celticminstrel> Or uh... I mean, is it a different font, or a different font size, or what?
20160226 07:31:22< ancestral> Different font
20160226 07:31:39< celticminstrel> Wasn't it supposed to be serif? I forget.
20160226 07:31:40< ancestral> Anyway, what I’m trying to say is
20160226 07:31:48< ancestral> No, sorry, the screenshot I had before
20160226 07:32:09< ancestral> That was changed to illustrate the difference
20160226 07:32:25< ancestral> This is like the top image (http://mproud.com/wesnoth/gameplay-fonts.png)
20160226 07:33:16< celticminstrel> Okay, so, for some reason these radiobuttons are now behaving like checkboxes...
20160226 07:34:13< ancestral> (This is the two currently; top bad, bottom better: http://mproud.com/wesnoth/wesnoth-mac.jpg)
20160226 07:34:25< ancestral> What I’m trying to say is, there’s wasted space here
20160226 07:34:50< ancestral> Lots of brown in that sidebar
20160226 07:35:39< celticminstrel> I guess, yeah. Some of that space is "reserved" for things this specific unit doesn't have, like traits or abilities.
20160226 07:36:30< vultraz> ancestral: bottom too thick
20160226 07:36:49< ancestral> Okay, let me take another series of screenshots, just a portion
20160226 07:37:20< celticminstrel> Whyyy are they behaving like checkboxes... :|
20160226 07:48:52< ancestral> vultraz, celticminstrel: http://mproud.com/wesnoth/cadell.png
20160226 07:49:17< vultraz> ancestral: something midway between the two
20160226 07:49:19< ancestral> Obviously, the text resulting in the ellipsis would get fixed
20160226 07:49:54< ancestral> vultraz: There’s barely anything available between the two
20160226 07:50:10< ancestral> Wesnoth is a game about numbers
20160226 07:50:41< ancestral> I don’t see why they need to be very small
20160226 07:50:43< celticminstrel> Is the right your 18/14/13/11 version?
20160226 07:51:12< ancestral> Yes
20160226 07:51:26< ancestral> Without tweaks to line breaks
20160226 07:51:30< ancestral> etc.
20160226 07:51:36< vultraz> celticminstrel: if you make the default font size 18 title size will be enormous
20160226 07:52:02< celticminstrel> So the unit name is 18?
20160226 07:52:08< ancestral> Yes
20160226 07:52:16< celticminstrel> And the range/damage-type is 11?
20160226 07:52:29-!- boucman_work [~jrosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20160226 07:52:36< ancestral> I think that’s 13 instead of 10
20160226 07:53:12< celticminstrel> Ah, so it must be the terrain name that's 11.
20160226 07:53:12< ancestral> Okay, let me ask a question
20160226 07:53:25< celticminstrel> And most of the other text I guess is 14.
20160226 07:53:30< ancestral> What is the most important things you look at when playing?
20160226 07:53:33< celticminstrel> Or are just a few things 14...
20160226 07:54:09< ancestral> (celticminstrel: I don’t remember all the different sizes atm sorry. You can look at default.cfg if you want.)
20160226 07:54:28< celticminstrel> My initial objection to 18 is not, in fact, valid, since it's not the normal font size but rather only used for a very few things.
20160226 07:55:14< ancestral> For me, it’s HP and attacks/damage
20160226 07:55:17< ancestral> Like
20160226 07:55:35< celticminstrel> I was about to say that most of the sidebar is important, but...
20160226 07:55:35< ancestral> That’s most important to me. And maybe the type of unit, and unit alignment bonus after that.
20160226 07:55:44< celticminstrel> Attacks are certainly more important than most of the rest.
20160226 07:55:51< celticminstrel> HP/XP/MP also.
20160226 07:55:57< celticminstrel> XP less so, I suppose.
20160226 07:56:02< ancestral> XP is sorta nice, but not always critical, unless it’s about to level
20160226 07:56:22< ancestral> Movement is whatever, because I can figure that out when I am about to move the unit
20160226 07:56:25< celticminstrel> Maybe alignment bonus.
20160226 07:56:34< ancestral> So weapons are important
20160226 07:56:37< celticminstrel> Oh right, status effects as well, but those aren't text.
20160226 07:57:27< celticminstrel> I don't think I often look at the terrain defense in the sidebar.
20160226 07:57:43< celticminstrel> Normally when I'm worried about defense, it's with respect to a target space.
20160226 07:57:59< ancestral> http://wesnoth.io/bestiary/spearman
20160226 07:58:13< celticminstrel> The numbers next to the time of day are super-unimportant, yet larger than pretty much everything else.
20160226 07:58:18< ancestral> I want to put some of the icons on the screen, if there’s spae
20160226 07:58:19< ancestral> *space
20160226 07:58:50< ancestral> celticminstrel: Agreed on thtat.
20160226 07:59:00< celticminstrel> Ehh... I dunno...
20160226 07:59:10< celticminstrel> Those icons are pretty big.
20160226 07:59:17< ancestral> Then we’d render them smaller
20160226 07:59:24< celticminstrel> I guess.
20160226 07:59:41< ancestral> Here’s the thing
20160226 07:59:46< ancestral> It’s a wall of text
20160226 07:59:53< ancestral> And it’s kinda maybe a little bit smallish
20160226 08:00:05< ancestral> Most good games don’t do this.
20160226 08:00:10-!- iceiceice [~chris@wesnoth/developer/iceiceice] has joined #wesnoth-dev
20160226 08:00:26< ancestral> Most good games use icons, or graphics
20160226 08:00:27< celticminstrel> There's plenty of space on the right, so if the icons were 1/2 or 1/3 the dimensions, then it would probably fit fine.
20160226 08:00:33< ancestral> And honestly
20160226 08:00:34< celticminstrel> Speaking of icons...
20160226 08:00:38< ancestral> The sidebar can be wider
20160226 08:00:46< ancestral> If anything, guess what
20160226 08:00:57< ancestral> Our monitors got wider in the last 10 years
20160226 08:01:12< celticminstrel> I'm not saying it can't be wider, but it doesn't need to be wider to fit attack icons.
20160226 08:01:18< celticminstrel> Probably.
20160226 08:01:53< iceiceice> ancestral, i think the best thing would be like
20160226 08:02:01< iceiceice> if the side bar stuff is shown in some kind of mouse over popup
20160226 08:02:05< iceiceice> next to the unit or something
20160226 08:02:18< ancestral> That’s been talked about before
20160226 08:02:20< iceiceice> i think thats more like what other games do
20160226 08:02:35< iceiceice> is it just not enoguh space?
20160226 08:02:37< ancestral> It would require more coding
20160226 08:02:55< ancestral> So feel free to go do it ;-)
20160226 08:03:04< iceiceice> i dont want to work in gui2 though
20160226 08:03:12< ancestral> ^ that’s not a good sign
20160226 08:03:39< celticminstrel> I don't think that could be done in GUI2.
20160226 08:03:44< ancestral> (not you, I just meant, if gui2 isn’t approachable… 😉 )
20160226 08:03:50< celticminstrel> GUI2 is exclusively modal dialogs.
20160226 08:03:56< ancestral> Ah
20160226 08:04:01< iceiceice> celticminstrel, that's what i thoguht but i think it also does tooltips?
20160226 08:04:02< ancestral> What does the tool tips?
20160226 08:04:03< iceiceice> i could be wrong though
20160226 08:04:14< iceiceice> i thought mordante had a special subsystem for that and that alone or soething
20160226 08:04:14< ancestral> GUI1?
20160226 08:04:18< iceiceice> maybe i'm wrong thouhg
20160226 08:04:20< celticminstrel> That's true, it does seem to have something about tooltips.
20160226 08:04:26< iceiceice> the only time i worked on gui2 really
20160226 08:04:29< iceiceice> was when i made certain dialogs
20160226 08:04:31< iceiceice> also
20160226 08:04:32< ancestral> Might be in reports.cpp
20160226 08:04:34< celticminstrel> Though they're not really tooltips in the conventional sense - they appear at the bottom of the screen.
20160226 08:04:47< iceiceice> when i made it so that you can click hyperlinks in text
20160226 08:04:50< ancestral> celticminstrel: Huh?
20160226 08:05:00< iceiceice> that required like 3 days of staring at gui2 code and trying to see through to the bottom
20160226 08:05:04< ancestral> Next time hover over the tod graphic
20160226 08:05:27< celticminstrel> Anyway, speaking of icons, there's these which are still unused: https://github.com/wesnoth/wesnoth/tree/master/images/icons/profiles
20160226 08:05:29< iceiceice> there's some really bizarre part of it that relies of the Formula AI stuff
20160226 08:05:45< celticminstrel> The set is incomplete, but...
20160226 08:06:01< celticminstrel> I think the damage types, at least, are complete.
20160226 08:06:11< ancestral> I like some of them
20160226 08:06:24< ancestral> ‘Am’ is worthless to some languages
20160226 08:06:49< celticminstrel> Yeah, I consider all the ones with letters to be placeholders - they're the reason I said it was incomplete.
20160226 08:06:58< ancestral> I just feel like
20160226 08:07:08< ancestral> Wesnoth is getting smaller
20160226 08:07:21< ancestral> I say that meaning comparatively-speaking
20160226 08:07:36< ancestral> Forgetting completely about hi-dpi/retina stuff
20160226 08:07:39< celticminstrel> The main ones there would be the damage types, ranged/melee, and maybe terrain defence.
20160226 08:07:42< celticminstrel> ^defense
20160226 08:07:44< ancestral> Smaller monitors have larger resolutions
20160226 08:09:02< celticminstrel> (Of course, if we're going to go and use icons for ranges and damage types, there needs to be a way for add-ons to specify those icons for new ranges and damage types.)
20160226 08:09:51< ancestral> (And larger monitors have even bigger resolutions)
20160226 08:11:34< celticminstrel> Of course, we also have mobiles tablets now, too.
20160226 08:11:51< celticminstrel> But, I guess those need a completely different theme anyway.
20160226 08:12:02< ancestral> No one is building from this repo for mobile
20160226 08:12:10< ancestral> afaik
20160226 08:12:12< celticminstrel> That's true too.
20160226 08:13:01< ancestral> I want Wesnoth to be more approachable to new players, and eliminating walls of small text is a good start
20160226 08:13:21< ancestral> At least, I feel like that might be the case
20160226 08:13:34< celticminstrel> Ohhh... this wasn't working because I forgot a small little & somewhere. >_>
20160226 08:13:59< ancestral> That zoom slider, for example
20160226 08:14:06< ancestral> Is barely usable
20160226 08:14:25< ancestral> It’s so small, it’s hard to tell what percent, and I don’t know what the tick marks mean
20160226 08:14:27< celticminstrel> Yeah, it needs to be less fine-grained.
20160226 08:14:39< celticminstrel> Last I looked it went up by increments of 1%.
20160226 08:14:57< ancestral> In my theme I’m playing around with, I abolished it
20160226 08:15:08< ancestral> and have + and – zoom buttons instead
20160226 08:15:59< celticminstrel> What's your increment?
20160226 08:16:09< ancestral> I can’t control that from the themes
20160226 08:16:17< celticminstrel> So uh...
20160226 08:16:31< ancestral> Whatever is normal for the zoom out and zoom in action
20160226 08:16:35< ancestral> 5% or something?
20160226 08:16:39< celticminstrel> Okay?
20160226 08:16:44< ancestral> (Probably too fine)
20160226 08:16:47 * celticminstrel has no idea.
20160226 08:17:00< celticminstrel> 5% is the minimum, but 25% would probably be quite sufficient.
20160226 08:18:22-!- iceiceice is now known as masked_man
20160226 08:19:56< ancestral> What is N()?
20160226 08:20:04< celticminstrel> Hm?
20160226 08:20:15< ancestral> N_()
20160226 08:20:25< celticminstrel> Oh, I think that's something about gettext.
20160226 08:20:40< celticminstrel> Translatable strings with plurals.
20160226 08:21:41< ancestral> Okay
20160226 08:22:12< celticminstrel> Okay, the dialog works, but the preferences aren't saved. Clearly I missed something somewhere.
20160226 08:22:59< ancestral> Searching for “zoom_at_min” in the project…
20160226 08:24:36< ancestral> MinZoom = 4; MaxZoom = 288
20160226 08:25:37< celticminstrel> Those are weird numbers...
20160226 08:26:04-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160226 08:27:26< ancestral> 	const int zoom_amount = 4;
20160226 08:28:09< ancestral> Default is 1.0
20160226 08:28:35< ancestral> I think each step is 4
20160226 08:28:48< ancestral> 72 steps to zoom in
20160226 08:29:04< ancestral> Hmm, maybe not
20160226 08:30:11-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 244 seconds]
20160226 08:31:29< ancestral> 4 is the steps. 288 is max. 4 is min. I don’t know what default is.
20160226 08:31:51< celticminstrel> Who came up with these numbers?
20160226 08:31:55< ancestral> Maybe 72?
20160226 08:32:00< celticminstrel> Why are they not multiples of 5?
20160226 08:32:02< ancestral> I might be the number of pixels wide
20160226 08:32:14< celticminstrel> Hmm, I suppose...
20160226 08:32:24< ancestral> It might be easier to scale in even numbers and such
20160226 08:32:41< celticminstrel> If it is the number of pixels, and 72 is the default, then it's referring to the size of a hex.
20160226 08:32:54< celticminstrel> I think hexes are 72x72, right?
20160226 08:34:00< ancestral> 17 … DEFAULT … 53ish
20160226 08:34:13< ancestral> 70 * 4
20160226 08:34:14< ancestral> Bingo
20160226 08:34:25< celticminstrel> ?
20160226 08:34:26< ancestral> Yes, default must be 72
20160226 08:34:36< ancestral> 17 zoom outs to the minimum (4)
20160226 08:34:47< ancestral> 53 or 54 zoom ins to the maximum (288)
20160226 08:34:47< irker372> wesnoth: Celtic Minstrel wesnoth:master 2e36ed230fa7 / / (4 files in 3 dirs): Use widget groups in the advanced graphics options dialog https://github.com/wesnoth/wesnoth/commit/2e36ed230fa77f07a9b098c48fa5c40b8965c1f7
20160226 08:34:52< celticminstrel> Okay.
20160226 08:35:11< ancestral> So really, when zooming in, the steps should be larger
20160226 08:35:19< ancestral> In other words, linear zoom makes less sense
20160226 08:35:35< celticminstrel> Hmm.
20160226 08:35:39< ancestral> Like, the difference from 50%-60% is more profound than from 150% to 160%
20160226 08:36:30< celticminstrel> The range is around 5% up to 400%.
20160226 08:36:44< celticminstrel> What sort of steps would you propose?
20160226 08:36:44< ancestral> A lot of graphics editors actually skip small increments at larger numbers
20160226 08:37:07< ancestral> Hmm
20160226 08:37:21< ancestral> Give me a moment to see if I can find the formula
20160226 08:38:31< ancestral> So right now it’s just n + 4
20160226 08:40:46-!- Kwandulin [~Miranda@p200300760F0BC585009A1BE2EEA60AA1.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160226 08:40:59< ancestral> I think we need a multiplication thing, like n*1.1
20160226 08:41:28< celticminstrel> I was expecting concrete values, actually.
20160226 08:41:30< ancestral> Maybe n * 1.25
20160226 08:41:48< ancestral> Well, linear is poor at higher values
20160226 08:41:51< celticminstrel> Why does git keep complaining about unstashed changes in the middle of a rebase... :|
20160226 08:41:52< ancestral> So okay
20160226 08:43:31< masked_man> ancestral, i would guess the main thing is you would sort of like the number to divide 72 evenly or somewhat evenly
20160226 08:43:44< masked_man> because otherwise you will get nonuniform blurring across the sprite i guess
20160226 08:44:15< ancestral> 72 80 90 102 116 132 150 170 192 216 242 270
20160226 08:44:39< celticminstrel> I was thinking concrete percentages, but I guess that works too.
20160226 08:44:43< masked_man> i mean the sprite should look much better at 144 than at 132 or 150
20160226 08:44:54< ancestral> True
20160226 08:45:10< ancestral> good point
20160226 08:45:16< ancestral> 72 144 288
20160226 08:45:23< ancestral> Okay let’s fill the rest in
20160226 08:45:51< ancestral> 72 108 144 180 288
20160226 08:46:11< ancestral> Maybe a little more granular? okay
20160226 08:46:30< masked_man> wait so is the idea you dont just want a linear progression?
20160226 08:46:37< masked_man> i mean you could just take steps of 72 / 4 
20160226 08:46:40< masked_man> or 72 / 8
20160226 08:46:43< ancestral> We do that currently
20160226 08:46:47< masked_man> hmm ok
20160226 08:46:59< ancestral> masked_man: Do you use GIMP or Photoshop or anything like that?
20160226 08:47:10-!- masked_man is now known as iceiceice
20160226 08:47:12< iceiceice> hehe sorry
20160226 08:47:21< ancestral> :-{
20160226 08:47:24< ancestral> :-P
20160226 08:47:31< iceiceice> i forgot i chnged my nick
20160226 08:48:01< iceiceice> i use gimp sometimes
20160226 08:48:04< iceiceice> and inkscape
20160226 08:48:07< iceiceice> i like inkscape
20160226 08:48:07< ancestral> Perfect
20160226 08:48:10< ancestral> Inkscape
20160226 08:48:12< ancestral> When you zoom in
20160226 08:48:16< ancestral> What are the zoom percentages?
20160226 08:48:29< ancestral> It’s not 1000 1100 1200 1300 1400 1500…
20160226 08:48:36< ancestral> Like
20160226 08:48:42< celticminstrel> GIMP is a bit terrible, but I use it because it's free.
20160226 08:48:45< ancestral> It’s probably 100% increments up until 1000
20160226 08:48:53< ancestral> And then from there it’s probably 1000 increments
20160226 08:48:57< iceiceice> ok when i boot up inkscape
20160226 08:48:59< iceiceice> it is at 35%
20160226 08:49:00< ancestral> amirite?
20160226 08:49:03< ancestral> Right but
20160226 08:49:04< iceiceice> then 50%
20160226 08:49:08< iceiceice> then 70%
20160226 08:49:11< iceiceice> then 99%
20160226 08:49:15< iceiceice> then 140%
20160226 08:49:18< iceiceice> then 198 %
20160226 08:49:21< iceiceice> then 280%
20160226 08:49:28< iceiceice> which seems... a bit crackheaded honestly
20160226 08:49:31< celticminstrel> Uhh, what.
20160226 08:49:36< celticminstrel> Those do seem really weird.
20160226 08:49:38< ancestral> 15, 20, 30, 50, 60, 90
20160226 08:49:38< iceiceice> finally 396%
20160226 08:49:46< celticminstrel> I happen to have GIMP open right now.
20160226 08:49:52< ancestral> Does anyone get what I’m trying to say herE?
20160226 08:50:03< ancestral> When you zoom in you want to get higher increments
20160226 08:50:14< celticminstrel> At a glance, it looks like it's going by powers of two.
20160226 08:50:42< ancestral> The difference between 100% and 120% is more significant than 1000% and 1020%
20160226 08:51:05< celticminstrel> 6.25%, 12.5%, 25%, 50%, 100%, 200%, 400%, 800%, 1600%
20160226 08:51:06< ancestral> If I’m at 1000% and I want to zoom in further
20160226 08:51:15< ancestral> I want to zoom much further in
20160226 08:51:18< ancestral> Not 20%
20160226 08:51:27< ancestral> (+20% I mean)
20160226 08:51:38< ancestral> celticminstrel: Right
20160226 08:51:46< iceiceice> so here's one thing you could do
20160226 08:52:18< iceiceice> you could say the sequence works in groups of 4
20160226 08:52:26< celticminstrel> So, something like 9 - 18 - 36 - 72 - 144 - 288, maybe.
20160226 08:52:39< iceiceice> yeah or you could do that
20160226 08:53:01< celticminstrel> iceiceice: Were you in the middle of saying something?
20160226 08:53:04< ancestral> Here’s what I suggest
20160226 08:53:05< ancestral> 18 21 24 30 36 45 54 72 90 108 126 144 180 216 288
20160226 08:53:25< ancestral> Increments 3 3 6 6 9 9 18 18 18 36 36 72
20160226 08:53:41< iceiceice> i think what i was going to say is too complicated
20160226 08:53:49< ancestral> Something along those lines
20160226 08:53:56< celticminstrel> 180 is 72*2.5.
20160226 08:54:01< ancestral> I can express that in an equation
20160226 08:54:07< celticminstrel> So 250%.
20160226 08:54:08< iceiceice> ancestral, one thing to consider is that, if you guys move to use the xBRZ for zooming mostly,
20160226 08:54:13< iceiceice> or a similar method,
20160226 08:54:19< iceiceice> that thing works only for integer multiples of the sprite
20160226 08:54:28< iceiceice> you can't actually do like "150%"
20160226 08:54:35< celticminstrel> Then you've got 125%, 150%, 175%.
20160226 08:54:46< ancestral> Well, 200% is 144 pixels wide. That is a big jump
20160226 08:54:54< iceiceice> yeah
20160226 08:55:08< ancestral> I mean, I might be okay with it
20160226 08:55:08< celticminstrel> iceiceice: What about scaling down then?
20160226 08:55:15< ancestral> But it sounds like some people want something in between
20160226 08:55:17< iceiceice> but non-fractional zooming of pixel art is inherently squirrely, imo
20160226 08:55:25< iceiceice> *non-integral i mean
20160226 08:55:42< celticminstrel> iceiceice: Also, if xBRZ doesn't work for fractions, what currently happens when you attempt to do that?
20160226 08:55:48< iceiceice> celticminstrel, when we scale things down, i think we just use an averaging method
20160226 08:55:58< iceiceice> like in that case its many pixels being compressed to a single one
20160226 08:56:05< iceiceice> so there's not that much ambiguity about how to do it
20160226 08:56:10< ancestral> Maybe even 7 steps is too many
20160226 08:56:11< celticminstrel> Ah.
20160226 08:56:20< ancestral> 50 is wayyyyyyyyyy too many
20160226 08:56:24< iceiceice> i mean yo could make some rule like "bold / bright colors count more when we zoom in"
20160226 08:56:31< ancestral> But 2 might be too few
20160226 08:56:35< iceiceice> but i think we just do a simple average
20160226 08:56:44< iceiceice> *zoom out i guess i meant to say
20160226 08:56:55< iceiceice> ancestral, the way it works right now in 1.13, unless someone changed it,
20160226 08:57:07< iceiceice> is basically it tried to use xBRZ if possible,
20160226 08:57:16< iceiceice> if you are exactly at 200% or 300% it doesnt do any linear filtering
20160226 08:57:22< iceiceice> if you are at 250%,
20160226 08:57:29< iceiceice> then what it does is, uses xBRZ to get to 200%
20160226 08:57:38< iceiceice> then linear filtering (at 125% scale I guess) to get to 250%
20160226 08:57:47< iceiceice> like, whatever is left over it uses linear filtering for
20160226 08:58:06< celticminstrel> Is that why there's xBRZ+Linear and also xBRZ+NN in the options?
20160226 08:58:10< iceiceice> acutally you can pick if the leftover is handled by linear or nearest neighbor
20160226 08:58:11< iceiceice> yeah
20160226 08:58:23< celticminstrel> So, if you choose the latter, it uses NN instead of linear to get the fractional part.
20160226 08:58:41< iceiceice> yes
20160226 08:58:47< vultraz> what are you guys talking about?
20160226 08:58:51< iceiceice> if you are at xBRZ+NN and zoomed at 150%, it doesn't use xBRZ at all
20160226 08:58:57< iceiceice> because doubling would be too much already
20160226 08:59:00< celticminstrel> vultraz: The map zooming.
20160226 08:59:04< vultraz> ah
20160226 08:59:12< ancestral> So if it was: {18} {24} {36} 54 [72] 108 [144] [288]
20160226 08:59:24< vultraz> doesn't ancestral want that to be fractional not arbitrary? 
20160226 09:00:03< ancestral> That would be 25% 33% 50% 75% 100% 150% 200% 400%
20160226 09:00:17< iceiceice> yeah that looks pretty good
20160226 09:00:19< iceiceice> imo
20160226 09:00:22< celticminstrel> So 25%/33%/50%/75%/100%/150%/200%/400%.
20160226 09:00:51< iceiceice> you could add "300%" we can render that nicely
20160226 09:00:54< iceiceice> up to you though
20160226 09:00:57< ancestral> Sure
20160226 09:01:00< ancestral> or remove 150%
20160226 09:01:05< ancestral> Depends what people want
20160226 09:01:08< celticminstrel> I think 150% is okay.
20160226 09:01:24< celticminstrel> How many is that if you have both... nine levels...
20160226 09:01:26< ancestral> 216 is 300%
20160226 09:01:38< ancestral> Yeah it would render nicely
20160226 09:02:21< celticminstrel> It would be ideal if the ticks on the zoom slider indicated the possible zooms, and the slider snapped into those locations.
20160226 09:02:29< ancestral> ^ yes
20160226 09:03:20< celticminstrel> BTW, iceiceice, did you say the sidebar tooltips are handled by GUI2? I assumed they weren't.
20160226 09:03:43< ancestral> reports.cpp
20160226 09:03:51< ancestral> Has some tooltips
20160226 09:04:21< celticminstrel> GUI2 tooltips (or what I was thinking of when that was brought up) are what you see in Preferences.
20160226 09:05:36< iceiceice> i dont know for sure
20160226 09:08:22< iceiceice> ancestral, before you said people talked about putting all the side bar info into a popup or something,
20160226 09:08:27< iceiceice> what was the argument against that?
20160226 09:08:44< ancestral> Hmm
20160226 09:08:53< ancestral> I think the work involved
20160226 09:08:57< iceiceice> ohh
20160226 09:09:00< iceiceice> ok now i see the log again
20160226 09:09:02< ancestral> But also, comparing the selected unit with others
20160226 09:09:03< ancestral> So like
20160226 09:09:14< ancestral> Now I remember
20160226 09:09:21< ancestral> iceiceice: It was more, I want to attack another unit
20160226 09:09:43< ancestral> I want to see the stats instead of going for the attack, clicking “stats” and looking
20160226 09:09:55< ancestral> Think Civilization or similar games
20160226 09:10:05< ancestral> It tells you the odds of winning on the main screen without any dialogs
20160226 09:10:13< iceiceice> yeah
20160226 09:10:19< iceiceice> i mean i like that in wesnoth you can see the stats
20160226 09:10:21< iceiceice> like in detail
20160226 09:10:27< iceiceice> in Civ its very opaque wtf those stats mean
20160226 09:10:27< ancestral> Right, and that’s great
20160226 09:10:30< ancestral> But
20160226 09:10:36< celticminstrel> I think there's a theme with two sidebars, one containing the selected unit and one containing the hovered unit.
20160226 09:10:37< ancestral> Sometimes players just want to know the odds of winning
20160226 09:10:47< ancestral> Without tons of stats, without clickingt
20160226 09:10:59< ancestral> celticminstrel: There is
20160226 09:11:01< ancestral> I’m not a fan
20160226 09:11:09< ancestral> It’s confusing to me
20160226 09:11:18< celticminstrel> I didn't really try it out, just looked at it briefly.
20160226 09:11:25< ancestral> But there is a [selected_unit_*] tag
20160226 09:11:28< iceiceice> so what would be ideal
20160226 09:11:28< iceiceice> like
20160226 09:11:37< ancestral> So theoretically someone could do something with that
20160226 09:11:42< iceiceice> when you click the unit, some buttons show up beneath it
20160226 09:11:47< iceiceice> with an icon for each attack type?
20160226 09:11:57< iceiceice> and you select the attack type 
20160226 09:11:59< iceiceice> and then click an enemy?
20160226 09:12:06< ancestral> Anything is possible
20160226 09:12:08< iceiceice> and then it happens?
20160226 09:12:16< iceiceice> or, you select stats somehow
20160226 09:12:22< ancestral> At a certain point
20160226 09:12:27< ancestral> This becomes Wesnoth 2.0 ;-)
20160226 09:12:47< ancestral> Or GUI3
20160226 09:17:55 * celticminstrel zzzzzz
20160226 09:19:16-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has quit [Quit: KABOOM! It seems that I have exploded. Please wait while I reinstall the universe.]
20160226 09:40:59-!- ancestral [~ancestral@97.116.184.84] has quit [Quit: i go nstuf kthxbai]
20160226 10:04:51-!- aidanhs [~aidanhs@81.4.110.234] has quit [Ping timeout: 248 seconds]
20160226 10:05:52-!- aidanhs [~aidanhs@81.4.110.234] has joined #wesnoth-dev
20160226 10:08:09-!- zombah [~zombah@2a02:28:3:1:214:4fff:fe47:5920] has joined #wesnoth-dev
20160226 10:16:52-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20160226 10:19:01-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 255 seconds]
20160226 10:19:02-!- wedge010 is now known as wedge009
20160226 10:39:20-!- mjs-de [~mjs-de@f049128037.adsl.alicedsl.de] has joined #wesnoth-dev
20160226 10:43:49-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has joined #wesnoth-dev
20160226 11:08:26-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160226 11:12:33-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 240 seconds]
20160226 11:35:02-!- irker372 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20160226 11:53:00< vultraz> dammit
20160226 11:53:12< vultraz> display::flip() is more important than I thought
20160226 11:55:43< vultraz> or
20160226 11:55:44< vultraz> hm
20160226 11:55:47< vultraz> maybe it's video::flip
20160226 11:55:57< vultraz> CVideo
20160226 12:13:52< vultraz> Aginor: this confuses me
20160226 12:14:07-!- horrowind [~Icedove@2a02:810a:8380:834:21b:fcff:fee3:c3ff] has joined #wesnoth-dev
20160226 12:18:04< vultraz> Aginor: I can't figure out where to put the calls to video.flip
20160226 12:24:07-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160226 12:30:32-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has joined #wesnoth-dev
20160226 12:39:21< iceiceice> why do you have to call that explicitly?
20160226 12:39:49< iceiceice> shouldn't it just be done once per frame by the main lop
20160226 12:39:51< iceiceice> *loop
20160226 12:46:09-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20160226 13:02:34< vultraz> iceiceice: elaborate
20160226 13:03:01< iceiceice> i guess i dont know how wesnoth's graphics engine really works, maybe its more complicated
20160226 13:03:45< iceiceice> why would you have to flip the display more than once per frame?
20160226 13:04:19< iceiceice> like, why shouldyou be thinking about where to put multiple calls to this thing
20160226 13:04:24< iceiceice> shouldn't there just be one call, in each main loop
20160226 13:04:36< iceiceice> but i dont know, maybe the gui needs a separate flip for some reason
20160226 13:04:37< vultraz> well video.flip in sdl2 basically only calls SDL_RenderPresent
20160226 13:04:52< vultraz> display.flip is a different thing but it also calls video.flip
20160226 13:05:02< vultraz> SDL_RenderPresent isn't supposed to be every frame... is it?
20160226 13:05:09-!- gfgtdf [~chatzilla@x50ab6f58.dyn.telefonica.de] has joined #wesnoth-dev
20160226 13:06:39< gfgtdf> iceiceice: afaik every gui1 dialog implements its own main loop that calls events::pump(), video.flip() adn all that stuff.
20160226 13:06:52< iceiceice> oh
20160226 13:12:07< vultraz> gfgtdf: do you think it would be possible to convert that progress bar that displays when you connect to the mp server to gui2
20160226 13:13:43< gfgtdf> vultraz: i i think that shoulds be possible (assuming that these is a prograssbar wiget in gui2).
20160226 13:13:54< gfgtdf> vultraz: even if not it most likeley wouldn be too hard ot make one
20160226 13:13:58< vultraz> gfgtdf: yes there is and a network transmission dialog
20160226 13:14:12< gfgtdf> vultraz: in gui2 ?
20160226 13:14:14< vultraz> gfgtdf: yes
20160226 13:14:44< vultraz> gfgtdf: but the current code is complicated and i can't figure out how it works
20160226 13:14:50< vultraz> the gui1 code*
20160226 13:15:44< vultraz> gfgtdf: since it has to do with network should i talk to loonycyborg maybe?
20160226 13:18:38< gfgtdf> vultraz: hmm the gui2 network transmission dialog uses boost asio
20160226 13:18:51< vultraz> oh?
20160226 13:18:57< vultraz> hm
20160226 13:18:59< vultraz> maybe loonycyborg is dealing with it already then
20160226 13:19:08< vultraz> I mean converting the connect progressbar
20160226 13:19:13< gfgtdf> vultraz: hmmm no mostlikely it is used to connect to the addon server
20160226 13:19:16< loonycyborg> I made gui2 transmission dialog
20160226 13:19:26< loonycyborg> it's for addons only though
20160226 13:19:44< loonycyborg> at least network part of it
20160226 13:19:51< loonycyborg> shadowm made ui
20160226 13:20:39< vultraz> loonycyborg: it can't be for any network transmission?
20160226 13:21:16< loonycyborg> it can be, but I'm not sure about exact details
20160226 13:21:41< gfgtdf> vultraz: if you want to make a gui2 trnamsission dialog version for the old network code you can mostlikeley copy the netowrk_transnission cod and replave the tnetwork_transmission::pump_monitor::process code with soemthing that calls network::receive_data liek in dialogs.cpp
20160226 13:35:17-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has joined #wesnoth-dev
20160226 13:51:24-!- horrowind [~Icedove@2a02:810a:8380:834:21b:fcff:fee3:c3ff] has quit [Quit: horrowind]
20160226 14:03:48-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20160226 14:41:50-!- gfgtdf [~chatzilla@x50ab6f58.dyn.telefonica.de] has quit [Quit: ChatZilla 0.9.92 [Firefox 44.0.2/20160210153822]]
20160226 14:54:12-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has quit [Quit: Connection closed for inactivity]
20160226 15:06:36-!- horrowind [~Icedove@2a02:810a:8380:834:21b:fcff:fee3:c3ff] has joined #wesnoth-dev
20160226 15:48:41-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: mattsc]
20160226 15:49:34-!- oldlaptop [~quassel@192.183.14.123] has quit [Ping timeout: 240 seconds]
20160226 16:08:04-!- Kwandulin [~Miranda@p200300760F0BC5B7B870075D8CD7A178.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160226 16:09:20-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20160226 16:22:57-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160226 16:23:01-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160226 16:23:07-!- Greg-Bog_ [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160226 16:24:12-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has joined #wesnoth-dev
20160226 16:24:12-!- prkc [~prkc@46.166.188.242] has quit [Remote host closed the connection]
20160226 16:26:03-!- iceiceice [~chris@wesnoth/developer/iceiceice] has quit [Ping timeout: 276 seconds]
20160226 16:28:03-!- ancestral [~ancestral@8.sub-70-197-237.myvzw.com] has joined #wesnoth-dev
20160226 16:30:53-!- Kwandulin [~Miranda@p200300760F0BC5B7B870075D8CD7A178.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160226 16:43:51-!- zombah [~zombah@2a02:28:3:1:214:4fff:fe47:5920] has quit [Quit: Leaving]
20160226 16:46:38-!- horrowind [~Icedove@2a02:810a:8380:834:21b:fcff:fee3:c3ff] has quit [Quit: horrowind]
20160226 16:48:49-!- ancestral [~ancestral@8.sub-70-197-237.myvzw.com] has quit [Quit: i go nstuf kthxbai]
20160226 17:13:41-!- nurupo [~nurupo.ga@unaffiliated/nurupo] has quit [Ping timeout: 248 seconds]
20160226 17:13:41-!- clavi [~clavi@163.172.10.77] has quit [Ping timeout: 248 seconds]
20160226 17:13:41-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has quit [Ping timeout: 248 seconds]
20160226 17:13:41-!- lobby [~wesnoth@wesnoth/bot/lobby] has quit [Ping timeout: 248 seconds]
20160226 17:13:52-!- lobby_ [~wesnoth@wesnoth/bot/lobby] has joined #wesnoth-dev
20160226 17:13:52-!- Topic for #wesnoth-dev: Wesnoth Developers Channel | >>> Want to help? Go here: http://r.wesnoth.org/t42911 (and thanks!) <<< | Logs: http://irclogs.wesnoth.org | Bug tracker: http://bugs.wesnoth.org
20160226 17:13:52-!- Topic set by shadowm [~ignacio@wesnoth/developer/shadowm] [Sun Jan 17 05:24:23 2016]
20160226 17:13:52[Users #wesnoth-dev]
20160226 17:13:52[ _laco          ] [ Elsi      ] [ lobby_      ] [ shikadibot] 
20160226 17:13:52[ aeonchild      ] [ esr       ] [ Lohengramm  ] [ Soliton   ] 
20160226 17:13:52[ aeth           ] [ fendrin   ] [ loonycyborg ] [ stikonas  ] 
20160226 17:13:52[ Aginor         ] [ Gambit    ] [ matthiaskrgr] [ TC01      ] 
20160226 17:13:52[ AI0867         ] [ Greg-Bog_ ] [ mattsc      ] [ TheJJ     ] 
20160226 17:13:52[ aidanhs        ] [ Greywhind ] [ midzer      ] [ timotei_  ] 
20160226 17:13:52[ Appleman1234   ] [ heirecka  ] [ minzbonbon  ] [ tomreyn   ] 
20160226 17:13:52[ boucman_work   ] [ higgins   ] [ mjs-de      ] [ vincent_c ] 
20160226 17:13:52[ bumbadadabum   ] [ Ivanovic  ] [ molgrum     ] [ vultraz   ] 
20160226 17:13:52[ celticminstrel ] [ iwaim__   ] [ new_one     ] [ wedge009  ] 
20160226 17:13:52[ clavi          ] [ janebot   ] [ nurupo      ] [ Yaiyan    ] 
20160226 17:13:52[ Crendgrim      ] [ Jetrel    ] [ pydsigner   ] [ zookeeper ] 
20160226 17:13:52[ crimson_penguin] [ Jetrel_bot] [ quentinp    ] [ {V}       ] 
20160226 17:13:52[ DDR            ] [ kidneb    ] [ Ravana_     ] 
20160226 17:13:52[ elias          ] [ knotwork__] [ Rhonda      ] 
20160226 17:13:52[ EliDupree      ] [ legoktm   ] [ Samual      ] 
20160226 17:13:52-!- Irssi: #wesnoth-dev: Total of 61 nicks [0 ops, 0 halfops, 0 voices, 61 normal]
20160226 17:14:10-!- allefant [~allefant@allegro/developer/allefant] has joined #wesnoth-dev
20160226 17:14:13-!- Channel #wesnoth-dev created Tue Jan 27 05:28:41 2009
20160226 17:14:58-!- crimson_pingvin [~crimson_p@ec2.happyspork.com] has joined #wesnoth-dev
20160226 17:15:18-!- Irssi: Join to #wesnoth-dev was synced in 96 secs
20160226 17:15:19-!- Kwandulin [~Miranda@p200300760F0BC5B7482386174087B4AF.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160226 17:20:31-!- clavii [~clavi@163.172.10.77] has joined #wesnoth-dev
20160226 17:20:32< celticminstrel> Ravana_: Invalid tokens in a filter formula doesn't seem to crash the game. 
20160226 17:20:39-!- Netsplit *.net <-> *.split quits: crimson_penguin, esr, elias, nurupo, Yaiyan, clavi
20160226 17:20:39-!- crimson_pingvin is now known as crimson_penguin
20160226 17:20:39-!- crimson_penguin [~crimson_p@ec2.happyspork.com] has quit [Changing host]
20160226 17:20:39-!- crimson_penguin [~crimson_p@wesnoth/developer/crimsonpenguin] has joined #wesnoth-dev
20160226 17:20:46-!- nurupo_ [~nurupo.ga@unaffiliated/nurupo] has joined #wesnoth-dev
20160226 17:20:54-!- nurupo_ is now known as nurupo
20160226 17:21:14< Ravana_> maybe, I have only dealt with that in 1.12
20160226 17:21:21< celticminstrel> Was there some other formula thing that lead to crashes? Maybe the crash bit was actually fixed.
20160226 17:21:30< celticminstrel> It does terminate processing of the current event, though.
20160226 17:22:25< Ravana_> the underlying cause was empty $variable that was used in formula calculation, making wesnoth crash with message that illegal unary operator
20160226 17:22:31-!- louis94 [~~louis94@91.178.240.113] has joined #wesnoth-dev
20160226 17:23:12-!- Netsplit over, joins: Yaiyan
20160226 17:23:36< celticminstrel> I see.
20160226 17:27:52-!- esr [~esr@wesnoth/developer/esr] has joined #wesnoth-dev
20160226 17:28:54< celticminstrel> Actually, it seems the Lua engine is capturing the exception thrown by the formula engine.
20160226 17:30:09-!- boucman_work [~jrosen@wesnoth/developer/boucman] has quit [Ping timeout: 244 seconds]
20160226 17:33:12-!- Greg-Bog_ [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160226 17:35:07-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160226 17:54:58-!- allefant is now known as elias
20160226 18:07:59-!- louis94 [~~louis94@91.178.240.113] has quit [Ping timeout: 240 seconds]
20160226 18:13:46-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160226 18:27:11-!- oldlaptop [~quassel@50-37-53-169.mskg.mi.frontiernet.net] has joined #wesnoth-dev
20160226 18:29:41-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160226 19:02:38-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Ping timeout: 244 seconds]
20160226 19:22:29< mattsc> celticminstrel: thanks again for all the proof-reading
20160226 19:22:53< mattsc> I agree that the avoidance of trailing preposition you corrected sounded awkward
20160226 19:23:44< mattsc> Also, I have tested the comment on using the [aspect] syntax we discussed yesterday again now, and it does indeed work, so I removed that comment again.
20160226 19:24:33< mattsc> Not sure what I did before, but I did indeed test it.  Must have screwed up the test somehow.
20160226 19:24:58-!- ancestral [~ancestral@63.92.240.233] has joined #wesnoth-dev
20160226 19:25:00< mattsc> Also Modifying_AI_Componetns and AI_Recruitment are now ready.
20160226 19:26:26< mattsc> AI_recruitment was a lot more work than I had expected!  Apparently I had not gone over it myself yet after it was written by the original author.  The way it is, I only kind of cleaned up the formatting and the wording, but didn’t make any significant changes overall.  I think it’s “good enough” the way it is.
20160226 19:26:54< mattsc> 4 pages to go ...
20160226 19:34:01-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160226 19:41:31-!- ancestral [~ancestral@63.92.240.233] has quit [Quit: i go nstuf kthxbai]
20160226 19:46:08< celticminstrel> The new recruitment CA is pretty complicated.
20160226 19:46:23< mattsc> yep
20160226 19:47:07< celticminstrel> BTW, does [recruitment_instructions] work already in 1.12?
20160226 19:47:14< mattsc> yes
20160226 19:47:16< celticminstrel> And [recruitment] already not work?
20160226 19:47:25< mattsc> (very first line of that page)
20160226 19:47:45< mattsc> [recruitment] still works, but has not been the default since Wesnoth 1.8
20160226 19:47:47< celticminstrel> Oh, heh. I hadn't loaded the page yet (I was talking from when I looked at it before).
20160226 19:48:47< mattsc> I think … maybe they still used the old recruitment even when the RCA AI became the default.
20160226 20:16:30-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: mattsc]
20160226 20:24:55-!- trewe [~trewe@bl20-20-15.dsl.telepac.pt] has joined #wesnoth-dev
20160226 20:32:05< zookeeper> umm. didn't asserts use to produce a stacktrace in the debug log or something?
20160226 20:32:22< celticminstrel> There's no reason whatsoever why asserts would do that.
20160226 20:32:41< celticminstrel> Unless you have some sort of debug build that changes how asserts work, I suppose.
20160226 20:33:02< zookeeper> oh wait, i'm mixing up stuff with another project entirely
20160226 20:33:03< celticminstrel> An assert is normally "this condition is false? Panic, exit, shut down!"
20160226 20:34:00< fendrin> I thought that asserts give you a stacktrace. And building without them is the default for speed reasons.
20160226 20:34:28< celticminstrel> Why would they give you a stacktrace?
20160226 20:34:46< celticminstrel> The C standard library provides no means of accessing stack frames.
20160226 20:34:55< celticminstrel> And assert is part of the C standard library.
20160226 20:35:14< celticminstrel> If you want a stack trace in C, you need to dig into POSIX ot WinAPI.
20160226 20:35:45< fendrin> I guess my IDE catches the assert and produces a stacktrace?
20160226 20:36:08< celticminstrel> Quite possibly.
20160226 20:36:20< celticminstrel> In a POSIX environment, an assert produces SIGABRT.
20160226 20:36:31< celticminstrel> The IDE can probably install a handler for that which gives you a stack trace.
20160226 20:37:00< fendrin> maybe I only get it in a managed debug run
20160226 20:37:06< celticminstrel> In XCode (and, I believe, MSVC), it breaks into the debugger. Not quite the same as giving a stack trace, but it does allow you to see the stack.
20160226 20:37:31< fendrin> Yes
20160226 20:37:38< fendrin> Eclipse is behaving similar.
20160226 20:37:56< celticminstrel> Eclipse... I haven't used that for awhile...
20160226 20:38:13< fendrin> not the best ide for c++
20160226 20:38:24< fendrin> But getting better.
20160226 20:38:44< celticminstrel> I recall its indexing was much too slow, but it did have one nice feature - the ability to preview the expansion of a macro.
20160226 20:39:06< celticminstrel> (At least one.)
20160226 20:39:23< fendrin> It is still a bit slow. 
20160226 20:39:32-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20160226 20:39:47< fendrin> Although modern machines make it fast enough.
20160226 20:40:05< celticminstrel> It's much, much worse if you have a lot of macros. If you're using Boost.Preprocessor for anything complicated in Eclipse, you might need to disable the indexing.
20160226 20:41:32< fendrin> During the last few versions I had problems with getting it setuped properly for Wesnoth.
20160226 20:42:08< fendrin> Neither the cmake nor the sconsolidator methods gave me a good result.
20160226 20:42:14 * celticminstrel notes that if you want to add -ed to setup, you need to double the P... but, I think, you can just use "setup" as the past form.
20160226 20:43:24-!- prkc [~prkc@563BD80A.dsl.pool.telekom.hu] has joined #wesnoth-dev
20160226 20:44:33< fendrin> cmake works but suffers from the cmake build system not longer being properly maintained.
20160226 20:45:30< fendrin> The sconsolidator plugin somehow chocks on Wesnoth. When scanning dependencies it fails to terminate and eats all the ram.
20160226 20:47:32-!- Kwandulin [~Miranda@p200300760F0BC5B7482386174087B4AF.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160226 20:55:34-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection]
20160226 20:55:41-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev
20160226 20:59:22< celticminstrel> I wonder what would happen if you used something like [modify_ai]action,path=delete,aspect[aggression]
20160226 21:05:55< celticminstrel> The composite aspect vs standard aspect thing is suddenly confusing.
20160226 21:06:13< celticminstrel> "Any aspect listed at AiWML which is not specifically pointed out to be a composite aspect (and there are very few of those), is a standard aspect."
20160226 21:07:01< celticminstrel> If a composite aspect is one that contains facets, doesn't that mean all of them are composite?
20160226 21:07:12< celticminstrel> mattsc: ^
20160226 21:13:41< celticminstrel> You know, it might be "cheating" for the AI to take enemy recruit lists into account. Is there actually any way for a player to view the recruit list of another player?
20160226 21:14:12< celticminstrel> In MP, of course, you can know the enemy's recruit list simply because there's a limited set of options, but in campaigns, it could be anything.
20160226 21:16:03-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20160226 21:16:03-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Read error: Connection reset by peer]
20160226 21:17:16< zookeeper> that seems like the most acceptable form of "cheating", really
20160226 21:17:37< celticminstrel> Hm?
20160226 21:18:07< zookeeper> i don't see what could possibly require clarification in that :p
20160226 21:18:23< celticminstrel> I'm wondering why you think that.
20160226 21:19:09< zookeeper> because it seems like it can't really make the player feel like the AI is cheating
20160226 21:19:21< celticminstrel> Hmm, maybe.
20160226 21:19:53< celticminstrel> It actually makes me wonder if there should be a way to view enemy recruit lists.
20160226 21:20:47< zookeeper> in limited contexts, sure, like unmodified faction recruit lists in MP
20160226 21:21:10< zookeeper> in campaigns, i wouldn't, because the lists can contain all sorts of stuff the player isn't supposed to see
20160226 21:21:16< celticminstrel> I see.
20160226 21:21:32< zookeeper> hack units, units which could spoil a surprise, etc
20160226 21:21:32< celticminstrel> Any examples?
20160226 21:21:36< celticminstrel> Ah.
20160226 21:22:03< zookeeper> ah you pretty much see what the enemy can recruit anyway after the first turn
20160226 21:22:09< celticminstrel> Well, the spoiling surprise bit could be solved by not putting the unit on the recruit list at the start of the scenario, though.
20160226 21:22:14< celticminstrel> I guess that's true.
20160226 21:22:41< zookeeper> and it can often change mid-scenario, so being able to look at it on turn 1 and then seeing some units which weren't shown suddenly appear on turn 3 wouldn't be very nice
20160226 21:23:10< celticminstrel> Even though your own recruit list could also change?
20160226 21:23:35< zookeeper> well if your own recruit list changes, you'll know because someone will tell you that you can now recruit horsemen
20160226 21:23:49< zookeeper> no one will tell you that the AI will start recruiting saurians only from turn 3 onwards for balance reasons
20160226 21:24:03< zookeeper> if you let the player see the list, then that's an implicit promise that the information is accurate until you're told that it's changed
20160226 21:24:13< celticminstrel> I see.
20160226 21:24:14< zookeeper> and that'd be annoying from a campaign author POV
20160226 21:25:13< zookeeper> so... that's why i wouldn't expose it at least in campaigns or when the side is not simply a multiplayer faction
20160226 21:25:33< celticminstrel> Easier not to bother exposing it, I guess.
20160226 21:27:57< celticminstrel> mattsc: "Instead, one has to create a [facet] for turns 1-2, a [facet] for turn 3 and another [facet] for turn 4-10 in order to achieve this behavior. "
20160226 21:28:13< celticminstrel> Couldn't you create one with turns=3 and one with turns=1-2,4-10?
20160226 21:32:43-!- iceiceice [~chris@ext-74.ias.edu] has joined #wesnoth-dev
20160226 21:32:43-!- iceiceice [~chris@ext-74.ias.edu] has quit [Changing host]
20160226 21:32:43-!- iceiceice [~chris@wesnoth/developer/iceiceice] has joined #wesnoth-dev
20160226 21:33:19-!- louis94 [~~louis94@91.178.240.113] has joined #wesnoth-dev
20160226 21:33:45< celticminstrel> I'm not sure I like the new recruitment aspects. The versatility and options and stuff are food, but there are some things about them...
20160226 21:33:49< celticminstrel> ^good
20160226 21:46:23-!- ancestral [~ancestral@63.92.240.233] has joined #wesnoth-dev
20160226 21:53:24-!- ypnos [~ypnos@lme51.informatik.uni-erlangen.de] has joined #wesnoth-dev
20160226 22:09:09-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Ping timeout: 250 seconds]
20160226 22:14:17-!- ancestral [~ancestral@63.92.240.233] has quit [Quit: i go nstuf kthxbai]
20160226 22:15:09-!- ancestral [~ancestral@63.92.240.233] has joined #wesnoth-dev
20160226 22:19:37-!- vultraz [~chatzilla@124.109.10.167] has joined #wesnoth-dev
20160226 22:19:37-!- vultraz [~chatzilla@124.109.10.167] has quit [Changing host]
20160226 22:19:37-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20160226 22:22:33-!- louis94 [~~louis94@91.178.240.113] has quit [Ping timeout: 240 seconds]
20160226 22:28:41 * mattsc reads what celticminstrel has written …
20160226 22:28:48-!- fendrin [~quassel@wesnoth/developer/fendrin] has quit [Read error: Connection reset by peer]
20160226 22:29:00 * Aginor wishes mattsc a good evening
20160226 22:29:51< mattsc> Thanks; and good afternoon to you.
20160226 22:29:59< vultraz> allos Aginor 
20160226 22:30:11< mattsc> Hmm, I guess it’s not afternoon there yet
20160226 22:30:27< celticminstrel> 'lo mattsc 
20160226 22:30:36< mattsc> celticminstrel: yes all aspects are internally composite aspects.
20160226 22:31:18< mattsc> That line should real: all aspects that are not specifically pointed out to be only definable as compisite aspects, …
20160226 22:31:26< mattsc> something like that
20160226 22:31:55< celticminstrel> So rather than composite/standard, you're trying to point out the ones that don't have a simplified syntax?
20160226 22:32:06< celticminstrel> Something like complex/simple.
20160226 22:32:34< mattsc> Well, the code uses composite and standard
20160226 22:32:56< mattsc> standard are the ones that can be defined using the simple syntax
20160226 22:33:02< Aginor> mattsc: it's technically still morning :D
20160226 22:33:08< celticminstrel> Hmm.
20160226 22:33:26< mattsc> Aginor: :) ; and it’s really still mid afternoon here
20160226 22:33:36-!- fabi [~quassel@176.0.66.187] has joined #wesnoth-dev
20160226 22:33:37< mattsc> celticminstrel: maybe I don’t understand what you are saying
20160226 22:33:37-!- fabi [~quassel@176.0.66.187] has quit [Changing host]
20160226 22:33:38-!- fabi [~quassel@wesnoth/developer/fendrin] has joined #wesnoth-dev
20160226 22:33:38< celticminstrel> Maybe I'm still just confused, but somehow I thought composite aspects were defined by [aspect] while standard aspects were defined by [facet].
20160226 22:33:39< Aginor> mattsc: sorry, I get confused :/
20160226 22:34:15< celticminstrel> Maybe it'll make more sense if I look through the code more.
20160226 22:34:55< mattsc> celticminstrel: Could you re-read this section and see if that makes sense: Wesnoth_AI_Framework#The_.5Bai.5D_Tag_.E2.80.94_Aspects
20160226 22:35:09< mattsc> https://wiki.wesnoth.org/Wesnoth_AI_Framework#The_.5Bai.5D_Tag_.E2.80.94_Aspects
20160226 22:35:28< mattsc> Aginor: so am I; I didn’t even notice at forst
20160226 22:35:30< mattsc> first
20160226 22:36:55< mattsc> celticminstrel: and then tell me what part of that you’re confused about so that I can try to describe it more clearly
20160226 22:37:03< celticminstrel> Aginor's appearance reminds me that I was also reviewing his PR.
20160226 22:37:23< Aginor> success!
20160226 22:37:28< celticminstrel> ?
20160226 22:37:34-!- Appleman1234 [~Appleman1@KD106161148078.au-net.ne.jp] has quit [Ping timeout: 244 seconds]
20160226 22:37:42< Aginor> with reminding you, celticminstrel :)
20160226 22:37:46< Aginor> without nagging
20160226 22:38:49< mattsc> celticminstrel, zookeeper: I generally have no issues with the AI cheating; neither subtly nor blatantly.
20160226 22:39:16< mattsc> Anything that makes it play a little better is good, in my opinion
20160226 22:39:33-!- fabi [~quassel@wesnoth/developer/fendrin] has quit [Ping timeout: 240 seconds]
20160226 22:40:02< mattsc> But I do know that not everybody agress with me on that and I’m not trying to convince anybody otherwise.
20160226 22:41:21< mattsc> celticminstrel: I acknowledge what you say about the new recruitment CA/aspects, but are you suggesting that something should be changed?  I was not really involved much with its development at all, or with the writing of that page, I just cleaned up the wording and formatting a little.
20160226 22:41:41< celticminstrel> I don't have any concrete suggestions as yet.
20160226 22:41:48< mattsc> okay
20160226 22:42:01< celticminstrel> Except maybe splitting limits into a separate aspect, but I'm unsure that actually makes sense.
20160226 22:42:05< vultraz> Aginor: do you have some time to help me with the flip() calls thing?
20160226 22:42:15< mattsc> I agree that sime things are … non-ideal, but compared to some other student projects, this is actually a pretty good one
20160226 22:42:21-!- ancestral [~ancestral@63.92.240.233] has quit [Quit: i go nstuf kthxbai]
20160226 22:42:56< celticminstrel> Also, http://www.designersnotebook.com/Columns/103_Bad_Game_Designer_X/psychica103_bad_game_designer_x.htm
20160226 22:43:15 * celticminstrel rereads the aspect description.
20160226 22:43:24< Aginor> vultraz: not right now, sorry. I'm about to go out shopping
20160226 22:43:31< vultraz> ok, no prob 
20160226 22:43:50< mattsc> celticminstrel: As for that comment on creating 3 aspects, I agree with you, if you can mix - and , in the syntax  (that comment was on the page like that already, I did not check whether it’s true)
20160226 22:44:01< celticminstrel> Ah, okay.
20160226 22:44:57< celticminstrel> The C++ code has a utility function for expanding lists of comma-separated ranges, so I would assume the AI used that function, thus allowing hyphens and commas to be mixed.
20160226 22:45:33< celticminstrel> Okay, so the section on aspects is using the term "standard aspect" for what I've been thinking of as "simple aspects".
20160226 22:45:40< mattsc> I’d assume so too; give me a few minutes and I’ll test it.
20160226 22:46:02< celticminstrel> But now, look at that full configuration example.
20160226 22:46:12< celticminstrel> The [aspect] tag has name=composite_aspect.
20160226 22:46:20< celticminstrel> And the [facet] tag has name=standard_aspect.
20160226 22:46:29< mattsc> celticminstrel: Ah, okay; that’s probably because previously the wiki used those interchangably in a couple places; I’ve cleaned that up by now
20160226 22:46:43< mattsc> yes, it does
20160226 22:47:19< mattsc> it’s all somewhat inconsistent
20160226 22:48:00< celticminstrel> That seems to imply that "standard" or "composite" isn't a label attached to specific aspects (eg aggression), but rather something about the structure. Actually, I wonder why the name key is needed at all (maybe it's an implementation detail). I'll look into that.
20160226 22:48:46-!- fabi [~quassel@176.0.66.187] has joined #wesnoth-dev
20160226 22:48:46-!- fabi [~quassel@176.0.66.187] has quit [Changing host]
20160226 22:48:46-!- fabi [~quassel@wesnoth/developer/fendrin] has joined #wesnoth-dev
20160226 22:48:48< mattsc> okay; I’ve wondered that too, this the comment I made about it on the wiki
20160226 22:49:17-!- mjs-de [~mjs-de@f049128037.adsl.alicedsl.de] has quit [Remote host closed the connection]
20160226 22:49:46< celticminstrel> I think perhaps it would be better if a [recruitment_instructions] facet automatically included the [recrut]importance=0 tag implicitly if no [recruit] tags are present.
20160226 22:50:16< celticminstrel> The way the [recruit] tags are handled also makes me think of how stages or candidate actions are processed.
20160226 22:51:22< mattsc> but then what do you do if you actually really do not want general recruiting, but only a couple units per turn?
20160226 22:51:31< mattsc> and yes, on the latter, my guess is that was the intent
20160226 22:52:01< celticminstrel> Well, like I said, the implicit one would be added only if there are no other [recruit] tags.
20160226 22:52:12< celticminstrel> So, just by adding a [recruit] tag, you override it.
20160226 22:52:19< celticminstrel> Though I'm not sure that's really the best way either.
20160226 22:52:33< zookeeper> mattsc, well, if you limit "cheating" to just giving the AI access to more information, then sure, people usually wouldn't have big issues with that. which is why we should use those quotes because it's not like the kind of cheating people usually associate with cheating AI's :P
20160226 22:53:14< zookeeper> as in, usually "cheating AI" means that they have higher chance to hit or make more damage than they should, or things like that
20160226 22:53:16< celticminstrel> zookeeper, mattsc: But if the AI was allowed to see through fog and shroud, I think people would complain (assuming they noticed).
20160226 22:53:42< mattsc> zookeeper: ah, yes, good point;  that’s not what I was talking about
20160226 22:54:13< mattsc> celticminstrel: yes, and it should not go straight for hidden units, I agree with that too
20160226 22:54:29< mattsc> We’ve actually had that happen at some point because something was broken, so we fixed that
20160226 22:54:57< mattsc> celticminstrel: you know what I do with Fred about the seeing through fog thing?
20160226 22:55:25-!- louis94 [~~louis94@91.178.240.113] has joined #wesnoth-dev
20160226 22:56:25< mattsc> celticminstrel: https://github.com/mattsc/AI-demos/blob/master/AI-demos/utils/fred_events.cfg#L27  :P
20160226 22:57:06< celticminstrel> Uh, okay.
20160226 22:57:16< celticminstrel> That's not AI-related though. :P
20160226 22:57:39< mattsc> No, but Fred sees through fog and so this is what I do to level the playing field
20160226 22:57:46< celticminstrel> Ah.
20160226 22:58:04< celticminstrel> That seems like a minor problem with Fred though.
20160226 22:59:08< mattsc> Phbt!  ;)
20160226 22:59:50< zookeeper> we have such a trove of AI stuff now that i wish i had none of these other things to deal with and could just start learning all that
20160226 22:59:55< mattsc> Anyways, back to the [recruit] tag, so you are talking about using [limit], but without actually specifying what to recruit?
20160226 23:00:10< celticminstrel> Yeah, basically like how the [recruitment] aspect worked.
20160226 23:00:19< mattsc> zookeeper: yeah …
20160226 23:00:27< celticminstrel> So you have [limit] tags but no [recrui][ tags.
20160226 23:00:30< mattsc> celticminstrel: that’s reasonable
20160226 23:00:32< celticminstrel> Gah, spare bracket.
20160226 23:01:03< mattsc> Also, I just tested using ‘turns=1-2,4-5’ and that works as expected
20160226 23:01:12< mattsc> So I’ll change that sentence on the wiki
20160226 23:03:09< mattsc> celticminstrel: done
20160226 23:03:43< mattsc> actually, I don’t like that
20160226 23:04:08< celticminstrel> My branch makes [leader_goal]id=blah both set the facet ID to blah and use blah for the auto-remove.
20160226 23:05:38< mattsc> celticminstrel: that sounds good to me  (there isn’t a problem with that, right?)
20160226 23:05:51< mattsc> Also, I changed that comment under that aspect example yet again
20160226 23:06:04< celticminstrel> Well, looking at it, I guess I need to make it not do so if id is set to * or a number, though.
20160226 23:06:24< mattsc> Ah, yes.
20160226 23:08:08< mattsc> Ah, rats, I had missed on of the then occurences (in place of than).  There were lots of them.  Thanks for fixing that.  
20160226 23:11:19< mattsc> celticminstrel, zookeeper: and just to be clear, I mean that I have no issues with the AI cheating (in the way we later clarified) *for the time being*.  I agree, of course, that the eventual goal should be an AI that does not cheat.  But for now, we’re a long way from that.
20160226 23:12:59< mattsc> celticminstrel: have I addressed all of your points or did I forget something?
20160226 23:13:14< celticminstrel> I think you addressed them all.
20160226 23:14:26< mattsc> okay, phew  ;)   Do we have any action items coming out of them?
20160226 23:14:39< celticminstrel> Action items?
20160226 23:14:45< celticminstrel> Like, things for me to do?
20160226 23:14:47< zookeeper> mattsc, yeah, and as far as i know, wesnoth is still quite impressive in the sense that it's a strategy game that the AI can play rather effectively without actual cheating as in bending the gamerules.
20160226 23:15:35< mattsc> celticminstrel: well, things for me to do is what I meant, really, but otherwise yes
20160226 23:15:46< celticminstrel> Ah. I don't think so, then.
20160226 23:16:25< mattsc> zookeeper: indeed; the more I work on AI development, the more impressed I am by the default AI
20160226 23:16:43< mattsc> celticminstrel: ok, thanks
20160226 23:27:30< vultraz> celticminstrel: random question: what would you think of automatically connecting to the primary MP server when wesnoth launches?
20160226 23:27:48< celticminstrel> That sounds totally pointless.
20160226 23:27:58< celticminstrel> Why would you do that?
20160226 23:28:02< Ravana_> isn't this already doable with command line?
20160226 23:29:21< vultraz> smoother unification between the mp and sp modes, since then you could eliminate the 'connecting to server' phase 
20160226 23:29:51< vultraz> from a user POV, at least
20160226 23:30:19< zookeeper> from a user POV i really appreciate when the game makes it clear when i'm online or offline
20160226 23:30:24< celticminstrel> A user wishing to launch Wesnoth to play a campaign should not be forced to join a server.
20160226 23:30:35< Aginor> no, you'll still have that connecting phase, you've just added it to the already huge startup time
20160226 23:30:55< zookeeper> and even more than that i appreciate that i don't have to think about how to respond to a firewall prompt which pops up the first time i run a game
20160226 23:30:58< iceiceice> vultraz, what if you want to connect to the alternate server
20160226 23:31:06< vultraz> Aginor: "huge"?
20160226 23:31:07< Aginor> there shouldn't be any talking to the network unless initiated by the user
20160226 23:31:11< vultraz> Aginor: it's like 2 seconds
20160226 23:31:15< Aginor> vultraz: it takes seconds!
20160226 23:31:17< celticminstrel> vultraz: The startup time is kinda long actually.
20160226 23:31:27< celticminstrel> It's probably at least 5 seconds on my computer.
20160226 23:31:27< iceiceice> vultraz, the reason most modern games do this is not "smoothness of interface"
20160226 23:31:42< iceiceice> its because they lock down the platform and don't want people to have 'alternate servers' not controlled by the developers
20160226 23:31:52< iceiceice> also they want to track your offline activity
20160226 23:31:52< celticminstrel> I haven't actually timed it (and it varies depending on whether my computer is currently experiencing paging difficulties).
20160226 23:32:18< vultraz> Aginor: are you aiming for instantaneous launch time?
20160226 23:32:38< iceiceice> vultraz, i think all the users would be opposed to this change
20160226 23:32:58< celticminstrel> Except maybe ones who do nothing but multiplayer.
20160226 23:33:14< Aginor> vultraz: not instentanious, but less would be nice
20160226 23:33:31< Aginor> why do we need to have a huge overhead in launch to show the title screen?
20160226 23:33:41< celticminstrel> And like Ravana_ said, if they use Windows or Linux that's easily set up by making a special shortcut (or whatever the Linux distribution's equivalent is).
20160226 23:33:50< Aginor> shouldn't the minimum viable readong of preferences and list of scenarious be enough?
20160226 23:34:05< celticminstrel> Well, let's see...
20160226 23:35:23-!- Appleman1234 [~Appleman1@KD106161141247.au-net.ne.jp] has joined #wesnoth-dev
20160226 23:36:12< zookeeper> well, there's certainly plenty of things which could probably be omitted at that stage
20160226 23:36:31< zookeeper> as far as WML goes, that is
20160226 23:36:44< Aginor> mmm
20160226 23:36:57< vultraz> probably, yes
20160226 23:37:02< vultraz> for example, terrain graphics
20160226 23:37:15< Aginor> I suspect that it all gets parsed and dealt with, title screen gets shown, then it gets parsed and delat with again when a map is loaded
20160226 23:37:32< vultraz> yup
20160226 23:37:37< zookeeper> you need campaigns (they're define-guarded, so not a problem), help pages and a bunch of stuff required for them...
20160226 23:37:55< celticminstrel> You don't need to load any of the WML in preparation for the title screen. Except tips.cfg.
20160226 23:38:21< zookeeper> celticminstrel, yes you do, because you can't pop up a loading screen when entering the campaign menu or opening the help
20160226 23:38:39< zookeeper> i mean you could but it'd be obnoxious :p
20160226 23:38:54< celticminstrel> Aginor: When loading a map, it's parsed with different defined constants, I believe.
20160226 23:39:14< Aginor> I have to get going now, done making list
20160226 23:39:17< zookeeper> data/_main.cfg already has a TITLE_SCREEN define thing going on, dunno if it's actually used for anything much
20160226 23:39:26< Aginor> I'd be happy to resume this discussion later
20160226 23:39:37< zookeeper> so could maybe just expand on how that is used
20160226 23:40:05< celticminstrel> Looks like the majority of load time is split between "initializing user interface" and "verifying/creating cache".
20160226 23:40:16< celticminstrel> There's also loading fonts and one or two other small things.
20160226 23:41:40< zookeeper> looks like currently the only thing TITLE_SCREEN is used to exclude are terrain graphics rules
20160226 23:41:46< celticminstrel> zookeeper: Campaigns and help don't need to be loaded in order to display the title screen, but you're right that the game should be prepared to present them almost as soon as the title screen is shown, so it's better if they're already loaded.
20160226 23:42:15< vultraz> why can';t they just load them in the backgrounds
20160226 23:42:19< vultraz> ackground
20160226 23:42:21< vultraz> BACKGROUND
20160226 23:42:24< celticminstrel> Maybe some of the loading could be delegated to a separate thread.
20160226 23:42:31< celticminstrel> ...which vultraz said before I had a chance to.
20160226 23:42:47< zookeeper> because that's hard to do
20160226 23:43:19< celticminstrel> It's not hard to do, but if you want to ensure there's no lag if they instantly go to help or campaigns once the title screen shows up, that might be hard.
20160226 23:43:38< zookeeper> i'll believe that when i see it
20160226 23:44:23< zookeeper> we're talking in context of the existing codebase after all
20160226 23:44:48< vultraz> the whole point of these discussions is the assumption that we will change the existing codebase :P
20160226 23:46:05< celticminstrel> Collect the relevant loading functionality into a function. Send that function off to a new thread which is kept track of in various places. When the data is actually needed, join the thread.
20160226 23:46:59< celticminstrel> The only thing I can see in _main.cfg that could maybe be omitted is themes.
20160226 23:47:09< celticminstrel> But I guess that does need to be loaded if you go to preferences.
20160226 23:47:24< celticminstrel> Oh, maybe AIs could be omitted too?
20160226 23:47:33< vultraz> at least half of the load time is spent on the UI
20160226 23:47:37< celticminstrel> AIs don't really take up much space though.
20160226 23:47:52< vultraz> even despite cutting out about half the widget WML, it didn't improve much.
20160226 23:48:44< celticminstrel> In other words, GUI2 is largely responsible for the long load time.
20160226 23:50:29< celticminstrel> mattsc: The ID of the RCA stage is in fact ai_default_rca::candidate_action_evaluation_loop
20160226 23:50:44< celticminstrel> (The testing_ai_default form is supported as "old syntax")
20160226 23:51:15< celticminstrel> And ai_default_rca is also the prefix for all the CAs.
20160226 23:51:32< mattsc> celticminstrel: what page are you talking about?
20160226 23:51:37< celticminstrel> (Which again are also supported in testing_ai_default as "old syntax")
20160226 23:51:46< vultraz> well, there's also the initial cache creation phase
20160226 23:51:50< celticminstrel> Uh, whichever page mentions the IDs. I think it's actually one of the legacy ones.
20160226 23:51:58< celticminstrel> I'm looking at the code here, though.
20160226 23:52:04< celticminstrel> vultraz: True.
20160226 23:52:14< mattsc> oh, yes, the legacy pages are partially out of date, I am not going to update those
20160226 23:52:24< celticminstrel> Hmm.
20160226 23:52:44< mattsc> the new documentation contains the link to the current version on the repository  (and btw, it’s the name, not the id)
20160226 23:53:00< mattsc> the id is main_loop
20160226 23:53:09< celticminstrel> Well, you did refer to the macro definition files, which use the newer syntax, so I guess that's fine.
20160226 23:53:30< mattsc> right, that’s what I meant up there, sorry
20160226 23:53:49< mattsc> and yes, if I had all the time in the world, I would update the legacy pages too, but I don’t  :|
20160226 23:54:24< vultraz> celticminstrel: in fact, the 'suggested length' for that stage has gone up
20160226 23:54:36< celticminstrel> Apparently there's something called [lua_goal].
20160226 23:54:36< vultraz> er, the GUI i mean
20160226 23:54:42< celticminstrel> vultraz: Huh?
20160226 23:54:56< celticminstrel> Uh, sorry. I mean [goal]name=lua_goal
20160226 23:55:22< vultraz> celticminstrel: each phase has a number bound to it
20160226 23:55:33< mattsc> celticminstrel: https://wiki.wesnoth.org/LuaAI#Goals_and_targets
20160226 23:55:51< vultraz> if you run the game with --log-info=loadscreen, you get the suggested values for those numbers
20160226 23:55:57< vultraz> init-gui is recommended at 27325
20160226 23:56:04< mattsc> I updated that section a couple days ago; but the page as a whole is not done yet
20160226 23:56:11< vultraz> loadscreen.cpp currently has it at 27089
20160226 23:56:22< zookeeper> personally i don't find the load times too bad. i mean they're of course long for what you might expect from a relatively simple 2D game, but if a campaign starts in <10s that's pretty good.
20160226 23:56:54< celticminstrel> mattsc: Any objection to linking there from the AiWML section on goals?
20160226 23:57:09< mattsc> celticminstrel: nope, not at all.  Good idea.
20160226 23:58:13< celticminstrel> There's also [goal]name=protect
20160226 23:58:22< celticminstrel> I'm not sure whether or not this is a synonym for protect_unit.
20160226 23:58:32< celticminstrel> Oh, wait.
20160226 23:58:40< celticminstrel> It's a synonym for protect_location.
20160226 23:58:57 * celticminstrel decides it's not worth documenting.
20160226 23:59:25< mattsc> ok
--- Log closed Sat Feb 27 00:00:05 2016