--- Log opened Wed Aug 31 00:00:14 2016
--- Day changed Wed Aug 31 2016
20160831 00:00:13< celticmistral> Maybe that was a slightly different config API method?
20160831 00:00:27< celticmistral> Honestly, the idea of having a magic "invalid" config is pretty dumb in my opinion.
20160831 00:00:41< vultraz> hm?
20160831 00:00:54< vultraz> what do you mean
20160831 00:01:11< celticmistral> Some (v)config functions return a special static "null config" object.
20160831 00:01:27-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160831 00:01:30< celticmistral> That's probably what !mod tests.
20160831 00:01:38< celticmistral> "this == &null_config"
20160831 00:02:14< celticmistral> vconfig has the "unconstructed" static object.
20160831 00:02:28< vultraz> that's... weird?
20160831 00:02:38< celticmistral> Yes.
20160831 00:02:40< vultraz> why not just use .empty() :|
20160831 00:02:42-!- TheJJ [~rofl@ipbcc36896.dynamic.kabel-deutschland.de] has quit [Ping timeout: 265 seconds]
20160831 00:02:51< celticmistral> That's not quite the point.
20160831 00:02:59< celticmistral> It's different from .empty().
20160831 00:03:00-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160831 00:03:11< celticmistral> It's a magic config that means "no config was here".
20160831 00:03:24< celticmistral> I guess the config equivalent of boost::blank
20160831 00:03:48< vultraz> I see
20160831 00:04:17< vultraz> well, irreverent i guess 
20160831 00:04:25< celticmistral> Irreverent?
20160831 00:04:39< vultraz> :|
20160831 00:04:45< vultraz> irrelevant 
20160831 00:05:47< celticmistral> Yeah, I guess.
20160831 00:06:02< vultraz> hm..
20160831 00:06:28< vultraz> actually, i probably don't need to set addons_outcome if mod not found..
20160831 00:07:09< vultraz> well...
20160831 00:07:37< vultraz> it already checks each addon against installed ones..
20160831 00:07:42< celticmistral> Probably NEEDS_DOWNLOAD, I'd say.
20160831 00:07:49< vultraz> ands mods will be addons...
20160831 00:07:57< vultraz> and if it's not found it becomes NEED_DOWNLOAD
20160831 00:07:59< celticmistral> So wait, is the check redundant?
20160831 00:08:17< vultraz> I'm trying to determine
20160831 00:08:38< vultraz> what do you think
20160831 00:09:03< vultraz> we still need to run the mod through check_addon_version_compatibility..
20160831 00:09:07< vultraz> so i guess it can stay
20160831 00:09:44< celticmistral> No, if the mod is not present at all, you don't need to check version. You can just download it.
20160831 00:10:07< vultraz> right
20160831 00:10:17< vultraz> so if the addon needed is not found...
20160831 00:10:23< vultraz> it's already flagged as missing
20160831 00:10:27< vultraz> and needs download
20160831 00:10:58< vultraz> yeah it should work as-is
20160831 00:11:12< vultraz> should test it, though
20160831 00:11:47-!- jswensen [~jswensen@s48075040.temp.wsu.edu] has quit [Quit: jswensen]
20160831 00:12:46< vultraz> hm
20160831 00:14:34< vultraz> just noticed the gui1 lobby has a slightly more organized game info view
20160831 00:14:50< vultraz> or informative
20160831 00:16:24< vultraz> hm
20160831 00:16:29< vultraz> a mod may be part of another addon
20160831 00:16:41< vultraz> meaning we cannot treat it as an individual addon 
20160831 00:16:50< vultraz> meaning the current behavior is indeed correct..
20160831 00:16:51< vultraz> I think
20160831 00:23:45-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160831 01:05:20< vultraz> if(a > b) { a = b; } is the same as a = min(a, b), right?
20160831 01:05:49< vultraz> unless my logic is totally flawed 
20160831 01:23:51< celticmistral> Uh.
20160831 01:24:33< celticmistral> Yeah, seems like it.
20160831 01:24:52< celticmistral> I suppose it could be marginally more efficient, but not to a level that anyone would need to worry about.
20160831 01:25:58< vultraz> which version would be more efficient?
20160831 01:26:46< celticmistral> Doesn't matter.
20160831 01:27:05< celticmistral> The compiler would probably optimize both to the same thing anyway.
20160831 01:30:13< shadowm> In which case you should aim for readability. foo = min(bar, baz) is far more expressive and readable than using multiple symbols and branching instructions.
20160831 01:31:39< vultraz> agreed
20160831 01:31:42< shadowm> That also happens to be why people should prefer using standard algorithms where it makes sense instead of inlining new nameless clones of them.
20160831 01:31:44< celticmistral> Good point.
20160831 01:32:07 * celticmistral has been using standard algorithms quite a bit recently.
20160831 01:32:15< celticmistral> (Not only in Wesnoth.)
20160831 01:32:41 * vultraz reads up on aggregate initialization
20160831 01:34:27< vultraz> so apparently.. if you have struct A{ string a, string b}; and then do A foo { "bar" };, then foo.a == "bar"?
20160831 01:36:38< vultraz> and if you were to switch the order of a and b, then b would be "bar"?
20160831 01:39:37< vultraz> that seems rather volatile, since you could break stuff by changing the order of the struct members, couldn't you?
20160831 01:41:27< vultraz> (though for the record, A foo = { foo.a = "bar" }; does build for me, but I don' t know if that's a good thing to do)
20160831 01:41:39< vultraz> celticminstrel: thoughts?
20160831 01:46:25-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20160831 01:46:43< vultraz> hm. "If the initializer clause is an expression, implicit conversions are allowed as per copy-initialization, except if they are narrowing (as in list-initialization) (since C++11)." So maybe A foo { foo.a = "bar" }; is indeed ok
20160831 01:51:15< celticmistral> What.
20160831 01:51:27< celticmistral> How does A foo = {foo.a = "bar"} compile.
20160831 01:51:39< celticmistral> There is a C99 syntax A foo = {.a = "bar"} mind you.
20160831 01:51:45< celticmistral> But that's not supported by MSVC in C++.
20160831 01:51:52< celticmistral> MSVC 2013 at least.
20160831 01:52:00< celticmistral> (I think it's supported by MSVC 2013 in C, though.)
20160831 01:52:18< celticmistral> You're correct that it's potentially volatile.
20160831 01:52:40< celticmistral> Was it just that syntax you wanted my opinion on?
20160831 01:52:48< vultraz> yes
20160831 01:53:12< celticmistral> Oh, right.
20160831 01:53:39< celticmistral> I can see why {foo.a = "bar"} might compile.
20160831 01:53:46< celticmistral> However, it wouldn't do what you expect.
20160831 01:54:04< celticmistral> In particular, if you did {foo.b = "bar"} instead, both elements would be set to "bar".
20160831 01:54:16< vultraz> I see
20160831 01:54:28< celticmistral> Because foo.a = "bar" is an expression that assigns "bar" to foo.a and then evaluates to foo.a.
20160831 01:54:35< vultraz> it seems odd there's no member specifier 
20160831 01:54:44-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160831 01:54:59< celticmistral> (It's also probably bad practice to refer to a variable in its own initialization list.)
20160831 01:55:50< celticmistral> C99 does have member specifiers – {.a = "bar"} or for arrays {[3] = "bar"} – but C++ I guess doesn't want to support C-style programming too much, or something.
20160831 01:59:33-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 240 seconds]
20160831 02:08:45-!- travis-ci [~travis-ci@ec2-54-211-221-115.compute-1.amazonaws.com] has joined #wesnoth-dev
20160831 02:08:46< travis-ci> wesnoth/wesnoth#10630 (master - 7ec92eb : Charles Dang): The build has errored.
20160831 02:08:46< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/156380354
20160831 02:08:46-!- travis-ci [~travis-ci@ec2-54-211-221-115.compute-1.amazonaws.com] has left #wesnoth-dev []
20160831 02:13:14< vultraz> hmmm
20160831 02:14:47< vultraz> the 'out of date' message doesn't appear
20160831 02:17:52-!- TheJJ [~rofl@ipbcc36896.dynamic.kabel-deutschland.de] has joined #wesnoth-dev
20160831 02:21:51< vultraz> hm
20160831 02:22:28< vultraz> the gui1 lobby has some specific data on mp campaigns..
20160831 02:22:31< vultraz> the gui2 one does not..
20160831 02:23:10< vultraz> ah yeah
20160831 02:23:12< vultraz> it's missing
20160831 02:23:25< vultraz> if you start an mp campaign, it shows "Unknown Scenario"
20160831 02:23:27< vultraz> ok
20160831 02:23:45-!- TheJJ [~rofl@ipbcc36896.dynamic.kabel-deutschland.de] has quit [Ping timeout: 265 seconds]
20160831 02:23:53< vultraz> for some reason it seems the version check wasn't in the campaigns section of the gui1 lobby, though..
20160831 02:37:29-!- irker488 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20160831 02:52:14< vultraz> GAH
20160831 02:56:26< vultraz> why doesn't it recognize this as a campaign
20160831 02:59:18< vultraz> let''s try the logic from gui1..
20160831 03:00:21< vultraz> nope
20160831 03:00:35< vultraz> for some reason, it doesn't detect it.
20160831 03:23:13< vultraz> this is rather frutsrating
20160831 03:24:43-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: So long and thanks for all the fish.]
20160831 04:17:48-!- fabi_ [~fabi@176.5.137.16] has quit [Remote host closed the connection]
20160831 04:18:20-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160831 04:22:54-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 260 seconds]
20160831 04:41:00< Aginor> g'day
20160831 06:01:52-!- celticmistral [~celticmin@unaffiliated/celticminstrel] has quit [Quit: And lo! The minstrel departs, to spread the music to the masses!]
20160831 06:06:35-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160831 06:11:07-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 255 seconds]
20160831 06:14:17-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has joined #wesnoth-dev
20160831 06:16:04-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20160831 06:37:22< Sirp> I have posted this topic for people interested in nominating wesnoth, inc board members: http://forum.wesnoth.org/viewtopic.php?f=2&t=44553
20160831 06:38:05-!- Kwandulin [~Miranda@p200300760F424183CD0B00633C1AB7D2.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160831 06:41:04< Aginor> Sirp: thank you, great to see things happening
20160831 06:41:47< shadowm> Can somebody post a link to the mailing list just in case im busy
20160831 06:44:16< Sirp> doing that now
20160831 06:45:52< Sirp> Also, for voting I tend to think only actual developers (people with svn commit rights) should be able to vote. Nominations from anyone is fine though.
20160831 06:46:46< celticminstrel> Wait, Sirp = Dave?
20160831 06:46:59 * celticminstrel had no idea.
20160831 06:47:02< shadowm> The credits say so, yes.
20160831 06:48:28< shadowm> The devs forum is currently open to replies from random people, so maybe vultraz or someone may want to make it a rule that people who aren't members of the Developers group replying to that thread will get banned.
20160831 06:48:46< Aginor> Sirp: I'd suggest you post about the voting process upfront, so that people know what to expect
20160831 06:49:40< celticminstrel> If you want only devs to be able to post in the thread, wouldn't locking it work?
20160831 06:50:02< celticminstrel> Devs can post in locked threads, right?
20160831 06:50:12< shadowm> Right now yes.
20160831 06:50:22< shadowm> May give off the impression that it's not to be read or replied to though.
20160831 06:50:31< celticminstrel> Maybe.
20160831 06:51:02< shadowm> Also wasn't serious about banning just deleting posts should suffice
20160831 06:59:01-!- atarocch [~atarocch@natmobil.sfa.se] has joined #wesnoth-dev
20160831 06:59:47-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has quit [Quit: And lo! The computer falls into a deep sleep, to awake again some other day!]
20160831 07:20:10< shadowm> I've revoked write access to the DD forum by Forum Regulars and unaffiliated users.
20160831 07:20:36< shadowm> Quite predictably someone without a group affiliation had to go and post to the thread.
20160831 07:40:07-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20160831 07:44:09-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 276 seconds]
20160831 07:44:09-!- wedge010 is now known as wedge009
20160831 07:50:29< vultraz> shadowm: should i edit his post where he said "SVN write access"
20160831 07:51:11< shadowm> Well, I think you should ask him first?
20160831 07:51:29< shadowm> I've always had a rule that I don't edit other people's posts without asking them unless they are obviously violating a rule.
20160831 07:51:48< shadowm> (Or a mod action was otherwise involved, e.g. splitting or merging threads.)
20160831 07:52:41< shadowm> If I didn't have this rule some people like 8680 would've spent most of their time fixing other people's grammatical or punctuation mistakes.
20160831 07:54:54-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160831 07:59:22-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 252 seconds]
20160831 08:15:57-!- TheJJ [~rofl@ipbcc36896.dynamic.kabel-deutschland.de] has joined #wesnoth-dev
20160831 08:20:11-!- TheJJ [~rofl@ipbcc36896.dynamic.kabel-deutschland.de] has quit [Ping timeout: 244 seconds]
20160831 08:26:33-!- nurupo [~nurupo.ga@unaffiliated/nurupo] has quit [Ping timeout: 240 seconds]
20160831 08:26:51< vultraz> hmmmmm...
20160831 08:27:32< vultraz> somehow this string is getting messed up...
20160831 08:27:35< vultraz> "map_info is: Default — 33×33 TEST STRING"
20160831 08:28:18< vultraz> or is that just utf8..
20160831 08:31:37< vultraz> i think i might know what's wrong
20160831 08:33:23< vultraz> ahhhh
20160831 08:33:25< vultraz> jesus
20160831 08:33:35< vultraz> took all afternoon to figure out it's a missing label :|
20160831 08:34:56-!- nurupo [~nurupo.ga@unaffiliated/nurupo] has joined #wesnoth-dev
20160831 08:37:39< vultraz> ok, so the question becomes which format do we want
20160831 08:38:04< vultraz> i suppose one should denote a campaign
20160831 08:43:35< zookeeper> ..?
20160831 08:44:37< vultraz> zookeeper: in the mp lobby, i suppose it should say whether a game is a campaign or not
20160831 08:45:00< zookeeper> seems most reasonable
20160831 09:02:45< vultraz> hmm
20160831 09:02:57< vultraz> for some reason, campaigns don't get addon validation...
20160831 09:12:50-!- horrowind [~Icedove@2a02:810a:8380:10a8:21b:fcff:fee3:c3ff] has joined #wesnoth-dev
20160831 09:18:08-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has quit [Quit: .]
20160831 09:43:10-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160831 09:45:50-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has joined #wesnoth-dev
20160831 09:47:30-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 244 seconds]
20160831 10:00:06-!- irker007 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20160831 10:00:06< irker007> wesnoth: Charles Dang wesnoth:master 9ee737454f39 / src/gui/dialogs/lobby/ (data.cpp data.hpp lobby.cpp): MP Lobby: some cleanup to addon requirement interface https://github.com/wesnoth/wesnoth/commit/9ee737454f399b1d030fdb84a1709d2d4a1df68a
20160831 10:00:07< irker007> wesnoth: Charles Dang wesnoth:master 3c3f2d71d160 / src/game_config_manager.cpp: Save addon info for campaigns too https://github.com/wesnoth/wesnoth/commit/3c3f2d71d160b967f5bc55dc5be61af1d6f3b835
20160831 10:00:08< irker007> wesnoth: Charles Dang wesnoth:master 8b15851d3dbb / src/gui/dialogs/lobby/data.cpp: MP Lobby: add correct display formatting and version validation for Campaigns https://github.com/wesnoth/wesnoth/commit/8b15851d3dbb986f218e9edefb14bd41995e830a
20160831 10:02:23< vultraz> absolutely no idea why addon info was excluded for campaigns :|
20160831 10:06:37< vultraz> also it seems that text filtering is done on string that's invisible to the user :|
20160831 10:25:16-!- Nobun [~nobun@host97-56-dynamic.47-79-r.retail.telecomitalia.it] has joined #wesnoth-dev
20160831 10:37:56-!- horrowind [~Icedove@2a02:810a:8380:10a8:21b:fcff:fee3:c3ff] has quit [Quit: horrowind]
20160831 10:57:53-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160831 11:02:31-!- fabi_ [~fabi@176.4.53.30] has joined #wesnoth-dev
20160831 11:02:51-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 265 seconds]
20160831 11:37:46-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160831 12:00:43-!- Duthlet [~Duthlet@dslb-188-104-247-024.188.104.pools.vodafone-ip.de] has joined #wesnoth-dev
20160831 12:12:11-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20160831 12:18:55-!- Shiki [~Shiki@141.39.226.227] has joined #wesnoth-dev
20160831 12:19:23-!- TheJJ [~rofl@ipbcc36896.dynamic.kabel-deutschland.de] has joined #wesnoth-dev
20160831 12:23:39-!- TheJJ [~rofl@ipbcc36896.dynamic.kabel-deutschland.de] has quit [Ping timeout: 258 seconds]
20160831 12:24:55-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160831 12:29:12-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 240 seconds]
20160831 12:46:25-!- Kwandulin [~Miranda@p200300760F424183CD0B00633C1AB7D2.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160831 12:58:23-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160831 13:07:01-!- Nobun [~nobun@host97-56-dynamic.47-79-r.retail.telecomitalia.it] has quit [Remote host closed the connection]
20160831 13:11:48-!- TheJJ [~rofl@ipbcc36896.dynamic.kabel-deutschland.de] has joined #wesnoth-dev
20160831 13:16:22-!- TheJJ [~rofl@ipbcc36896.dynamic.kabel-deutschland.de] has quit [Ping timeout: 255 seconds]
20160831 13:23:51-!- Kwandulin [~Miranda@p200300760F4241832865B9D38F7857E6.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160831 13:29:05-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20160831 13:34:12-!- irker007 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20160831 13:41:59-!- irker510 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20160831 13:41:59< irker510> wesnoth: Charles Dang wesnoth:master b726e9fe1dee / / (4 files in 2 dirs): MP Lobby: improved game row layout https://github.com/wesnoth/wesnoth/commit/b726e9fe1dee75df6ff7ba831d667772e662ea51
20160831 13:44:32-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20160831 13:57:14-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has quit [Quit: .]
20160831 14:07:31-!- Kwandulin [~Miranda@p200300760F4241832865B9D38F7857E6.dip0.t-ipconnect.de] has quit [Quit: Kwandulin]
20160831 14:12:07-!- Kwandulin [~Miranda@p200300760F4241832865B9D38F7857E6.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160831 14:31:56< irker510> wesnoth: Charles Dang wesnoth:master 4654f245f14d / src/ (addon/manager_ui.cpp gui/dialogs/addon/uninstall_list.cpp): Fixed some dialogs not undrawing properly in the addons manage workflow https://github.com/wesnoth/wesnoth/commit/4654f245f14d5b8f61080238d3114d242db5d36f
20160831 14:32:33< mattsc> vultraz: src/gui/dialogs/lobby/data.cpp:441:51: error: missing field 'outcome' initializer [-Werror,-Wmissing-field-initializers]
20160831 14:32:45< vultraz> wha
20160831 14:32:49< mattsc> is that knwon and being worked on?
20160831 14:32:53< mattsc> I guess not ...
20160831 14:33:12< mattsc> Did you add any new files or something that Xcode does not yet know about?
20160831 14:33:24< vultraz> sounds like it was cases by a change just earlier
20160831 14:34:03< vultraz> aggregate initialization on a struct..
20160831 14:34:48< vultraz> how come it doesn't like it :|
20160831 14:35:25-!- Kwandulin [~Miranda@p200300760F4241832865B9D38F7857E6.dip0.t-ipconnect.de] has quit [Quit: Kwandulin]
20160831 14:36:14-!- travis-ci [~travis-ci@ec2-54-80-189-138.compute-1.amazonaws.com] has joined #wesnoth-dev
20160831 14:36:15< travis-ci> wesnoth/wesnoth#10632 (master - b726e9f : Charles Dang): The build has errored.
20160831 14:36:15< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/156520920
20160831 14:36:15-!- travis-ci [~travis-ci@ec2-54-80-189-138.compute-1.amazonaws.com] has left #wesnoth-dev []
20160831 14:37:10< vultraz> oh come ON :|
20160831 14:37:15 * vultraz sighs
20160831 14:37:33< vultraz> mattsc: no, that's not the issue
20160831 14:37:54< vultraz> issue is your and travis's compiler deciding it doesn't like the syntax
20160831 14:38:39< mattsc> vultraz: I am sure there is a reason why those compilers do that
20160831 14:38:53< vultraz> what version is yours?
20160831 14:39:19< mattsc> It’s a warning, so I can turn it off, but there’s a reason why we have treat warnings as errors enabled.
20160831 14:39:34< mattsc> Latest Xcode, whatever clang that uses.
20160831 14:39:36< mattsc> Let me check.
20160831 14:40:18< mattsc> Well, those are the stupid Apple versions … : Apple LLVM version 7.3.0 (clang-703.0.31)
20160831 14:47:42< irker510> wesnoth: Charles Dang wesnoth:master 8e5c16dbf693 / src/gui/dialogs/lobby/data.cpp: Specify all members in aggregate initialization https://github.com/wesnoth/wesnoth/commit/8e5c16dbf69383de4d92ae162a52a389450baa00
20160831 14:47:45< vultraz> mattsc: ^ does that fix it
20160831 14:51:44-!- atarocch [~atarocch@natmobil.sfa.se] has quit [Ping timeout: 260 seconds]
20160831 15:28:12-!- travis-ci [~travis-ci@ec2-54-211-221-115.compute-1.amazonaws.com] has joined #wesnoth-dev
20160831 15:28:13< travis-ci> wesnoth/wesnoth#10633 (master - 4654f24 : Charles Dang): The build has errored.
20160831 15:28:13< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/156535935
20160831 15:28:13-!- travis-ci [~travis-ci@ec2-54-211-221-115.compute-1.amazonaws.com] has left #wesnoth-dev []
20160831 15:30:42-!- atarocch [~atarocch@88.131.217.34] has joined #wesnoth-dev
20160831 15:35:31-!- Kwandulin [~Miranda@p200300760F42418358DA76A496CE7659.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160831 15:37:21< mattsc> vultraz: It fixes this error, yes.  Thanks.  [I don’t have time to hang around to wait for compiling to finish right now.]
20160831 15:42:59-!- VultCave [~chatzilla@124.109.10.167] has joined #wesnoth-dev
20160831 15:43:04-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Read error: Connection reset by peer]
20160831 15:43:21-!- VultCave is now known as vultraz
20160831 15:44:08-!- JyrkiVesterinen [~JyrkiVest@78-27-126-82.bb.dnainternet.fi] has joined #wesnoth-dev
20160831 15:50:24< mattsc> Now I get unused variable warnings in multiplayer_wait.cpp.  leader_pane_position and leader_pane_border.
20160831 15:50:39< mattsc> Got to be off though.  TTYAL.
20160831 15:50:44-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: So long and thanks for all the fish.]
20160831 15:52:57-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160831 15:54:31-!- horrowind [~Icedove@2a02:810a:8380:10a8:21b:fcff:fee3:c3ff] has joined #wesnoth-dev
20160831 15:56:22-!- fabi [~fabi@176.6.19.230] has joined #wesnoth-dev
20160831 15:58:08-!- vincent_c [~bip@vcheng.org] has quit [Quit: Coyote finally caught me]
20160831 15:58:46-!- vincent_c [~bip@vcheng.org] has joined #wesnoth-dev
20160831 15:59:43-!- fabi_ [~fabi@176.4.53.30] has quit [Read error: Connection reset by peer]
20160831 16:13:50-!- Shiki [~Shiki@141.39.226.227] has quit [Remote host closed the connection]
20160831 16:13:59-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has joined #wesnoth-dev
20160831 16:15:56-!- travis-ci [~travis-ci@ec2-54-80-189-138.compute-1.amazonaws.com] has joined #wesnoth-dev
20160831 16:15:57< travis-ci> wesnoth/wesnoth#10634 (master - 8e5c16d : Charles Dang): The build has errored.
20160831 16:15:57< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/156540312
20160831 16:15:57-!- travis-ci [~travis-ci@ec2-54-80-189-138.compute-1.amazonaws.com] has left #wesnoth-dev []
20160831 16:18:17-!- Kwandulin [~Miranda@p200300760F42418358DA76A496CE7659.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160831 16:23:42< celticminstrel> Note that = {0} is a conventional way to say "initialize all members to 0".
20160831 16:23:55-!- TheJJ [~rofl@ipbcc36896.dynamic.kabel-deutschland.de] has joined #wesnoth-dev
20160831 16:24:16 * celticminstrel just noticed that latest commit and felt like mentioning it since it's somewhat related.
20160831 16:28:07-!- TheJJ [~rofl@ipbcc36896.dynamic.kabel-deutschland.de] has quit [Ping timeout: 252 seconds]
20160831 16:41:30-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has joined #wesnoth-dev
20160831 16:43:02< tad_> zookeeper: I rebased all the HttT PRs (on general principle). I pushed an update to the TSG PR for all the changes except the big S05 conversation/objective/victory-paths changes.
20160831 16:46:32-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has quit [Quit: Page closed]
20160831 16:47:56< Soliton> loonycyborg: trunk server froze. attached with gdb and produced a corefile you can check out: ~/core.26428_wesnothd-trunk-git-1.13.5-136-gf321e71
20160831 16:52:58-!- EliDupree [~quassel@idupree.com] has quit [Ping timeout: 250 seconds]
20160831 16:55:13-!- EliDupree [~quassel@idupree.com] has joined #wesnoth-dev
20160831 16:55:57< celticminstrel> Matrix placement policy now works perfectly when horizontal grow is disabled! \o/
20160831 16:56:12< celticminstrel> Will probably need some additional fixup to work with horizontal grow though.
20160831 16:56:20-!- Kwandulin [~Miranda@p200300760F4241835C56635F93942045.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160831 17:02:16< celticminstrel> Should also probably ensure all cells in a row are the same height.
20160831 17:02:26< celticminstrel> But I'll leave that as a TODO note for now.
20160831 17:02:39< loonycyborg> Soliton: how exactly did it froze? with 100% cpu use?
20160831 17:02:48-!- horrowind [~Icedove@2a02:810a:8380:10a8:21b:fcff:fee3:c3ff] has quit [Quit: horrowind]
20160831 17:03:14-!- ChipmunkV [~vova@static-89-94-113-91.axione.abo.bbox.fr] has joined #wesnoth-dev
20160831 17:05:49< Soliton> loonycyborg: didn't check. ps output: wesnoth  26428  0.5  0.8 198676 53712 pts/0    R    Aug04 229:54 bin/wesnothd-trunk
20160831 17:07:42< loonycyborg> bt in that core file is  from simple_wml::uncompress_buffer function
20160831 17:08:43< loonycyborg> that could be an archive bomb, but I'm pretty sure I saw safeguards for that in simple_wml..
20160831 17:08:51-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160831 17:10:15-!- hk238 [~kvirc@unaffiliated/hk238] has joined #wesnoth-dev
20160831 17:17:21< irker510> wesnoth: ln-zookeeper wesnoth:master 0c945f87ab01 / src/network.cpp: Fix column misalignment in netstats report https://github.com/wesnoth/wesnoth/commit/0c945f87ab0137e398bbc97a089e4f4c3e12d535
20160831 17:18:43< Soliton> you could ask vultraz what he did. last thing in the log was him creating a game.
20160831 17:19:36< zookeeper> :]
20160831 17:19:57< loonycyborg> vultraz: did you end that session normally?
20160831 17:20:45< loonycyborg> or got a hang/disconnect?
20160831 17:26:01-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160831 17:30:31-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160831 17:38:35< Soliton> loonycyborg: looks like it wasn't hanging completely. i've connected several times with gdb and got different back traces...
20160831 17:40:32< Soliton> so the corefile probably doesn't help that much.
20160831 17:48:32-!- hk238 [~kvirc@unaffiliated/hk238] has quit [Quit: http://www.kvirc.net/ 4.9.1 Aria]
20160831 17:53:40-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20160831 17:55:19< vultraz> loonycyborg: uh.... yes. but I was running multiple sessions at once
20160831 17:58:12< vultraz> basically i was doing testing on the new lobby
20160831 17:58:48< vultraz> and probably spamming some log somewhere in the process :P
20160831 18:01:50-!- edgrey [~edgrey@178.204.82.169] has joined #wesnoth-dev
20160831 18:15:12-!- horrowind [~Icedove@2a02:810a:8380:10a8:21b:fcff:fee3:c3ff] has joined #wesnoth-dev
20160831 18:23:08< irker510> wesnoth: Gregory A Lundberg wesnoth:master fcbb72d8d06c / data/lua/wml-tags.lua: Fix bug [role] missing role= https://github.com/wesnoth/wesnoth/commit/fcbb72d8d06cc4c7b162af4eabcd88725d43367e
20160831 18:23:10< irker510> wesnoth: Charles Dang wesnoth:master a7bb98a4c819 / data/lua/wml-tags.lua: Merge pull request #761 from GregoryLundberg/GL_role_missing_role https://github.com/wesnoth/wesnoth/commit/a7bb98a4c819b70ec298b735767721489a733750
20160831 18:27:51< celticminstrel> Uhhh, what?
20160831 18:27:57< celticminstrel> Why did you merge that?
20160831 18:28:13< vultraz> his argument was convincing
20160831 18:28:31< celticminstrel> Doesn't seem that convincing to me.
20160831 18:29:29< celticminstrel> [role] is for assigning roles, not recalling units/
20160831 18:32:16-!- fabi [~fabi@176.6.19.230] has quit [Ping timeout: 255 seconds]
20160831 18:33:37< vultraz> blah, lobby will filter on stuff like ' - ' characters :|
20160831 18:33:59< celticminstrel> Not a whole lot you can do about that.
20160831 18:39:39< irker510> wesnoth: Charles Dang wesnoth:master d64842930ec0 / data/advanced_preferences.cfg: Changed Color Cursors default to Yes, to match the program behavior https://github.com/wesnoth/wesnoth/commit/d64842930ec0341e9153461f5fd203480217cd02
20160831 18:39:42< irker510> wesnoth: Charles Dang wesnoth:master 8c68d2a253d2 / data/gui/window/lobby_main.cfg: MP Lobby: few subtle layout improvements https://github.com/wesnoth/wesnoth/commit/8c68d2a253d29e2db955864357e9bdfab1781aa5
20160831 18:40:56 * celticminstrel wonders if zookeeper has an opinion on that.
20160831 18:44:27< vultraz> hm?
20160831 18:44:31< vultraz> the role thing?
20160831 18:44:46< celticminstrel> Yes.
20160831 18:47:28< vultraz> celticminstrel: do you think it's worth displaying the name of the difficulty an mp campaign is being played at
20160831 18:47:41-!- Shiki [~Shiki@141.39.226.227] has joined #wesnoth-dev
20160831 18:48:00< celticminstrel> Sure.
20160831 18:48:15< vultraz> :/
20160831 18:48:33< celticminstrel> ...why that face.
20160831 18:49:01< vultraz> need to copy all the compatibility code from campaign_difficulty 
20160831 18:49:16< zookeeper> celticminstrel, vultraz, uh so what does that commit do? just suppress an error if your [role] is missing role=? ehh...
20160831 18:49:24< zookeeper> seems harmless, i guess?
20160831 18:49:46< celticminstrel> I dunno, I think an error on missing role= is quite logical.
20160831 18:49:54< celticminstrel> Though it should probably be more descriptive than it was.
20160831 18:50:05< zookeeper> although one could argue that the error should only be suppressed if you actually have any of that other recall-related stuff
20160831 18:50:11< vultraz> I wonder if I can make it a static class member..
20160831 18:50:39< zookeeper> because if you only write [role] type=Spearman [/role] then that's obviously baloney and an warning (not an error, i'd think) would be in order
20160831 18:50:44< celticminstrel> So you don't have a problem with using [role] to recall the highest-level unit?
20160831 18:51:52< zookeeper> well if someone wants to do it then who am i to have a problem with it
20160831 18:52:38< zookeeper> hrhm
20160831 18:53:15< zookeeper> although that's not really a sound rationale for changing any existing behavior
20160831 18:53:36< zookeeper> at least his example should have included giving them an actual role just for show :p
20160831 18:55:50< zookeeper> could type= in [recall] be made to prioritize matching units in order, like [role] does? that would be handy.
20160831 18:56:21< celticminstrel> I guess it could.
20160831 18:56:52< celticminstrel> I suppose with a lot of effort that could even be made true of SUF in general.
20160831 18:57:10< celticminstrel> I don't think the latter is worth it though.
20160831 18:57:12< zookeeper> i dunno about that, it could have quite a bit of unforeseen consequences
20160831 18:57:24< zookeeper> but in [recall] it'd make a lot of sense i'd think
20160831 18:58:00< celticminstrel> Well, [recall] is still C++ though. (Not that that actually makes a difference.)
20160831 18:59:00< celticminstrel> (Just that, if it was Lua, it and [role] could probably share some code if they were both doing the same type-priority thing.)
20160831 18:59:46< celticminstrel> Hmm, come to think of it, the implementation of [role] means that it iterates over all units on the map for each possible type.
20160831 19:00:14< celticminstrel> Surely it would be more efficient to filter for all units of any allowed type, then sort them to match the type order given.
20160831 19:00:20< celticminstrel> (With a stable sort, of course.)
20160831 19:00:35< celticminstrel> I wonder if Lua has a built-in stable sort.
20160831 19:01:31< JyrkiVesterinen> No, it doesn't.
20160831 19:03:38< celticminstrel> Yeah, I discovered that through Google just now. :/
20160831 19:06:06< celticminstrel> Doesn't look like it can be made stable with a carefully crafted comparator, either.
20160831 19:06:38-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160831 19:06:56< JyrkiVesterinen> You can just implement insertion sort or something. It's not that hard.
20160831 19:07:15< celticminstrel> I don't really want to implement my own stable sort, but even if it just gathered all possible units first, then looped through those once per possible type, I think that would be more efficient than what it's currently doing.
20160831 19:07:38< celticminstrel> I guess it's not that hard, it's just, tedious or something.
20160831 19:09:43< vultraz> celticminstrel: re the difficulty, I think we only need the actual name (ie, Hard, Normal) not the description as well,  like gui1 has?
20160831 19:10:02< celticminstrel> I don't know off the top of my head which sorts are stable, but given you suggested it, I guess insertion is.
20160831 19:10:05< vultraz> ie, having it say "Lord(Hard)" really means nothing. All we care about is Hard.
20160831 19:10:24< celticminstrel> Though I can't remember the difference between insertion and selection either...
20160831 19:10:28< celticminstrel> vultraz: Sure, that makes sense.
20160831 19:10:42-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160831 19:11:18< celticminstrel> (Or is stability not a feature of the underlying algorithm? I don't think my classes ever discussed stability.)
20160831 19:11:43< vultraz> blah, moving the conversion code out-of-class now makes the difficulty dialog crash
20160831 19:12:14< celticminstrel> What are you talking about?
20160831 19:13:03< vultraz> i split the code from the tcampaign_difficulty constructor into an out-of-class function so i could use it in the lobby data without copying the code.
20160831 19:13:40< celticminstrel> I guess you overlooked something.
20160831 19:13:45< vultraz> I wonder if passing the reference provided as a constructor argument to the out-of-class function is bad..
20160831 19:14:30< celticminstrel> There's not reason for it to bad unless that out-of-class function stores the reference somewhere.
20160831 19:14:41< celticminstrel> (Somewhere global, to be specific.)
20160831 19:15:12< vultraz> no, it's just accessed..
20160831 19:15:14< vultraz> hm
20160831 19:15:21 * vultraz runs debugger
20160831 19:15:49-!- fabi [~fabi@176.0.92.102] has joined #wesnoth-dev
20160831 19:16:23-!- mjs-de [~mjs-de@x4e31e2c8.dyn.telefonica.de] has joined #wesnoth-dev
20160831 19:18:52< vultraz> why.. is the crash in config::swap
20160831 19:18:53< vultraz> I don
20160831 19:18:55< vultraz> 't use swap :|
20160831 19:19:13< celticminstrel> Backtrace.
20160831 19:19:25< celticminstrel> Swap is probably called from the config's copy or move constructor.
20160831 19:19:30< celticminstrel> ^and/or
20160831 19:20:11-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160831 19:20:31< vultraz> let's try taking a copy...
20160831 19:20:52-!- ChipmunkV [~vova@static-89-94-113-91.axione.abo.bbox.fr] has quit [Quit: ChipmunkV]
20160831 19:21:23< celticminstrel> I think that crash probably indicates that that's already happening...
20160831 19:21:51< vultraz> ah wait
20160831 19:21:53< vultraz> found it
20160831 19:21:56< vultraz> dumb error
20160831 19:22:19< vultraz> i changed the function to return a config instead of operate on a reference and forgot to add a return statement :P
20160831 19:22:55< vultraz> (since I really hate the kind of thing where you do config foo; func(foo); just to fill it up :| )
20160831 19:23:11< vultraz> (much rather do config foo = func())
20160831 19:23:34< celticminstrel> Yeah, if you're generating the config, that makes sense.
20160831 19:23:52< celticminstrel> If the goal is operating on an existing config, then passing by reference is more efficient, though probably only marginally...
20160831 19:24:35< vultraz> in both usecases the existing config is either empty or being created
20160831 19:25:51< vultraz> random: can you append the contents of a stringsteam to another stream?
20160831 19:25:59< celticminstrel> Yes.
20160831 19:26:15< celticminstrel> In fact, you can append the contents of any input stream to any other stream.
20160831 19:26:24< celticminstrel> Or something like that.
20160831 19:26:39< celticminstrel> Maybe any output stream.
20160831 19:26:58< celticminstrel> ANYWAY, the way to do it is steam1 << stream2.rdbuf();
20160831 19:27:13< celticminstrel> ^stream1
20160831 19:27:53< vultraz> ahh
20160831 19:27:55< vultraz> thanks
20160831 19:28:34< celticminstrel> Ugh, I can't see a way to make this matrix placement work with horizontal grow...
20160831 19:28:51< celticminstrel> I guess I'll leave it with the extra empty space for now.
20160831 19:30:26-!- Kwandulin [~Miranda@p200300760F4241835C56635F93942045.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160831 19:34:01< celticminstrel> One of the possible resolutions doesn't show its aspect ratio...
20160831 19:35:00< vultraz> 3 don't
20160831 19:35:12< celticminstrel> Any idea why?
20160831 19:35:46< celticminstrel> Which ones don't show it?
20160831 19:36:10-!- Shiki [~Shiki@141.39.226.227] has quit [Quit: Verlassend]
20160831 19:36:18< vultraz> there's a condition to showing:
20160831 19:36:22< vultraz> 		const int div = boost::math::gcd(res.first, res.second);
20160831 19:36:23< vultraz> 		const int ratio[2] = {res.first/div, res.second/div};
20160831 19:36:25< vultraz> 		if(ratio[0] <= 10 || ratio[1] <= 10) {
20160831 19:36:34< celticminstrel> Why?
20160831 19:36:44 * vultraz shrugs
20160831 19:37:02< celticminstrel> I thought you added that.
20160831 19:37:08< vultraz> it was there before
20160831 19:37:19< celticminstrel> Just unused?
20160831 19:37:24< vultraz> no, used
20160831 19:40:03-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20160831 19:43:42-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 276 seconds]
20160831 19:43:42-!- wedge010 is now known as wedge009
20160831 19:44:19< celticminstrel> The C plural gettext macro is _n, right?
20160831 19:44:35< celticminstrel> Or n_?
20160831 19:44:39< celticminstrel> Something like that.
20160831 19:45:19< celticminstrel> I hate it when XCode sees that I clicked twice very quickly on different nodes and decides it's a double-click and collapses the parent node instead.
20160831 19:45:49< vultraz> _n
20160831 19:47:46< celticminstrel> Geh, don't use assert to compare selector and pager item counts!
20160831 19:47:55< vultraz> what?
20160831 19:48:06< celticminstrel> I'm changing it to VALIDATE.
20160831 19:48:09< vultraz> sure
20160831 19:48:19< celticminstrel> It's a condition that could easily be false if someone decided to add a new prefs page.
20160831 19:49:18< celticminstrel> Or if someone decided to reskin GUI2, which is probably unlikely but not impossible.
20160831 19:51:05< celticminstrel> I think VALIDATE_WITH_DEV_MESSAGE isn't needed since it's a condition that a non-dev is unlikely to encounter.
20160831 19:52:38-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160831 19:57:10-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 244 seconds]
20160831 19:57:51-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160831 19:58:19< irker510> wesnoth: Charles Dang wesnoth:master 312a80e55ec3 / src/gui/dialogs/ (campaign_difficulty.cpp campaign_difficulty.hpp): Campaign Difficulty: split constructor code into an out-of-class function https://github.com/wesnoth/wesnoth/commit/312a80e55ec3fc43b546909f479f9633a48076c9
20160831 19:58:22< irker510> wesnoth: Charles Dang wesnoth:master c08c55b704a0 / src/gui/dialogs/lobby/data.cpp: MP Lobby: cleanup and improvements to game data https://github.com/wesnoth/wesnoth/commit/c08c55b704a09273d9b8b24a73f9cdcd44e023de
20160831 19:58:52< vultraz> might eliminate map_info altogether
20160831 19:59:00< vultraz> and text filter individual elements
20160831 20:00:26< vultraz> as far as I can tell, it wouldn't be any more expensive than filtering on a string comprised of all of them
20160831 20:00:33< celticminstrel> Oh come on, why did you randomly change variable names?
20160831 20:00:43< vultraz> I didn't?
20160831 20:00:53< celticminstrel> difficulty_list -> difficulty_opts
20160831 20:01:02< vultraz> .. I did?
20160831 20:01:18< celticminstrel> Hmm, wait a second, what's really going on there...
20160831 20:01:31< celticminstrel> Ah, no, I misread the diff, sorrt.
20160831 20:01:33< celticminstrel> ^sorry
20160831 20:01:46< celticminstrel> (There's still difficulties_ -> result, but that one's more understandable.)
20160831 20:06:47-!- mjs-de [~mjs-de@x4e31e2c8.dyn.telefonica.de] has quit [Remote host closed the connection]
20160831 20:11:49< irker510> wesnoth: Charles Dang wesnoth:master 4308baac805e / data/gui/window/mp_create_game.cfg src/gui/dialogs/multiplayer/mp_create_game.cpp: MP Create: hide Mods header if none are available https://github.com/wesnoth/wesnoth/commit/4308baac805ebc1f15558fe5239acd04335e9de5
20160831 20:18:16< vultraz> ok, what else needs to be done..
20160831 20:18:34-!- fabi [~fabi@176.0.92.102] has quit [Ping timeout: 252 seconds]
20160831 20:23:54< vultraz> hm, what happened to the vision tooltip
20160831 20:28:16< celticminstrel> The what now?
20160831 20:28:37< celticminstrel> I should have three commits coming soon, BTW. As soon as I actually do the third one.
20160831 20:29:17< celticminstrel> One of them fixes the matrix placement, but I need to explicitly call invalidate_layout for it to work properly with horizontal grow, which is annoying.
20160831 20:29:37< vultraz> :|
20160831 20:30:11< celticminstrel> (Without horizontal grow, it works properly.)
20160831 20:30:16< celticminstrel> Still, at least it works.
20160831 20:30:28< celticminstrel> I put in TODO/FIXME notes about the relevant problem.
20160831 20:31:13< vultraz> i shall hold off on this commit until you push
20160831 20:31:37< celticminstrel> If it's already done, then go ahead.
20160831 20:31:46< celticminstrel> Since I haven't even done the third change yet.
20160831 20:32:03< celticminstrel> Unless it touches generator.cpp, in which case you can wait, :P
20160831 20:33:03-!- travis-ci [~travis-ci@ec2-54-80-161-132.compute-1.amazonaws.com] has joined #wesnoth-dev
20160831 20:33:04< travis-ci> wesnoth/wesnoth#10638 (master - 8c68d2a : Charles Dang): The build has errored.
20160831 20:33:04< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/156602567
20160831 20:33:04-!- travis-ci [~travis-ci@ec2-54-80-161-132.compute-1.amazonaws.com] has left #wesnoth-dev []
20160831 20:33:24< irker510> wesnoth: Charles Dang wesnoth:master 2f905128564f / src/gui/dialogs/lobby/lobby.cpp: MP Lobby: fixup add_tooltip_data https://github.com/wesnoth/wesnoth/commit/2f905128564f72d57d8c530795166234f8ef68b0
20160831 20:36:04< vultraz> i honestly didn't realize earlier how much work the lobby had needed :P
20160831 20:36:12< vultraz> but i think it has all the relevant features
20160831 20:36:13< vultraz> now
20160831 20:36:44< celticminstrel> Sigh, it doesn't work with the fixed-width linked group now... :|
20160831 20:36:55< celticminstrel> I guess for now I'll disable the linked group.
20160831 20:37:50-!- travis-ci [~travis-ci@ec2-54-211-221-115.compute-1.amazonaws.com] has joined #wesnoth-dev
20160831 20:37:51< travis-ci> wesnoth/wesnoth#10637 (master - a7bb98a : Charles Dang): The build has errored.
20160831 20:37:51< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/156598291
20160831 20:37:51-!- travis-ci [~travis-ci@ec2-54-211-221-115.compute-1.amazonaws.com] has left #wesnoth-dev []
20160831 20:38:02< celticminstrel> I guess your problem was that emplace/insert doesn't do anything if the key already exists.
20160831 20:40:33< vultraz> only way we'll be sure it's all good is to have users test it
20160831 20:45:27-!- prkc [~prkc@192.40.89.19] has quit [Ping timeout: 276 seconds]
20160831 20:47:06< mattsc> vultraz, or whoever: here’s the next one
20160831 20:47:09< mattsc> “src/game_initialization/multiplayer_wait.cpp:58:16: error: unused variable 'leader_pane_position’”
20160831 20:47:26< mattsc> and the same in the next line for leader_pane_border
20160831 20:47:33< vultraz> your compiler is really, really strict
20160831 20:47:38< irker510> wesnoth: Jyrki Vesterinen wesnoth:master 88ca728080de / join.lua src/gui/dialogs/dialog.cpp: Skip the GUI2 FLG dialog in the automatic multiplayer test https://github.com/wesnoth/wesnoth/commit/88ca728080de150dc659d5b8100e040def0f61bd
20160831 20:47:38< vultraz> even stricter than travis
20160831 20:47:40< irker510> wesnoth: Jyrki Vesterinen wesnoth:master 446de553bd3c / src/playturn_network_adapter.cpp: Fix UB in the playturn_network_adapter class https://github.com/wesnoth/wesnoth/commit/446de553bd3c4fb54c57fcb1b1f614fac4b68f84
20160831 20:47:42< irker510> wesnoth: Jyrki Vesterinen wesnoth:master 6016bdf2f249 / join.lua: Fix two bugs causing multiplayer tests to fail https://github.com/wesnoth/wesnoth/commit/6016bdf2f249e8a65769fbe44e0d63c29110c6d7
20160831 20:48:08< JyrkiVesterinen> ...and with those four fixes, I finally got MP tests to pass reliably on my machine.
20160831 20:48:09< mattsc> I’m just using Xcode with whatever settings people have set it up with to use.  ;)
20160831 20:48:36< JyrkiVesterinen> Hopefully they'll pass in Travis as well. They have been broken too long.
20160831 20:48:53< vultraz> JyrkiVesterinen: does that fix the gui2 tests too?
20160831 20:49:05< JyrkiVesterinen> No, I haven't done anything to them.
20160831 20:51:13< celticminstrel> JyrkiVesterinen: About that aliasing warning on the reinterpret_cast, do you think I should use a union instead of reinterpret_cast in both cases? I could just also suppress the warning with pragmas.
20160831 20:51:35< celticminstrel> I think using a union would guarantee that the memory is aligned properly though, which might be good.
20160831 20:51:41< JyrkiVesterinen> Yes, union is a good option.
20160831 20:51:44< vultraz> JyrkiVesterinen: any way you could look into the crash in the gui2 tests on join({})?
20160831 20:51:46< celticminstrel> 'aky
20160831 20:51:48< mattsc> vultraz: remember, I barely know the difference between compiling and linking  ;)
20160831 20:51:49< celticminstrel> ^'kay
20160831 20:52:12< celticminstrel> vultraz: It's not the compiler.
20160831 20:52:14< JyrkiVesterinen> vultraz, can you give more information about that crash?
20160831 20:52:18< celticminstrel> It's the options.
20160831 20:52:44< celticminstrel> mattsc could get around this issue by disabling the "warn about unused variables" option in the project settings.
20160831 20:53:21< vultraz> JyrkiVesterinen: ensure you have the experimental mp lobby selected, run the gui2 test (utils/travis/mp_test_executor_gui2.sh I think the name is), and wait for Join to try to join the game. It will crash.
20160831 20:53:23< mattsc> celticminstrel: I know — but it seems we’re doing this a lot lately, which seems … unideal
20160831 20:53:36< celticminstrel> Probably because vultraz does not have that warning enabled.
20160831 20:54:05< celticminstrel> I think it is enabled in scons though, so vultraz should probably figure out how to enable it in CodeBlocks.
20160831 20:54:10< vultraz> JyrkiVesterinen: if, in the lua plugin file (there are separate ones for gui2), if you call context.join() (no empty table), it won't crash but will not join the game
20160831 20:54:22< vultraz> JyrkiVesterinen: with the empty table it does crash :/
20160831 20:54:33< mattsc> celticminstrel: A “while” (probably a couple years) ago, the Xcode project simply used -Werror stricly, without exceptions.  That’s not even close any more these days.
20160831 20:54:44< celticminstrel> JyrkiVesterinen: Note that although that's a shell-script, the three Wesnoth calls should work just fine from a normal Windows command prompt.
20160831 20:55:11< celticminstrel> (Though of course you'd need three command prompts open to do that.)
20160831 20:55:15< JyrkiVesterinen> Okay, sure. Crashes are fairly easy to debug. :)
20160831 20:55:22< vultraz> er.. hm. 
20160831 20:55:23< JyrkiVesterinen> I'll likely get around to it later this week.
20160831 20:55:28< vultraz> i just tested and now HOST crashes :|
20160831 20:55:30< vultraz> blah
20160831 20:55:35 * vultraz gives up
20160831 20:56:46< celticminstrel> Oh hey, Jyrki added a plugins context to every GUI2 dialog.
20160831 20:56:49< irker510> wesnoth: Charles Dang wesnoth:master c4e3838d56eb / src/game_initialization/multiplayer_wait.cpp: Cleaned up unused variables https://github.com/wesnoth/wesnoth/commit/c4e3838d56ebf25ad7b7f1ea985f9edf6eaa50ef
20160831 20:56:55< vultraz> mattsc: ^
20160831 20:57:10< vultraz> oh huh
20160831 20:57:12< vultraz> he did
20160831 20:57:16< vultraz> I wonder if this is useful..
20160831 20:57:27< celticminstrel> Could it be that that gets in the way of the explicit plugins context in the lobby?
20160831 20:58:08< mattsc> vultraz: thank you
20160831 20:58:23< celticminstrel> Well, since he said he'd look into the crash with plugins in the GUI2 library, I guess he'll find a way to fix that if it does.
20160831 20:58:47< celticminstrel> JyrkiVesterinen: BTW, "ensure you have the GUI2 lobby selected" means enabling it in advanced prefs, in case you weren't aware.
20160831 20:59:10< celticminstrel> vultraz: I think it's very useful.
20160831 20:59:11< JyrkiVesterinen> I didn't even think that adding a plugins context to all dialogs would cause trouble.
20160831 20:59:25< JyrkiVesterinen> Are plugins even used outside the automatic MP tests?
20160831 20:59:34< vultraz> no
20160831 20:59:41< vultraz> well, the mp dialogs, you mean
20160831 21:00:10< celticminstrel> vultraz: No, he means the tests. :P
20160831 21:00:20< vultraz> well both are true
20160831 21:01:36< vultraz> hm
20160831 21:02:47< celticminstrel> JyrkiVesterinen: The reason I thought it might cause trouble is because the MP Lobby is a GUI2 dialog which installs a plugin context in pre_show, so at the very least the plugin script would need to be aware of the possibility of that context, and be able to skip through it without closing the dialog.
20160831 21:03:01< celticminstrel> MP Create is likewise a GUI2 dialog that does this.
20160831 21:03:41< celticminstrel> I guess you could test if a dialog has its own plugins context using dynamic_cast...
20160831 21:04:08< JyrkiVesterinen> Well, on the other hand it's useful for test scripts to be able to close an arbitrary dialog.
20160831 21:04:13< mattsc> Ooo, mising era_blank.cfg
20160831 21:04:14< celticminstrel> Yes, indeed.
20160831 21:04:20< mattsc> *missing
20160831 21:04:24< celticminstrel> mattsc: I already fixed that locally.
20160831 21:04:44< vultraz> ok, Host crashes if the UI isn't on Scenarios...
20160831 21:04:45< mattsc> celticminstrel: I’ll do so here too  (I actually do know how to do that, surprise surprise)
20160831 21:04:46< vultraz> wat :|
20160831 21:05:10< celticminstrel> vultraz: Maybe it should call context.set_type?
20160831 21:05:20< vultraz> ?
20160831 21:05:21< celticminstrel> I dunno.
20160831 21:05:29< vultraz> it does
20160831 21:05:33< celticminstrel> Ah.
20160831 21:05:41< vultraz> but there's no ui interaction
20160831 21:05:43< vultraz> remember
20160831 21:06:31< celticminstrel> But it does remember what the UI was last set to from prefs.
20160831 21:06:57< mattsc> celticminstrel: could you add copying libpcre.1.dylib to the Frameworks dir too, please?
20160831 21:07:08< celticminstrel> That wasn't there already?
20160831 21:07:12< celticminstrel> I guess I can do that.
20160831 21:07:12< mattsc> nope
20160831 21:10:08< celticminstrel> So this is weird. _("strike", "strikes", 1) returns "strike*".
20160831 21:18:00-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160831 21:18:25< celticminstrel> Oh, I see the problem. I didn't check for npos.
20160831 21:18:48-!- prkc [~prkc@gateway/vpn/privateinternetaccess/prkc] has joined #wesnoth-dev
20160831 21:22:07-!- JyrkiVesterinen [~JyrkiVest@78-27-126-82.bb.dnainternet.fi] has quit [Quit: Going to bed]
20160831 21:23:17-!- travis-ci [~travis-ci@ec2-54-158-100-40.compute-1.amazonaws.com] has joined #wesnoth-dev
20160831 21:23:18< travis-ci> wesnoth/wesnoth#10639 (master - c08c55b : Charles Dang): The build has errored.
20160831 21:23:18< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/156623164
20160831 21:23:18-!- travis-ci [~travis-ci@ec2-54-158-100-40.compute-1.amazonaws.com] has left #wesnoth-dev []
20160831 21:28:54-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160831 21:33:28< mattsc> Sigh …
20160831 21:33:41< mattsc> Now I get eras.cfg not found in _main.cfg …
20160831 21:33:58 * zookeeper blinks
20160831 21:34:06< mattsc> If I remove that line, MP games don’t show and ears (obviously) and any campaign segfaults upon start
20160831 21:34:26< mattsc> s/and ears/any ears
20160831 21:34:31< mattsc> grrrrr
20160831 21:34:35< vultraz> i moved era.cfg
20160831 21:34:36< mattsc> ears = eras
20160831 21:34:48< vultraz> it was in data/multiplayer/eras.cfg
20160831 21:34:49< mattsc> did you start Wesnoth after you did that?
20160831 21:34:51< vultraz> now it's in data/
20160831 21:35:00< vultraz> and yes, I updated the include path]
20160831 21:35:57< zookeeper> ...why would you move it out of multiplayer/ ?
20160831 21:36:11< mattsc> vultraz: “error preprocessor: Macro/file 'eras.cfg' is missing at _main.cfg:36"
20160831 21:36:18< zookeeper> it's, like... a multiplayer thing
20160831 21:36:23< mattsc> anyways, I am going back to 1.12 for now, this is taking way too much time
20160831 21:36:30< vultraz> zookeeper: sp loads them too
20160831 21:37:05-!- prkc [~prkc@gateway/vpn/privateinternetaccess/prkc] has quit [Remote host closed the connection]
20160831 21:37:32< zookeeper> well should we consider the multiplayer dir to mean "this stuff is not loaded except in MP"?
20160831 21:38:00< vultraz>  yes
20160831 21:38:07< zookeeper> eras are a multiplayer thing by definition
20160831 21:38:08< mattsc> Hmm, the file did not get copied for me…
20160831 21:38:18< zookeeper> whether they're loaded elsewhere too or not (why are they?)
20160831 21:38:36< mattsc> One of those quirks of Xcode?  I guess I could try a clean build…
20160831 21:38:42< mattsc> Sigh.
20160831 21:39:56< celticminstrel> vultraz: Can you explain why update_details calls on_tab_select in MP Create?
20160831 21:40:37< vultraz> celticminstrel: to update the details in the tabs
20160831 21:40:38< celticminstrel> mattsc: Uhh, that's probably my fault, since awhile ago I made XCode copy those files explicitly rather than implicitly.
20160831 21:40:48-!- horrowind [~Icedove@2a02:810a:8380:10a8:21b:fcff:fee3:c3ff] has quit [Quit: horrowind]
20160831 21:40:50< celticminstrel> vultraz: Okay, and is there any reason why it needs to be called last?
20160831 21:41:02< mattsc> celticminstrel: oh …  why?
20160831 21:41:13< vultraz> not really but it's best
20160831 21:41:21< celticminstrel> mattsc: Because I felt it improved the progress indicator during the copy stage, if I recall correctly.
20160831 21:41:31< celticminstrel> vultraz: Why is it best?
20160831 21:41:44< vultraz> since it ensures all details are already set
20160831 21:41:46< mattsc> celticminstrel: I see
20160831 21:41:58< vultraz> when it updates 
20160831 21:42:12< celticminstrel> vultraz: The reason I'm asking is that I'm getting a missing widget in update_details, presumably because it's in a hidden tab.
20160831 21:42:14< vultraz> i suppose in practice it just needs to be called after resetting the config_engine_
20160831 21:42:34< vultraz> celticminstrel: what widget?
20160831 21:42:39< celticminstrel> game_title
20160831 21:42:45< celticminstrel> 800x600
20160831 21:43:06< celticminstrel> I can move the on_tab_select call to just after the config engine is reset.
20160831 21:43:11< vultraz> oh yeah
20160831 21:43:13< celticminstrel> Hopefully that'd fix it.
20160831 21:43:15< vultraz> that's a new thing
20160831 21:43:22< vultraz> I didn't add it to the smaller resolutions
20160831 21:43:23< celticminstrel> Huh?
20160831 21:43:27< celticminstrel> Oh, okay.
20160831 21:43:32< celticminstrel> I guess that means I need to add it then.
20160831 21:43:41< celticminstrel> Well, time to revert to GUI1 so I don't have to recompile.
20160831 21:43:44< vultraz> it's a label that goes above the minimap
20160831 21:43:55-!- prkc [~prkc@gateway/vpn/privateinternetaccess/prkc] has joined #wesnoth-dev
20160831 21:44:17-!- Appleman1234 [~Appleman1@119.104.100.118] has quit [Ping timeout: 244 seconds]
20160831 21:44:45< celticminstrel> Heh, the mods section looks really bad - the name doesn't fit, and part of it is outside the gold selection rect.
20160831 21:44:52< celticminstrel> (On GUI1, so nothing to care about.)
20160831 21:45:19< celticminstrel> BTW vultraz, the behaviour of Cancel at MP Connect differs if GUI2 lobby is enabled.
20160831 21:45:42< celticminstrel> With GUI1 lobby, it returns to the config screen, but with GUI2 lobby, it boots to titlescreen.
20160831 21:45:55< celticminstrel> Ack wait, I forgot to set a turn limit.
20160831 21:46:18< celticminstrel> There's a way to set a turn limit from Lua/WML, right?
20160831 21:46:41< vultraz> yes
20160831 21:46:45< vultraz> turns = 
20160831 21:46:51< celticminstrel> What?
20160831 21:46:56< vultraz> ..what?
20160831 21:47:07< celticminstrel> I mean from the command console.
20160831 21:47:11< vultraz> oh
20160831 21:47:13< vultraz> no
20160831 21:47:17< vultraz> you can set the current turn though
20160831 21:47:21< celticminstrel> ie, something that would work in an event.
20160831 21:47:27< vultraz> ;turn=num
20160831 21:47:39< celticminstrel> Isn't there a WML tag that sets the limit?
20160831 21:48:08< Ravana_> https://wiki.wesnoth.org/DirectActionsWML#.5Bmodify_turns.5D
20160831 21:48:29< vultraz> celticminstrel: i assume you used Local Game?
20160831 21:48:40< celticminstrel> Yeah, why?
20160831 21:48:47< vultraz> hm
20160831 21:48:49< vultraz> ok
20160831 21:50:48< celticminstrel> This would be easier if I had a command-line tool that renders and pages HTML...
20160831 21:50:50< zookeeper> vultraz, Fix duplicate loading of eras in mp
20160831 21:50:59< vultraz> yes?
20160831 21:51:00< zookeeper> what caused duplicate loading?
20160831 21:51:07< celticminstrel> Ah, the key is value. Okay then!
20160831 21:51:27< zookeeper> and why wasn't the duplicate loading fixed instead of moving the files?
20160831 21:51:52< celticminstrel> An excellent question.
20160831 21:52:22< vultraz> Moving the file was a logical part of fixing it
20160831 21:52:28 * celticminstrel does think that the eras are better left in multiplayer/ even if thy're included from outside that.
20160831 21:53:13< zookeeper> if something loads a file/dir twice, then no a logical part of a fix is not to move the file/dir elsewhere
20160831 21:53:13< vultraz> era.cfg was included from multiplayer/_main.cfg
20160831 21:53:24< vultraz> it was  also from data/_main.cfg
20160831 21:53:32< Aginor> morning
20160831 21:53:37< vultraz> when you entered mp, the file was now loaded twice
20160831 21:53:54< vultraz> i deleted the include in multiplayer/_main.cfg, since it was already loaded outside of mp
20160831 21:54:10< vultraz> then, since it was loaded outside of mp, I moved the file since it's now no longer just for mp
20160831 21:54:46-!- fabi [~fabi@176.0.92.102] has joined #wesnoth-dev
20160831 21:55:07< zookeeper> "<vultraz> it was  also from data/_main.cfg" <- only in SP
20160831 21:55:58< vultraz> why distinguish?
20160831 21:56:06< vultraz> it's loaded for both then, and now
20160831 21:56:12< vultraz> except now the macros aren't defined twice
20160831 21:56:23< vultraz> meaning no preprocessor warnings
20160831 21:57:32< zookeeper> because if data/_main.cfg loads eras.cfg only in SP and in MP data/_main.cfg loads data/multiplayer/_main.cfg which loads eras.cfg, then how can eras.cfg be getting loaded twice?
20160831 21:58:14< vultraz> it's likely the preprocessor doesn't do a full rebuild when entering mp or something
20160831 21:58:22< vultraz> somehow, the macros were redefined
20160831 21:58:37< vultraz> i'm not exactly sure why, but this fixes it and the behavior is more logical
20160831 21:58:43< vultraz> why is this such a problem :|
20160831 21:59:33< celticminstrel> Well, it seems like this fixed the markup?
20160831 21:59:35< zookeeper> this is just sanity-checking
20160831 21:59:44< celticminstrel> I thought there'd be something else, but guess not.
20160831 22:00:09< vultraz> zookeeper: are we so entrenched in our ways that we cannot move a *file* without debate :|
20160831 22:00:29< celticminstrel> [Aug 31@5:53:13pm] zookeeper: if something loads a file/dir twice, then no a logical part of a fix is not to move the file/dir elsewhere
20160831 22:00:29-!- Appleman1234 [~Appleman1@KD119104106244.au-net.ne.jp] has joined #wesnoth-dev
20160831 22:00:30< celticminstrel> This ^
20160831 22:00:55< celticminstrel> I don't feel strongly about it, but to me, eras are logically part of multiplayer (even if included outside it) and thus belong in multiplayer/.
20160831 22:00:56< vultraz> "then no a [] is not"??
20160831 22:01:08< celticminstrel> ??
20160831 22:01:26< vultraz> " then no a logical part of a fix is not to"
20160831 22:01:49< celticminstrel> I don't get what you're asking/
20160831 22:01:57< zookeeper> vultraz, no, but when you move a file from its logical place to another one in what seems like an attempt to workaround some bug that you don't even mention anywhere as you do so, then sanity-checking is in order :>
20160831 22:02:06< celticminstrel> The logical fix would've been to just remove the include in multiplayer/
20160831 22:02:08-!- Duthlet [~Duthlet@dslb-188-104-247-024.188.104.pools.vodafone-ip.de] has quit [Quit: leaving]
20160831 22:02:25< celticminstrel> Or wrap the include outside of multiplayer/ in an #ifndef MULTIPLAYER
20160831 22:03:23< vultraz> [09:02:06]	celticminstrel	The logical fix would've been to just remove the include in multiplayer/
20160831 22:03:24< vultraz> yes
20160831 22:03:26< vultraz> did so
20160831 22:03:41< celticminstrel> Right, I mean the logical fix would be to do that and nothing else.
20160831 22:03:43< vultraz> I moved it because it seemed logical outside since it's used by both :|
20160831 22:03:48< vultraz> that's *different*
20160831 22:03:49< vultraz> not a workaround
20160831 22:03:51< celticminstrel> Moving the file isn't a logical part of the fix.
20160831 22:03:55< vultraz> yes
20160831 22:03:58< vultraz> it's *extra*
20160831 22:04:08< celticminstrel> Eras are logically part of multiplayer, so it's weird for them to be outside.
20160831 22:04:20< celticminstrel> (Why are they included in _main.cfg, anyway?)
20160831 22:04:30< vultraz> sp/mp unification, remember
20160831 22:05:04< zookeeper> i thought sp/mp unification was a technical endeavour
20160831 22:08:08< celticminstrel> Well, whatever.
20160831 22:08:20< celticminstrel> Still disagree on the logic of it, but don't care enough to want to reverse it.
20160831 22:08:55< zookeeper> i mean it's simply a bit baffling that you think that the fact that eras.cfg needs to be included from outside multiplayer (still i don't know why that is; presumably there is some reason) means that the file is no longer logically a multiplayer thing and that thus it should be moved out of data/multiplayer/. but i don't care much about internal file locations like that either.
20160831 22:09:44-!- travis-ci [~travis-ci@ec2-54-205-13-9.compute-1.amazonaws.com] has joined #wesnoth-dev
20160831 22:09:45< travis-ci> wesnoth/wesnoth#10640 (master - 4308baa : Charles Dang): The build has errored.
20160831 22:09:45< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/156626420
20160831 22:09:45-!- travis-ci [~travis-ci@ec2-54-205-13-9.compute-1.amazonaws.com] has left #wesnoth-dev []
20160831 22:11:11< mattsc> celticminstrel: could you add that to your Xcode updates too, please?
20160831 22:11:19< celticminstrel> Travis hasn't yet reached Jyrki's push, right?
20160831 22:11:36< celticminstrel> mattsc: Yeah. Adding eras.cfg, adding libpcre, anything else?
20160831 22:11:42< zookeeper> it also seems like the kind of thing that someone would notice, think that it'd make more sense in data/multiplayer/, and move it back, again thinking they've made things more logical and neat.
20160831 22:11:50< celticminstrel> And removing era_blank.cfg has already been done.
20160831 22:11:51< mattsc> celticminstrel: eras.lua also
20160831 22:12:05< celticminstrel> Eh? That's not in lua/?
20160831 22:12:16< mattsc> nope, it’s in data/ now
20160831 22:12:28< celticminstrel> I think that should be in lua/.
20160831 22:12:35< vultraz> it was alongside eras.cfg
20160831 22:12:41< vultraz> so i moved it alongside it
20160831 22:12:50< celticminstrel> Hmm.
20160831 22:14:02< vultraz> [08:37:32]	zookeeper	well should we consider the multiplayer dir to mean "this stuff is not loaded except in MP"?
20160831 22:14:05< vultraz> this was my logic
20160831 22:14:12< vultraz> mp-exclusive stuff
20160831 22:14:51< zookeeper> anyway, the relevant commit that introduced the era loading in SP seems to be c2e1d71fc25
20160831 22:16:46< vultraz> and?
20160831 22:17:03< zookeeper> well it's relevant for the duplicate loading
20160831 22:17:46< zookeeper> and i don't really understand what it does and how it's supposed to work
20160831 22:21:27< vultraz> I hope you're not going to move the files back
20160831 22:23:44< celticminstrel> It'd be great if they were moved back, but I can't be bothered to do it myself.
20160831 22:24:41< celticminstrel> Gonna make game_title default instead of default_large on the 800x600 resolution.
20160831 22:26:18< vultraz> obv
20160831 22:26:59< celticminstrel> Otherwise I get a slight vertical scrollbar.
20160831 22:30:17-!- celmin [~celticmin@unaffiliated/celticminstrel] has joined #wesnoth-dev
20160831 22:30:59< mattsc> zookeeper: I have a question on what to name a tag for the Assassin Micro AI
20160831 22:31:52< mattsc> I’m finally getting around to adding the [avoid] tag, but I think the best way of doing it is to not make it a strict avoiding of the locations defined by the SLF.
20160831 22:32:30< mattsc> Essentially, the AI adds a big penalty to those hexes and will never go there, _except_ when it does not have another choice.
20160831 22:33:03< mattsc> But to avoid (no pun intended) confusion, I think this should not be called [avoid] then.
20160831 22:33:10< mattsc> I can’t come up with a better name though.
20160831 22:33:19< zookeeper> hmh, okay. sounds reasonable since you can still use other AI controls to make them _never_ go there, right?
20160831 22:33:27< mattsc> [soft_avoid] seems … weird
20160831 22:33:40< zookeeper> [dislike] :>
20160831 22:33:44< celmin> At what point does it not have another choice?
20160831 22:33:58< mattsc> Well, I am intending to add a key stiket_avoid=yes/no to do that, yes
20160831 22:34:17< mattsc> I can’t type today!
20160831 22:34:22< mattsc> *strict_avoid
20160831 22:34:26< zookeeper> a mighty typo
20160831 22:34:40< mattsc> I don’t think I ever woke up today  :P
20160831 22:35:15< celmin> vultraz: Is there any reason to always show eras and mods side-by-side, rather than mods below eras?
20160831 22:35:23< vultraz> space
20160831 22:35:36< vultraz> more space for a longer mod list
20160831 22:35:37< zookeeper> i do wonder the same as celmin though, when could it possibly not have another choice?
20160831 22:35:50< mattsc> celmin: it’s mostly to make thing easier on both the scenario designer and the AI coder (i.e. me) to make the behavior easily exploitable by the player
20160831 22:36:13< celmin> vultraz: I see, hmm. (And the 800x600 version wouldn't work otherwise, as well.)
20160831 22:36:33< celmin> But it seems to be the mods list that's causing the 1024x768 version to have a horizontal scrollbar...
20160831 22:37:05< mattsc> Let’s say I have forest-hiding assassins that I want to sneak up through the forest north of the enemy keep and let the aboud the desert in the south.
20160831 22:37:27< mattsc> So I set them to avoid all non-forest terrain.
20160831 22:37:45< mattsc> All the player has to do is move two hexes into the desert, and they will never get there.
20160831 22:37:58< mattsc> And yes, I know the argument with the radius=6 around the enemy leader thing.
20160831 22:38:07< mattsc> So I move my leader 8 hexes into the desert.
20160831 22:38:45< mattsc> It puts a lot of burden on the scenario designer to make sure that something like that is not possible.
20160831 22:39:19< celmin> So basically, they have no other choice only when the leader is not adjacent to a non-avoided tile?
20160831 22:39:22< mattsc> By contrast, if I don’t make it impossible, but give a high penalty for all non-forest hexes, then the AI will chose the path mostly through forest, with the fewest hexes across desert.
20160831 22:39:34< mattsc> Right.
20160831 22:39:52< mattsc> Which is what I would call the desired behavior anyway.
20160831 22:40:11< zookeeper> mattsc, okay... so if my leader is 8 hexes into the desert, _but_ there is a 1-hex path of forest leading to it which just happens to be blocked by a unit... then will the assassin go through the desert, or does it ignore the blocked path?
20160831 22:40:12< celmin> Possibly also only if it's in reach, or at least the unit is on the closest non-avoided tile to the leader.
20160831 22:40:38< mattsc> zookeeper: it ignores the blocked path
20160831 22:41:02< mattsc> blocked = impossible; desert = undesirable
20160831 22:42:04< celmin> So does that mean it'll go through the desert?
20160831 22:42:12< mattsc> yes
20160831 22:42:46< mattsc> But there will be an option to switch to completely avoiding the desert, if that’s what’s wanted
20160831 22:43:34< mattsc> That will essentially disable the AI though, if there’s no path to the target.
20160831 22:43:37< celmin> Hmm, disabling horizontal grow on the era menu button seems to have crashed the game at startup...
20160831 22:44:02< celmin> Oh, wait, I guess the crash is a transient thing unrelated to that.
20160831 22:44:15< mattsc> zookeeper: anyway, [dislike] is an option that I had thought of too, but it sounds a bit strange to me also
20160831 22:44:29< celmin> Seems like the loading screen still has some threading issues?
20160831 22:44:35< mattsc> [unprefer]  :P
20160831 22:44:42< celmin> Unless this is actually preferable.
20160831 22:45:22 * celmin thinks that [avoid] kind of makes sense though - the connotation of the word is "I won't go there unless I have to", and it's the rest of the AI that treats it unusually.
20160831 22:45:37< celmin> …that use of "unusually" is weird.
20160831 22:45:46< zookeeper> skirt, fudge, shirk, shun, avert? :p
20160831 22:45:59< mattsc> celmin: yes, but I want to avoid confusion with how [avoid] is used elsewhere
20160831 22:46:30< mattsc> shun, that’s not bad, maybe …
20160831 22:47:36< zookeeper> mattsc, anyway... it sounds reasonable, i'm just wondering about possible consequences.
20160831 22:47:38< mattsc> celmin: that’s why I am having an issue with naming it, because I think that avoid is actually the perfect word for it
20160831 22:48:44< zookeeper> well, invert it: have a [prefer] tag for defining the "allowed" locations
20160831 22:48:50< zookeeper> problem solved :P
20160831 22:49:05< zookeeper> (dunno if that would work)
20160831 22:49:31-!- travis-ci [~travis-ci@ec2-54-158-100-40.compute-1.amazonaws.com] has joined #wesnoth-dev
20160831 22:49:32< travis-ci> wesnoth/wesnoth#10641 (master - 2f90512 : Charles Dang): The build has errored.
20160831 22:49:32< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/156631705
20160831 22:49:32-!- travis-ci [~travis-ci@ec2-54-158-100-40.compute-1.amazonaws.com] has left #wesnoth-dev []
20160831 22:49:41< mattsc> zookeeper: hmm ...
20160831 22:50:49< mattsc> That’s interesting.
20160831 22:51:51 * mattsc is thinking about the technical implementation of a potential [prefer] tag…
20160831 22:52:31< mattsc> I’m not sure if custom path finding works with non-integer costs
20160831 22:52:50< mattsc> But if not, there’s a way around that too
20160831 22:52:56< celmin> Why is that relevant?
20160831 22:53:58< celmin> Also, it seems like something easily looked up in the source.
20160831 22:54:20< mattsc> because my gut reaction is to implement preferred hexes by reducing their movement cost
20160831 22:54:44< celmin> I see.
20160831 22:55:08< mattsc> but of course it can be done by increasing the cost for all other hexes instead
20160831 22:55:58< mattsc> zookeeper: I’ll take a step back and think about this.  That’s an interesting suggestion.
20160831 22:56:34< celmin> vultraz: Is grow_factor irrelevant if horizontal_grow is false?
20160831 22:56:43< vultraz> not sure
20160831 22:56:50< vultraz> but I think it is
20160831 22:56:52< vultraz> relevant
20160831 22:57:04-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has joined #wesnoth-dev
20160831 22:57:31< mattsc> From the wiki: “The function should return a floating-point value that is the cost for entering the given tile. This cost should be greater or equal to one.”
20160831 22:57:43< zookeeper> mattsc, remind me: if the assassin can't find any safe multi-turn path to the target, will it just tread water and move to the closest safe location, or will it try to reach the target even through unsafe locations if that's the only choice?
20160831 22:58:10< celmin> The code clamps the return value to a minimum of 1.
20160831 22:58:14< zookeeper> i could test but this is faster... :p
20160831 22:58:23< celmin> (In src/scripting/lua_pathfind_cost_calculator.hpp)
20160831 22:58:35< mattsc> zookeeper: If the only options are non-safe paths, it will take the “safest” of those.
20160831 22:59:07< mattsc> But it will prefer a long detour before taking any of those.
20160831 22:59:35< mattsc> That could be modified or configurable too.
20160831 22:59:47< zookeeper> hmh, okay. sounds good as-is for now.
20160831 23:00:10< mattsc> celmin: yes, that’s what the wiki says too (see above)
20160831 23:00:15< mattsc> thanks
20160831 23:00:46< zookeeper> is "safest" determined by the regular "lowest chance to die or lowest potential for damage" formulas?
20160831 23:02:32< celmin> What? Why does the A* search take a cost calculator by pointer if it's going to assume that it's not nullptr?
20160831 23:03:11< celmin> I don't see any reason why the A* search would fail with costs <1...
20160831 23:03:30< mattsc> zookeeper: had to look that up; it’s a somewhat ad-hoc equation, but one that makes some sense based on my intuition from working with AIs
20160831 23:03:49-!- edgrey [~edgrey@178.204.82.169] has quit [Ping timeout: 258 seconds]
20160831 23:04:11< celmin> Maybe someone with better understanding of the algorithm would know whether that's the case in general.
20160831 23:04:33< mattsc> So, it’s the maximum damage that can be dealt by all enemies on a hex, multiplied by the hit_chance for the unit on the hex, and then scaled by the max_hitpoints of the units.
20160831 23:05:00< celmin> In other words, I don't see any reason not to remove the restriction that the cost must be >= 1.
20160831 23:05:24< zookeeper> mattsc, ok, just curious. as long as it somehow takes the relative danger of different enemy units into account i'm sure it's fine...
20160831 23:05:43< mattsc> zookeeper: yes, it does
20160831 23:05:46< celmin> Ohh, now it says it's assumed >= 1. Hmm. Why?
20160831 23:06:28< mattsc> I also multiply that by an arbitrary factor of 5 at the end … 
20160831 23:07:28< mattsc> celmin: I only know how an A* search works in principle, but not much about the practical implications.
20160831 23:08:46< mattsc> The whole custom cost thing does not work very well for Wesnoth anyway, because it does not take (not) entering a 2 movecost hex with only 1 MP remaining into account correctly etc.
20160831 23:09:58< mattsc> zookeeper: so what’s taken into acount is: damage the enemies can deal (taking resistance etc. into account), the chance_to_hit for the AI unit, and what fraction of the unit’s max hitpoints that is.
20160831 23:11:00< mattsc> In other words, it will prefer high-HP units that are resistant to the enemy weapons on hexes where they have strong defense.
20160831 23:11:24< mattsc> And most of all it will prefer hexes where the enemy can’t deal much, or even better: no, damage
20160831 23:12:20< mattsc> I guess that should be chance-to-BE-hit 3 lines up there
20160831 23:23:17< irker510> wesnoth: ln-zookeeper wesnoth:master 991e72aaef32 / data/core/sounds/ambient/ (morning.ogg night.ogg): Lowered the volume of the night and morning sounds https://github.com/wesnoth/wesnoth/commit/991e72aaef321570504fd569c811bacf645a7797
20160831 23:27:06< vultraz> y u do dis?
20160831 23:29:11< zookeeper> because so many people have complained about them for years
20160831 23:29:21< vultraz> oh?
20160831 23:29:37< fabi> yeah, they are nice somehow. But too loud.
20160831 23:34:34< celmin> Makes sense to me. I've kinda thought they were a bit loud too.
20160831 23:35:35< celmin> It seems like there is a reason for A* to ensure costs >= 1, though I'm not clear on exactly what it is.
20160831 23:35:40< shadowm> I still think some variations would help.
20160831 23:35:47< shadowm> (Re ToD transition sounds.)
20160831 23:37:27< zookeeper> i still agree but sadly i find the prospect too tricky to be taken on lightly
20160831 23:47:39-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20160831 23:53:13< celmin> Getting ready to push finally.
20160831 23:58:20 * shadowm wasn't aware zookeeper agreed.
20160831 23:59:39-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160831 23:59:55< zookeeper> i thought i agreed the last time you brought it up? well, maybe i misremember.
--- Log closed Thu Sep 01 00:00:21 2016