--- Log opened Thu Sep 01 00:00:21 2016
20160901 00:02:43-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160901 00:04:24-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has joined #wesnoth-dev
20160901 00:04:32-!- travis-ci [~travis-ci@ec2-54-80-161-132.compute-1.amazonaws.com] has joined #wesnoth-dev
20160901 00:04:33< travis-ci> wesnoth/wesnoth#10642 (master - 6016bdf : Jyrki Vesterinen): The build has errored.
20160901 00:04:33< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/156635125
20160901 00:04:33-!- travis-ci [~travis-ci@ec2-54-80-161-132.compute-1.amazonaws.com] has left #wesnoth-dev []
20160901 00:06:22< tad_> Looking for guidance on fixing an error concerning [terrain_type]. I can make it go away but I'm concerned it might have unintended side effects or may indicate a deeper problem.
20160901 00:06:53< irker510> wesnoth: Charles Dang wesnoth:master 02497800560b / src/tests/gui/test_gui2.cpp: Fixup tests https://github.com/wesnoth/wesnoth/commit/02497800560b536f7a1a7f3832e5f9179c3e9771
20160901 00:06:56< zookeeper> tad_, had no time for the TSG PR today, but tomorrow...
20160901 00:07:22 * zookeeper was just about to go ->
20160901 00:07:26< tad_> zookeeper: No rush. I'm beginning another pass at DM ..
20160901 00:07:26< vultraz> oh, celmin was gonna push
20160901 00:07:28< vultraz> sorry :(
20160901 00:07:38< celmin> :(
20160901 00:07:56< celmin> Well, I can just rebase, but since I touched the XCode project it's a bit of a pain.
20160901 00:08:22< tad_> The issue is Delfador's Memoirs and message on the screen concerning "The map cannot be loaded." due to invalid terrain types.
20160901 00:08:27< celmin> I guess I'll try harder to fix MP Create 1024x768 before pushing, then.
20160901 00:09:12< tad_> It appears that the error appears on Load panel. Probably generating or loading the preview map.
20160901 00:09:34< vultraz> oh yeah
20160901 00:09:36< vultraz> that happens sometimes
20160901 00:09:40< celmin> Sounds like the scope is incorrect.
20160901 00:09:49< tad_> In _main.cfg it loads [terrain_type]
20160901 00:09:51< celmin> DM is not in scope, so it can't display the custom terrains.
20160901 00:10:02< celmin> I think what it should do is just treat custom terrains as void.
20160901 00:10:02< tad_> Right. I can snub the error, tho
20160901 00:10:54< tad_> If I move [terrain_type] outside the #ifdef's so it ALWAYS loads the errors go away because Load panel can now find them.
20160901 00:11:17< tad_> This is consistent with going into the editor, which always sees the DM custom terrains.
20160901 00:11:46-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [Ping timeout: 252 seconds]
20160901 00:11:52< tad_> But does that not mean that the DM custom terrains can collide with some UMC custom terrain? And is that bad or we-dont-care?
20160901 00:12:05-!- atarocch [~atarocch@88.131.217.34] has quit [Remote host closed the connection]
20160901 00:12:36< celmin> Yeah, that's a problem.
20160901 00:12:48< tad_> I looked at the C++ can can see where I'd do celmin's suggestion but I can't figure out how to ignore the errors there.
20160901 00:13:12-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160901 00:13:34< tad_> I do, however, has a PR in the works which will, at least, make it so it only complains once per bad tile instead of twice and a non-suppressible error in addition
20160901 00:14:52< tad_> celmin: OK, so moving the [terrain_type] outside the #ifdef protected area is bad.
20160901 00:18:53< tad_> I was looking at src/terrain/type_data.cpp, function try_merge_terrains (c. line 156) and the if (..) return false and thinking it needs what celmin suggested with a twist .. "If it's a UMC-reserved type [YZyz], then use a default replacement, otherwise return false" but I don't know how to test the first character of the type (they look like integers there to me) or what an appropriate replacement would be.
20160901 00:19:30< celmin> Huh, integers?
20160901 00:20:06< celmin> I seem to recall there was a length limit of 4… wonder if that could explain why...
20160901 00:20:08< tad_> Well, I might have read the code wrong but it looks like the 2-to-4 characters are stored in an int32
20160901 00:20:32< tad_> Which is no problem for C but a bite-your-ass in C++ ...
20160901 00:21:20< tad_> So, yeah, that's probably why the 2-to-4 requirement.
20160901 00:21:22< celmin> I don't think it's really a problem in C++ either?
20160901 00:22:22-!- vultraz [~chatzilla@124.109.10.167] has quit [Ping timeout: 255 seconds]
20160901 00:22:27< tad_> Well, sure, I can cast away the const-ness and cast it as a [4] char array but (a) it's endian-dependent and (b) when I tried the compiler went all pedantic on me
20160901 00:23:36< tad_> The endian issue is probably the old have-to-be-portable in me from way back and may not be an issue here.
20160901 00:24:44< celmin> The endian-dependency doesn't really matter as long as it's not written to disk in that format though.
20160901 00:26:19< celmin> Git is not letting me revert a commit. ><
20160901 00:27:20< celmin> It says "All conflicts fixed; run git revert --continue", but no matter how many times I do what it says, nothing changes.
20160901 00:27:22< fabi> tad_: Iirc the custom terrain in DM only served the purpose to hack some game mechanics.
20160901 00:27:46< tad_> I hit that. I ended up aborting the revert and starting over.
20160901 00:28:05< celmin> Tried that too.
20160901 00:28:11< fabi> tad_: If you rework DM removing the terrain should be fine.
20160901 00:28:22< tad_> fabi: I could code around the custom tiles and eliminate them. But I wanted to track down why the errors in the first place.
20160901 00:28:43< celmin> And checkout on the touched file is also not working… :|
20160901 00:29:09< tad_> fabi: But, yeah, I always felt it was sorta hack-ish the way it's done.
20160901 00:30:18< celmin> Oh, I think I must've already reverted it and accidentally committed it. Grrr.
20160901 00:30:25< celmin> That means I need to interactive rebase it. :|
20160901 00:30:48< tad_> Ah .. you forgot the -n to not apply the revert ...
20160901 00:31:57< tad_> On the issue of [terrain_type] I just put up a PR to reduce the number of errors on screen about it. That can go in now, or it can wait in case we decide to do something more about it in the engine.
20160901 00:33:20-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20160901 00:34:53< vultraz> tad_: will you be adding more to 762?
20160901 00:35:25< tad_> [19:31] <tad_> On the issue of [terrain_type] I just put up a PR to reduce the number of errors on screen about it. That can go in now, or it can wait in case we decide to do something more about it in the engine.
20160901 00:35:36< tad_> Up to you.
20160901 00:36:37< irker510> wesnoth: Gregory A Lundberg wesnoth:master 0a131f873c5b / src/map/map.cpp: Fix bug: No need to say it twice https://github.com/wesnoth/wesnoth/commit/0a131f873c5bd294cbd649d8919d63cb8b21fb18
20160901 00:36:39< irker510> wesnoth: Charles Dang wesnoth:master 019f9ab7576a / src/map/map.cpp: Merge pull request #762 from GregoryLundberg/GL_no_map_error_for_UMC_tiles https://github.com/wesnoth/wesnoth/commit/019f9ab7576a03ea7dbbd2da4ed66e711d50aae4
20160901 00:36:49< tad_> I think I'm going to take fabi's suggestion and recode to eliminate the custom terrain in DM anyway. So the issue of whether to complain about UMC custom types can wait.
20160901 00:38:02< celmin> tad_: No, I did use --no-commit, but I had a;ready reverted it earlier.
20160901 00:39:01< tad_> celmin: tbh I usually mess up reverts myself and have learned to fork a test branch whenever I want to check out one.
20160901 00:40:30< celmin> The problem is that revert stages the reversion, so I accidentally committed it along with something unrelated.
20160901 00:40:50 * tad_ nods.
20160901 00:41:10< celmin> So now I need to rebase and fix that.
20160901 00:41:16< fabi> Huh, that was hard. But now I have only 4 lines left which are wrong transcompiled in a whole campaign.
20160901 00:41:36< tad_> Like I said I usually mess up. checkout -b tells me I have uncommitted work and that usually stops me ..
20160901 00:42:17< celmin> That recent commit from vultraz gives me deja vu.
20160901 00:42:24< vultraz> what?
20160901 00:42:26< celmin> I thought I saw the exact same commit earlier today.
20160901 00:42:45< tad_> What, the "attempt to placate Travis" commits? He's done a few with the same message ...
20160901 00:42:54< celmin> The content of the commit.
20160901 00:42:57< celmin> The fixup tests one.
20160901 00:44:09< celmin> Why am I getting an "It is now Brackdon_Cawr's turn" popup?
20160901 00:44:16< celmin> I don't think that's in my scenario...
20160901 00:45:33< Aginor> 12:42 < tad_> What, the "attempt to placate Travis" commits? He's done a few  with the same message ...
20160901 00:45:58< tad_> hmm?
20160901 00:45:59< Aginor> tad_: different compiler/set of flags than what travis uses
20160901 00:46:19< Aginor> so it works for him but travis refuses to compile it because of stricter flags
20160901 00:46:52< tad_> I don't read 'em .. Travis isn't my thing. I just notices the same message a day or two apart.
20160901 00:47:39< Aginor> tad_: travis is our CI system
20160901 00:48:08< celmin> What does CI stand for, anyway?
20160901 00:48:10< tad_> Oh, I know what it is. I just don't look at it.
20160901 00:48:17< tad_> Continuous Integration.
20160901 00:48:34< celmin> Ah.
20160901 00:48:40< tad_> The idea is as we make changes, Travis checks it compiles, runs tests, and reports 
20160901 00:49:36-!- travis-ci [~travis-ci@ec2-54-158-100-40.compute-1.amazonaws.com] has joined #wesnoth-dev
20160901 00:49:37< travis-ci> wesnoth/wesnoth#10643 (master - c4e3838 : Charles Dang): The build has errored.
20160901 00:49:37< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/156637740
20160901 00:49:37-!- travis-ci [~travis-ci@ec2-54-158-100-40.compute-1.amazonaws.com] has left #wesnoth-dev []
20160901 00:49:38< tad_> We talked about tests for campaigns yesterday .. if we could do that we'd certainly want to have Travis run them
20160901 00:50:00< fabi> ^
20160901 00:51:32< Aginor> that'd be great
20160901 00:51:51< Aginor> and while we're wishing for stuff, I'd also always want master to be passing travis
20160901 00:52:07< Aginor> instead of being in a bad state the majority of the time :)
20160901 00:52:28< Aginor> although maybe I'm just looking at unfortunate times
20160901 00:52:45< tad_> I usually ignore Travis' complaints because it compiles fine for me or I pop in here and we get it fixed.
20160901 00:53:45< tad_> Travis is great if you're doing TDD (test-driven design) but this project is so old it's going to be a real pain to get it into a test-able state ...
20160901 00:54:04-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160901 00:54:05< Aginor> CI != TDD though
20160901 00:54:15< Aginor> CI allows us to always be releasable
20160901 00:54:22< Aginor> which is really not the same as TDD
20160901 00:54:29< tad_> Nope but it sure helps with TDD
20160901 00:55:47< celmin> Aginor: There've been efforts to fix the Travis issues already, so it's not like people are just ignoring it.
20160901 00:56:22< Aginor> celmin: I know, and I didn't mean for it to come of as a snide remark, which it unfortunately did
20160901 00:56:46< Aginor> celmin: at the same time, I think it might be good to do more work in feature branches that can be merged back when they are stable and passing
20160901 00:57:05< celmin> True.
20160901 00:57:14< fabi> I think all the gui2 work could be done in a branch
20160901 00:57:24< fabi> and merged after a steam release
20160901 00:57:26< Aginor> tad_: I disagree, if you're doing TDD you should be running your tests even before you push to your main repo ;)
20160901 00:57:47< celmin> fabi: Pretty sure the idea is for all the GUI rewrites to be included in the Steam release.
20160901 00:57:59< fabi> That idea is not good.
20160901 00:58:21 * Aginor shrugs
20160901 00:58:36< celmin> I don't see a problem with it.
20160901 00:58:36< Aginor> we should just release 1.12 on steam, do dev as the beta channel
20160901 00:58:37-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 265 seconds]
20160901 00:58:46< Aginor> updated weekly or something
20160901 00:59:11< fabi> Aginor: I agree. 1.12 on steam.
20160901 00:59:44< fabi> I don't even see how the project prepares for that steam release.
20160901 01:00:45< Aginor> fabi: I know as much as you do :)
20160901 01:00:47< tad_> Aginor: I'd point out that automating the TDD tests in the main repo is a fall back so lazy programmers don't slip one in ...
20160901 01:01:30< irker510> wesnoth: Celtic Minstrel wesnoth:master d9111620bd67 / projectfiles/Xcode/Wesnoth.xcodeproj/project.pbxproj: Update XCode project https://github.com/wesnoth/wesnoth/commit/d9111620bd6718de10a39844091ee8ca24551e6c
20160901 01:01:32< irker510> wesnoth: Celtic Minstrel wesnoth:master 3ddb497b13d8 / data/eras.lua: Fix typo https://github.com/wesnoth/wesnoth/commit/3ddb497b13d8ef9c799273ecd9fdf705eaee4455
20160901 01:01:34< irker510> wesnoth: Celtic Minstrel wesnoth:master 8b839e88ad80 / data/gui/window/mp_create_game/ (small_resolution.cfg tiny_resolution.cfg): Fix MP Create on smaller resolutions https://github.com/wesnoth/wesnoth/commit/8b839e88ad804315843a91ff539e54b4142a934a
20160901 01:01:36< irker510> wesnoth: Celtic Minstrel wesnoth:master 3ff03cc66bef / / (3 files in 3 dirs): Fix matrix placement policy (mostly) https://github.com/wesnoth/wesnoth/commit/3ff03cc66bef5b87017110416a2a6b632b37da8b
20160901 01:01:38< irker510> wesnoth: Celtic Minstrel wesnoth:master 3f7db7f8789e / data/lua/wml-tags.lua: Give error on missing role= in [role] without [auto_recall] https://github.com/wesnoth/wesnoth/commit/3f7db7f8789e6d3b2091890edc755bea49b08009
20160901 01:01:40< irker510> wesnoth: Celtic Minstrel wesnoth:master f24ba1da508e / src/ (15 files in 9 dirs): Use reference instead of pointer for pathfinding cost calculator, since it canno https://github.com/wesnoth/wesnoth/commit/f24ba1da508e580168ed6ebfd9517219eee31692
20160901 01:01:42< irker510> wesnoth: Celtic Minstrel wesnoth:master 43bad8cc13de / src/gui/ (dialogs/preferences_dialog.cpp widgets/window.cpp): Some error message improvements https://github.com/wesnoth/wesnoth/commit/43bad8cc13dea5ab3470ce2fdc6b0b7031723049
20160901 01:01:44< irker510> wesnoth: Celtic Minstrel wesnoth:master 742b4b0be3c1 / data/lua/wml/objectives.lua src/tstring.cpp: Fix an issue with translatable strings that included a possible plural https://github.com/wesnoth/wesnoth/commit/742b4b0be3c182569d1633bbd9b54a691a9b6a24
20160901 01:02:12< tad_> When I started working here I'd seen the comments about Steam and thought "cool" then I saw how bad the UI was and throught "boy, they broke it and are nowhere near where they need to be"
20160901 01:02:34< celmin> When you say "how bad the UI was" do you mean "relative to 1.12"?
20160901 01:02:34< tad_> But things have gotted a **LOT** better in the past months 
20160901 01:02:38< celmin> Or just overall?
20160901 01:02:50-!- travis-ci [~travis-ci@ec2-54-80-161-132.compute-1.amazonaws.com] has joined #wesnoth-dev
20160901 01:02:51< travis-ci> wesnoth/wesnoth#10644 (master - 991e72a : ln-zookeeper): The build has errored.
20160901 01:02:52< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/156670998
20160901 01:02:52-!- travis-ci [~travis-ci@ec2-54-80-161-132.compute-1.amazonaws.com] has left #wesnoth-dev []
20160901 01:03:06< celmin> Why is the build erroring anyway? MP tests still stalling?
20160901 01:03:42< Aginor> tad_: I fully agree with that statement (about CI complementing TDD nicely)
20160901 01:03:47< tad_> Yeah .. when I started it was unreadable panels, missing events .. real hard to use unless you were careful. But, as I said, it's vastly improved
20160901 01:05:32< tad_> As to doing 1.12 on Steam and doing a Beta channel .. my only suggestion would be, if you do that, do NOT commit to a schedule for the Beta channel. Push it up when there's a consensus it's OK and not just because it's Friday
20160901 01:06:06< Aginor> push whenever there's a new dev release I'd say
20160901 01:06:53< tad_> We could do more of those. 1.13.5+dev is already way better than 1.13.5 and we have a number of milestones for 1.13.6 still to go
20160901 01:08:39< vultraz> Aginor, fabi, tad_: we are not doing 1.12 on steam, nor are we pushing back the UI work until after a Steam release. That's final.
20160901 01:09:29< vultraz> fabi: and frankly, your assertion that it's "not good" to make the UI as nice as possible for the Steam release is ridiculous.
20160901 01:09:29< fabi> vultraz: Does the steam green light expire some day?
20160901 01:09:35< vultraz> NO
20160901 01:09:37< vultraz> No*
20160901 01:09:39< vultraz> sorry, caps
20160901 01:10:10< fabi> I would focus on complete different stuff and not risk making everything very unstable.
20160901 01:11:42< vultraz> We have around 6 months 
20160901 01:11:48< vultraz> That's sufficient time.
20160901 01:11:57< fabi> Then you should already be frozzen.
20160901 01:11:57< vultraz> If you want to help make stuff stable, do so
20160901 01:12:06< vultraz> Don't sit there preaching. 
20160901 01:12:19< tad_> My only comment about the Green light is that it's been announced on the forums. So I expect in a few months we'll start seeing "When will you make an end?" comments there.
20160901 01:12:52< vultraz> tad_: and we've also announced a tentative release date of early 2017
20160901 01:13:00< celmin> 6 months in advance seems far to early for a feature freeze.
20160901 01:13:03< vultraz> if people think that it'll happen earlier it's not our problem.
20160901 01:13:05< celmin> 1 month would be more reasonable.
20160901 01:13:30< fabi> hmmm
20160901 01:13:32< fabi> well
20160901 01:13:36< fabi> I tried.
20160901 01:13:37< celmin> vultraz: So, how's the matrix / t_string fix for you?
20160901 01:13:44< vultraz> celmin: building now
20160901 01:13:51< tad_> vultraz: agreement. The issue is setting expectations and meeting them. If it was set for early 2017 then we're find until March or so ...
20160901 01:14:18< vultraz> tad_: we'll make it
20160901 01:14:27< vultraz> why do you think we're all working so hard :)
20160901 01:14:35< vultraz> (you included)
20160901 01:14:35< celmin> I guess we should look at feature-freeze in January or February, then.
20160901 01:14:52< celmin> And put out a release candidate not too long after.
20160901 01:15:22< tad_> vultraz: two months ago I'd have said "no way" but you and celmin have been really making GREAT progress on the UI. I'm impressed.
20160901 01:15:55< tad_> I'd propose a goal date of Jan 15th for feature freeze.
20160901 01:16:15< vultraz> something in the neighborhood
20160901 01:16:19< tad_> I won't be nearly through all the campaigns by then, but I should have the noob-level stuff done.
20160901 01:17:58< vultraz> ya know, I last year I did some significant cleanup of a bunch of campaigns.
20160901 01:18:06< vultraz> I'm surprised how much there's still be be done.
20160901 01:18:18< vultraz> Then again, I focused mostly on code.
20160901 01:18:40< vultraz> My major undertaking was the crapfest that is Northern Rebirth, and I never made it past halfway through :|
20160901 01:21:11< tad_> I find myself looking at stuff and wondering if I really need to fix it. Take DM .. I just spotted that Leollyn's death is NOT always a defeat-condition and am using type-tree (TY Celmin) to fix it. But, if it's been like that forever and nobody's spotted it is it really a problem?
20160901 01:21:14< fabi> vultraz: You did not seem to have much interrest in integrating my work into wesnoth either.
20160901 01:21:27< vultraz> What work?
20160901 01:21:32< vultraz> You haven't showed us anything :|
20160901 01:21:32< fabi> WSL
20160901 01:21:49< tad_> fabi: what is "wsl"?
20160901 01:22:00< fabi> tad_: The WML successor I am working on.
20160901 01:22:01< vultraz> We cannot be expected to take something and just "plug it in"
20160901 01:22:09< vultraz> If you want it in, you'll have to help.
20160901 01:22:48< tad_> Oh .. well that would be nice .. I have a LOT of ideas but have been avoiding them because I don't want to step away from cleaning up the campaigns as they are now.
20160901 01:23:23< fabi> tad_: I already asked you in private. But you seem to not look very often :-)
20160901 01:23:41< fabi> tad_: I am glad for every feedback  I can get.
20160901 01:24:10< tad_> I do tend to not notice the little button appearing when someone send a private on irc.
20160901 01:24:15< celmin> I'd say nobody spotting something does not by itself indicate that it's not a problem.
20160901 01:24:37< celmin> And I'm not interested in WSL at all, to be honest.
20160901 01:26:27< tad_> I've not looked at fabi's work and don't want to be side-tracked. But I'd love to see something more Lua-ish for the bulk of the functional code. To my mind, the big problem is doing [event] in a C-ish / Lua-ish kinda way but without all that setup and tear-down work we have to do now.
20160901 01:27:36< fabi> That is exactly what I am aiming at.
20160901 01:28:42< tad_> Did you do a BNF for it? I suppose can take a couple hours and review the syntax.
20160901 01:29:08< fabi> Well, I just wrote one.
20160901 01:29:14< fabi> It is not perfect yet.
20160901 01:29:25< vultraz> celmin: confirmed better matrix placement and fix for the tstring thing
20160901 01:29:27< fabi> But I can already translate AOI.
20160901 01:29:30< tad_> Did you use a tool like lex/yacc or something?
20160901 01:29:44< vultraz> celmin: these should really be in a linked group though
20160901 01:30:45< celmin> vultraz: I know.
20160901 01:30:46< fabi> tad_: No. I just use lua.
20160901 01:31:00< celmin> I commented that out because the placement didn't work when they were in a linked group.
20160901 01:31:24-!- EliDupree [~quassel@idupree.com] has quit [Ping timeout: 258 seconds]
20160901 01:31:25< fabi> tad_: The language I translate WML into is Moonscript.
20160901 01:31:43< fabi> tad_: Moonscript is to Lua what coffeescript is to javascript.
20160901 01:32:07< fabi> tad_: Thus a umc designer can also use lua directly.
20160901 01:32:17< celmin> Rather than WSL I'd prefer just adding code= to [event] and similar places.
20160901 01:32:35< celmin> So that you can write the entire event in Lua without using any ActionWML.
20160901 01:33:10< celmin> There'd also need to be a way to pass a Lua function to wesnoth.add_event (I think that's what it's called).
20160901 01:33:27-!- EliDupree [~quassel@idupree.com] has joined #wesnoth-dev
20160901 01:34:05< vultraz> celmin: ok, well it's good progress
20160901 01:34:18< tad_> celmin: The issue is, if we're going to do a language on top of WML (which is what I'd do) or on top of Lua .. it needs to be all-encompasing and avoid the bumps in the road going straight Lua has now. AND it needs to be a single language. The UMC people won't 'buy' having a 'third' language but we can sell an 'change and never switch back' language.
20160901 01:34:45< fabi> ^
20160901 01:34:51< vultraz> celmin: once it's done i'll probably add a custom toggle panel definition 
20160901 01:35:15< celmin> I think it would be good to drop ActionWML.
20160901 01:35:20< celmin> I don't think it's good to drop WML.
20160901 01:35:33< celmin> With a few exceptions, WML does its job very well.
20160901 01:35:42< tad_> Which is why my concept was to 'compile to WML'
20160901 01:35:45< celmin> (The main exceptions I can think of are unit animations and terrain graphics.)
20160901 01:37:55< tad_> WML works but it's a pain to use. 'Too much sugar for a dime.' having to do all that [] and [/] stuff. BUT .. and this is where Fabi's idea might hit a hiccup .. it has GOT to be able to seamlessly work with what we have now. So it needs to be able to mix a unit.cfg and a _main.cfg with a moonlight scenario ... 
20160901 01:38:38< celmin> I think Fabi's idea is basically "take the WML and compile it to this new thing".
20160901 01:39:03< tad_> Dunno. Haven't looked.
20160901 01:39:13< celmin> But I honestly don't see how the new thing is significantly better than WML.
20160901 01:39:25< celmin> I don't really mind the [] and [/] stuff that much.
20160901 01:39:33< celmin> Maybe because I've done some web stuff.
20160901 01:39:53< celmin> (Though the closing tags are often optional there, but they're never optional in WML.)
20160901 01:40:30< tad_> Well, HTML was designed as a markup language. WML is a markup-ish language but doesn't mark up anything.
20160901 01:40:30< celmin> We're at 65601 commits, huh.
20160901 01:40:46< celmin> Already past the point where it'll fit in a 2-byte integer.
20160901 01:41:20< celmin> WML is a data definition language. I think its basis is more in conf-files than XML.
20160901 01:41:37< celmin> (That could explain the use of square brackets, for example)
20160901 01:41:55< celmin> This is why it's not great for code (eg ActionWML).
20160901 01:42:33< celmin> BTW, you said UMC people won't buy having a third language, but there's already three - WML, WFL, Lua.
20160901 01:43:28< tad_> "We have 99 problems, so I wrote this cool front-end. Now we have 100 problems."
20160901 01:43:35< celmin> Haha
20160901 01:43:53< celmin> (Why front-end?)
20160901 01:44:04< tad_> a <whatever>
20160901 01:44:55< tad_> It's a mantra you see a lot in the Javascript area .. everyone does a new 'framework' and none of them do what you want but all are 'close-enough'
20160901 01:48:32-!- travis-ci [~travis-ci@ec2-54-205-13-9.compute-1.amazonaws.com] has joined #wesnoth-dev
20160901 01:48:33< travis-ci> wesnoth/wesnoth#10645 (master - 0249780 : Charles Dang): The build has errored.
20160901 01:48:33< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/156677590
20160901 01:48:33-!- travis-ci [~travis-ci@ec2-54-205-13-9.compute-1.amazonaws.com] has left #wesnoth-dev []
20160901 01:49:33< celmin> XCode uses <#…#> to indicate placeholders in autocompletion...
20160901 01:50:11< vultraz> ah, find_widget failed
20160901 01:50:13< tad_> GTG. Wife is out and needs me to come fetch her.
20160901 01:50:16< vultraz> must be that game_name thing
20160901 01:50:16 * tad_ waves
20160901 01:50:20-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has left #wesnoth-dev []
20160901 01:50:33< celmin> vultraz: Hm?
20160901 01:51:03< vultraz> celmin: travis hasn't reached your commit where you fixed mp create 
20160901 01:51:09< celmin> Ah.
20160901 01:51:24< celmin> Right, Travis tests the dialogs at all resolutions, I think.
20160901 01:51:31< celmin> Something like that.
20160901 01:51:45< celmin> All defined resolutions, I mean.
20160901 01:52:09< celmin> Wonder if that includes having a separate instance for any widget whose definition contains multiple resolutions...
20160901 01:53:33< vultraz> btw, any comments on my update to the game display in the lobby?
20160901 01:54:09< vultraz> I moved the era label to a line and explicitly labeled it "Era"... added a list of active mods... moved the map size to the right using the map icon gui1 had and made it small
20160901 01:54:47< celmin> I don't think I've even seen the game display in the lobby.
20160901 01:54:57< celmin> Since usually there are none when I connect.
20160901 01:55:07< vultraz> thou must create one :P
20160901 01:55:12< celmin> Exactly.
20160901 01:55:27< celmin> Which means I need two instances.
20160901 01:57:15< vultraz> i shall get you a screenshot
20160901 01:58:40< vultraz> celmin: https://drive.google.com/file/d/0B-mR9s8FduLLLWNwQmFjd0pPY1E/view?usp=sharing
20160901 01:59:52< vultraz> it's rather closer to the old gui1 display, but displays the info in a more compact fashion
20160901 02:00:58< shadowm> Reminder that you people should be considering nominations for the Wesnoth Inc board.
20160901 02:02:19< vultraz> I assume you are also pondering your consideration?
20160901 02:02:27< celmin> That layout makes me a little uneasy.
20160901 02:02:34< celmin> How does it look at 800x600?
20160901 02:03:23< vultraz> bLAH
20160901 02:03:33< vultraz> need to update low res versions of the lobby :|
20160901 02:03:35< vultraz> :| : | :| :| :|
20160901 02:04:20< celmin> There's only one low resolution version for the lobby.
20160901 02:05:05< shadowm> vultraz: Yeah, I'm going to vote for me of course.
20160901 02:05:05< vultraz> blah
20160901 02:05:09< vultraz> window scrollbars
20160901 02:05:23< shadowm> I don't think I'm allowed to break character here.
20160901 02:05:33< vultraz> list of mods can get too long
20160901 02:06:06< vultraz> perhaps i should indeed move the icons back to the right of row 2...
20160901 02:06:11< vultraz> and the mods list can slider under it..
20160901 02:06:51< celmin> Keep in mind that you can make the game info layout be different at the lower resolution.
20160901 02:07:08< celmin> So that it takes more height rather than width.
20160901 02:07:24< celmin> (For example)
20160901 02:09:33< vultraz> hm
20160901 02:09:44< vultraz> this combobox causes a window scrollbar even without a game
20160901 02:12:31< irker510> wesnoth: Charles Dang wesnoth:master f3e63ebc9637 / data/gui/window/lobby_main.cfg: MP Lobby: added Replay Options combobox to small resolution layout https://github.com/wesnoth/wesnoth/commit/f3e63ebc96376db8ae800b9cd9015fe5302b831f
20160901 02:12:53< vultraz> maybe should have macroized that
20160901 02:18:51< mattsc> celmin: just did a clean build after your commits and libpcre is missing again (or still) for me
20160901 02:19:37< celmin> mattsc: Is it listed in the Frameworks group at the bottom?
20160901 02:20:29< mattsc> celmin: no
20160901 02:20:40< celmin> …I'm sure I added it though.
20160901 02:21:01< mattsc> There are 34 things total to be copied, starting with Growl, then lots of dylibs, then SDL2
20160901 02:21:17< celmin> I'm talking about the Frameworks group in the left sidebar.
20160901 02:21:34< celmin> In the files tab.
20160901 02:21:39< mattsc> Ah, it’s listed for wesnothd, but not for Wesnoth
20160901 02:21:57< celmin> Are you looking at the files tab, or the target phases?
20160901 02:22:21< mattsc> No, this is under Build Phases, copy frameworks
20160901 02:22:27< mattsc> yes, it is listed in the files tab
20160901 02:22:34< celmin> Okay, feel free to add that and commit it, then.
20160901 02:22:42< celmin> And any other XCode changes you wanted to commit.
20160901 02:22:58< celmin> There was something else, right?
20160901 02:23:21< celmin> If possible, add it at its alphabetical location. :)
20160901 02:28:14< mattsc> not possible
20160901 02:28:23< mattsc> and eras.cfg is missing again …
20160901 02:28:29< celmin> I thought you could drag-and-drop it wherever...
20160901 02:28:31< mattsc> probably eras.lua as well
20160901 02:28:58< celmin> I definitely added eras.* to the project, but perhaps it didn't add them to the Copy Files stage, or perhaps it added them to the wrong Copy Files stage.
20160901 02:31:06< mattsc> Hmm, live and learn.
20160901 02:31:16< mattsc> I didn’t know the drag/drop thing worked like that ...
20160901 02:32:09< mattsc> I did see the lib and the two era files in your commit, but they aren’t there in the data directory.
20160901 02:32:41< mattsc> And that’s after cleaning the build just to be sure it would do everything.
20160901 02:33:06< celmin> Check the Copy Files build phases, then.
20160901 02:33:16< celmin> It might've been added to the wrong phase, or not at all.
20160901 02:35:42< mattsc> Right, it’s not there.
20160901 02:35:47< mattsc> Well, wasn’t.  It is now.
20160901 02:36:16< mattsc> It appears to be working for me now, just testing a few things.
20160901 02:37:04< celmin> No hurry. I still haven't opened the XCode project after that rebase.
20160901 02:38:15< irker510> wesnoth: mattsc wesnoth:master c296a679908c / projectfiles/Xcode/Wesnoth.xcodeproj/project.pbxproj: Update Xcode project https://github.com/wesnoth/wesnoth/commit/c296a679908c5b107715010dd40997e8eeb43c7b
20160901 02:38:34-!- travis-ci [~travis-ci@ec2-54-80-161-132.compute-1.amazonaws.com] has joined #wesnoth-dev
20160901 02:38:34< mattsc> celmin: ^ that works for me
20160901 02:38:35< travis-ci> wesnoth/wesnoth#10647 (master - 019f9ab : Charles Dang): The build has errored.
20160901 02:38:35< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/156681986
20160901 02:38:35-!- travis-ci [~travis-ci@ec2-54-80-161-132.compute-1.amazonaws.com] has left #wesnoth-dev []
20160901 02:41:12< celmin> mattsc: So, eras.* wasn't added to an incorrect copy files stage?
20160901 02:41:53-!- Bonobo [~Bonobo@2001:44b8:254:3200:ddf8:b9cf:5c99:641b] has joined #wesnoth-dev
20160901 02:45:13-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has quit [Quit: Connection closed for inactivity]
20160901 02:47:24< mattsc> celmin: not in any of the places where I looked
20160901 02:48:32< mattsc> Oh, of course now I see it.  Blargh.
20160901 02:48:39< celmin> Heh...
20160901 02:48:48< mattsc> It’s added to ‘Copy Bundle Resources’
20160901 02:48:57< mattsc> I thought I’d looked there.
20160901 02:48:57< celmin> That's what I suspected might've happened.
20160901 02:49:14< mattsc> Should I remove it from there?
20160901 02:49:57< mattsc> Btw, in the meny bar, it’s called Xcode on my computer, not XCode.  I don’t know which one is right, or whether they changed it or something.
20160901 02:50:09< celmin> Mine is Xcode too.
20160901 02:50:15< celmin> I dunno why I always say XCode.
20160901 02:50:40< mattsc> It makes more sense as XCode to me, but …
20160901 02:50:42< mattsc> anyway
20160901 02:56:06< celmin> s/say/type/
20160901 02:56:32< mattsc> hehe, I like how you pronounce the capital C
20160901 02:56:48< celmin> ;p
20160901 02:58:08< mattsc> Btw, I am currently doing another clean build after removing the eras files from ‘copy bundle resources’.
20160901 02:58:20< mattsc> I’ll commit the change after I make sure everything’s working
20160901 02:58:31< celmin> Okay.
20160901 02:58:54< celmin> I started working on a different project, so there's no hurry,
20160901 02:59:37< mattsc> Alright.
20160901 03:14:34< irker510> wesnoth: mattsc wesnoth:master dedb51d2f269 / projectfiles/Xcode/Wesnoth.xcodeproj/project.pbxproj: Another Xcode project update https://github.com/wesnoth/wesnoth/commit/dedb51d2f269bcedd7d1078ebeeb8d28e9e70e92
20160901 03:22:45-!- travis-ci [~travis-ci@ec2-54-158-100-40.compute-1.amazonaws.com] has joined #wesnoth-dev
20160901 03:22:46< travis-ci> wesnoth/wesnoth#10648 (master - 742b4b0 : Celtic Minstrel): The build has errored.
20160901 03:22:46< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/156685452
20160901 03:22:46-!- travis-ci [~travis-ci@ec2-54-158-100-40.compute-1.amazonaws.com] has left #wesnoth-dev []
20160901 03:25:13< celmin> Still erroring… :|
20160901 03:29:33< vultraz> wait for MY lobby fix commit :P
20160901 03:41:26-!- hk238 [~kvirc@unaffiliated/hk238] has joined #wesnoth-dev
20160901 03:41:38-!- hk238 [~kvirc@unaffiliated/hk238] has quit [Max SendQ exceeded]
20160901 03:57:03-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection]
20160901 03:57:09-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev
20160901 03:57:41-!- travis-ci [~travis-ci@ec2-54-205-13-9.compute-1.amazonaws.com] has joined #wesnoth-dev
20160901 03:57:42< travis-ci> wesnoth/wesnoth#10649 (master - f3e63eb : Charles Dang): The build has errored.
20160901 03:57:42< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/156694460
20160901 03:57:42-!- travis-ci [~travis-ci@ec2-54-205-13-9.compute-1.amazonaws.com] has left #wesnoth-dev []
20160901 03:58:30-!- hk238 [~kvirc@t224.ip7.netikka.fi] has joined #wesnoth-dev
20160901 03:59:37< celmin> vultraz: Looks like it's still erroring though.
20160901 04:02:03-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160901 04:03:32-!- JyrkiVesterinen [~JyrkiVest@87-100-133-184.bb.dnainternet.fi] has joined #wesnoth-dev
20160901 04:06:01-!- hk238 [~kvirc@t224.ip7.netikka.fi] has quit [Quit: http://www.kvirc.net/ 4.9.1 Aria]
20160901 04:06:30< vultraz> ah, but more pass now
20160901 04:06:33-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 240 seconds]
20160901 04:06:54-!- RatArmy [~RatArmy@240f:b3:88e3:1:221:85ff:fec0:94bc] has joined #wesnoth-dev
20160901 04:08:40< vultraz> looks like host never hosts :|
20160901 04:15:23-!- EliDupree [~quassel@idupree.com] has quit [Remote host closed the connection]
20160901 04:15:33-!- EliDupree [~quassel@idupree.com] has joined #wesnoth-dev
20160901 04:18:00< irker510> wesnoth: Charles Dang wesnoth:master db2381806dc8 / data/gui/window/lobby_main.cfg: MP Lobby: enable game list horizontal scrollbar for now https://github.com/wesnoth/wesnoth/commit/db2381806dc88181e5ced842a63d9501e7fc94ac
20160901 04:18:45< vultraz> really wish we had something like a 'fadeout' mode where we could set certain area to crop and display their contents as fading out :/
20160901 04:18:54< celmin> BTW, did you do anything about re-enabling esc yet?
20160901 04:19:14< vultraz> no
20160901 04:19:19< celmin> Or I could do it.
20160901 04:19:36< celmin> Not sure what you mean by this fadeout thing.
20160901 04:20:07< vultraz> doesn't even really have to be a fadeout
20160901 04:20:20< vultraz> but imagine something where we could say...
20160901 04:20:40< vultraz> 'ok, if this label is going to cause the listbox to grow too much, try to truncate it with ellipses until it fits"
20160901 04:21:16-!- Nobun-irssi [~user@host158-62-dynamic.19-79-r.retail.telecomitalia.it] has joined #wesnoth-dev
20160901 04:21:20< celmin> The vertical_list placement policy probably needs a bit of work.
20160901 04:21:20< vultraz> if we had anura's capabilities we could have the text fade out to transparency instead, but you get the idea.
20160901 04:22:19< Nobun-irssi> Hi celmin. Elvish Hunter told me about the new feature of supporting plural forms on translations. I'd like to talk about it to you when you have time
20160901 04:22:21< celmin> I think what you describe may also be possible right now by using the layout invalidation blocker, setting the label text, and calling calculate_best_size until you get something that fits.
20160901 04:22:48-!- Nobun [~nobun@host158-62-dynamic.19-79-r.retail.telecomitalia.it] has joined #wesnoth-dev
20160901 04:22:55< celmin> (I think it'd work at the dialog point, not sure if it can be done right in the generator.)
20160901 04:23:08-!- Nobun-irssi [~user@host158-62-dynamic.19-79-r.retail.telecomitalia.it] has quit [Client Quit]
20160901 04:23:29< celmin> Hi Nobun.
20160901 04:23:53< Nobun> Hi celmin :)
20160901 04:24:07< celmin> Currently this feature is only supported in Lua.
20160901 04:24:16< celmin> I'm not sure yet how to support it in WML.
20160901 04:24:58< Nobun> yes I read. I have an idea how to add it on WML, but I'd like to understand some things, before
20160901 04:25:06< celmin> The regex for Lua would need to match _("singular string", "plural string"
20160901 04:25:30< Nobun> this is not the only issue, unluckly
20160901 04:25:59< Nobun> however I'd like  to ask you if adding plural forms on _ is the only solution
20160901 04:26:37< Nobun> it's not only a question related to wmlxgettext parsing, but also for end-user usage
20160901 04:26:39< celmin> Well, in Lua, _ is a variable, so it would be possible to conventionally use a different variable name for plural forms.
20160901 04:27:00< celmin> (Though enforcing it might be hard.)
20160901 04:27:12< celmin> (But I don't think conventions need to be enforced.)
20160901 04:27:44< Nobun> well... the _ variable was invented for two reason, on my point of view
20160901 04:27:47< celmin> For me, using _ is the most logical thing to do in the Lua code. In WML the situation is completely different though.
20160901 04:27:51< celmin> Okay, go ahead.
20160901 04:28:21< Nobun> the first reason is... having the same marker on Lua and WML for 'this is a translatable string'
20160901 04:29:01< Nobun> I think that using _ as a real function (even if it is actually a function I know) can introduce some confusion on UMC authors
20160901 04:29:58< Nobun> the second reason was for wmlxgettext parsing, I thing... but this second problem can be solved thinking together a good regex to manage the new syntax
20160901 04:30:11< Nobun> s/I thinkg/I think
20160901 04:30:49< Nobun> what do you think about, celmin ?
20160901 04:31:47< Nobun> (I have to admit I could be too concerned... after all I never learnt how to actually use lua in wesnoth addons)
20160901 04:32:21< celmin> My reasoning for using _() is that the convention for Lua is to "declare" a textdomain with "local _ = wesnoth.textdomain 'blah'", and that produces a callable userdata (basically a function) which looks up strings in that domain. So, to me, it makes sense to use the same function for singular and plural strings. In C, of course, there are different functions; but C doesn't support optional parameters, so that's absolutely required.
20160901 04:32:55< celmin> If you preferred to use _n() for plurals, as in C, you could easily do "local _n = _" after declaring your textdomain.
20160901 04:33:20< celmin> But to me, that seems a bit pointless.
20160901 04:33:21< Nobun> hmm understood
20160901 04:34:02< Nobun> yeah I got your point.
20160901 04:34:44-!- travis-ci [~travis-ci@ec2-54-158-100-40.compute-1.amazonaws.com] has joined #wesnoth-dev
20160901 04:34:46< travis-ci> wesnoth/wesnoth#10650 (master - c296a67 : mattsc): The build has errored.
20160901 04:34:46< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/156697591
20160901 04:34:46-!- travis-ci [~travis-ci@ec2-54-158-100-40.compute-1.amazonaws.com] has left #wesnoth-dev []
20160901 04:35:16< Nobun> I have also a question about it. How plural forms actually work on wesnoth lua code?
20160901 04:35:43< Nobun> I mean... If I remember well the function requires three parameter (singular form, plural form, int counter)
20160901 04:35:46< celmin> Did E_H point you to the bug report about plurals?
20160901 04:35:51< celticminstrel> Yes, that's correct.
20160901 04:36:24< Nobun> yes, celticminstrel... he did (even if I didn't save the link)
20160901 04:36:24< celticminstrel> In the C++ code, there's a macro _n("singular", "plural", count)
20160901 04:37:03< Nobun> ok but... my question is... how can you sure that plural form will actually apply when you use it on WML / Lua code?
20160901 04:37:14< Nobun> the problem is how 'count' value is collected
20160901 04:37:44< celmin> The count value doesn't need to be collected by wmlxgettext.
20160901 04:37:54< Nobun> obliouvsly
20160901 04:38:05< celmin> Then maybe I misunderstood your question...
20160901 04:38:26-!- Lohengramm [sid1929@gateway/web/irccloud.com/x-fndpcoyxtcbslavt] has quit [Read error: Connection reset by peer]
20160901 04:39:27< Nobun> I mean... before working on wmlxgettext adding the support to plural form, perhaps a test with fake lua code and fake pot file (edinting pot manually with the two-three sentences containing plural on test.lua) actually can switch from singular to plural form
20160901 04:39:59< Nobun> I mean... a test to check if the plural form actually works on game
20160901 04:40:02< celmin> Ah.
20160901 04:40:26< celmin> I've tested that already. It's only used in one place at the moment - data/lua/wml/objectives.lua, line 71 I think.
20160901 04:40:59-!- Lohengramm [sid1929@gateway/web/irccloud.com/x-uipslyqolroypnyn] has joined #wesnoth-dev
20160901 04:41:06< Nobun> After all supporting this on wmlxgettext will require a lot of work... adding regex rule is one (and probably the most important) of the steps required, but the code require a deeper manipulation
20160901 04:41:09< celmin> When I set a turn limit in game and skip turns to the end, the objectives are correctly updated to say "this turn left" instead of "n turns left".
20160901 04:41:24< celmin> So, the plural form does work in the game.
20160901 04:42:58< Nobun> Ok. So we can go ahead.
20160901 04:42:58< celmin> (And I tested that with an edited po-file for en_GB, which made that pair of strings a singular/plural form.)
20160901 04:42:58< celmin> (Originally they were separate strings.)
20160901 04:42:58< celmin> (Though somehow it even works on en_US.)
20160901 04:42:58< celmin> BTW, you said you had an idea for plural forms in WML?
20160901 04:43:04< Nobun> celmin: plural forms, can be translated in all languages :)
20160901 04:43:18< Nobun> yes... I have an idea about it, but not sure it can work in actual game
20160901 04:43:24< celmin> I see.
20160901 04:43:49< Nobun> I was thinking mainly a method wich can be parsed to wmlxgettext and on a same time could be not too hard for end-users
20160901 04:44:04< Nobun> I thinking about a 'plural table'
20160901 04:44:25< Nobun> something like
20160901 04:44:52< Nobun> Wait a moment... I will use dpaste
20160901 04:47:15< Nobun> celmin... I think we could add it something like this:  http://dpaste.com/29Q537P
20160901 04:48:39-!- RatArmy [~RatArmy@240f:b3:88e3:1:221:85ff:fec0:94bc] has quit [Quit: Leaving]
20160901 04:48:55< Nobun> where the first sentence could omit the _
20160901 04:51:42< Nobun> this could be possible since wmlxgettext could change state when the NewTag (we must think about a better name, obliouvsly) is reached... more or less like it currently happen on lua tag (it does not change state, but the rule for << >> to start and end lua block will work only if it is inside a lua tag)
20160901 04:53:00-!- Kwandulin [~Miranda@p200300760F42415E5966D161A0AF96BE.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160901 04:53:06< Nobun> I also thinking about introducing a versioning on wmlxgettext (a --version option) and giving version number like:  2016.09.01.py3
20160901 04:53:08< celmin> Hmm. I dunno.
20160901 04:54:08-!- JyrkiVesterinen [~JyrkiVest@87-100-133-184.bb.dnainternet.fi] has quit [Quit: Going somewhere]
20160901 04:54:32< Nobun> the main concern on this purposal is not about wmlxgettext collect... but about if it works
20160901 04:55:23< Nobun> on wmlxgettext side this solution is optimal, becouse on a new tag (without nested tags) I can add a dedicated state
20160901 04:57:14< Nobun> however, celmin: I have another question about the plural form lua support:
20160901 04:58:07< Nobun> can we force somehow a defined syntax for sentences inside _ ( )  ?
20160901 04:58:27< Nobun> I explain my question
20160901 04:58:53< Nobun> currently wmlxgettext support all kind of sentences with _ 
20160901 04:59:25< Nobun> _ ' sentence 1'   _ "sentence 2"  _ [===[ sentence 3 with arbitrary number of equals ]===]
20160901 05:00:33< Nobun> allowing all those forms (and multi-line version) will could be an headache when inside the _ (....
20160901 05:14:08-!- travis-ci [~travis-ci@ec2-54-205-13-9.compute-1.amazonaws.com] has joined #wesnoth-dev
20160901 05:14:09< travis-ci> wesnoth/wesnoth#10651 (master - dedb51d : mattsc): The build has errored.
20160901 05:14:09< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/156701452
20160901 05:14:09-!- travis-ci [~travis-ci@ec2-54-205-13-9.compute-1.amazonaws.com] has left #wesnoth-dev []
20160901 05:15:11< Nobun> celmin, celticminstrel  ?
20160901 05:15:32-!- ancestral [~ancestral@75-168-189-115.mpls.qwest.net] has joined #wesnoth-dev
20160901 05:16:04< celmin> Hi ancestral 
20160901 05:16:19< ancestral> Hi there!
20160901 05:16:46< celmin> ancestral: Do you have any objections to me setting -fsanitize=address for debug builds in the XCode project?
20160901 05:17:17< celmin> Nobun: Yeah, you're right. I think we don't need to worry about bracketed sentences, at least.
20160901 05:17:19< ancestral> I have no clue what that is
20160901 05:17:23< celmin> It's probably good to support single quotes.
20160901 05:17:25< ancestral> I trust your judgment, however, so go ahead
20160901 05:17:56< celmin> ancestral: http://clang.llvm.org/docs/AddressSanitizer.html#usage
20160901 05:18:04< Nobun> celmin: and how about multi-line strings on _ (   ?
20160901 05:18:27< celmin> You mean "" type strings that go across multiple lines?
20160901 05:18:33< Nobun> yes
20160901 05:18:46< celmin> I'm not sure, honestly.
20160901 05:19:34< Nobun> usually plural forms should be used in short sentences, but we can't know what a UMC author could need
20160901 05:19:51< celmin> It would of course be great to support all possible types of string, but I don't think it's high priority to do so.
20160901 05:20:14< celmin> It should at least probably support _("singular string", <new line> "plural string"
20160901 05:20:26< celmin> ie, the two strings on separate lines
20160901 05:20:29< Nobun> well this is not a problem. celmin 
20160901 05:20:52< Nobun> however if I can think a solution that will works with all forms and multi-line I will implement it
20160901 05:20:54< celmin> ancestral: Any opinions after looking at that?
20160901 05:21:13< celmin> Nobun: So about WML, the solution I was thinking of is similar to yours, but wrapped in a macro.
20160901 05:21:42< celmin> However, I'm uncertain about that sort of implementation, because it's less generic. I'm not sure how it would work outside of ActionWML.
20160901 05:21:43< Nobun> celmin: a macro could be harder to parse for wmlxgettext
20160901 05:21:48< celmin> Really?
20160901 05:21:53< Nobun> yes
20160901 05:21:56< celmin> Why?
20160901 05:22:20< Nobun> becouse macro are considered 'normal code' on wmlxgettext (like it happened on perl wmlxgettext)
20160901 05:22:40< celmin> Huh?
20160901 05:22:44< ancestral> celmin: So the benefit is… faster better stronger builds?
20160901 05:22:51< Nobun> this means, on wmlxgettext, that wmlxgettext thinks it he is parsing the tag where macro is found
20160901 05:23:03< celmin> ancestral: Not faster, but potentially easier to catch bad memory accesses.
20160901 05:24:37< Nobun> this will introduce a situation where the sentence with plural form is commented with the informations about tag wich contains the macro... wich is something that could be not good
20160901 05:25:16< celmin> I'm not sure I understand what you mean.
20160901 05:25:19< ancestral> Cool, go for it
20160901 05:25:44< Nobun> celmin... I try to make an example
20160901 05:29:59< Nobun> celmin: this is not a good example, but it could explain better what I mean
20160901 05:30:01< Nobun> http://dpaste.com/2H1K34Q
20160901 05:30:45< Nobun> as you can see the comment remember all the details about the tag where the macro where found
20160901 05:31:28< celmin> And you consider that to be good, or bad?
20160901 05:31:45< Nobun> in this case the tag is a scenario... but it could be another tag with all its values... wich can be confusing becouse can say details that are not actually related to the sentence
20160901 05:31:57< Nobun> I consider it a bad thing
20160901 05:32:28< Nobun> it is a similar problem wich happened with perl wmlxgettext when it adds extra comment wich are wrong
20160901 05:32:51< celmin> I think it could be a good thing though, at least in some cases.
20160901 05:33:34< celmin> One problem with the tag-based WML plurals is that you can't concatenate it like you can with regular plural forms.
20160901 05:34:05< Nobun> well not completely true
20160901 05:34:27< Nobun> if we add a parameter variable we can consider the WML plural as a variable
20160901 05:34:50< Nobun> wich can be expandend in the actual sentence in singular or plural form
20160901 05:35:14< celmin> Do you mean using [set_variable] to assign the plural to a variable?
20160901 05:36:11< Nobun> I mean something like [store_unit] wich creates internally a variable (without [set_variable]) when using variable=... parameter
20160901 05:36:42< celmin> So you're suggesting something like a [plural] tag that stores the plural string in a variable?
20160901 05:37:28< Nobun> yes... if the solution is not considered too tricky
20160901 05:37:48< Nobun> after all I can understand that having a lot of variables this way could be not a good thing
20160901 05:37:49< celmin> Actually, that's probably the easiest way to support plurals in WML.
20160901 05:37:58< celmin> But it wouldn't work in [scenario].
20160901 05:38:08< celmin> It would only work in [event] or [command].
20160901 05:38:31< Nobun> yes and there is also another downside I didn't think about, and I see only now when we are talking...
20160901 05:38:35< celmin> (Well, and other similar tags - [then], [else], [do], [case], [default].)
20160901 05:39:12< Nobun> ... the other downside is.... a UMC author could want to use the [plural] tag converting it in a macro
20160901 05:39:23< Nobun> wich will break the wmlxgettext parsing
20160901 05:39:57< Nobun> so probably the WML macro could be the best solution for WML (but will be a nightmare for wmlxgettext parsing)
20160901 05:41:58< celmin> There is one problem with the tag/macro idea - how to determine the textdomain.
20160901 05:43:35< Nobun> I'm not completely able to figure out the problem you are mentioning...
20160901 05:44:35< celmin> And suddenly I thought of a way to solve that...
20160901 05:45:03< celmin> My initial concept of the [plural] tag had a textdomain key, specifying the domain explicitly.
20160901 05:45:36< celmin> And it also didn't have a normal translation mark on the singular or plural strings.
20160901 05:45:42< Nobun> yes... this could be a nice thing if you implement [plural] tag as variable
20160901 05:45:51< celmin> Your example on the other hand had no textdomain, but included a translation mark on the singular form.
20160901 05:46:10< Nobun> well... actually translation mark can be omitted
20160901 05:46:23< celmin> The translation mark means that the "current textdomain" will be captured, so maybe there's a way to make use of that...
20160901 05:47:08< celmin> If we went with a macro, I'd have the macro insert the translation marks.
20160901 05:47:30< celmin> So {PLURAL "a" "b" $count} expands to something like [plural]singular=_"a" etc
20160901 05:48:35< celmin> I'd have to expand on the API in order to implement this [plural] tag in Lua (or, I think, even in C++).
20160901 05:49:04< Nobun> before doing that we should be think about it a moment
20160901 05:49:33< celmin> Yeah, we should consider several possibilities.
20160901 05:50:16< Nobun> exactly. After all we must be sure to choose a good solution for all the needs (both for WML and for wmlxgettext)
20160901 05:50:24< celmin> One possible down-side of the plural tag/macro is that I have a hard time imagining it working outside of an event… though, I also can't think of any reason why someone would want to use it outside of an event… except maybe race names.
20160901 05:50:52< Nobun> before starting to actually work at it... coming back later will risk to vaste work on c++ / lua you did
20160901 05:51:03< celmin> Right.
20160901 05:51:27< celmin> I'm pretty confident on the Lua implementation that I already did, but I'm very uncertain about how to do it in WML.
20160901 05:52:04< celmin> One option would be to update the WML parser to understand syntax very similar to what I implemented for Lua… but outside of an event, where does the count come from?
20160901 05:52:26< Nobun> yes. I agree. the lua solution is currently good... I had some doubts about it, but I got your points about why you implemented it in that way
20160901 05:53:40< Nobun> After all, coming back to wmlxgettext support, the biggest work is not adding the regexps (wich can be a long and complex work too), but changing all the code around pywmlx code to add low-level support
20160901 05:54:51< Nobun> so when I will be able to add support for lua code already existing, the longer work will be done
20160901 05:56:08< Nobun> I have to warn that I will need a lot of time to add this support (I need to change all the logic around pot sentence internal management, in order to add a general support for plural forms)
20160901 05:56:31-!- travis-ci [~travis-ci@ec2-54-80-161-132.compute-1.amazonaws.com] has joined #wesnoth-dev
20160901 05:56:32< travis-ci> wesnoth/wesnoth#10652 (master - db23818 : Charles Dang): The build has errored.
20160901 05:56:32< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/156709518
20160901 05:56:32-!- travis-ci [~travis-ci@ec2-54-80-161-132.compute-1.amazonaws.com] has left #wesnoth-dev []
20160901 05:56:42< celmin> Ah, I can understand that, yeah.
20160901 05:57:00< celmin> The syntax for plurals BTW is:
20160901 05:57:08< celmin> msgid "singular"
20160901 05:57:15< celmin> msgid_plural "plural"
20160901 05:57:23< celmin> msgstr[0] "singular translated"
20160901 05:57:28< celmin> msgstr[1] "plural translated"
20160901 05:57:38< Nobun> thank celmin 
20160901 05:57:51< celmin> I'm not quite sure how it looks in the pot-files, but you could do a search for msgid_plural in eg wesnoth.pot.
20160901 05:58:41< Nobun> well... no problem... pot format is documented ;)
20160901 05:59:21< celmin> It's basically the same format. I'm just not sure whether you'd see msgstr "" or msgstr[0] "" or both the latter and msgstr[1] ""
20160901 06:00:16< celmin> As long as you can finish in less than 6 months, I think it'll be fine. :P
20160901 06:00:29< Nobun> if you agree I will try to work on this feature on my own... If you want to delegate this work to an official wesnoth developer is fine for me (I can be too slow for your needs), but I think the best solution is that I do this work, as long as I am the one wich probably is most confortable with the source code I wrote
20160901 06:00:30< celmin> There's no hurry right now.
20160901 06:00:45< celmin> Go ahead and work at it if you want.
20160901 06:01:22< Nobun> yeah... I cannot guarantee you anything at the moment. So you can also assign this task to another person in the meantime
20160901 06:01:45< Nobun> but I will start to work at this in any case
20160901 06:01:51< Nobun> if you agree
20160901 06:02:04< celmin> I don't mind.
20160901 06:03:03< celmin> I guess it won't be ready for 1.13.6 probably, but that's fine. It'd be great if it could be ready for 1.13.7, and I'd consider it necessary for it to be ready for 1.14.
20160901 06:03:04< Nobun> however... I will introduce a changelog for wmlxgettext (on pywmlx directory) and a version number  YYYY.MM.DD.py3
20160901 06:05:08< Nobun> this becouse the online documentation usually follows only the last wmlxgettext release, wich changed something in the command line on some versions
20160901 06:05:31< Nobun> (mainly with text color support)
20160901 06:05:57-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160901 06:06:46< Nobun> ok... I will try to inform you about the work progress while developing the support...
20160901 06:07:22< Nobun> what is the best way? A PM to you? write down here in #wesnoth-dev?
20160901 06:07:44< celmin> Either is fine.
20160901 06:08:14< Nobun> ok
20160901 06:10:05-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 244 seconds]
20160901 06:11:16-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has joined #wesnoth-dev
20160901 06:12:05< Nobun> celmin: https://www.gnu.org/software/gettext/manual/html_node/PO-Files.html  ---> go at line where it is written   " A different kind of entries is used for translations which involve plural forms."
20160901 06:13:13< Nobun> I'm not sure how to determine [0] [1] or whatever else, and if more than two version could be allowed
20160901 06:13:51< celmin> Adding the c-format flag seems like a good idea if it can be determined that the string is being passed to string.format() … though that might be prohibitively hard.
20160901 06:15:38< celmin> Or lua-format I guess.
20160901 06:15:54< Nobun> lua-format does not exist
20160901 06:16:01< Nobun> as far as I know
20160901 06:16:09< celmin> Well, I found it on the page you linked, though.
20160901 06:16:16< Nobun> ah ok
20160901 06:16:28< Nobun> however... I would avoid to use flags
20160901 06:16:35 * celmin shrugs.
20160901 06:16:46< Nobun> but my question is related to msgstr [X] id number
20160901 06:16:55< celmin> I have no idea how gettext knows how many plural forms are allowed.
20160901 06:17:11< celmin> In my testing with en_GB, it ignored all but the first two (ie [0] and [1]).
20160901 06:17:52< Nobun> what was the counter value when singular (0) and when plural (1) ?
20160901 06:18:21< celmin> The value in square brackets is one less than the counter value.
20160901 06:18:28-!- atarocch [~atarocch@natmobil.sfa.se] has joined #wesnoth-dev
20160901 06:18:34< celmin> So, a count of 1 uses msgstr[0], and a count of 2 or more uses msgstr[1].
20160901 06:19:51< Nobun> hmmm
20160901 06:20:03< Nobun> celmin can you do a test for me?
20160901 06:20:13< celmin> What kind of test?
20160901 06:20:23-!- Ravana_ [~Ravana@unaffiliated/ravana/x-2327071] has quit [Disconnected by services]
20160901 06:20:34< Nobun> can you try to test count 1 (singular) count 2 (dual) count 3 plural on a fake translation?
20160901 06:20:50< Nobun> there are some languages like greek that allows dual before plural
20160901 06:20:52-!- Ravana_ [~Ravana@unaffiliated/ravana/x-2327071] has joined #wesnoth-dev
20160901 06:20:54< celmin> I know.
20160901 06:21:13< celmin> It didn't work when I edited the en_GB po-file to have singular-plural-dual.
20160901 06:21:21< Nobun> if msgstr[1] and msgstr[2] will work on translation without changing the code, than we are sure
20160901 06:21:33< celmin> I'm not sure how gettext decides how many forms are allowed.
20160901 06:22:13< Nobun> en_GB should be the language source... right?
20160901 06:22:20< celmin> Hm?
20160901 06:22:22< Nobun> or is it the translation?
20160901 06:22:30< celmin> en_US is the source language.
20160901 06:22:35< Nobun> ok
20160901 06:22:40< celmin> For some reason there is an en_GB "translation".
20160901 06:23:20< Nobun> so when you try to use msgstr[0] msgstr[1] and msgstr[2] wich ones actually worked?
20160901 06:23:34< celmin> In en_GB, it ignored all but the first two.
20160901 06:24:17< Nobun> so having counter = 100, for example... it uses msgstr[1] and not msgstr[2] if I understood correctly
20160901 06:25:21< celmin> Yeah. But, I think it somehow depends on the target language or something.
20160901 06:27:46< celmin> Oh, I found it. It depends on something in the header entry at the top of the po-file.
20160901 06:28:29< celmin> "Plural-Forms: nplurals=2; plural=n == 1 ? 0 : 1;"
20160901 06:29:02< celmin> So that basically says, "if n is 1, use msgstr[0]; for any other n, use msgstr[1]"
20160901 06:32:04< Nobun> yes I read... but I am not sure what this nplural is... it is something that seems c code not pot file...
20160901 06:32:16< Nobun> or perhaps I read badly
20160901 06:32:33< celmin> It's sort of a limited C expression, yes.
20160901 06:32:57< Nobun> what page are you reading? (not sure if we are reading the same thing)
20160901 06:33:04< celmin> https://www.gnu.org/software/gettext/manual/html_node/Plural-forms.html
20160901 06:34:48< Nobun> ah ok... I read this page before, but I readed badly... you are right
20160901 06:34:56< Nobun> that nplus is a parameter inside the pot header
20160901 06:35:43< Nobun> wich is a parameter that currently wmlxgettext don't write... and probably it is good it doesn't write it
20160901 06:36:14< Nobun> probably this is something that a translator might want to write manually... but it seems somewhat advanced
20160901 06:37:16< celmin> From here, it seems that the expectation would be for wmlxgettext to write the pot-file as if there were only two plural forms:
20160901 06:37:17< celmin> https://www.gnu.org/software/gettext/manual/html_node/Translating-plural-forms.html#Translating-plural-forms
20160901 06:38:38< Nobun> yeah... wich is the most standard way
20160901 06:38:54-!- ancestral [~ancestral@75-168-189-115.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20160901 06:39:25-!- celmin [~celticmin@unaffiliated/celticminstrel] has quit [Quit: And lo! The minstrel departs, to spread the music to the masses!]
20160901 06:52:05-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20160901 06:52:47-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has quit [Quit: .]
20160901 07:07:15< Nobun> celticminstrel: se ci sei ancora: ho testato msginit ed in effetti lui ti crea un nuovo pot con la corretta sintassi per la plural form supportata nella lingua
20160901 07:07:28< celticminstrel> ???
20160901 07:07:46< Nobun> ho fatto questo test... ho preso un pot a caso tra quelli contenuti in wesnoth (wesnoth/en_GO.po)
20160901 07:08:09< Nobun> poi ho controllato i valori 'locale' installati nel mio sistema con locale -a
20160901 07:08:16< Nobun> ah sorry
20160901 07:08:23< Nobun> I was talking in Italian :P
20160901 07:08:32< Nobun> now I will translate... sorry for trouble
20160901 07:09:20< Nobun> I tested msgint and I can confirm that it creates a new po file with the right syntax for the plural form required by a specific language (in my case italian)
20160901 07:09:44< Nobun> then I checked locale values installed on my system with 'locale -a'
20160901 07:10:20< Nobun> then I create a test pot file with msginit starting from a wesnoth pot file (po/wesnoth/en_GB.po)
20160901 07:10:52< Nobun> with  misginit -i en_GB.po -o test.po -l it_IT.utf8
20160901 07:11:13< Nobun> (that one is the value of italian locale installed on my system)
20160901 07:11:55< Nobun> and the new created pot file (test.po) contains the right value for plural    (  "Plural-Forms: nplurals=2; plural=(n != 1);\n" )
20160901 07:12:30< Nobun> infact the italian language, like english, has singular form if value is equal to one, and plural form for all other numbers
20160901 07:13:07< Nobun> so probably wmlxgettext don't require to add a specific plural form value in header pot instructions...
20160901 07:13:15< Nobun> what do you think about it?
20160901 07:13:30< celticminstrel> Yeah, I don't think wmlxgettext needs to worry about the number of plural forms.
20160901 07:14:16< Nobun> sorry for the fact I previosly spoke in Italian... for a moment I forgot I was talking on an international irc channel :P
20160901 07:15:14< celticminstrel> I was able to recognize it as Italian, but not much more. Maybe a hint of the meaning.
20160901 07:16:07< Nobun> yeah... well my fault 
20160901 07:28:47-!- Nobun [~nobun@host158-62-dynamic.19-79-r.retail.telecomitalia.it] has quit [Remote host closed the connection]
20160901 07:34:12-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has joined #wesnoth-dev
20160901 07:40:04-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20160901 07:43:14-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 260 seconds]
20160901 07:43:14-!- wedge010 is now known as wedge009
20160901 07:50:24< wedge009> celticminstrel: Hello, if you have time, could you look at https://gna.org/bugs/?25026, please? You wrote the iterator stuff in config.Xpp which seems to be causing this bug.
20160901 07:50:54< wedge009> I'm still having a poke around, but I don't really understand what's going on.
20160901 07:51:31< celticminstrel> It's almost 4 AM, so I won't be poking at it tonight, but I'll look at the report at least.
20160901 07:51:59< wedge009> No worries, whenever you feel ready. In fact, go to bed now, 4 am is normally when I get up. You shouldn't be sleeping then. :p
20160901 07:52:15< celticminstrel> Yeah, I know, my sleep schedule is all over the place.
20160901 07:52:52< celticminstrel> Segmentation fault, huh...
20160901 07:53:31< celticminstrel> Oh wow, it's September.
20160901 07:53:53< wedge009> Oh, welcome. :p
20160901 07:53:59< wedge009> Glad you could join us.
20160901 07:54:16-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160901 07:54:38< wedge009> Trying gfgtdf's suggestion, see what happens.
20160901 07:56:05< celticminstrel> With a segmentation fault in config::const_all_children_iterator::operator*, my first instinct would be that you're dereferencing an end iterator.
20160901 07:56:49< wedge009> i_->index seems to be bad.
20160901 07:56:51< celticminstrel> Though, access after delete is certainly also a possibility.
20160901 07:57:07< wedge009> Not sure where the end iterator comes in.
20160901 07:57:10< wedge009> Anyway, go to bed.
20160901 07:57:14< wedge009> This can wait.
20160901 07:57:27< celticminstrel> i_->index is probably an implementation detail of your iterators.
20160901 07:57:48< celticminstrel> i_ should be, uhh... probably a vector iterator into the ordered_children array.
20160901 07:58:11< celticminstrel> I don't think I can offer any further insight without actually building though.
20160901 07:58:57-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 258 seconds]
20160901 07:59:02< celticminstrel> While it's certainly possible that it's a result of my config changes, they were mostly copy-pasting code, so it seems a little unlikely.
20160901 07:59:06-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has quit [Quit: And lo! The computer falls into a deep sleep, to awake again some other day!]
20160901 07:59:33< vultraz> wedge009: I'm looking into https://gna.org/bugs/?24512
20160901 07:59:53< wedge009> Okay. Need help?
20160901 08:00:02< vultraz> nah, I think I know the bug
20160901 08:00:32< wedge009> Okay, happy to test if/when you think you have a fix ready.
20160901 08:04:02-!- Kwandulin [~Miranda@p200300760F42415E5966D161A0AF96BE.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160901 08:07:13-!- TheJJ [~rofl@ipbcc36896.dynamic.kabel-deutschland.de] has joined #wesnoth-dev
20160901 08:12:34-!- TheJJ [~rofl@ipbcc36896.dynamic.kabel-deutschland.de] has quit [Ping timeout: 265 seconds]
20160901 08:15:09-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160901 08:19:12-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 240 seconds]
20160901 08:20:32< irker510> wesnoth: Wedge009 wesnoth:master 3f10cc72ad79 / src/units/attack_type.hpp: Possible fix for bad iterator in config object (bug #25026). https://github.com/wesnoth/wesnoth/commit/3f10cc72ad7959ea0ebf79a320fe93779590a63c
20160901 08:21:16< wedge009> celticminstrel: ^
20160901 08:39:22-!- Kwandulin [~Miranda@p200300760F42415EF86978F1F258F3EA.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160901 08:39:53< vultraz> issue identified 
20160901 08:40:02< vultraz> now what's a fix that works..
20160901 08:57:18-!- prkc_ [~prkc@gateway/vpn/privateinternetaccess/prkc] has joined #wesnoth-dev
20160901 08:57:23-!- prkc_ [~prkc@gateway/vpn/privateinternetaccess/prkc] has quit [Remote host closed the connection]
20160901 08:58:26-!- prkc [~prkc@gateway/vpn/privateinternetaccess/prkc] has quit [Remote host closed the connection]
20160901 08:58:39-!- prkc [~prkc@gateway/vpn/privateinternetaccess/prkc] has joined #wesnoth-dev
20160901 08:58:55-!- Netsplit *.net <-> *.split quits: irker510, crimson_penguin, DDR, esr, _laco, Soliton
20160901 09:00:57< vultraz> wedge009: seems irker has split, but I just pushed a fix. Please confirm.
20160901 09:03:10-!- Netsplit over, joins: Soliton, irker510, esr, DDR, crimson_penguin, _laco
20160901 09:09:46-!- _laco [~laco@static.183.80.201.138.clients.your-server.de] has quit [K-Lined]
20160901 09:10:35-!- horrowind [~Icedove@2a02:810a:8380:10a8:21b:fcff:fee3:c3ff] has joined #wesnoth-dev
20160901 09:12:13-!- TheJJ [~rofl@ipbcc36896.dynamic.kabel-deutschland.de] has joined #wesnoth-dev
20160901 09:15:04-!- travis-ci [~travis-ci@ec2-54-205-13-9.compute-1.amazonaws.com] has joined #wesnoth-dev
20160901 09:15:05< travis-ci> wesnoth/wesnoth#10653 (master - 3f10cc7 : Wedge009): The build has errored.
20160901 09:15:05< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/156742096
20160901 09:15:05-!- travis-ci [~travis-ci@ec2-54-205-13-9.compute-1.amazonaws.com] has left #wesnoth-dev []
20160901 09:17:36-!- TheJJ [~rofl@ipbcc36896.dynamic.kabel-deutschland.de] has quit [Ping timeout: 250 seconds]
20160901 09:19:25-!- Duthlet [~Duthlet@dslb-188-104-247-024.188.104.pools.vodafone-ip.de] has joined #wesnoth-dev
20160901 09:24:33-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160901 09:28:56-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 265 seconds]
20160901 09:44:08-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160901 10:04:47-!- travis-ci [~travis-ci@ec2-54-158-100-40.compute-1.amazonaws.com] has joined #wesnoth-dev
20160901 10:04:48< travis-ci> wesnoth/wesnoth#10654 (master - 92b0b5d : Charles Dang): The build has errored.
20160901 10:04:48< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/156750601
20160901 10:04:48-!- travis-ci [~travis-ci@ec2-54-158-100-40.compute-1.amazonaws.com] has left #wesnoth-dev []
20160901 10:24:14-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Read error: Connection timed out]
20160901 10:29:12< wedge009> vultraz: Thanks, change makes sense and the bug is resolved.
20160901 10:33:12-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20160901 10:39:10-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160901 10:41:24< vultraz> hm
20160901 10:41:40< vultraz> JyrkiVesterinen: looks like the mp test hangs after a game on some builds...
20160901 10:42:22< vultraz> host logs off but join sits around
20160901 10:44:12< vultraz> idling in lobby waiting for game
20160901 10:44:24< vultraz> but the game is over..
20160901 10:44:36-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 276 seconds]
20160901 10:50:35-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20160901 10:52:33-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 265 seconds]
20160901 10:52:56-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Read error: Connection timed out]
20160901 10:53:22< irker510> wesnoth: Charles Dang wesnoth:master aef6248d3467 / images/misc/map.png: New map size icon by myself https://github.com/wesnoth/wesnoth/commit/aef6248d3467671e91cebc1dcc9fec87e80789c4
20160901 10:54:01-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20160901 10:55:09-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 260 seconds]
20160901 10:55:13-!- TheJJ [~rofl@ipbcc36896.dynamic.kabel-deutschland.de] has joined #wesnoth-dev
20160901 10:56:30-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20160901 10:57:11-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Read error: Connection reset by peer]
20160901 10:58:04-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20160901 11:00:12-!- TheJJ [~rofl@ipbcc36896.dynamic.kabel-deutschland.de] has quit [Ping timeout: 276 seconds]
20160901 11:20:39-!- horrowind [~Icedove@2a02:810a:8380:10a8:21b:fcff:fee3:c3ff] has quit [Quit: horrowind]
20160901 11:30:32< JyrkiVesterinen> vultraz: Yes, I read the Travis logs in the morning.
20160901 11:30:45< JyrkiVesterinen> I think this new problem is actually /caused/ by https://github.com/wesnoth/wesnoth/commit/6016bdf2f249e8a65769fbe44e0d63c29110c6d7 .
20160901 11:31:24< JyrkiVesterinen> It seems that, for whatever reason, in Travis the join script doesn't get the MP wait screen between the scenarios.
20160901 11:31:37< JyrkiVesterinen> Whereas it did get the screen in my local tests.
20160901 11:32:29< JyrkiVesterinen> I have been thinking that I may have to resort to print debugging here.
20160901 11:33:29< JyrkiVesterinen> In other words, to create a branch where the game logs much more information in Travis MP tests, to find the point where the behavior diverges in Travis and in my computer.
20160901 11:33:51< JyrkiVesterinen> So that I can find out why the behavior is different and unify it.
20160901 11:34:26< JyrkiVesterinen> Of course, that kind of print debugging is going to be S L O W . . .
20160901 11:35:52< vultraz> oh deer :(
20160901 11:47:29-!- travis-ci [~travis-ci@ec2-54-158-100-40.compute-1.amazonaws.com] has joined #wesnoth-dev
20160901 11:47:30< travis-ci> wesnoth/wesnoth#10655 (master - aef6248 : Charles Dang): The build has errored.
20160901 11:47:30< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/156774661
20160901 11:47:30-!- travis-ci [~travis-ci@ec2-54-158-100-40.compute-1.amazonaws.com] has left #wesnoth-dev []
20160901 11:47:32-!- TheJJ [~rofl@ipbcc36896.dynamic.kabel-deutschland.de] has joined #wesnoth-dev
20160901 12:00:30-!- horrowind [~Icedove@2a02:810a:8380:10a8:21b:fcff:fee3:c3ff] has joined #wesnoth-dev
20160901 12:14:26-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160901 12:18:52-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 252 seconds]
20160901 12:36:35-!- Kwandulin [~Miranda@p200300760F42415EF86978F1F258F3EA.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160901 12:42:59-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has joined #wesnoth-dev
20160901 13:27:42< mattsc> zookeeper: Using a [prefer] tag and functionality for the assassin AI works very nicely.
20160901 13:27:56< mattsc> I’ve set it up on the HttT-S8 map for the assassin squad to prefer going through the northern part of the map and that’s what they do.  But if I move the target into the water in the southeast, they’ll head straight for it instead as there is no preferred route in that case.
20160901 13:28:37-!- fabi [~fabi@176.0.92.102] has quit [Read error: Connection reset by peer]
20160901 13:38:32< zookeeper> mattsc, ok, cool. so what's the plan WRT actually incorporating it? will you put it in and then i can tweak the map and parameters, or..?
20160901 13:39:48< irker510> wesnoth: Charles Dang wesnoth:master 75d60ca1844a / data/gui/window/game_load.cfg: Game Load: ensure game list toggle panel grows to fill space horizontally https://github.com/wesnoth/wesnoth/commit/75d60ca1844a0e9124f08050f125d5edcaf2054f
20160901 13:39:51< irker510> wesnoth: Charles Dang wesnoth:master b1e97e68e235 / data/gui/window/game_stats.cfg src/gui/dialogs/game_stats.cpp: Game Stats: use list_data for tab bar https://github.com/wesnoth/wesnoth/commit/b1e97e68e23556a184542cfd51b5ce8acfd81cc9
20160901 13:41:17< mattsc> zookeeper: I am planning to turn it into a Micro AI (today, if I can find the time), and then I think you can put it in together with the tweaks.
20160901 13:41:29< zookeeper> sounds good
20160901 13:48:21-!- Nobun [~nobun@host66-25-dynamic.50-82-r.retail.telecomitalia.it] has joined #wesnoth-dev
20160901 13:55:54-!- fabi [~fabi@176.0.57.109] has joined #wesnoth-dev
20160901 14:03:46< irker510> wesnoth: Charles Dang wesnoth:master c3a89396fc2b / src/game_initialization/singleplayer.cpp: Cleaned up SP game initialization code https://github.com/wesnoth/wesnoth/commit/c3a89396fc2b8529703180918214077be0b2b2bb
20160901 14:16:37-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has quit [Quit: .]
20160901 14:24:35-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has joined #wesnoth-dev
20160901 14:29:28-!- hk238 [~kvirc@unaffiliated/hk238] has joined #wesnoth-dev
20160901 14:43:38-!- Kwandulin [~Miranda@p200300760F42415E55DB25DA040CEADD.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160901 14:56:14< mattsc> celticminstrel: with your new organization scheme of the MAIs in maid-defs/, should the assassin AI be under misc.cfg or have its own file?
20160901 14:56:44< Nobun> celticminstrel: is an acceptable solution if we don't support _ ( " translatable string") anymore on lua?
20160901 14:56:44< mattsc> It consists of 2 CAs, one for the assassin quad moves and then a simple_attack CA
20160901 14:56:59< Nobun> I mean for translatable strings WITHOUT plural form
20160901 14:57:27< celticminstrel> Heh, maid.
20160901 14:57:55< mattsc> Oh well ...
20160901 14:58:03-!- atarocch [~atarocch@natmobil.sfa.se] has quit [Ping timeout: 240 seconds]
20160901 14:58:11< celticminstrel> mattsc: If it doesn't have a lot of additional options, I think it makes sense to put it in misc.cfg.
20160901 14:59:00< celticminstrel> Nobun: Well, we're not dropping support for that in the engine (it would be basically impossible)...
20160901 14:59:32< mattsc> celticminstrel: Okay.  I have no personal opinion on that, I just didn’t quite understand what your organization scheme was.  For example, why is patrol.cfg its own file?
20160901 14:59:47< celticminstrel> I don't remember.
20160901 14:59:55< mattsc> But it really doesn’t matter to me.  So I’ll put it into misc.
20160901 15:00:06< Nobun> celticminstrel: obliouvsly I'm talking about wmlxgettext collecting rules
20160901 15:00:21< celticminstrel> I think there was some general move towards putting related MAIs together and giving ones with larger definitions their own file, but I don't remember the logic in any detail.
20160901 15:00:42< Nobun> the current regexp match also if the the translatable string is _ (" ")
20160901 15:00:54< mattsc> Right, that’s how it looks.  Anyways, onward.
20160901 15:01:10< celticminstrel> Nobun: I wasn't aware it was already supported, but anyway, I can kind of see your problem.
20160901 15:01:13< Nobun> but it is not designed to capture correctly the _ ( "string")
20160901 15:01:39-!- Appleman1234 [~Appleman1@KD119104106244.au-net.ne.jp] has quit [Ping timeout: 264 seconds]
20160901 15:01:50< celticminstrel> Although a regex matching the full thing can easily distinguish _("") and _("","", based on the presence of a comma, I guess a regex matching just the prefix doesn't have that info.
20160901 15:02:13< Nobun> celticminstrel: this is true only if the string with plural is single line
20160901 15:02:24< celticminstrel> So if you can't defer determination of whether it's a singular or plural until after the first string is parsed, then sure, I guess it's fine to drop support for _("")
20160901 15:02:33< Nobun> but if it is multi-line it is impossible to distinguish
20160901 15:02:52< celticminstrel> I wouldn't say it's impossible to distinguish.
20160901 15:03:06< Nobun> with the current code of wmlxgettext is so
20160901 15:03:08< celticminstrel> You just don't know whether you're parsing a plural form until you reach the end of the first string.
20160901 15:03:19< celticminstrel> So it's certainly possible to distinguish.
20160901 15:03:33< Nobun> becouse, in order to save time, it parse regular expression line by line, it does not make a regular expression parsing multiple lines at a time
20160901 15:03:44< celticminstrel> But I don't expect WML authors to be using that syntax, so it's not high priority to support it.
20160901 15:04:03< celticminstrel> Right, I'm aware of that, but even with that limitation I would say it is possible to distinguish.
20160901 15:04:48< celticminstrel> vultraz: I see you added a missing header comment - maybe scour the source dir for other files that are missing one?
20160901 15:04:56< Nobun> yeah,.. however the current regexp, even if valid also for _ (" ") is not designed to collect that case, so I assume that _ ( "quote") is currently assumed as non-translatable string
20160901 15:06:08< celticminstrel> I suppose _{"", "", count} could be supported instead, though I think that's kinda silly.
20160901 15:06:40< Nobun> celticminstrel: it will not added
20160901 15:06:51< celticminstrel> ?
20160901 15:07:13< Nobun> wmlxgettext is still design to drop empty string, wich must NOT be added on pot file format
20160901 15:07:39< Nobun> so if _ "", "", count is collected, it is dropped becouse msgid is empty
20160901 15:08:12< Nobun> after all msgid "" is reserved for pot header informations
20160901 15:08:16< celticminstrel> No, that was just an example to show syntax.
20160901 15:08:40< celticminstrel> I mean like _{"singular", "plural", count} with curly braces instead of parentheses.
20160901 15:08:44< Nobun> ah ok... a snapshot for WML
20160901 15:08:47< celticminstrel> But I think that's a bit silly.
20160901 15:08:51 * vultraz has found another thing he needs to convert to gui2 :|
20160901 15:09:02< celticminstrel> And that's still for Lua.
20160901 15:09:22< vultraz> need to split the options management from MP Create into a utility class and make a new options-only dialog for SP campaigns to launch
20160901 15:09:35< Nobun> yes, understood celticminstrel 
20160901 15:09:51< celticminstrel> vultraz: I'm pretty sure I brought that up when you were working on the options in MP Create.
20160901 15:10:03< celticminstrel> BTW, we still need to make them preserve the designer's ordering, right?
20160901 15:10:24< Nobun> on wmlxgettext side it is the same having _( or having _{ ... both in good and in bad expects of regexp parsing
20160901 15:11:08< vultraz> celticminstrel: you did?
20160901 15:11:17< celticminstrel> Yeah, pretty much I guess (only difference is the latter wouldn't conflict with _("")).
20160901 15:11:25< celticminstrel> vultraz: I recall something to that effect, yes.
20160901 15:11:40-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has joined #wesnoth-dev
20160901 15:12:09< vultraz> well, it shouldn't be too hard
20160901 15:12:22< Nobun> yeah but as I said the current regexp for _ "", even if evalued true on _ ("") is not designed for that case, so probably the sentence will be considered non-translatable in the current code (wich means it will be not added on pot file, if I am right)
20160901 15:12:39< vultraz> we basically recreated the mp::options classes
20160901 15:14:37< celticminstrel> vultraz: You could probably split out the option handling from MP Create into a utility class.
20160901 15:14:53< vultraz> [02:09:21]	vultraz	need to split the options management from MP Create into a utility class  [..]
20160901 15:14:55< vultraz> :|
20160901 15:14:57< celticminstrel> XD
20160901 15:15:31< celticminstrel> I'm thinking a utility class that handles all the widgets, though.
20160901 15:15:48< tad_> celticminstrel: Tested 'Journeyman Mage' again, still no workie. So changed 'Mage' to 'White Mage' and you get ONE guess and it had best be correct ...
20160901 15:15:49< celticminstrel> So, it takes ... probably the window as input, and handles everything to do with the options panel.
20160901 15:16:09< vultraz> either the window or tree view, yes
20160901 15:17:24< vultraz> I think it'd just need a window and a create_engine 
20160901 15:17:46< celticminstrel> Yeah, you could pass the tree view as well, that would work well.
20160901 15:17:50< celticminstrel> ^instead
20160901 15:19:37< zookeeper> so the unit_tree= key matches if the type of the unit being matched exists anywhere in the advancement tree(s) that the given type belongs to?
20160901 15:21:18< celticminstrel> Not quite. It matches that type and anything that that type can advance to.
20160901 15:21:32< celticminstrel> So if you said type_tree=White Mage you'd only get white mages and mages of light.
20160901 15:21:34< zookeeper> s/unit_tree/type_tree
20160901 15:21:37< zookeeper> right
20160901 15:22:25< tad_> celticminstrel: If you're discussing 'type_tree' .. I just tested and if you say "White Mage" you don't get any matches.
20160901 15:22:38< celticminstrel> Huh...
20160901 15:22:43< tad_> Yep.
20160901 15:22:47< celticminstrel> I'll have to run some tests.
20160901 15:22:51< tad_> It's the embedded space, maybe?
20160901 15:23:09< tad_> "Mage" matches "Mage" and "White Mage"
20160901 15:23:48< tad_> But I thought .. is it "Journeyman Mage", or that it's custom? So I changed it to "White Mage" and no workie
20160901 15:24:04< zookeeper> i'm not sure what the potential usecases are supposed to be, so it's hard to say how it should work. the main use would seem to be to check whether two units are "related", in which case the current check-advancements-only method wouldn't work (or, you'd have to check it two different ways for it to work).
20160901 15:24:24-!- travis-ci [~travis-ci@ec2-54-80-161-132.compute-1.amazonaws.com] has joined #wesnoth-dev
20160901 15:24:25< travis-ci> wesnoth/wesnoth#10657 (master - c3a8939 : Charles Dang): The build has errored.
20160901 15:24:25< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/156820016
20160901 15:24:25-!- travis-ci [~travis-ci@ec2-54-80-161-132.compute-1.amazonaws.com] has left #wesnoth-dev []
20160901 15:25:00< tad_> The primary use I'm putting it to is checking "Is it this, or any advancement" to fix bugs like a message type=Spearman not getting a message if the Spearman advanced
20160901 15:28:47< zookeeper> yeah but if you actually specify Spearman in your code then you can just as easily write all the types there
20160901 15:29:33< tad_> Which is tedious, error-prone and not future-proof.
20160901 15:30:48< zookeeper> sure, and avoiding that is a perfectly nice use for the feature :J i'm just not sure that it shouldn't check whether the _whole_ unit tree of the matched unit contains the given type
20160901 15:30:53< fabi> hi tad_
20160901 15:31:04< tad_> fabi: hello
20160901 15:31:11< celticminstrel> Ah, I wonder if it could be the space...
20160901 15:31:27< celticminstrel> Pretty sure it uses the same split function as type though...
20160901 15:31:46-!- atarocch [~atarocch@88.131.217.34] has joined #wesnoth-dev
20160901 15:31:51< tad_> zookeeper: Actually, would tend to agree. No reason not to have the option to have "Royal Guard" match to "Spearman" .. but only as an option.
20160901 15:32:32< celticminstrel> Well, to have type_tree=Royal Guard match Spearman, we would need to iterate through all unit types and check if there's an advancement path to Royal Guard.
20160901 15:32:33-!- prkc [~prkc@gateway/vpn/privateinternetaccess/prkc] has quit [Ping timeout: 240 seconds]
20160901 15:32:52< tad_> celticminstrel: I was about to check that in the C++ .. needs to split on \s*,\s*
20160901 15:33:14< celticminstrel> Yeah, there's a utility function that handles the splitting.
20160901 15:33:51< zookeeper> celticminstrel, yeah i thought it might
20160901 15:34:04< celticminstrel> While I do think matching "units that can advance to this" could be useful, it's significantly less efficient than matching "units this can advance to".
20160901 15:34:08< tad_> celticminstrel: Or add an advances_from attribute to all units, allowing a list.
20160901 15:34:38< celticminstrel> The help screen does list the reverse direction, I think...
20160901 15:34:46< zookeeper> it does
20160901 15:35:15< Ravana_> and help screen has massive delays to start up when there are over 2000 units
20160901 15:35:16< tad_> The help system probably builds the entire tree and can sythesize the back-references
20160901 15:35:24< celticminstrel> Maybe.
20160901 15:36:35< tad_> So: matching a unit: (1) exact match, (2) include all advances_to, (3) include all advances_from, (4) include all both
20160901 15:37:31< celticminstrel> advances_from is currently a tag which adds the current unit to another unit's advances_to.
20160901 15:37:56 * tad_ gaks and sputters
20160901 15:37:58< tad_> ok
20160901 15:38:26< celticminstrel> (It's intended for grafting custom units into the default unit tree.)
20160901 15:38:46< zookeeper> celticminstrel, wait wait wait. why would you need to check all units? can't you just walk up the advancement tree of both the matched unit and the type_tree type, and see if one is found in the other?
20160901 15:39:09< tad_> that would do it.
20160901 15:39:18< celticminstrel> So you're proposing adding an actual advances_from key in the config? (That feels a bit redundant.) Or just adding advance_from when the unit is built?
20160901 15:39:35< celticminstrel> zookeeper: Oh... yes, now that you mention it, I guess I was overthinking it.
20160901 15:40:22< zookeeper> yeah i didn't think of that initially either
20160901 15:40:26< tad_> Well, you're going from O(n) to O(2n) but who's to notice, we're not talking 60 million units in the tree
20160901 15:40:47-!- hk238 [~kvirc@unaffiliated/hk238] has quit [Quit: http://www.kvirc.net/ 4.9.1 Aria]
20160901 15:40:54< zookeeper> but... it stands to reason that the only units you need to check are the ones provided
20160901 15:41:05< celticminstrel> Yeah.
20160901 15:42:02< vultraz> ok, class split
20160901 15:42:22< vultraz> t'was rather simple
20160901 15:42:34< celticminstrel> So, should I make type_tree also match things that can advance to the given units, or use a separate key for that?
20160901 15:43:53< tad_> celticminstrel: there are four cases, all valid. Need either two or four keys to handle them. I'd vote '2' .. type= and type_mactch=Exact|Parents|Children|All
20160901 15:44:37< tad_> type_match=exact|to|from|any default exact
20160901 15:45:21< celticminstrel> So you're saying I should remove type_tree.
20160901 15:45:44< tad_> 'change' is the word I'd use but, yeah
20160901 15:46:11< tad_> First, let's get it matching "White Mage" tho .. then you can re-use the work
20160901 15:47:45-!- prkc [~prkc@192.40.89.18] has joined #wesnoth-dev
20160901 15:47:46< zookeeper> well sure there are four theoretical cases, but are these actually useful in practise enough to warrant such a complicated mess?
20160901 15:48:36< celticminstrel> My thought was either: 1) Keep type_tree as is and add a new key for the "children/from" behaviour. 2) Change type_tree to the "all" behaviour.
20160901 15:48:40< tad_> zookeeper: probably yes .. maybe not .. never underestimate the capacity of your user to surprise you
20160901 15:49:11< celticminstrel> The latter is less flexible, since it doesn't allow the parents/to or children/from behaviour.
20160901 15:49:34< celticminstrel> The former allows "all" behaviour by using both keys.
20160901 15:49:57< tad_> celticminstrel: don't wory about breakage. at this point I'm prolly the only to try it and not that much
20160901 15:50:19< celticminstrel> I'm certainly not worried about breaking people that used type_tree, if that's what you mean.
20160901 15:50:35-!- Nobun [~nobun@host66-25-dynamic.50-82-r.retail.telecomitalia.it] has quit [Quit: Salve a tutti]
20160901 15:50:52< zookeeper> tad_, sure but the important bit is to make things possible to do... catering to esoteric usecases that one person will use in one add-on to save them 10 extra lines of code is not worth it. if a feature allows something to be done that was not possible before, but the esoteric usecases require a bit of extra code on top of that then it's basically not a problem at all.
20160901 15:50:57< tad_> celticminstrel: Step back and ask "What is easiest to remember and get right when I've not used the feature in a year and barely remember it?"
20160901 15:51:23< zookeeper> now, i'm not sure whether just making it check "all" would allow all the other cases to be simulated in a reasonably simple manner
20160901 15:52:00< tad_> zookeeper: I can make up use cases as examples. What I can't do is say someone would actually use them.
20160901 15:52:14< tad_> zookeeper: And we're talking about design-time effort.
20160901 15:52:58< tad_> A general rule-of-thumb. If it cost N to fix at design-time, it costs 10N to fix in beta-test and 100N to fix after public release ...
20160901 15:53:24< celticminstrel> Which of the three (from, to, all) do you think is the most likely for people to want?
20160901 15:53:34< celticminstrel> If the answer is "all", then the three-key method is a bad choice.
20160901 15:53:40< zookeeper> hard to say
20160901 15:53:42< tad_> The fourth -- "exact" the default
20160901 15:53:53< celticminstrel> No, that one's excluded from the question.
20160901 15:54:15< tad_> Just error on key given but missing a value
20160901 15:54:22< celticminstrel> Huh?
20160901 15:55:22< tad_> type=Swordsman / type_tree=all|from|to|[exact] .. just error if type_tree is there but has no value. exact is current usage so it's the most common
20160901 15:55:56-!- vincent_c [~bip@vcheng.org] has quit [Quit: Coyote finally caught me]
20160901 15:56:33< zookeeper> if it only used "all", how would you write a SUF which matches if unit A is an advancement of unit B?
20160901 15:56:37< tad_> If you're asking which to check first, as an optimization .. matched so no need to do the other .. I'll take 'to' .. look up the chain first
20160901 15:57:07-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Read error: Connection reset by peer]
20160901 15:57:08< zookeeper> type_tree=$A.type level=... uh i guess you can't do lessthan/greaterthan things there
20160901 15:57:10-!- vincent_c [~bip@vcheng.org] has joined #wesnoth-dev
20160901 15:57:25 * zookeeper scratches head but also goes afk for a while
20160901 15:58:14-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20160901 15:58:47< tad_> 'if ((tree == 'all' or tree=='to') and search()) or ((tree=='all' or tree=='from') and search())
20160901 16:00:35-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160901 16:00:37< zookeeper> i'd suggest two keys, advances_from= and advances_to=, but those are the sames as used for defining advancements with different semantics (no trees) so it'd be confusing
20160901 16:01:36< tad_> If we're talking about what to put in the SUF, it's a question of whether to have 2, 3 or 4 keys there to express the cases.
20160901 16:01:56< celticminstrel> I think I would expect advances_to to only match direct advancements.
20160901 16:02:08-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Read error: Connection reset by peer]
20160901 16:02:18-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160901 16:03:33-!- Appleman1234 [~Appleman1@KD119104106244.au-net.ne.jp] has joined #wesnoth-dev
20160901 16:03:37< tad_> If we use 2 keys (type= and tree=) we can handle all possible. If we use 3 (type= advto= advfrom=) we need to worry about combining is 2 or more aregiven
20160901 16:04:13-!- Bonobo [~Bonobo@2001:44b8:254:3200:ddf8:b9cf:5c99:641b] has quit [Ping timeout: 255 seconds]
20160901 16:04:26< vultraz> hmm..
20160901 16:04:32-!- Bonobo [~Bonobo@2001:44b8:254:3200:74f3:a595:6f15:b6e7] has joined #wesnoth-dev
20160901 16:04:39< vultraz> I don't  have a window argument until pre_show...
20160901 16:04:40< vultraz> so...
20160901 16:05:01< vultraz> I guess I need to use a unique_ptr of this class instead of an object directly
20160901 16:05:10< tad_> My vote is SUF uses type=<exact match> and we simply add type_tree=all|from|to|exact and avoid the issue of combining
20160901 16:05:10< celticminstrel> Yeah, that's what I was just about to suggest.
20160901 16:05:14< celticminstrel> vultraz: ^
20160901 16:05:23< celticminstrel> vultraz: And reset() it in post_show.
20160901 16:05:34< celticminstrel> Well, I guess that's optional though.
20160901 16:05:56< celticminstrel> Since if for some reason you showed the dialog again, it would be re-reset in the pre_show; and no matter what it'll be reset in the destructor.
20160901 16:09:52< Ravana_> level filter is done with formula=
20160901 16:10:03< celticminstrel> ?
20160901 16:11:14< celticminstrel> I suppose, if type_tree was always "all" behaviour, you could get "to" / "from" with something like formula="level >= 2"...
20160901 16:11:53< celticminstrel> I still haven't gotten an answer on which of "all", "to", "from" is most likely desirable.
20160901 16:12:11< Ravana_> that would rely on unit not advancing to unit of its own level
20160901 16:12:13< tad_> Why does it matter?
20160901 16:12:17< celticminstrel> (Well, the preceding does assume that level always increases on advancement.)
20160901 16:13:00< Ravana_> I guess "to" would be used most
20160901 16:13:13-!- Bonobo [~Bonobo@2001:44b8:254:3200:74f3:a595:6f15:b6e7] has quit [Ping timeout: 255 seconds]
20160901 16:13:22< celticminstrel> So, "match any units that this unit could advance to".
20160901 16:13:30< celticminstrel> s/units/types/
20160901 16:13:39< celticminstrel> ^s/unit/type/
20160901 16:13:54< tad_> Istm you're asking "If type_tree is present which option should be the default if it's missing a value" and the current usage is "exact" 
20160901 16:14:45< tad_> If you're asking which case to test first, for short-circuit, then "to" is my vote.
20160901 16:14:46< celticminstrel> That statement assumes your proposed change is made.
20160901 16:15:03< celticminstrel> But the question is intended to decide which proposed change to make.
20160901 16:15:10< tad_> Ah.
20160901 16:15:51< tad_> I like fewer keys so I'd suggest type_tree determine the search scope and type determine the starting point.
20160901 16:16:03< vultraz> hmm
20160901 16:16:08< vultraz> widget notbeing found :|
20160901 16:16:23< celticminstrel> Seems like the current behaviour of type_tree does represent what people feel to be the most wanted use case.
20160901 16:16:36< celticminstrel> vultraz: Which widget? The tree view, or something under it?
20160901 16:16:59< vultraz> the tree
20160901 16:17:12< celticminstrel> You're not trying to find it after you set the pager, are you?
20160901 16:17:39< vultraz> ahhh
20160901 16:17:41< vultraz> good catch
20160901 16:19:06< vultraz> t'is works :D
20160901 16:19:23< celticminstrel> That sentence makes no grammatical sense.
20160901 16:19:30-!- Kwandulin [~Miranda@p200300760F42415E55DB25DA040CEADD.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160901 16:19:35< celticminstrel> (Given that it expands to "it is works")
20160901 16:20:07< vultraz> this*
20160901 16:20:09 * tad_ looks for the 'slap-celmin' emoticon
20160901 16:20:16 * celticminstrel reels.
20160901 16:20:53< tad_> Learn to speak typonese.
20160901 16:21:13< celticminstrel> What is this magical language of which you speak.
20160901 16:21:41< tad_> teh on which maks ths readalbe
20160901 16:21:53< celticminstrel> Ergh...
20160901 16:23:20< irker510> wesnoth: Charles Dang wesnoth:master 725df26dd27f / / (7 files in 3 dirs): MP Create: split options handling into a helper class https://github.com/wesnoth/wesnoth/commit/725df26dd27f690288027929f2cde510b874464f
20160901 16:23:23< vultraz> celticminstrel: thou must... update XCode!
20160901 16:23:32< celticminstrel> Yeah, sure.
20160901 16:24:29< celticminstrel> Why do I get this really annoying "It is now Brackdon_Cawr's turn" popup?
20160901 16:24:34< vultraz> One knows code is good code when it can be moved around with 0 refactoring
20160901 16:24:42< celticminstrel> Is that related to moving eras?
20160901 16:24:49< vultraz> :| :| :| :| :|
20160901 16:24:51< celticminstrel> No, that couldn't be right.
20160901 16:25:01< celticminstrel> It doesn't seem to be in my scenario.
20160901 16:25:14< vultraz> oh, I see what you're saying
20160901 16:25:19< celticminstrel> (Couldn't be right because this is in MP to begin with.)
20160901 16:25:30< vultraz> one always gets that popup when its your turn
20160901 16:25:40< celticminstrel> I never got it before a few days ago.
20160901 16:26:03< vultraz> maybe your notifications weren't working?
20160901 16:26:04< tad_> Isn't there a preference to stop the popup?
20160901 16:26:05< celticminstrel> Pfft, Unit Create has "Male" selected for a shyde. Maybe I should fix this. :P
20160901 16:26:16< celticminstrel> Notifications? What notifications?
20160901 16:26:31< vultraz> popup notifications?
20160901 16:26:40< celticminstrel> I don't have Growl.
20160901 16:26:44< celticminstrel> If that's what you mean.
20160901 16:26:55< vultraz> no idea :P
20160901 16:27:10< vultraz> also, Unit Create doesn't disable invalid gender options for the selected unit
20160901 16:27:17< vultraz> you might want to implement that
20160901 16:27:17< celticminstrel> Growl is a notification service for older versions of OSX.
20160901 16:27:25< celticminstrel> What preference is this?
20160901 16:28:23< celticminstrel> Maybe Turn Dialog?
20160901 16:28:49< vultraz> yes
20160901 16:29:08< celticminstrel> I don't remember enabling that.
20160901 16:29:15< celticminstrel> Oh well, it's off now.
20160901 16:29:30< celticminstrel> Testing type_tree as it currently stands.
20160901 16:29:38< celticminstrel> I have a shaman, a druid, and a shyde to match against.
20160901 16:29:45< vultraz> GUI1 options manager code: 598 lines. Our GUI2 options manager: 262
20160901 16:30:26< celticminstrel> Actually, I think I need to eat something before continuing. Skipped breakfast.
20160901 16:31:00< vultraz> (just occurred to me i could probably add the code to get the config of options there too)
20160901 16:31:27< vultraz> *proceeds to do so*
20160901 16:32:48< celticminstrel> In theory that means the SP config only needs to have that tree and a couple of buttons.
20160901 16:33:01< celticminstrel> So it should be really, really easy to implement.
20160901 16:33:08< celticminstrel> (The dialog I mean)
20160901 16:33:16< vultraz> Indeed
20160901 16:36:13< irker510> wesnoth: Charles Dang wesnoth:master 0498934b38be / src/gui/dialogs/multiplayer/ (mp_create_game.cpp mp_options_helper.cpp mp_options_helper.hpp): MP Create: move options config getter to mp_options_helper https://github.com/wesnoth/wesnoth/commit/0498934b38bef24c174b97e3855fc6eebce435d6
20160901 16:36:31-!- JyrkiVesterinen [~JyrkiVest@87-100-143-151.bb.dnainternet.fi] has joined #wesnoth-dev
20160901 16:36:42< vultraz> Plus the WML for the tree is already written :D
20160901 16:40:00-!- edgrey [~edgrey@178.205.216.224] has joined #wesnoth-dev
20160901 16:55:07-!- travis-ci [~travis-ci@ec2-54-80-161-132.compute-1.amazonaws.com] has joined #wesnoth-dev
20160901 16:55:08< travis-ci> wesnoth/wesnoth#10658 (master - 725df26 : Charles Dang): The build failed.
20160901 16:55:08< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/156859886
20160901 16:55:08-!- travis-ci [~travis-ci@ec2-54-80-161-132.compute-1.amazonaws.com] has left #wesnoth-dev []
20160901 16:59:07< celticminstrel> Indeed.
20160901 16:59:30< vultraz> oh deer
20160901 16:59:36< vultraz> why is travis complaining now
20160901 16:59:53< vultraz> ah, unused parameter 
20160901 16:59:56< vultraz> will fix
20160901 16:59:57-!- hk238 [~kvirc@unaffiliated/hk238] has joined #wesnoth-dev
20160901 17:01:59< celticminstrel> Uuugh, XCode crashed while debugging again... :|
20160901 17:02:35-!- ChipmunkV [~vova@static-89-94-113-91.axione.abo.bbox.fr] has joined #wesnoth-dev
20160901 17:13:20-!- travis-ci [~travis-ci@ec2-54-158-100-40.compute-1.amazonaws.com] has joined #wesnoth-dev
20160901 17:13:21< travis-ci> wesnoth/wesnoth#10659 (master - 0498934 : Charles Dang): The build failed.
20160901 17:13:21< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/156863395
20160901 17:13:21-!- travis-ci [~travis-ci@ec2-54-158-100-40.compute-1.amazonaws.com] has left #wesnoth-dev []
20160901 17:15:00< celticminstrel> tad_: I think the issue here might be that type_tree currently matches advancements of the given unit, but not the given unit itself.
20160901 17:15:45< celticminstrel> type_tree=Mage is not matching a mage, and type_tree=White Mage is only matching a Mage of Light, for example.
20160901 17:15:56< celticminstrel> (I have one of every unit in the tree on the map.
20160901 17:15:58< celticminstrel> )
20160901 17:18:03< tad_> Well when I tested it matched "Mage" and all advancements but was not matching "White Mage" .. like the space was the issue, not the specific unit or its level
20160901 17:18:33< celticminstrel> It didn't match Mage when I tested just now.
20160901 17:23:16< tad_> I'm on Battle for Wesnoth v1.13.5+dev (db23818-Clean)
20160901 17:23:21< tad_> Let me re-test.
20160901 17:25:32< tad_> You are correct.
20160901 17:26:23-!- higgins [~higgins@68.ip-149-56-14.net] has quit [Quit: Leaving]
20160901 17:30:18< celticminstrel> Okay, so that should be easy to fix then, but there's also the question of whether to expand the behaviour to units that advance to the given units.
20160901 17:32:39-!- Duthlet [~Duthlet@dslb-188-104-247-024.188.104.pools.vodafone-ip.de] has quit [Quit: leaving]
20160901 17:47:12< tad_> I would handle it using tags .. I thought we'd talked it out and it was simply a matter of choosing how you wanted to do the options.
20160901 17:47:27< celticminstrel> Tags?
20160901 17:47:52< tad_> It's jsut a choice of testing A>B or B>A .. tags/attributes/however-you-wanna-do-it-whachamacallits
20160901 17:48:50-!- hk238 [~kvirc@unaffiliated/hk238] has quit [Quit: http://www.kvirc.net/ 4.9.1 Aria]
20160901 17:54:19< vultraz> hmmm
20160901 17:54:30< vultraz> dialog shows up, but options list be empty...
20160901 17:57:49< vultraz> gah :|
20160901 18:00:35-!- higgins [~higgins@68.ip-149-56-14.net] has joined #wesnoth-dev
20160901 18:01:07< mattsc> zookeeper: https://wiki.wesnoth.org/Micro_AIs#Assassin_Squad_Micro_AI_.28ai_type.3Dassassin.29
20160901 18:01:34< mattsc> The MAI is not actually committed yet, but you can let me know if you see anything “objectionalbe” in the described behavior.
20160901 18:06:21< vultraz> ok, so looks like it needs to be called before prepare_for_new_level
20160901 18:07:48< celticminstrel> mattsc: Why does it require a single target unit? (Although I can't say it doesn't make sense - that's generally the point of assassins anyway.)
20160901 18:09:22< tad_> I found the description fairly understandable. The questions which popped into my mind when reading it were all answered by thinking "probably way to hard to do, this will work as described"
20160901 18:10:21< mattsc> celticminstrel: yes, because that’s usually the point of assassins and because I don’t want to deal with having to figure out the logic of distributing assasins across several targets.
20160901 18:10:37< mattsc> I one wants to have several targets, you can always use several instances of the MAI.
20160901 18:10:39< celticminstrel> I see, alright then.
20160901 18:10:45< celticminstrel> And yeah, I was thinking that too.
20160901 18:11:00< celticminstrel> (Which pushes the work of distributing the assassins onto the designer. :P )
20160901 18:11:10< mattsc> Yup.
20160901 18:11:32< mattsc> If there’s ever a strong need for adding that functionality, it can be added at some point.
20160901 18:12:06< mattsc> I guess I would have only needed to use the word ‘add’ once in that sentence.
20160901 18:13:06< tad_> Might be an idea .. grouped-ai give it a SUF for actors and another SUF/SLF for targets and have it choose the best actor for a target or the best target for an actor
20160901 18:13:34< tad_> but pushing the work to the designer is a fine solution
20160901 18:14:17< mattsc> tad_: on your first statement: without knowing what you mean specifically, in general I agree with the statement.  One of the things I’ve learned with AI design is that fairly simple approaches usually work just as well in-game as something much more complex.
20160901 18:14:22< mattsc> Most of the time, not always.
20160901 18:14:56< mattsc> tad_: The goto MAI does that.
20160901 18:15:15< mattsc> (that’s concerning your second comment)
20160901 18:15:36< mattsc> But there the selection criterion is simply “choose the closest” goal.
20160901 18:15:50< mattsc> Unless unique_goals=true.
20160901 18:16:21< mattsc> The point being, I have the mechanism in place already, and all it would, in principle, need is a different rating function.
20160901 18:17:31< mattsc> But then, as we all seem to agree, I’d prefer to leave that to the scenario designer for the time being, unless somebody comes up with a usecase that shows a compelling reason to do so in the AI.
20160901 18:17:43< mattsc> … which somebody probably will, eventually.
20160901 18:18:59-!- Kwandulin [~Miranda@p200300760F42415EA48B89F53BAEA9AC.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160901 18:19:25< vultraz> it works :D
20160901 18:35:14< zookeeper> mattsc, the description seems good. as for multiple targets, if it can be done in the scenario then that's good already.
20160901 18:35:29< zookeeper> although i do wonder why is the target identified by id instead of SUF?
20160901 18:36:01< zookeeper> is it just to make it faster or something?
20160901 18:37:29< mattsc> zookeeper: it’s simply to force it to be a single unit
20160901 18:38:13< mattsc> It’s easy to make it a SUF and then simply choose whichever unit is first encountered.
20160901 18:38:24< mattsc> I just didn’t see the point in doing so.
20160901 18:38:39< mattsc> But I don’t really have an opinion one way or another.
20160901 18:39:16< zookeeper> right. in practise i'd expect the target to always be a unit with a known id so it's not a practical problem, it just seemed a bit unusual somehow.
20160901 18:44:21-!- mjs-de [~mjs-de@x5ce33565.dyn.telefonica.de] has joined #wesnoth-dev
20160901 18:52:02-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has quit [Ping timeout: 264 seconds]
20160901 18:56:00-!- fabi [~fabi@176.0.57.109] has quit [Ping timeout: 276 seconds]
20160901 19:11:41-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160901 19:21:51-!- tad__ [add94167@gateway/web/freenode/ip.173.217.65.103] has joined #wesnoth-dev
20160901 19:30:37-!- tad__ [add94167@gateway/web/freenode/ip.173.217.65.103] has quit [Quit: Page closed]
20160901 19:31:19-!- tad__ [add94167@gateway/web/freenode/ip.173.217.65.103] has joined #wesnoth-dev
20160901 19:34:54< tad__> celticminstrel: So what's the decision on the type_tree? Between my system locking, grandkids showing with great-grands in tow, android tablets needing resetting and lunches I've lost track of what was decided
20160901 19:35:59< celticminstrel> I too was distracted by lunch.
20160901 19:36:33-!- irker510 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20160901 19:37:22< tad__> Things calmed down here. Now I just have to deal with great-grandma wanting a baby of her own .. which is about 20 years past not-happening but I have to deal with it ...
20160901 19:40:02-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20160901 19:40:05< tad__> OK. I'll keep an eye on the commits and test the changes when they appear. In the meantime, I'm adding a fall-back to get the few changes I have working correctly and proceeding with my code-read on DM to update that PR.
20160901 19:42:51-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 265 seconds]
20160901 19:42:51-!- wedge010 is now known as wedge009
20160901 19:46:08-!- ChipmunkV [~vova@static-89-94-113-91.axione.abo.bbox.fr] has quit [Quit: ChipmunkV]
20160901 19:46:50-!- tad__ [add94167@gateway/web/freenode/ip.173.217.65.103] has quit [Quit: Page closed]
20160901 19:47:58-!- Kwandulin [~Miranda@p200300760F42415EA48B89F53BAEA9AC.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160901 19:55:56-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160901 19:59:01-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has joined #wesnoth-dev
20160901 19:59:07< bumbadadabum> hi
20160901 19:59:55< bumbadadabum> I have a new laptop now, so I'll be able to go back to hopefully developing more
20160901 20:01:00-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 276 seconds]
20160901 20:10:53-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160901 20:39:09-!- tad__ [add94167@gateway/web/freenode/ip.173.217.65.103] has joined #wesnoth-dev
20160901 20:40:08< tad__> vultraz: you around? I'm getting undefined reference errors linking wesnoth .. am I missing something from master?
20160901 20:44:02-!- fabi [~fabi@176.0.57.109] has joined #wesnoth-dev
20160901 20:45:57< tad__> To get a clean make I had to rebase to upstream/master then git reset --hard HEAD~2
20160901 20:50:38-!- tad__ [add94167@gateway/web/freenode/ip.173.217.65.103] has quit [Quit: Page closed]
20160901 20:58:18-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has quit [Ping timeout: 250 seconds]
20160901 21:00:52-!- JyrkiVesterinen [~JyrkiVest@87-100-143-151.bb.dnainternet.fi] has quit [Quit: .]
20160901 21:08:25-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160901 21:11:06-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has joined #wesnoth-dev
20160901 21:13:43-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160901 21:19:57-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160901 21:20:23-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160901 21:20:41-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160901 21:21:07-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160901 21:21:07-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160901 21:21:34-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160901 21:21:48-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160901 21:22:04-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160901 21:30:23-!- irker185 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20160901 21:30:23< irker185> wesnoth: Celtic Minstrel wesnoth:master 6bd2e67f2680 / projectfiles/Xcode/Wesnoth.xcodeproj/project.pbxproj: Update XCode project https://github.com/wesnoth/wesnoth/commit/6bd2e67f268045ec42f671c1a69ced11b1d1cc88
20160901 21:30:43< celticminstrel> I can't figure out how to fix the scrollbar in the 1024x768 lobby. :|
20160901 21:31:10< celticminstrel> Removing all the grow factors from era/mods stuff doesn't have any effect, so I suspect the culprit might be in one of the other two pages...
20160901 21:31:52< irker185> wesnoth: Celtic Minstrel wesnoth:master e0834cf58ba8 / src/units/filter.cpp: Minor fix to SUF type_tree https://github.com/wesnoth/wesnoth/commit/e0834cf58ba8b24b178e5ebc241ecf10b585887a
20160901 21:36:41-!- celmin [~celticmin@unaffiliated/celticminstrel] has joined #wesnoth-dev
20160901 21:39:14-!- RatArmy [~RatArmy@240f:b3:88e3:1:221:85ff:fec0:94bc] has joined #wesnoth-dev
20160901 21:41:22< irker185> wesnoth: mattsc wesnoth:master f7d9d747c056 / / (3 files in 3 dirs): New Micro AI: Assassin Squad AI https://github.com/wesnoth/wesnoth/commit/f7d9d747c056187d514beb4f9945906b95f81172
20160901 21:41:48< mattsc> zookeeper: ^  (and I realize that it is getting a bit late on your side)
20160901 21:42:45< mattsc> I also changed target_id -> [filter_second], both in the AI and on the wiki (with the comment that it must match a single unit on the wiki)
20160901 21:43:35< celmin> Apparently the technical reason for the >= 1 requirement is something like… if the cost function could return < 1, then the A* search could underestimate the cost of a path.
20160901 21:44:02< zookeeper> mattsc, cool stuff
20160901 21:44:04< celmin> Which could cause the search to have more false starts or something.
20160901 21:44:13 * celmin had a discussion about it in another channel yesterday.
20160901 21:44:33-!- Appleman1234 [~Appleman1@KD119104106244.au-net.ne.jp] has quit [Ping timeout: 240 seconds]
20160901 21:45:04< mattsc> zookeeper: also, you should be able to use the example on the wiki page almost literally for HttT S8
20160901 21:45:09< mattsc> celmin: interesting
20160901 21:46:41< mattsc> celmin: one of the problems (characteristics, maybe) of the kind of custom cost function I am using is that it is not matched to actual movement cost of the unit anyway.
20160901 21:46:52< mattsc> So that means I actually need to use a two-step process.
20160901 21:47:01< celmin> Hm?
20160901 21:47:18< mattsc> 1. Search for the “best” path; 2. Find how far along the path the unit can actually advance.
20160901 21:47:37< mattsc> So the issue you describe up there might or might not be a problem for this application.
20160901 21:48:13< mattsc> But in any case, I made sure cost is always >=1 on all hexes (by adding a penalty to non-preferred hexes rather than giving a bonus for preferred ones)
20160901 21:48:36< celmin> I feel like there should be away to do it in one step, but I can't think of anything, so maybe it's my imagination.
20160901 21:49:12< irker185> wesnoth: Celtic Minstrel wesnoth:master cc01b38f00af / changelog: Update changelog https://github.com/wesnoth/wesnoth/commit/cc01b38f00af943419a6b9a98918fbc652ae8722
20160901 21:49:14< mattsc> If you can come up with something, please let me know.  But I have not found anything that works (for the kind of application I want to do)
20160901 21:49:28< mattsc> At least not anything simple enough to be worth the effort.
20160901 21:49:47< bumbadadabum> bumbadadabum> Is there anything I can do?
20160901 21:49:47< bumbadadabum> <bumbadadabum> I'm kinda out of it at the moment
20160901 21:49:50-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has joined #wesnoth-dev
20160901 21:49:52< bumbadadabum> asked this in wesnoth by accident
20160901 21:49:54< bumbadadabum> oops
20160901 21:49:59< celmin> Yes, I think there are things you can do...
20160901 21:50:03< celmin> Any preferences?
20160901 21:51:14< bumbadadabum> something lua?
20160901 21:51:30< bumbadadabum> I also might try to finish the EI facelift starting next week
20160901 21:53:33-!- mjs-de [~mjs-de@x5ce33565.dyn.telefonica.de] has quit [Remote host closed the connection]
20160901 21:53:37< mattsc> zookeeper: if/when you put this into HttT, do you also want to mention it here: https://forums.wesnoth.org/viewtopic.php?p=589715#p589715
20160901 21:53:55< mattsc> (that’s not important right now, but if I don’t mention it I will forget)
20160901 21:54:33< celmin> Uhh, not quite sure what you could do that's pure Lua…
20160901 21:54:43< zookeeper> mattsc, ah yes, i had already forgotten about it. might as well mention it there, sure.
20160901 21:55:09< celmin> There's been some discussion of changing [modify_unit] to use the unit proxy for everything rather than storing and unstoring, but I haven't yet decided whether that's actually a good idea.
20160901 21:55:14-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has quit [Quit: Connection closed for inactivity]
20160901 21:56:32< celmin> I don't consider it any kind of priority, but I suppose you could pick a WML tag and port it to Lua… or one of the poorly-ported ones, such as [animate_unit]...
20160901 21:57:34< celmin> Anyone else know of Lua stuff that needs doing?
20160901 21:57:35-!- Appleman1234 [~Appleman1@KD119104104134.au-net.ne.jp] has joined #wesnoth-dev
20160901 21:57:41< mattsc> zookeeper: okay; I’ll remind you of that again later then  (assuming I remember myself)
20160901 21:58:58< zookeeper> perfect :P
20160901 22:00:16< mattsc> celmin, bumbadadabum: the micro AIs need to be checked whether they take invisible units into account correctly; I am sure there are some for which this should be added.  That’s Lua.
20160901 22:00:16-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has quit [Ping timeout: 250 seconds]
20160901 22:01:57< celmin> …does wesnoth.game_events.on_mouse_action exist?
20160901 22:02:03< celmin> It's mentioned here on the EasyCoding page.
20160901 22:02:37-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160901 22:02:39< celmin> mattsc: This should probably be updated BTW: https://wiki.wesnoth.org/EasyCoding#Improvements_to_AI
20160901 22:04:02< mattsc> celmin: yeah; but on quick reading it looks like it’s really only the first bullet that needs to be deleted
20160901 22:05:47< celmin> Maybe
20160901 22:08:26< celmin> vultraz: First bullet here? https://wiki.wesnoth.org/EasyCoding#GUI1_features
20160901 22:08:29-!- RatArmy [~RatArmy@240f:b3:88e3:1:221:85ff:fec0:94bc] has quit [Quit: Leaving]
20160901 22:08:31< celmin> (Ignore the GUI1 part)
20160901 22:08:50-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has quit [Ping timeout: 264 seconds]
20160901 22:09:11-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has joined #wesnoth-dev
20160901 22:15:42< vultraz> celmin: possibly, possibly
20160901 22:16:02-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has quit [Ping timeout: 264 seconds]
20160901 22:16:18< celmin> Possibly what?
20160901 22:16:26< celmin> Oh, the link?
20160901 22:16:54< vultraz> yes
20160901 22:16:58< vultraz> there's also "Make a header for the add-on when we organize scenarios by add-on in the MP create dialog"
20160901 22:17:17< celmin> Yeah, that would require changing the listbox to a treeview, I think.
20160901 22:17:34< celmin> Would also complicate the plugins quite a bit...
20160901 22:18:13< celmin> …wait, is it even possible to organize scenarios by the add-on at the moment?
20160901 22:18:18< vultraz> no
20160901 22:18:25< celmin> What about in GUI1?
20160901 22:18:29< vultraz> no
20160901 22:18:56< vultraz> I do have an idea, though
20160901 22:19:01< vultraz> maybe something similar to
20160901 22:19:03< vultraz> the grid listbox
20160901 22:19:04< vultraz> in hotkeys?
20160901 22:19:25< celmin> How so?
20160901 22:20:13< vultraz> maybe put a list of installed addons somewhere... where you can click to show/hide that addon's related content
20160901 22:20:34< celmin> Ah.
20160901 22:20:42-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has joined #wesnoth-dev
20160901 22:20:45< celmin> Well, that'd be problematic on the lower resolutions, but...
20160901 22:21:00< vultraz> anyway, not doing thta right now
20160901 22:21:06< vultraz> let me commit this sp thing
20160901 22:21:09< tad_> gui/dialogs/multiplayer/mp_create_game.cpp:142: undefined reference to `gui2::tmp_options_helper::tmp_options_helper
20160901 22:21:14< tad_> + 2 more
20160901 22:21:20-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has joined #wesnoth-dev
20160901 22:21:22< celmin> If you use the four-argument form of find_widget, passing false as the fourth parameter, and account for the possibility of it returning nullptr, then it would be possible to support it on larger resolutions but leave it out on smaller ones.
20160901 22:21:34< tad_> Do I need to edit the makefile someplace to add the file?
20160901 22:21:38< celmin> vultraz: Did you forget to update the CMakeLists?
20160901 22:21:48< vultraz> hm... no I did not..
20160901 22:21:56< celmin> Did not update, or did not forget? :P
20160901 22:21:59-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160901 22:22:11< vultraz> latter
20160901 22:22:16< bumbadadabum> I was about to say wesnoth doesn't build
20160901 22:22:26-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160901 22:22:28< tad_> So what file to I add where to make?
20160901 22:22:51< celmin> Well, the relevant file is src/gui/dialogs/multiplayer/mp_options_helper.?pp
20160901 22:22:52-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160901 22:23:04< celmin> I guess .cpp, you don't include headers in CMake/scons
20160901 22:23:11< bumbadadabum> mp_create_game.cpp:(.text+0x303e): undefined reference to `gui2::tmp_options_helper::update_options_list()'
20160901 22:23:19< bumbadadabum> this kills my wesnoth compilation
20160901 22:23:48-!- Greg-Bog_ [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160901 22:23:48-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Read error: Connection reset by peer]
20160901 22:23:52< celmin> From what vultraz said, it sounds like the file is already there?
20160901 22:23:57< celmin> Maybe in the wrong place though?
20160901 22:24:02< tad_> I have both the hpp and cpp
20160901 22:24:07< vultraz> hm
20160901 22:24:09< vultraz> hang on
20160901 22:24:20< vultraz> ah yes
20160901 22:24:21< celmin> By "there" I mean in the CMakeLists / Sconscript.
20160901 22:24:22< vultraz> I screwed up
20160901 22:24:26< vultraz> I included the .hpp
20160901 22:24:29< vultraz> in cmake/scons
20160901 22:24:30< celmin> Ah.
20160901 22:24:31< vultraz> I will fix
20160901 22:24:55< celmin> That's why it's a link error rather than a compile error, then.
20160901 22:25:04< celmin> Wait no, it's not.
20160901 22:25:06< celmin> Never mind.
20160901 22:25:15< celmin> (It'd be a link error either way.)
20160901 22:26:04-!- horrowind [~Icedove@2a02:810a:8380:10a8:21b:fcff:fee3:c3ff] has quit [Quit: horrowind]
20160901 22:26:24-!- travis-ci [~travis-ci@ec2-54-162-36-37.compute-1.amazonaws.com] has joined #wesnoth-dev
20160901 22:26:25< travis-ci> wesnoth/wesnoth#10661 (master - e0834cf : Celtic Minstrel): The build is still failing.
20160901 22:26:25< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/156938311
20160901 22:26:25-!- travis-ci [~travis-ci@ec2-54-162-36-37.compute-1.amazonaws.com] has left #wesnoth-dev []
20160901 22:27:04-!- atarocch [~atarocch@88.131.217.34] has quit [Remote host closed the connection]
20160901 22:27:51-!- atarocch [~atarocch@88.131.217.34] has joined #wesnoth-dev
20160901 22:29:10< celmin> Did anyone ever fix recursive macros?
20160901 22:30:57< irker185> wesnoth: Charles Dang wesnoth:master faed98a65103 / / (11 files in 7 dirs): Added standalone options configuration dialog for GUI2 https://github.com/wesnoth/wesnoth/commit/faed98a6510308e36475f0cebfedcdbfc67088b5
20160901 22:31:00< irker185> wesnoth: Charles Dang wesnoth:master f78679317241 / src/ (4 files in 2 dirs): Fixup compilation https://github.com/wesnoth/wesnoth/commit/f786793172411707c9cc323ad2204a0648e766e8
20160901 22:31:13< celmin> Guh, now I have to update XCode again huh.
20160901 22:31:14< vultraz> Now you need to update XCode... again :D
20160901 22:31:28< celmin> Could you update changelogs already?
20160901 22:34:27< vultraz> small benefit of my change: the settings screen will pop up before the loadingscree, now :D
20160901 22:34:35-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has quit [Quit: Page closed]
20160901 22:36:26< irker185> wesnoth: Celtic Minstrel wesnoth:master 9bc92e657bee / projectfiles/Xcode/Wesnoth.xcodeproj/project.pbxproj: Update XCode project https://github.com/wesnoth/wesnoth/commit/9bc92e657bee213f9c963ab498a58ea3125c5112
20160901 22:36:28< irker185> wesnoth: Celtic Minstrel wesnoth:master 60d543cf3beb / projectfiles/Xcode/Wesnoth.xcodeproj/project.pbxproj: XCode: Use address sanitizer in debug builds https://github.com/wesnoth/wesnoth/commit/60d543cf3beb2a3d07b2231778de5d33edace6fc
20160901 22:36:54-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has joined #wesnoth-dev
20160901 22:37:09< tad_> undefined reference to `gui2::tsp_options_configure::tsp_options_configure
20160901 22:37:22< tad_> undefined reference to `vtable for gui2::tsp_options_configure'
20160901 22:37:22 * vultraz flails
20160901 22:37:24< celmin> What?
20160901 22:37:32< vultraz> did
20160901 22:37:34< vultraz> I pull
20160901 22:37:35< vultraz> the same mistake
20160901 22:37:37< vultraz> twice :|
20160901 22:37:38< vultraz> :|
20160901 22:37:40< vultraz> :| :| :|
20160901 22:37:52 * vultraz needs.. moar coffee
20160901 22:39:38< celmin> This can be marked done now, right? https://wiki.wesnoth.org/NotSoEasyCoding#GUI2_menu https://gna.org/bugs/?22820
20160901 22:39:42< irker185> wesnoth: Charles Dang wesnoth:master 0cf8a2e71a4b / src/ (CMakeLists.txt SConscript): I am an idiot https://github.com/wesnoth/wesnoth/commit/0cf8a2e71a4bcd89f7f3b02234b047cf345027cf
20160901 22:39:48< vultraz> tad_: ^
20160901 22:40:20< celmin> I imagine those floating GUI1 buttons / textboxes in the in-game UI could be replaced by floating GUI2 buttons / textboxes.
20160901 22:40:33< tad_> brb.
20160901 22:40:34< celmin> vultraz: Your commit message is… I don't know.
20160901 22:40:37-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has quit [Client Quit]
20160901 22:41:02< vultraz> and yes, those can be marked as done
20160901 22:41:27< vultraz> ill remove the entry from the wiki
20160901 22:41:51< celmin> I was about to do that.
20160901 22:41:59< celmin> But sure, whatever, go ahead.
20160901 22:42:36< vultraz> I wonder how we would implement the tab_container widget 
20160901 22:42:51< celmin> Probably much the same way you implemented the unit preview pane.
20160901 22:43:03< vultraz> right
20160901 22:43:19< vultraz> there is a benefit to having the tab bar sometimes removed from the pages, though...
20160901 22:43:23< vultraz> such as in mp creat
20160901 22:43:25< vultraz> e
20160901 22:43:52< celmin> Well, those places probably would just not use the tab_container.
20160901 22:44:56 * celmin is currently in the middle of a full rebuild thanks to that XCode change.
20160901 22:45:08< vultraz> I should find a better place for that options tree view wml than where i put it...
20160901 22:45:18< celmin> Meh.
20160901 22:45:29< vultraz> since I discovered something I didn't notice last night :|
20160901 22:45:46-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has joined #wesnoth-dev
20160901 22:46:06< tad_> clean. and I saw you fixed the no-used warning, too
20160901 22:46:15< vultraz> I did, yes
20160901 22:46:34< vultraz> wish I knew how to enable that warning for myslef
20160901 22:46:55< vultraz> I already have Wall
20160901 22:47:15-!- edgrey [~edgrey@178.205.216.224] has quit [Ping timeout: 244 seconds]
20160901 22:47:34< vultraz> hm, there's a place to do so..
20160901 22:47:45< vultraz> but it already has Wunused..
20160901 22:48:16< vultraz> maybe it's actually Wunused-variable?
20160901 22:48:34< celmin> I'd expect Wunused to include that, but maybe not?
20160901 22:49:23< tad_> -Wall s/b be named -Walmost-all but IIRC it includes unused variables
20160901 22:49:39< vultraz> (and is it 'variable' or 'value'?)
20160901 22:49:55< vultraz> (oh, there are both)
20160901 22:50:07< vultraz> and the gcc docs says Wall includes it :|
20160901 22:50:09< vultraz> blah
20160901 22:50:29< celmin> There's also Wextra which includes most of the warnings not included by Wall (some of which are useless).
20160901 22:51:05< celmin> And then clang added Weverything which includes all the warnings excluded by Wextra / Wall.
20160901 22:51:18< vultraz> ahh
20160901 22:51:19< celmin> It's quite a mess really.
20160901 22:51:21< vultraz> it does not include wunused-parameter
20160901 22:51:30< celmin> Aha.
20160901 22:51:39< vultraz> but you'd think Wunused would include that...
20160901 22:51:51< celmin> You'd think so!
20160901 22:52:21< celmin> But no, they can't add it because people build with Werror and their builds could break.
20160901 22:52:44< celmin> That's why Wall isn't really all warnings anymore.
20160901 22:52:47< tad_> There are other static analysis tools we could add after clang .. if we want to be really bothered by warnings ...
20160901 22:53:11< celmin> vultraz is using gcc though
20160901 22:53:45< celmin> He could switch to clang, I guess.
20160901 22:53:51< tad_> Gcc is good. It has some issues but its pretty good nonetheless
20160901 22:54:26< celmin> I dunno if CodeBlocks explicitly supports clang, but it doesn't really need to.
20160901 22:54:36< tad_> There was a discussion about static analysis tools on reddit a while ago and my take-away is the people who complain about false-positives are the ones who need to pay attention to them :)
20160901 22:54:49< zookeeper> tad_, ok, didn't manage to handle the TSG PR today either, so... tomorrow :p
20160901 22:55:35< tad_> zookeeper: s'ok .. you're in charge I just follow orders :P .. 'sides I'm peering at DM now 
20160901 22:56:19< irker185> wesnoth: Charles Dang wesnoth:master 423fc10437dc / projectfiles/CodeBlocks/ (wesnoth.cbp wesnothd.cbp): CodeBlocks: explicitly added -Wunused-parameter flag https://github.com/wesnoth/wesnoth/commit/423fc10437dcc80a5a6ea4930e4d63833c0276cb
20160901 22:56:21< celmin> Death March!
20160901 22:56:24< vultraz> now I get warned :D
20160901 22:56:35-!- travis-ci [~travis-ci@ec2-54-162-36-37.compute-1.amazonaws.com] has joined #wesnoth-dev
20160901 22:56:36< travis-ci> wesnoth/wesnoth#10660 (master - 6bd2e67 : Celtic Minstrel): The build is still failing.
20160901 22:56:36< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/156937903
20160901 22:56:36-!- travis-ci [~travis-ci@ec2-54-162-36-37.compute-1.amazonaws.com] has left #wesnoth-dev []
20160901 22:56:46< tad_> If I wanted to get upset I'd point at the HttT mess .. but I made it so I'd best not ...
20160901 22:56:58< celmin> Heh
20160901 22:57:38< celmin> I wish the add-on manager PR would hurry up and get merged so I can do stuff with it.
20160901 22:57:57< vultraz> bug loonycyborg?
20160901 22:58:14< celmin> Well, it's not exactly his fault that it's unmerged, is it?
20160901 22:58:26< celmin> Generally the person who makes the PR is not responsible for merging it.
20160901 22:58:30< vultraz> well, the "review" is not coming.
20160901 22:59:00< celmin> vultraz: So, just to remind me… you didn't convert the random map settins dialog, right?
20160901 22:59:07< vultraz> not yet
20160901 22:59:14< celmin> Should I leave it on my TODO for now?
20160901 22:59:28< vultraz> I was going to, but if you're eying it it's simple
20160901 22:59:41< celmin> I'm not specifically eyeing it at the moment, just cleaning up my TODO list.
20160901 23:00:20< vultraz> what are you working on next?
20160901 23:00:36< celmin> Not sure. Might go back to the wml_tag_porting branch.
20160901 23:00:46< celmin> And work on a unit animation API.
20160901 23:01:05< celmin> Still unsure whether it should mirror the C++ API or do something different, though.
20160901 23:01:19< vultraz> perhaps you could remove those 'connecting to server' popups in favor of loading screen stages?
20160901 23:01:36< celmin> Perhaps, where is that exactly?
20160901 23:01:46< celmin> Should I search for tmp_lobby?
20160901 23:01:55< celmin> Or whatever that dialog's called.
20160901 23:02:13< vultraz> tnetwork_transmission::wesnothd_dialog
20160901 23:02:52< vultraz> only possible caveat is that the loading screen would appear even when it's not reloading config..
20160901 23:02:59< vultraz> (ie, connect to server, log out, reconnect)
20160901 23:03:46< vultraz> but right now having 5 popups of various contents is just ridiculous :|
20160901 23:04:53< tad_> Oh .. just out of curiousity .. what's the Lua stuff in master top-level for? intentionally there?
20160901 23:05:24< vultraz> yes
20160901 23:05:30< tad_> ok
20160901 23:05:33< tad_> np
20160901 23:05:58-!- celticmistral [~celticmin@unaffiliated/celticminstrel] has joined #wesnoth-dev
20160901 23:06:03-!- celmin [~celticmin@unaffiliated/celticminstrel] has quit [Read error: Connection reset by peer]
20160901 23:06:03-!- celticmistral is now known as celmin
20160901 23:07:57< celmin> tad_: host.lua and join.lua are plugins used by the MP test.
20160901 23:08:27-!- atarocch [~atarocch@88.131.217.34] has quit [Remote host closed the connection]
20160901 23:08:50< tad_> celmin: OK. Was just curious.
20160901 23:10:03< celmin> There are 26 instances of tnetwork_transmission… wonder if I can remove them all...
20160901 23:10:15< celmin> Not entirely sure if I can remove any of them, mind you.
20160901 23:14:40< celmin> tnetwork_transmission seems to install a pump monitor.
20160901 23:14:53< celmin> Maybe it would be better for the MP lobby and create dialogs to do that too, instead of a timer.
20160901 23:15:07-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has quit [Ping timeout: 258 seconds]
20160901 23:16:07< vultraz> oh?
20160901 23:16:11< vultraz> how would one do so?
20160901 23:16:23< vultraz> I'm all for such a thing
20160901 23:16:37< vultraz> since a 4 second delay on a refresh is rather... meh
20160901 23:16:44-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20160901 23:16:52< celmin> Not quite sure.
20160901 23:17:08< celmin> A pump monitor is called at the end of the event processing loop, BTW.
20160901 23:17:18< celmin> So, once per call to events::pump()
20160901 23:18:32< celmin> Maybe inheriting from events::pump_monitor is enough to register oneself as a pump monitor? (That and declaring the void process(events::pump_info&) function.)
20160901 23:19:10< celmin> Why is this connection_data a child class of tnetwork_transmission?
20160901 23:19:16< irker185> wesnoth: Charles Dang wesnoth:master fd53599f7732 / data/gui/ (7 files in 3 dirs): Moved custom_options_tree.cfg to a better location https://github.com/wesnoth/wesnoth/commit/fd53599f77329a414b3c94b6bbe273bba4a3dbbd
20160901 23:19:26< vultraz> (WML only, no c++ changes)
20160901 23:19:40< celmin> It seems lie the connection_data is the absolutely essential aspect of this, while tnetwork_transmission is just the GUI presentation.
20160901 23:19:44< celmin> ^seems like
20160901 23:21:31< vultraz> yeah, we just want the gui representation to change
20160901 23:21:41< celmin> Okay, so I bet that, if I change the line in wesnothd_dialog() that constructs and shows a tnetwork_transmission to a call to the loadscreen stage selection, that it would "just work".
20160901 23:22:21-!- travis-ci [~travis-ci@ec2-54-90-255-214.compute-1.amazonaws.com] has joined #wesnoth-dev
20160901 23:22:22< travis-ci> wesnoth/wesnoth#10663 (master - cc01b38 : Celtic Minstrel): The build is still failing.
20160901 23:22:22< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/156942124
20160901 23:22:22-!- travis-ci [~travis-ci@ec2-54-90-255-214.compute-1.amazonaws.com] has left #wesnoth-dev []
20160901 23:23:19< celmin> So, it's also used for the addon server, and in MP wait...
20160901 23:24:01< vultraz> addon server use should obviously be kept
20160901 23:24:05< celmin> ?
20160901 23:24:09< celmin> Why?
20160901 23:24:20< vultraz> we don't what a loading screen when connecting to the addon server, I think :P
20160901 23:24:22< vultraz> want
20160901 23:24:29< celmin> Why not?
20160901 23:24:36-!- gfgtdf [~chatzilla@x4e369e2d.dyn.telefonica.de] has joined #wesnoth-dev
20160901 23:25:04< vultraz> that's overkill, I think
20160901 23:25:08< celmin> Why?
20160901 23:25:37< vultraz> because.. it's a rather distracting change
20160901 23:25:53< celmin> I'm pretty sure that you can't even see the loading screen if it displays for a short enough time.
20160901 23:26:34< gfgtdf> wedge009: if you have commit acces you should usually also be able to set bugs to 'fixed' yourself  i'd think?
20160901 23:27:04< vultraz> he's not on gna
20160901 23:27:06 * tad_ tested, Tad approved. The fixes to type_tree get it working for my use cases. We probably still want the other two use cases but I'm happy with it now.
20160901 23:27:25< gfgtdf> vultraz: he sure has a gna account.
20160901 23:27:36< vultraz> notpart of the wesnoth group
20160901 23:27:52< vultraz> and i dont know how to add him :|
20160901 23:27:56< gfgtdf> shadowm: ^
20160901 23:27:56< vultraz> damn gna :|
20160901 23:28:49< vultraz> hmm
20160901 23:28:58< vultraz> I noticed with my new sp config options dialog
20160901 23:29:03< vultraz> youg et it even if there are no options
20160901 23:29:11< vultraz> to set
20160901 23:29:33< shadowm> What do you need me to do, whoever?
20160901 23:29:50< vultraz> shadowm: add wedge to the gna wesnoth group >_>
20160901 23:30:06< shadowm> What's your own account name there?
20160901 23:30:27< vultraz> mine is vultraz
20160901 23:30:29< celmin> Pretty sure it's wedg009?
20160901 23:30:33< celmin> wedge009
20160901 23:30:48< celmin> Well yeah, vultraz's is that, but wedge009 is the one who needs adding.
20160901 23:30:57< celmin> I've seen him commenting on issues.
20160901 23:30:59-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [Ping timeout: 265 seconds]
20160901 23:31:04< vultraz> perhaps he wants to make me and admin so I can do it? :P
20160901 23:31:10< celmin> Perhaps.
20160901 23:31:15< shadowm> Who is Chris C?
20160901 23:31:27< shadowm> (I know Christopher Beck is iceiceice. That's not Chris C.)
20160901 23:31:28< vultraz> who?
20160901 23:31:40< shadowm> I am asking who is Chris C, vultraz.
20160901 23:31:45< vultraz> I do not know
20160901 23:31:52< celmin> Where is this.
20160901 23:32:13< gfgtdf> vultraz: this is the code that makes skops the dialog in the gui2 verrsion https://github.com/wesnoth/wesnoth/blob/master/src/game_initialization/multiplayer_configure.cpp#L243
20160901 23:32:18< gfgtdf> skips*
20160901 23:32:31< vultraz> hm
20160901 23:32:40< shadowm> vultraz: You are now admin.
20160901 23:32:55< vultraz> guess I need to add a has_options member to the new options_manager
20160901 23:33:08< celmin> ?
20160901 23:33:22< shadowm> vultraz: Main -> Manage Members, first section from above.
20160901 23:33:44< shadowm> You'll notice that wedge009 is pending and you can easily approve him there instead of manually adding him at the very bottom.
20160901 23:33:51< shadowm> There's also this Chris C person.
20160901 23:34:03< vultraz> Dunno who that is.
20160901 23:34:08< vultraz> Anyway, wedge009 approved
20160901 23:35:25< gfgtdf> shadowm: here isa bugrpeort form cris c https://gna.org/bugs/?24943
20160901 23:35:45< gfgtdf> not sure if that helps
20160901 23:37:35< tad_>  Member since 30JUL2016 three total reports.
20160901 23:37:57< tad_> Probably jsut clicked a button as an experiment
20160901 23:39:41-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160901 23:39:44 * vultraz curses
20160901 23:41:11< vultraz> one cannot close a dialog from pre_show :| :| :|
20160901 23:41:54< shadowm> Why would you do that?
20160901 23:42:50< vultraz> bypass dialog if it's not needed
20160901 23:43:33< vultraz> guess I should come up with a proper conditional for not showing it :/
20160901 23:45:03< tad_> How did you get to it? Would it not make more sense to check there and not even try?
20160901 23:45:31< vultraz> I'm pondering the best check 
20160901 23:46:43< tad_> For example if it's from an element, disable the element .. 
20160901 23:46:56< tad_> Checking at the callee is a fine backstop, tho
20160901 23:47:27-!- travis-ci [~travis-ci@ec2-54-162-36-37.compute-1.amazonaws.com] has joined #wesnoth-dev
20160901 23:47:28< travis-ci> wesnoth/wesnoth#10664 (master - f786793 : Charles Dang): The build is still failing.
20160901 23:47:28< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/156951097
20160901 23:47:28-!- travis-ci [~travis-ci@ec2-54-162-36-37.compute-1.amazonaws.com] has left #wesnoth-dev []
20160901 23:54:07< shadowm> The dialog's caller should decide whether to display the dialog, not the dialog itself.
20160901 23:54:33< vultraz> true
20160901 23:54:35< shadowm> The execute() and display() static method were also intended to provide a standard mechanism to hide this.
20160901 23:55:25 * shadowm shudders to think of how many more misguided code design decisions must've been made since March.
20160901 23:57:02< vultraz> No misguided code decisions have been made.
20160901 23:57:22< shadowm> You just were in the process of making one. :p
20160901 23:58:10< vultraz> I tried a "simple" method reminiscent of the old one. Didn't work. Now trying a proper one
20160901 23:58:16 * vultraz googles std::set
--- Log closed Fri Sep 02 00:00:33 2016