--- Log opened Mon Sep 19 00:00:17 2016
20160919 00:00:46< tad_> mattsc: Also check while playing S29 .. I don't think you get it if you use Menu|Back to... either
20160919 00:00:49< shadowm> gfgtdf: I'd like think we can all agree to not remove things that don't get in anyone's way.
20160919 00:01:24< tad_> mattsc: My notes say only when using Load for a turn, not for scenario (checked: no version in them) and not for Back to ..
20160919 00:01:26< shadowm> Otherwise someone might also decide to start removing devs.
20160919 00:01:37-!- prkc [~prkc@46.166.138.140] has quit [Ping timeout: 272 seconds]
20160919 00:01:38< shadowm> Especially old ones.
20160919 00:01:48< celticminstrel> mattsc: This suggests that the version key is appearing when the AI context is converted to a config.
20160919 00:01:50< mattsc> tad_: Okay.  So it’s likely along the lines of what celticminstrel suspected; it somehow gets put into the saves correctly.
20160919 00:02:12 * tad_ shrugs.
20160919 00:02:23-!- fabi_ [~fabi@176.5.123.186] has quit [Ping timeout: 244 seconds]
20160919 00:02:28< mattsc> celticminstrel: yes; that’s a better way of expressing what I was trying to say
20160919 00:02:55< tad_> I'd have to look but I think that's the last of the messages I could not get to go away when cleaning up HttT, TSG, DM, AToTB ...
20160919 00:02:57< celticminstrel> Okay, so the fix is probably to delete src/ai/manager.cpp:221.
20160919 00:03:13< celticminstrel> I could do that through the website right now, I guess.
20160919 00:03:57< tad_> celticminstrel: The question is why only on Load panel, only for turns makes sense but why not on 'Back to...' .. that's why I set it aside to worry about later
20160919 00:04:13< celticminstrel> tad_: Huh?
20160919 00:04:50< tad_> The warnings and errors about version only occur when you use Load. They do not occur of you load the same turn-save using Menu|Back to
20160919 00:04:53-!- rwr3000 [45888259@gateway/web/freenode/ip.69.136.130.89] has joined #wesnoth-dev
20160919 00:05:18-!- rwr3000 [45888259@gateway/web/freenode/ip.69.136.130.89] has quit [Client Quit]
20160919 00:05:22< celticminstrel> I dunno why that would be.
20160919 00:05:36< celticminstrel> It's not a reason to not delete that line yet, is it?
20160919 00:05:37< mattsc> celticminstrel: I just looked into the savegame and ‘version=10703’ does appear in it.  So you are very likely correct.
20160919 00:05:44< tad_> celticminstrel: Same here. So I noted the errors and set the issue aside.
20160919 00:06:57< tad_> Personally, all I care about is that the console (stdout / stderr) stay clean and not indicate any problems.
20160919 00:06:57< mattsc> And a grep through src/ shows that that is the only place in the code where 10703 appears.
20160919 00:07:11< celticminstrel> Okay, I'll nuke it then.
20160919 00:07:22< mattsc> Sounds good.
20160919 00:07:41< tad_> No issues looking for it and not finding it, I trust.
20160919 00:07:52< irker634> wesnoth: Celtic Minstrel wesnoth:master 7b40d87f2844 / src/ai/manager.cpp: Don't save removed [ai]version= key in save files https://github.com/wesnoth/wesnoth/commit/7b40d87f2844975cd7eedcfe38aba41bd988a46f
20160919 00:08:23< tad_> Of course, mushroom clouds work, too.
20160919 00:10:55< mattsc> celticminstrel, tad_: makes the messages go away for me!
20160919 00:11:24< tad_> I'm sure it will.
20160919 00:12:06< tad_> I'm doing a make and, of course, it's making almost everyrthing. I really need to set up a separate local repo for tracking master and building ...
20160919 00:12:17< celticminstrel> Of course, you'll still get it if the game you loaded was from before the fix. O+:P
20160919 00:12:21< celticminstrel> ^ :P
20160919 00:12:33 * celticminstrel accidentally hit = instead of delete. <_<
20160919 00:13:18-!- travis-ci [~travis-ci@ec2-50-17-23-137.compute-1.amazonaws.com] has joined #wesnoth-dev
20160919 00:13:19< travis-ci> wesnoth/wesnoth#11066 (fix-tunnel-issues - 738491f : mattsc): The build is still failing.
20160919 00:13:19< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/160913569
20160919 00:13:19-!- travis-ci [~travis-ci@ec2-50-17-23-137.compute-1.amazonaws.com] has left #wesnoth-dev []
20160919 00:15:36-!- prkc [~prkc@46.166.188.213] has joined #wesnoth-dev
20160919 00:17:05< tad_> celmin: ya'know if you get bored or frustrated I have a few little UI issues and an engine bug or two on my maybe-someday list.
20160919 00:18:15< irker634> wesnoth: Celtic Minstrel wesnoth:fix-travis 457bec3ff895 / projectfiles/VC12/ (liblua.vcxproj liblua.vcxproj.filters): MSVC: Exclude lua.cpp (the standalone interpreter) https://github.com/wesnoth/wesnoth/commit/457bec3ff895cfa4bb156f03a3146c997a7fe1e3
20160919 00:18:17< irker634> wesnoth: Celtic Minstrel wesnoth:fix-travis 46634a9f31ef / src/gui/core/window_builder.cpp: Partial revert of 1816e1147f553c075b73c668117fb9833c5987c6 https://github.com/wesnoth/wesnoth/commit/46634a9f31ef3787811beb11db18bf30af30e463
20160919 00:18:27< irker634> wesnoth: Celtic Minstrel wesnoth:master 457bec3ff895 / projectfiles/VC12/ (liblua.vcxproj liblua.vcxproj.filters): MSVC: Exclude lua.cpp (the standalone interpreter) https://github.com/wesnoth/wesnoth/commit/457bec3ff895cfa4bb156f03a3146c997a7fe1e3
20160919 00:18:55< celticminstrel> (Making it a branch so as to avoid a string of "attempt to fix Travis" commits.)
20160919 00:23:56< celmin> I wonder if I can set this thing to not flash "Activity!" at me.
20160919 00:24:24< celmin> Oh hey, it has a whole list of options.
20160919 00:25:25< celmin> Looks like mattsc's build is failing, quite possibly due to the thing gfgtdf pointed out.
20160919 00:25:38< celmin> (Which would be my fault, if correct.)
20160919 00:32:16< celticminstrel> I wonder if there are any other functions to which the [[noreturn]] attribute should be applied.
20160919 00:32:44< celticminstrel> I applied it to wml_exception, and tjailbreak_exception::execute is a clear candidate, but I wonder if there are more...
20160919 00:34:42< tad_> Probably a lot of places. The problem is that over-use can degrade the output of the optimizer.
20160919 00:35:30< celticminstrel> Probably only functions that do nothing but throw, or functions that do nothing but assert.
20160919 00:35:50< celticminstrel> Which doesn't seem like it'd be "a lot of places"?
20160919 00:36:01< mattsc> Can I make it pass somehow without flooding everyboidy with commit messages here again?  The branch is missing 713711a2ccacf0, which is probably the problem.
20160919 00:37:06< celticminstrel> If I understand correctly, the compiler can't tell that it'll never return if the definition is off in a different translation unit.
20160919 00:37:31< celticminstrel> mattsc: Probably not, as that commit didn't actually fix it.
20160919 00:37:45< celticminstrel> (It fixes the build, but the tests are also broken.)
20160919 00:40:59-!- Bonobo [~Bonobo@2001:44b8:254:3200:d0c5:5136:c250:14b1] has joined #wesnoth-dev
20160919 00:51:05< mattsc> celticminstrel: I see.
20160919 00:51:34< mattsc> But if the build did fix it, I’d do a rebase followed by … another force push?
20160919 00:54:00-!- travis-ci [~travis-ci@ec2-54-158-59-140.compute-1.amazonaws.com] has joined #wesnoth-dev
20160919 00:54:01< travis-ci> wesnoth/wesnoth#11068 (master - 7b40d87 : Celtic Minstrel): The build is still failing.
20160919 00:54:01< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/160916019
20160919 00:54:01-!- travis-ci [~travis-ci@ec2-54-158-59-140.compute-1.amazonaws.com] has left #wesnoth-dev []
20160919 01:08:30-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has quit [Quit: Page closed]
20160919 01:17:48< celticminstrel> mattsc: Yes; in that case you'd probably rebase with 'git pull --rebase origin master'
20160919 01:18:54< mattsc> celticminstrel: okay; that only gets my local copy of the branch updated though, right?
20160919 01:20:05< celticminstrel> Correct.
20160919 01:20:18< celticminstrel> Then you 'git push --force' to update the remote.
20160919 01:20:54< irker634> wesnoth: Celtic Minstrel wesnoth:master 46634a9f31ef / src/gui/core/window_builder.cpp: Partial revert of 1816e1147f553c075b73c668117fb9833c5987c6 https://github.com/wesnoth/wesnoth/commit/46634a9f31ef3787811beb11db18bf30af30e463
20160919 01:21:10< mattsc> Right.  And if I did the push without the force, it’d be a mege again, like earlier today?
20160919 01:21:17< mattsc> *merge
20160919 01:21:19< celticminstrel> No, it would reject.
20160919 01:21:28< mattsc> Oh.
20160919 01:21:30< celticminstrel> It merges when you 'git pull' without specifying --rebase.
20160919 01:21:42< mattsc> Oh, alright.
20160919 01:22:21< celticminstrel> The above commit fixes the build (Travis didn't announce it for whatever reason, but the website says it passed), so you could do that now if you wish.
20160919 01:22:49 * mattsc needs to reboot quickly to get the external drive to be recognized …
20160919 01:22:56-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: mattsc]
20160919 01:23:49-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20160919 01:24:14< mattsc> wrong window; back here ...
20160919 01:24:29< Aginor> that was quick
20160919 01:24:40< mattsc> SSDs are nice  :)
20160919 01:24:58< celticminstrel> It takes me like 10-15 minutes to reboot. :/
20160919 01:25:12< mattsc> celticminstrel: okay, I’ll try that
20160919 01:25:12< celticminstrel> Possibly longer if Firefox is open.
20160919 01:25:23-!- gfgtdf_ [~chatzilla@x4e368db9.dyn.telefonica.de] has joined #wesnoth-dev
20160919 01:25:32< celticminstrel> Well, unless I killed it, I guess.
20160919 01:25:35-!- gfgtdf_ [~chatzilla@x4e368db9.dyn.telefonica.de] has quit [Client Quit]
20160919 01:26:36< celticminstrel> It'd be great if someone could figure out why 4663419f is really needed. There's probably something very wrong somewhere.
20160919 01:26:47< celticminstrel> Although it could also just be a compiler bug.
20160919 01:27:05-!- gfgtdf [~chatzilla@x4e363f91.dyn.telefonica.de] has quit [Ping timeout: 244 seconds]
20160919 01:28:10< celticminstrel> Menus are working, except that when you select something, nothing happens. :|
20160919 01:28:24< irker634> wesnoth: mattsc wesnoth:fix-tunnel-issues ae19588a37ce / src/pathfind/ (pathfind.cpp teleport.cpp teleport.hpp): Tunnels: fix vision not being propagated through tunnels https://github.com/wesnoth/wesnoth/commit/ae19588a37ce901a03ad17cefb6b2664209e7e2d
20160919 01:28:26< irker634> wesnoth: mattsc wesnoth:fix-tunnel-issues f6016500705e / src/pathfind/ (teleport.cpp teleport.hpp): [tunnel] tag: add new optional parameter allow_vision=yes/no https://github.com/wesnoth/wesnoth/commit/f6016500705e4aac15327db1e1a996490bf54427
20160919 01:29:17< mattsc> celticminstrel: so I did what you said, and this ^ is what happened.  Is that okay like that?
20160919 01:29:28< celticminstrel> That seems fine?
20160919 01:29:39< mattsc> I mean, are those messages supposed to show up?
20160919 01:29:47< celticminstrel> Unless I'm missing something?
20160919 01:29:55< celticminstrel> I'm not sure what you're asking, I guess.
20160919 01:30:53< mattsc> Those two commits did not change, but the branch was rebased.  So is irker supposed to show them here when I rebase without changing them?
20160919 01:31:35< celticminstrel> Yes, that's expected. From git's perspective, they did change because they now have a different parent commit.
20160919 01:31:35< mattsc> I’m just a bit paranoid at the moment  :P
20160919 01:31:49< mattsc> Yeah, okay, that makes sense.
20160919 01:32:11< mattsc> Okay, so we’ll see if this passes travis now.
20160919 01:32:29-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20160919 01:42:34-!- travis-ci [~travis-ci@ec2-54-158-59-140.compute-1.amazonaws.com] has joined #wesnoth-dev
20160919 01:42:35< travis-ci> wesnoth/wesnoth#11070 (master - 457bec3 : Celtic Minstrel): The build is still failing.
20160919 01:42:35< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/160916916
20160919 01:42:35-!- travis-ci [~travis-ci@ec2-54-158-59-140.compute-1.amazonaws.com] has left #wesnoth-dev []
20160919 01:42:45-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has joined #wesnoth-dev
20160919 01:42:54< celticminstrel> Assuming that failure was the "remove lua.cpp" one.
20160919 01:43:07< celticminstrel> Yeah.
20160919 01:43:33< tad_> mattsc: Here's one I cannot reproduce but see every now and then: error ai/actions: Return value of AI ACTION was not checked. This may cause bugs!
20160919 01:44:10< tad_> Methinks the bark is worse than the bite for that message.
20160919 01:47:46< mattsc> tad_: yeah, that one’s been around every since I started meddling with the AI and I have never seen it cause any problems.
20160919 01:48:08< mattsc> I have it marked on my list as something to look into eventually, but it’s never made ti to the top of the list so far.
20160919 01:51:10< tad_> Yep. "Maybe-someday" but maybe reduce it from error to warning and drop the scary second sentence?
20160919 01:51:38< mattsc> Something like that, yes.
20160919 02:02:29-!- Greg-Boggs [~greg_bogg@2601:1c2:901:e170:ccf:8c91:43d2:fe86] has joined #wesnoth-dev
20160919 02:07:30-!- travis-ci [~travis-ci@ec2-50-17-23-137.compute-1.amazonaws.com] has joined #wesnoth-dev
20160919 02:07:31< travis-ci> wesnoth/wesnoth#11071 (master - 46634a9 : Celtic Minstrel): The build was fixed.
20160919 02:07:31< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/160922174
20160919 02:07:31-!- travis-ci [~travis-ci@ec2-50-17-23-137.compute-1.amazonaws.com] has left #wesnoth-dev []
20160919 02:17:39-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20160919 02:35:01-!- Greg-Boggs [~greg_bogg@2601:1c2:901:e170:ccf:8c91:43d2:fe86] has quit [Remote host closed the connection]
20160919 02:37:00< vultraz> celticminstrel: ok, so I think I can't use the FLG dialog unless I remove that leader dropdown and replace it with a list.
20160919 02:37:15< celticminstrel> Huh? Why?
20160919 02:37:16< vultraz> I guess I'll have to use a regular list for now..
20160919 02:37:24< celticminstrel> What's wrong with using the dropdown?
20160919 02:37:29< vultraz> it's really annoying to do 2 clicks to select a leader
20160919 02:37:45< vultraz> and if it annoys me in testing it will annoy the players
20160919 02:39:05< vultraz> if I make it a list and save selection then the number of clicks stays the same from the old version
20160919 02:41:18< vultraz> ie, 2 total
20160919 02:41:21< vultraz> instead of 3 total
20160919 02:41:51< celticminstrel> I seriously don't see why you're so worried about reducing the number of clicks by one, but if that's what you want to do, I don't really care that much.
20160919 02:44:18< tad_> The number of clicks isn't really a problem unless it's click-to-select then click-to-accept-selection then click-for-something-else .. which is what it sounds like. It is annoying to have the UI not understand a click-to-select really meant SELECT and not prepare-to-select
20160919 02:45:38< vultraz> well, doubeclick is click-to-select
20160919 02:46:14< celticminstrel> I'm... not sure what you're talking about.
20160919 02:47:02< vultraz> he's saying it's annoying to have to select something and then click OK
20160919 02:48:11-!- Greg-Boggs [~greg_bogg@2601:1c2:901:e170:ccf:8c91:43d2:fe86] has joined #wesnoth-dev
20160919 02:49:07< tad_> I don't mind "OK" if there are several things to review or it's a non-reversible operation.
20160919 02:51:24< tad_> What I hate is a button to popup a selection, click to select, click OK to go back. Realize you did it wrong, rinse, repeat.  It's better to click, select, oops click, select and go on
20160919 02:51:37< celticminstrel> Originally it was non-reversible, but that seems to be changing. There are however three things to select (which must be done in order).
20160919 02:52:02< celticminstrel> Well, in order is not 100% required if you're checking various options, maybe (it'll probably remember your gender selection if possible).
20160919 02:52:44< tad_> "in order" might be a problem, then. But "OK" to accept them when you've got all three set (and no errors from your choices) is good.
20160919 02:55:25< tad_> But if it's a radio for gender, and a drop-down for type (one-click to drop down, one more to select) and then OK, that's fine. It's the click-dropdown click-select click-rollup stuff I hate.
20160919 02:57:54< celticminstrel> "in order" is mainly because the available gender choices depend on your leader choice, and also because your leader options depend on the faction choice..
20160919 02:58:06< celticminstrel> ...I guess mainly is the second and also is the first, really.
20160919 02:58:18 * tad_ nods.
20160919 02:59:17< tad_> So you select a female, then select a faction/type where female is not allowed: either change it to male, or change it to not-known and disable the OK button.
20160919 02:59:32< tad_> I'd probably to the latter: disable the OK button.
20160919 02:59:41-!- celmin [461b1bda@gateway/web/freenode/ip.70.27.27.218] has quit [Quit: Page closed]
20160919 03:05:05< shadowm> That sounds inconvenient.
20160919 03:06:04< vultraz> the radio button is supposed to automatically change 
20160919 03:06:12< vultraz> it's not working very well, though
20160919 03:10:36< tad_> My philosophy is to have the user do it, not force a workflow, but prevent advancing on error. If I have a status line, I leave stuff wrong, display a message, and disable OK.
20160919 03:12:47< shadowm> I really can't see why the user shouldn't be allowed to proceed without two additional clicks if they select a single-gender unit.
20160919 03:13:09< shadowm> It's a game and a cosmetic choice, not a fire hazard.
20160919 03:15:33< shadowm> vultraz: "if it annoys me in testing it will annoy the players" -- You do realize that this is something you just said, right?
20160919 03:16:12< shadowm> If it didn't have your name attached to it I'd have assumed someone was making up a strawman.
20160919 03:17:07< vultraz> ... what?
20160919 03:17:22< shadowm> What applies to you needs not apply to everyone else.
20160919 03:17:36< celticminstrel> ^
20160919 03:17:50< shadowm> Personally, I've never heard of anyone complaining about the side setup UI.
20160919 03:18:10< shadowm> And I've heard a lot of complaints about the multiplayer UI otherwise.
20160919 03:25:56-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: So long and thanks for all the fish.]
20160919 03:32:36-!- Greg-Boggs [~greg_bogg@2601:1c2:901:e170:ccf:8c91:43d2:fe86] has quit [Remote host closed the connection]
20160919 04:03:40-!- JyrkiVesterinen [~jyrki@87-100-171-86.bb.dnainternet.fi] has joined #wesnoth-dev
20160919 04:15:52-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has quit [Quit: Page closed]
20160919 04:21:52-!- fabi [~fabi@176.7.128.129] has quit [Read error: Connection reset by peer]
20160919 04:21:59-!- fabi [~fabi@176.7.128.129] has joined #wesnoth-dev
20160919 04:28:44-!- irker634 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20160919 04:47:53-!- enchi [enchilado@defocus/yummy/enchilado] has quit [Ping timeout: 272 seconds]
20160919 04:49:36-!- enchi [enchilado@defocus/yummy/enchilado] has joined #wesnoth-dev
20160919 04:54:08-!- JyrkiVesterinen [~jyrki@87-100-171-86.bb.dnainternet.fi] has quit [Quit: Konversation terminated!]
20160919 04:58:26-!- TC02 [~quassel@venus.arosser.com] has quit [Ping timeout: 244 seconds]
20160919 05:05:36< vultraz> grr
20160919 05:05:41< vultraz> keep getting crashes..
20160919 05:06:28< vultraz> seems to be operating on a nullptr or something but that doesn't make sense
20160919 05:12:06< vultraz> ohh
20160919 05:12:10< vultraz> I got it
20160919 05:12:37< vultraz> was looking in the wrong place
20160919 05:31:45< vultraz> celticminstrel, gfgtdf: progress! https://drive.google.com/file/d/0B-mR9s8FduLLZVR3Y2lwSURocWM/view?usp=sharing
20160919 05:34:38< celticminstrel> Looks pretty good, though not entirely happy about the double-arrow... it seems kinda unintuitive.
20160919 05:36:17< shadowm> It looks like an interactive element, which it is not, surely?
20160919 05:36:38< vultraz> it's a button.
20160919 05:36:49< shadowm> Okay, let me rephrase that.
20160919 05:37:07< shadowm> It has the color scheme of icons used in interactive elements, but otherwise it doesn't look like an interactive element.
20160919 05:38:15< vultraz> hm
20160919 05:39:02< shadowm> (Also, the fact that you praise cheap modern UI designs and still insist on adding horizontal or vertical lines everywhere seems contradictory.)
20160919 05:41:20< shadowm> Also, I'm sure it _looks_ interactive when hovered, but... look, you like rubbing your love for touchscreens in my face at every opportunity, so I don't need to complete that line of thought.
20160919 05:41:52< shadowm> Also.
20160919 05:42:30< shadowm> What does this look like when expanded?
20160919 05:43:18< shadowm> Worth noting that changing leader faction and unit type is the most common action of all that are performed on this screen other than clicking I'm Ready.
20160919 05:44:37< vultraz> it bringsu p the flg dialog
20160919 05:44:59< shadowm> I haven't seen it. Link?
20160919 05:45:09< celticminstrel> The normal faction select dialog.
20160919 05:45:21< celticminstrel> (Though I imagine link would still be helpful since it changed a bit.)
20160919 05:59:00< vultraz> in a dota match will get one later
20160919 05:59:59-!- Kwandulin [~Miranda@p200300760F2C71FC4CDBDFE65995AD87.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160919 06:11:06-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has joined #wesnoth-dev
20160919 06:18:06-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has quit [Read error: Connection reset by peer]
20160919 06:18:12-!- celmin [~celmin@unaffiliated/celticminstrel] has joined #wesnoth-dev
20160919 06:30:14< vultraz> shadowm: https://drive.google.com/file/d/0B-mR9s8FduLLU2V5TURBTzlsUnM/view?usp=sharing
20160919 06:44:37< vultraz> shadowm: it's the screen previously shown when a user connected to a game
20160919 06:45:15< vultraz> shadowm: i was going to make it so the host/player see almost the same screen and both can use that dialog to change f/l/g
20160919 06:45:30< vultraz> shadowm: thoughts?
20160919 06:50:18-!- atarocch [~atarocch@93.56.160.28] has joined #wesnoth-dev
20160919 06:52:39-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has quit [Quit: .]
20160919 06:55:13-!- Kwandulin [~Miranda@p200300760F2C71FC4CDBDFE65995AD87.dip0.t-ipconnect.de] has quit [Ping timeout: 255 seconds]
20160919 07:19:21-!- boucman_work [~boucman@bob75-2-81-56-46-209.fbx.proxad.net] has joined #wesnoth-dev
20160919 07:40:58-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20160919 07:43:10< celmin> Is the inability to switch maps in the editor (using the Window menu) an existing bug or something I've introduced locally?
20160919 07:44:04< vultraz> celmin: it's just you
20160919 07:46:02< zookeeper> could we at least finally agree that the old scrollbars looked nicer?
20160919 07:46:48< vultraz> No
20160919 07:46:50< vultraz> :)
20160919 07:47:19< celmin> No strong preference, but if I were given the choice I'd probably choose the old ones.
20160919 07:47:34< vultraz> Likewise to "did the old font look nicer" and "did the old logo look better"
20160919 07:48:04< vultraz> the old ones were more decorative, yes
20160919 07:48:10< vultraz> but they took too much space
20160919 07:48:29< celmin> I think a GUI2 theme could be implemented to restore the old scrollbars. >_>
20160919 07:49:10< vultraz> NO.
20160919 07:49:13< vultraz> Oops
20160919 07:49:14< vultraz> No*
20160919 07:49:40< celmin> Obviously that would be "opt-int". :P
20160919 07:49:51< vultraz> Still no.
20160919 07:50:05< celmin> You're no fun. :P
20160919 07:50:18< vultraz> No old UI elements are returning.
20160919 07:50:35< zookeeper> they took just enough space that you didn't need to snipe to hit them
20160919 07:50:43< zookeeper> the new ones look terrible anyway
20160919 07:50:53< vultraz> :|
20160919 07:51:12 * vultraz mutters darkly
20160919 07:52:34< vultraz> You people want to drag wesnoth back to the 1.12 aesthetic.
20160919 07:52:49< shadowm> New isn't inherently better, you know.
20160919 07:52:55< celmin> ^
20160919 07:53:07< celmin> I wonder if all that controller stuff is an indication that there was some attempt to use MVC for the main game.
20160919 07:53:21< vultraz> Look at 1.12 and master side-by-side and tell me 1.12 looks better
20160919 07:53:29< shadowm> I can come up with all kinds of new stupid ideas, implement them, and their objective wrongness wouldn't be nullified.
20160919 07:53:35< vultraz> Not any specific element
20160919 07:53:41< vultraz> The general aesthetic
20160919 07:53:54< shadowm> It has a more usable font size and family, for starters.
20160919 07:53:58< celmin> The controller hierarchy (obviously) as the "controller". The display and/or display_context hierarchy as the "view" and/or "model".
20160919 07:54:15< shadowm> Okay, maybe not 'more usable', just 'less cringeworthy'.
20160919 07:54:28< vultraz> :|
20160919 07:54:41< vultraz> why do I even bother.
20160919 07:54:54< shadowm> Because you like forcing your decisions on other people. :p
20160919 07:55:06< zookeeper> ^
20160919 07:55:09< shadowm> It's an efficient approach, certainly.
20160919 07:55:14< celmin> Lato also has a totally different problem, though I guess DejaSans did partially share it.
20160919 07:55:23< celmin> ^DejaVuSans
20160919 07:55:34< celmin> The lack of certain Unicode ranges.
20160919 07:58:49< zookeeper> 1.12 has much nicer dialogs what with the drop shadows and rounded borders. i guess i don't care that much about the font difference. nicer scrollbars, nicer in-game top bar, no gargantuan padding between in-game menu elements.
20160919 07:59:14< shadowm> The new scrollbars need some very minor improvements but are otherwise good.
20160919 07:59:28< celmin> 1.12 loading screen has a progress bar, which is nicer.
20160919 07:59:40< celmin> Scrollbars are about even, honestly.
20160919 07:59:59< celmin> I suppose saving space is a decent rationale for the new ones.
20160919 08:00:23< shadowm> However, I suspect the new scrollbars make the UI touch-unfriendly until GUI2 gains actual touch gestures support, on account of their thinness.
20160919 08:00:25< celmin> zookeeper: Gargantuan padding? o.O
20160919 08:00:42< celmin> shadowm: I'm pretty sure the buttons already made it kinda touch-unfriendly?
20160919 08:00:44< shadowm> (Correct me if I'm wrong. I don't feel like getting out that crappy laptop right now.)
20160919 08:01:35< shadowm> celmin: It's probably still relatively easier to tap on regular buttons than drag that scrollbar.
20160919 08:02:07< celmin> I suppose.
20160919 08:02:17< celmin> ...wait, why am I celmin'd?
20160919 08:02:40< celmin> Whatever, going to sleep anyway.
20160919 08:02:45< vultraz> zookeeper: I plan to add drop shadows to dialogs again as soon as the framework is fixed
20160919 08:02:46-!- celmin [~celmin@unaffiliated/celticminstrel] has quit [Quit: And lo! The computer falls into a deep sleep, to awake again some other day!]
20160919 08:02:55< zookeeper> oh? well that's nice
20160919 08:03:03< vultraz> zookeeper: but how in hell was the topbar "nicer"
20160919 08:03:14< vultraz> because certain elements were in black boxes??
20160919 08:03:42< vultraz> "no gargantuan padding" well excuse me for making the menu items feel less cramped!
20160919 08:03:43< zookeeper> the elements were more defined, yes
20160919 08:04:33< shadowm> I gave you a suggestion for the theme UI elements but you decided to go overboard.
20160919 08:04:41< shadowm> I said "no borders", not "no backgrounds".
20160919 08:04:55< vultraz> it would look incredibly silly without borders
20160919 08:04:58< shadowm> Although right now that isn't relevant since I see that the boxes for most of the elements are back, and the menu buttons look like buttons again.
20160919 08:05:06< vultraz> yes
20160919 08:05:14< vultraz> I reverted the no-image button change
20160919 08:05:18< vultraz> since that was a mistake.
20160919 08:05:45< vultraz> and reverted most of the element boxes, yes
20160919 08:05:50< vultraz> but those in the topbar were not necessary
20160919 08:06:05< shadowm> (Not relevant for me, that is. I don't really miss the backgrounds for the top bar items right now, although it does look generally wonky anyhow.)
20160919 08:07:03< shadowm> <vultraz> it would look incredibly silly without borders
20160919 08:07:22< shadowm> Going back to the random horizontal and vertical bars all over your designs, people nowadays use backgrounds without borders.
20160919 08:07:45< shadowm> Generally of different contrasting colors in order to make them still usable for people with color vision defficiencies.
20160919 08:07:54< vultraz> sure
20160919 08:08:16< shadowm> (Or crappy screens. And crappy cheap phone screens are probably a good part of the reason everyone and their dog makes crappy cheap flat designs these days.)
20160919 08:08:36< vultraz> but that'd require additional tweaking besides just stripping the borders off the boxes
20160919 08:08:44< vultraz> (as for my lines, I try to use them sparingly)
20160919 08:08:45< shadowm> (Crappy as in lacking a sufficiently wide color gamut.)
20160919 08:09:19< vultraz> (I only have 2 in the lobby in order to define the gamelist since celmin insisted such a thing was neeed)
20160919 08:09:21< vultraz> needed*
20160919 08:09:44< vultraz> besides that, one in Create, one in Attack, and one in Game Stats
20160919 08:09:52< vultraz> could possibly remove the last
20160919 08:10:04< vultraz> and the one in staging might go
20160919 08:10:08-!- Kwandulin [~Miranda@p200300760F2C712FFC9FB41C890313DA.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160919 08:11:50-!- boucman_work [~boucman@bob75-2-81-56-46-209.fbx.proxad.net] has quit [Ping timeout: 265 seconds]
20160919 08:11:50< vultraz> i would also not be opposed to the return of a very small progress bar in the loading screen
20160919 08:12:00< vultraz> as long as it were very small
20160919 08:13:12< vultraz> but I do *not* appreciate this backwards attitude of hanging on to the past even if it's flawed.
20160919 08:14:59< vultraz> as dave and jet said, this project has the unfortunate tendency to shoot down new ideas.
20160919 08:15:32< vultraz> I reject this philosophy and instead go ahead an implement new ideas anyway.
20160919 08:16:43< vultraz> (I suppose that makes me sound like fabi o_O )
20160919 08:17:54< vultraz> but the point still stands that blindly proposing a total reversal of changes (like zookeeper demending the old scrollbars back) is *not* productive in any way.
20160919 08:18:00< zookeeper> there's also no need to hang on to the same flawed rhetoric and expecting that arbitrarily labeling a change as "progress" somehow is supposed to make it immune to criticism.
20160919 08:18:23< vultraz> No, it's not immune to criticism.
20160919 08:18:31< vultraz> And I have heeded criticism where it was valid.
20160919 08:18:47< vultraz> Ie, the menu buttons and the rest of the element bgs/borders
20160919 08:18:58< vultraz> Or the inclusion of up/down arrows in my minimal scrollbar design.
20160919 08:20:00< vultraz> shadowm says the new scrollbars need a few small tweaks
20160919 08:20:23< vultraz> I'll gladly hear them and will likely implement them
20160919 08:20:30< shadowm> I already communicated them to you once.
20160919 08:20:40< zookeeper> no one has argued that any of the new stuff is bad by virtue of being different than the old. you're just arbitrarily calling some criticism "valid" and some as "wanting to drag wesnoth back to <whatever>".
20160919 08:21:30< vultraz> If you support reverting to the 1.12 aesthetic then yes, you're dragging wesnoth back.
20160919 08:22:04< vultraz> If instead, you provide constructive suggestions to continue to improve, then that's valid and useful.
20160919 08:22:06< shadowm> I was hoping that the inclusion of a pun at the end of that mini-essay of mine would make it memorable and it wouldn't be forgotten -- alas, it seems I was incorrect.
20160919 08:22:38< shadowm> No-one's saying "revert this crap now".
20160919 08:22:48< zookeeper> ^
20160919 08:23:13< shadowm> People are just comparing the changes to the previous style and trying to point out what could be relearned from the old style.
20160919 08:23:45< shadowm> When someone says X is better than Y it doesn't necessarily mean you can't take both and produce Z.
20160919 08:24:00< vultraz> (I will, however, admit I might be a little passionate about this since I've put a lot of work into these various changes and to have zookeeper come in and say "can we all agree the old stuff looked better" is just rather disheartening :| )
20160919 08:24:45< shadowm> I'll also say that I feel the 1.10 style was better in some regards to both 1.12 and master.
20160919 08:25:13< shadowm> Mainly about some dubious choices regarding saturation and contrast.
20160919 08:25:45< shadowm> And also about texture -- I'm not a fan of wood textures and this also goes for zookeeper's map background.
20160919 08:26:15< vultraz> For the record, lordbob is going to produce us a new wooden map background, at my request.
20160919 08:26:19< zookeeper> vultraz, s/stuff/scrollbars. there's plenty of perfectly good stuff
20160919 08:26:22< shadowm> But other people seem to like the wooden appearance and I guess that's okay even if I can't really stand it.
20160919 08:26:39< zookeeper> he is?
20160919 08:26:43< vultraz> Yes
20160919 08:26:46< zookeeper> when was someone going to tell me?
20160919 08:26:52< shadowm> vultraz: Yeah, no, it's not about the textures' quality.
20160919 08:26:58< shadowm> It's about the fact that they are wooden .
20160919 08:27:28< vultraz> Yes, but I needed to mention the quality will improve.
20160919 08:27:44< shadowm> Like you can sell me your delicious meat-based dishes however you want but they are still meat-based therefore they're not for me.
20160919 08:28:16< vultraz> heh
20160919 08:28:25< shadowm> zookeeper: Probably in the commit message.
20160919 08:28:42< shadowm> It seems like that's the new style of doing things around here.
20160919 08:28:44< zookeeper> shadowm, so do you think there's a background material that would work?
20160919 08:29:03< vultraz> zookeeper: I will gladly tweak the scrollbar if you give me some suggestions on what to do.
20160919 08:29:06-!- boucman_work [~boucman@193.56.60.161] has joined #wesnoth-dev
20160919 08:29:11< shadowm> zookeeper: That's a good question. Does it have to be realistic?
20160919 08:29:31< zookeeper> not photorealistic since the maps arguably aren't quite that either
20160919 08:29:39< zookeeper> but... somewhat realistic? :p
20160919 08:29:58< shadowm> The old synthetic noise background worked for me, I know that much.
20160919 08:30:22< zookeeper> the uh what
20160919 08:30:31< vultraz> personally I'd replace the titlescreen background with one of lordbob's paintings with a parallax effect
20160919 08:30:34< shadowm> Oh, you mean for the maps? I was thinking about the in-game themable UI.
20160919 08:30:35< vultraz> But that can't be done
20160919 08:31:35< zookeeper> shadowm, ...we have a wood texture somewhere else?
20160919 08:31:45< shadowm> The in-game theme UI panels?
20160919 08:31:53< vultraz> shadowm: that's not.. really wood
20160919 08:31:57< zookeeper> hrhm
20160919 08:32:11< zookeeper> doesn't associate with wood for me. if i had to pick a physical material, i guess leather or something
20160919 08:32:22< vultraz> yeah, something like that
20160919 08:32:35< shadowm> Meh, close enough in color scheme.
20160919 08:32:37< zookeeper> it's pretty ambiguous i think
20160919 08:32:46< shadowm> Especially given the borders etc.
20160919 08:33:23< zookeeper> so you object more to color than texture?
20160919 08:33:45< shadowm> Hard to say about the maps since that involves some additional requirements on the presentation. You obviously couldn't have, for example, a felt tablecloth and have it make sense.
20160919 08:33:49< shadowm> Yes.
20160919 08:34:10< shadowm> If the map was hanging on a wall we'd have additional materials to choose from, such as brick or clay.
20160919 08:34:20< shadowm> *from
20160919 08:35:51< vultraz> Wouldn't work well
20160919 08:36:42< zookeeper> some kind of stone could still work
20160919 08:37:56< zookeeper> i'm open to suggestions on that front, since i'm a bit displeased with the wood myself
20160919 08:38:19< vultraz> I'm mostly displeased because it's low-quality
20160919 08:38:28< vultraz> which is why I asked lordbob to make a new one
20160919 08:38:35< zookeeper> for free?
20160919 08:38:49< vultraz> yeah
20160919 08:38:59< zookeeper> okay
20160919 08:39:36< zookeeper> when was this and, again, why do you do stuff like that in secret?
20160919 08:41:22< zookeeper> obviously i don't object to asking lordbob to make a new one, but i also can't know whether he knows the specs
20160919 08:43:43< vultraz> month or so ago
20160919 08:44:37< vultraz>  dunno, guess I didn't think it was that important.
20160919 08:45:52< shadowm> Proper communication is paramount.
20160919 08:46:37< shadowm> Especially in light of the background of that first link in the topic.
20160919 08:52:50-!- irker644 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20160919 08:52:50< irker644> wesnoth: Wedge009 wesnoth:master 21f01b0e5cdd / po/wesnoth/eo.po: Stop Esperanto date formatting crashing in Windows. https://github.com/wesnoth/wesnoth/commit/21f01b0e5cdd7ab5177d42cec2c422a5717dd53b
20160919 08:59:30< shadowm> Just refactored my file dialog implementation.
20160919 09:01:19 * shadowm crosses fingers.
20160919 09:02:26-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has joined #wesnoth-dev
20160919 09:04:12 * shadowm *eye twitch*
20160919 09:04:48< shadowm> I keep forgetting that normalize_path() does not resolve dot entries.
20160919 09:06:59< shadowm> "Returns the absolute path of a file." Lies.
20160919 09:07:30< JyrkiVesterinen> Well, technically the path is absolute if it doesn't depend on current directory. ;)
20160919 09:07:45< shadowm> Whoever wrote the BFS version of normalize_path for 1.12 made the same mistaken assumption as me.
20160919 09:08:28< shadowm> i.e. didn't realize that BFS makes a distinction between canonical and absolute paths.
20160919 09:08:37< JyrkiVesterinen> But it would indeed be useful if Boost.Filesystem offered a way to convert a path to absolute and resolve dot entries, but ignore symbolic links (or reparse points on Windows).
20160919 09:08:46< shadowm> canonical.
20160919 09:08:58< shadowm> Literally canonical().
20160919 09:09:37< shadowm> The big problem with it is that it requires all components to exist and errors out if they don't -- which makes sense, since dot entries *are* directory entries rather than magical keywords.
20160919 09:10:04< shadowm> But that makes my code slightly more convoluted.
20160919 09:10:29< JyrkiVesterinen> I think that at least some platforms enforce that dot entries are really magical keywords.
20160919 09:10:56< shadowm> Applications aren't supposed to know or rely on the implementation details of dot entries.
20160919 09:11:19< JyrkiVesterinen> It's not an implementation detail if the OS guarantees it.
20160919 09:11:35< JyrkiVesterinen> In that case the application can resolve dot entries without hitting the disk.
20160919 09:12:44< JyrkiVesterinen> Look at this. mkfs.vfat considers that dot entries which point to wrong places are filesystem errors and fixes them: https://linux.die.net/man/8/fsck.vfat
20160919 09:12:56< JyrkiVesterinen> "Directory . does not point to parent directory. The start pointer is adjusted."
20160919 09:12:58< shadowm> The POSIX documentation for readdir() says "If entries for dot or dot-dot exist, one entry shall be returned for dot and one entry shall be returned for dot-dot; otherwise, they shall not be returned."
20160919 09:13:13< shadowm> That word alone makes the difference.
20160919 09:13:26< shadowm> Yes, MS-DOS at least requires dot entries to exist in the FAT.
20160919 09:15:20< shadowm> Anyway, consdering that POSIX already has provisions for . and .. not always existing, it makes perfect sense that BFS will not make assumptions about them -- otherwise it'd not be true to POSIX.
20160919 09:15:48< JyrkiVesterinen> Hmm, yes, I see the point.
20160919 09:15:49< shadowm> Irrespective of the fact that IIRC Linux's VFS architecture synthesizes them for all directories.
20160919 09:17:28< shadowm> I guess I'll just add a third-parameter to normalize_path() instead of adding something silly like canonicalize_path().
20160919 09:17:33< shadowm> s/-/ /
20160919 09:25:25< shadowm> And now I get to check for errors in places I didn't plan for. Whoop-de-doo.
20160919 09:28:52< shadowm> OK, I guess it's not so bad on account of the refactored code.
20160919 09:29:33< zookeeper> humm. something's changed WRT how singleunit gender is randomized if no gender= is provided? because i'm randomly getting a female kaleh (not in master).
20160919 09:30:16< zookeeper> or maybe it's always happened for units whose type allows both genders and i've just been lucky enough before to not notice.
20160919 09:30:42< shadowm> Since when does Kaleh's type allow both genders?
20160919 09:31:17< zookeeper> since i've made him and nym be of the same Desert Youth type
20160919 09:31:38< shadowm> ... And yet that isn't master, somehow.
20160919 09:31:57< zookeeper> yes it's in my fork only, as is all my UtBS work
20160919 09:33:49< zookeeper> which is what i'll have to start figuring out what to do with
20160919 09:34:32< shadowm> Probably set the gender list differently for each variation or whatever you are trying to do
20160919 09:35:03< shadowm> Or just acknowledge that you aren't ever going to have Desert Youths that aren't either Kaleh or Nym and make their attributes specific to their gender?
20160919 09:35:22< zookeeper> oh it's not a problem
20160919 09:35:34< zookeeper> i was just wondering why it only apparently started happening now
20160919 09:37:56< shadowm> Greeeat, my factoring seems to work but I appear to be losing track of data somewhere.
20160919 09:38:58< shadowm> O
20160919 09:40:05< zookeeper> durr. Parse error when parsing help text: 'corrupted original file'
20160919 09:40:57< shadowm> I think I just had a case of "not having code results in not executing code", outside the dialog even.
20160919 09:44:09< shadowm> My refactored code... works.
20160919 09:44:24< shadowm>     wesnoth: src/gui/dialogs/file_dialog.cpp:272: bool gui2::tfile_dialog::process_submit_common(gui2::twindow&, const string&): Assertion `false && "Unimplemented selection mode or semantics"' failed.
20160919 09:44:28< shadowm> Most of the time.
20160919 09:45:46< shadowm> A missing break.
20160919 09:46:31< shadowm> Even though I was specifically keeping track of the switch structure, that somehow slipped past.
20160919 09:46:41< zookeeper> ah, the help parser choked on an em dash in an ANSI-encoded file.
20160919 09:46:51< shadowm> I'm speaking syntax now. This must be what being vultraz feels like.
20160919 09:50:15< shadowm> And now I have a case of not initializing a POD-type member variable.
20160919 09:50:49< shadowm> Because our SCons configuration doesn't include the magic flag that emits warnings on missing members in constructor initialization lists.
20160919 09:51:05< shadowm> Thanks, whoever.
20160919 09:54:30 * zookeeper is uneasy about pushing a rebase of a fork branch
20160919 09:54:34< zookeeper> what if it lost some of my stuff?
20160919 09:55:31-!- shadowm [~ignacio@wesnoth/developer/shadowm] has quit [Ping timeout: 244 seconds]
20160919 09:56:57-!- shadowm [~ignacio@wesnoth/developer/shadowm] has joined #wesnoth-dev
20160919 10:02:53< zookeeper> also i think the only way to get around my uneasiness about pushing this to master is to keep it parallel with the old version until such a time as it's actually finished.
20160919 10:04:21< zookeeper> ah, [option] in [campaign] is a thing now... i wonder if i could use that.
20160919 10:04:30< zookeeper> s/option/options
20160919 10:05:30< zookeeper> need to abuse the preprocessor so i guess not.
20160919 10:12:21< shadowm> Phew, I think I finally got all the POSIX border cases right.
20160919 10:12:31< shadowm> Now... it's Windows' turn.
20160919 10:14:06-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20160919 10:15:19< shadowm> Ugh.
20160919 10:16:01< shadowm> Already hit a massive regression in the directory tree traversal logic.
20160919 10:16:42-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 276 seconds]
20160919 10:16:43-!- wedge010 is now known as wedge009
20160919 10:16:46< shadowm> Going into a parent that can't be opened for whatever reason causes a dialog exit and the caller gets the failed path back. -.-
20160919 10:17:50< shadowm> And entering \ does nothing.
20160919 10:18:02< shadowm> Entering a drive letter optionally followed by \ does though.
20160919 10:18:40< shadowm> Oh well, seems to work otherwise.
20160919 10:19:05< shadowm> So I might finally publish this tomorrow or so and not feel like it's utter garbage.
20160919 10:19:33< zookeeper> i suppose [difficulty] define= can't take a list
20160919 10:19:42< zookeeper> which would have been convenient just now, but oh well
20160919 10:21:10< vultraz> zookeeper: it does not
20160919 10:21:22< vultraz> zookeeper: it could be made to do so, I suppose. What's the usecase?
20160919 10:21:31< shadowm>    debug gui/dialogs/file_dialog: register_new_selection(): new selection '..' is an object that is not a directory. Full path: C:\users\shadowm\My Documents
20160919 10:21:34< shadowm> Waaait a second.
20160919 10:21:50< zookeeper> vultraz, i'll spare you the burden of knowledge
20160919 10:21:55< zookeeper> because you wouldn't like it
20160919 10:21:56< shadowm> I guess Wine believes it's a file?
20160919 10:22:11< vultraz> as such i will assume it's not something I'd approve of :P
20160919 10:22:26< shadowm> Actually, maybe it's reporting it as a symbolic link rather than a directory.
20160919 10:22:49< shadowm> Instead of following the symbolic link.
20160919 10:23:29< shadowm> Maybe I needn't worry about this border case then since it's pretty idiotic already.
20160919 10:24:38< shadowm> Although it makes me somewhat concerned that I might see strange behavior on Windows as well with exotic filesystem objects.
20160919 10:25:28< zookeeper> vultraz, well basically i think i'll incorporate the new UtBS stuff as three new options in the difficulty selection, so that's the way you choose whether to play the old or new until 1.14. so a new difficulties would be EASY,NEW and NORMAL,NEW and HARD,NEW so i could determine whether it's old or new by checking just one symbol.
20160919 10:25:38< shadowm> So I'll focus on fixing \ for now and ignore the other issue since it's really beyond the scope of the dialog.
20160919 10:25:42< zookeeper> s/difficulties/difficulty
20160919 10:26:02< vultraz> zookeeper: .... why...
20160919 10:26:17< zookeeper> because it's the only way
20160919 10:26:59< vultraz> A, if this is such a huge change, just give it a new campaign entry and a separate folder
20160919 10:27:09< vultraz> B, if not, why maintain 2 versions in 1.13?
20160919 10:27:26< zookeeper> well having them simply be separate campaigns technically is the other possibility
20160919 10:27:55< zookeeper> but i think it's nicer to present them as close together as possible
20160919 10:28:33< vultraz> you do realize that unless such a thing is in 1.14, almost no one will care where you put it?
20160919 10:29:29< vultraz> I do not recommend maintaining the code with preprocessor directives if it's a huge change
20160919 10:29:34< vultraz> it's a huge pain
20160919 10:29:57< zookeeper> sure, i'm counting on those outside of "almost no one". besides 1.13 will receive more and more testers the closer it gets to 1.14.
20160919 10:30:22< zookeeper> and i'm just currently investigating how much of a pain it would be.
20160919 10:31:33< vultraz> a very big pain
20160919 10:32:16< zookeeper> we'll see
20160919 10:32:56< vultraz> unless the changes to utbs are not as extensive as i have been led to believe by the fact that it's been going on for more than a year
20160919 10:34:14< vultraz> but I also spoke to LB at the end of last month and he said he'll be returning to utbs soon
20160919 10:34:17< vultraz> so, we'll see
20160919 10:34:46< vultraz> (actually should be around now. I'll have to check with him again)
20160919 10:35:16< zookeeper> he already posted somewhere that he'll be returning to it
20160919 10:36:57< shadowm> Oh dear GODS.
20160919 10:37:18< shadowm> Monospace font text labels look AWFUL now.
20160919 10:37:48< vultraz> really?
20160919 10:37:52< vultraz> i thought they looked fine
20160919 10:38:48< shadowm> https://dl.dropboxusercontent.com/u/21371130/screenshots/890908908.png
20160919 10:39:10< shadowm> Why is it bold? Why is it larger than the variable-width font?
20160919 10:39:43< shadowm> (And how did I manage to trigger this old cache bug in the first place.)
20160919 10:40:42< vultraz> I honestly have no idea why it looks the way it does
20160919 10:40:59< shadowm> I feel very strongly about this since this dialog is my baby.
20160919 10:41:04-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160919 10:41:09< vultraz> Lato doesn't have a dedicated monospace file
20160919 10:41:20< vultraz> Yet it still comes up with monospace somewhere to display
20160919 10:41:28< vultraz> But it works so I don't question.
20160919 10:41:36< shadowm> It's like I come home one day to find that the babysitter has dipped my child in a bucket of paint.
20160919 10:42:27< shadowm>     family_order_monospace=_ "DejaVu Sans Mono"
20160919 10:42:44< vultraz> wait what>
20160919 10:43:05< vultraz> oh yeah.. that..
20160919 10:43:11< vultraz> .... wait ... why did we leave that?
20160919 10:43:14< vultraz> I don't remember 
20160919 10:43:20< vultraz> . why does that even work
20160919 10:43:23< shadowm> I guess the problem is that someone is using an excessively large font size for it AND a weird hinting mode.
20160919 10:43:25-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has joined #wesnoth-dev
20160919 10:43:54< vultraz> all the font sizes are bigger, yes..
20160919 10:43:59< shadowm> vultraz: BEcause 1) the files are still there (AS THEY SHOULD BE), and 2) the font is installed in my system so it can still be used in the event it's removed frm the game files.
20160919 10:44:02< vultraz> but they don;t differ by ttyle
20160919 10:44:20< vultraz> style
20160919 10:44:28< vultraz> we don't have a ttyle class :P
20160919 10:46:25< shadowm> Whatever.
20160919 10:47:06< shadowm> I must focus on my task so I can go back to being on a break. I can't get distracted with people breaking my old crap.
20160919 11:03:59-!- Duthlet [~Duthlet@dslb-146-060-179-135.146.060.pools.vodafone-ip.de] has joined #wesnoth-dev
20160919 11:05:36< zookeeper> well the UtBS hack is in fact _easy_. it's just ugly because i'd need to add one preprocessor symbol hack to core to enable it to work. but i don't see that as a problem since it's short-term temporary.
20160919 11:06:02< zookeeper> usually ugliness is bad because it's harder to maintain, but that's not a concern here so...
20160919 11:11:18< zookeeper> nevermind, i can confine that to UtBS _main.cfg after all
20160919 11:11:43< JyrkiVesterinen> "Short-term temporary". Famous last words. There is way too much software out there that was supposed to be "short-term temporary" 10 years ago...
20160919 11:12:26< JyrkiVesterinen> Excel sheets being used as databases, for example. Reading The Daily WTF is a good way to find out about those projects.
20160919 11:15:24< zookeeper> JyrkiVesterinen, sure but in this case it's not just something that is supposed to get dealt with sometime soon, it's something that's simply unacceptable for a stable release so there's no option to leave it lingering.
20160919 11:15:53< zookeeper> well, i'm not sure if it'd really be unacceptable or not, but that's the mindset i'm operating at anyway
20160919 11:16:10< JyrkiVesterinen> I wasn't really objecting. Just a general notice.
20160919 11:16:16< zookeeper> fair enough
20160919 11:45:52< matthiaskrgr> oh wow, the new gui looks amazing! 
20160919 11:46:22< vultraz> to which area are you referring?
20160919 11:47:04< matthiaskrgr> a lot of stuf has changed since I checked last time :) 
20160919 11:47:09< matthiaskrgr> that was maybe a semester ago or so 
20160919 11:47:54< vultraz> are you on master or 1.13.5?
20160919 11:48:34< matthiaskrgr> git 
20160919 11:49:06< vultraz> ah, good
20160919 11:49:40< matthiaskrgr> funny, I terminated the game but its still running with 100% cpu now 
20160919 11:49:46< matthiaskrgr> ah there
20160919 11:50:12< matthiaskrgr> might have been the leak sanitizer
20160919 11:50:32< vultraz> have you seen the new MP Create screen?
20160919 11:50:39< irker644> wesnoth: gfgtdf wesnoth:master 8508b42bed28 / src/ (format_time_summary.cpp gettext.hpp gettext_boost.cpp save_index.cpp): use boost locale for strftime https://github.com/wesnoth/wesnoth/commit/8508b42bed28a956aa5aa8a120cb17ed9bf8404f
20160919 11:51:46< matthiaskrgr> yes, it looks very good :)
20160919 11:51:56< matthiaskrgr> there is some UB though :P 
20160919 11:52:01< vultraz> oh?
20160919 11:52:24< matthiaskrgr> http://pastebin.com/B8RWuyAE  
20160919 11:53:17< vultraz> I have no idea how to fix this :P
20160919 11:53:20< matthiaskrgr> it happens in the very moment when I click "create game" 
20160919 11:53:26< matthiaskrgr> I can get you a trace of that, one sec
20160919 11:55:17< matthiaskrgr> http://pastebin.com/YNgG2Nz8  
20160919 11:55:53< matthiaskrgr> (if you compile with -fsanitize=undefined, you can    export UBSAN_OPTIONS=print_stacktrace=1    to get UB stacktraces) 
20160919 11:57:48< vultraz> what's this about image not fitting on canvas? 
20160919 11:58:18< matthiaskrgr> maybe it was the window resolution
20160919 11:58:36-!- travis-ci [~travis-ci@ec2-54-158-59-140.compute-1.amazonaws.com] has joined #wesnoth-dev
20160919 11:58:37< travis-ci> gfgtdf/wesnoth-old#695 (fix_24927 - 94c2497 : gfgtdf): The build is still failing.
20160919 11:58:37< travis-ci> Build details : https://travis-ci.org/gfgtdf/wesnoth-old/builds/161002177
20160919 11:58:37-!- travis-ci [~travis-ci@ec2-54-158-59-140.compute-1.amazonaws.com] has left #wesnoth-dev []
20160919 11:59:03< vultraz> you do see the mp create background image right?
20160919 12:00:15< vultraz> huh, the ub is in the node destructor..
20160919 12:00:17< vultraz> this is odd
20160919 12:00:32< matthiaskrgr> oh this is funny
20160919 12:00:39< matthiaskrgr> so in the create game screen
20160919 12:00:52< matthiaskrgr> when the window/Screen size is too small
20160919 12:01:12< matthiaskrgr> the map description sentence is not wrapped at all but extended fully
20160919 12:01:19< vultraz> oh yeah
20160919 12:01:20< vultraz> known
20160919 12:01:27< matthiaskrgr> ok
20160919 12:01:41< vultraz> our ui toolkit just gives up and decides 'fuck it, we're showing ALL the text' :P
20160919 12:01:49< matthiaskrgr> hehe
20160919 12:02:19< JyrkiVesterinen> I see what is going on with the undefined behavior problem.
20160919 12:02:30< JyrkiVesterinen> Ttree_view inherits tscrollbar_container.
20160919 12:02:54< JyrkiVesterinen> When the tree view is destroyed, ~tscrollbar_container() is called.
20160919 12:03:07< JyrkiVesterinen> It destroys the tree view nodes.
20160919 12:03:23< JyrkiVesterinen> Tree view node destructor attempts to access the tree view itself.
20160919 12:03:38< JyrkiVesterinen> But, at that point, the tree view is already being destroyed.
20160919 12:03:58< JyrkiVesterinen> It's no longer a ttree_view. It has decayed into a tscrollbar_container.
20160919 12:04:07< matthiaskrgr> there are also a couple of memleaks here and there
20160919 12:04:22< matthiaskrgr> do you want me to flood GNA with that? :P
20160919 12:04:45< vultraz> how do you glean all that from a paste o_O
20160919 12:04:50< vultraz> and how would one fix it
20160919 12:04:54< matthiaskrgr> not that one 
20160919 12:04:57< matthiaskrgr> in general I mean 
20160919 12:05:01< vultraz> and yes, please, post any memleak bugs
20160919 12:05:02< JyrkiVesterinen> From the paste and the source code. 8-)
20160919 12:05:07< matthiaskrgr> oh thta was to JyrkiVesterinen 
20160919 12:06:27< JyrkiVesterinen> I suggest adding a destructor to ttree_view. The ttree_view destructor can destroy the tree nodes safely.
20160919 12:07:04< JyrkiVesterinen> Then the tscrollbar_container destructor needs to take into account that ttree_view may already have destroyed them.
20160919 12:07:32< vultraz> could you do this maybe
20160919 12:08:00< JyrkiVesterinen> Yeah. I'll add it to my todo list.
20160919 12:08:14< vultraz> :D thanks
20160919 12:08:21< matthiaskrgr> :>
20160919 12:09:10< vultraz> gfgtdf: no return in save_info::format_time_local() now
20160919 12:26:22-!- travis-ci [~travis-ci@ec2-54-242-255-5.compute-1.amazonaws.com] has joined #wesnoth-dev
20160919 12:26:23< travis-ci> wesnoth/wesnoth#11077 (master - 8508b42 : gfgtdf): The build was broken.
20160919 12:26:23< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/161008302
20160919 12:26:23-!- travis-ci [~travis-ci@ec2-54-242-255-5.compute-1.amazonaws.com] has left #wesnoth-dev []
20160919 12:27:39-!- travis-ci [~travis-ci@ec2-54-161-106-73.compute-1.amazonaws.com] has joined #wesnoth-dev
20160919 12:27:47< travis-ci> gfgtdf/wesnoth-old#696 (fix_24927 - 8508b42 : gfgtdf): The build is still failing.
20160919 12:27:47< travis-ci> Build details : https://travis-ci.org/gfgtdf/wesnoth-old/builds/161008333
20160919 12:27:47-!- travis-ci [~travis-ci@ec2-54-161-106-73.compute-1.amazonaws.com] has left #wesnoth-dev []
20160919 12:33:45-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has quit [Ping timeout: 250 seconds]
20160919 12:34:43< zookeeper> the UtBS thing via difficulty selection seems to work just fine, and all the hackiness is nicely confined to _main.cfg.
20160919 12:35:11< zookeeper> not that i should be surprised because i had done the same difficulty hack before with the new carryover system
20160919 12:39:44-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20160919 12:40:27-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160919 12:43:04-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Client Quit]
20160919 12:45:46-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160919 12:46:08-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Client Quit]
20160919 12:47:53-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160919 12:51:52-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20160919 12:51:59< matthiaskrgr> heh
20160919 12:52:00< matthiaskrgr> /home/matthias/vcs/github/wesnoth/src/game_initialization/multiplayer_connect.cpp:134:1: note: variable tracking size limit exceeded with -fvar-tracking-assignments, retrying without
20160919 12:52:17< irker644> wesnoth: Charles Dang wesnoth:master 3e90310038b2 / src/save_index.cpp: Fixed warning about no return https://github.com/wesnoth/wesnoth/commit/3e90310038b22942ad2b81b1522fe359b7fa7c0a
20160919 12:55:33-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160919 12:59:26-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20160919 12:59:40-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160919 12:59:49-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20160919 13:07:12-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20160919 13:08:58-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160919 13:09:45< matthiaskrgr> wasn't there some kind of internal tests ?
20160919 13:09:57< vultraz> hm?
20160919 13:10:04< matthiaskrgr> some test scenary
20160919 13:10:07< vultraz> of what kind?
20160919 13:10:13< matthiaskrgr> scenario
20160919 13:10:29-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has joined #wesnoth-dev
20160919 13:10:36< matthiaskrgr> it tested different kind of events and dialogs etc iirc
20160919 13:10:40< vultraz> there are test scenarios, yes
20160919 13:10:48< vultraz> wesnoth.exe -t
20160919 13:11:14< matthiaskrgr> ah yes, thanks 
20160919 13:11:34-!- Kwandulin [~Miranda@p200300760F2C712FFC9FB41C890313DA.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160919 13:25:22-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has joined #wesnoth-dev
20160919 13:26:59-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has quit [Client Quit]
20160919 13:31:17-!- travis-ci [~travis-ci@ec2-54-158-59-140.compute-1.amazonaws.com] has joined #wesnoth-dev
20160919 13:31:18< travis-ci> wesnoth/wesnoth#11078 (master - 3e90310 : Charles Dang): The build was fixed.
20160919 13:31:18< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/161022846
20160919 13:31:18-!- travis-ci [~travis-ci@ec2-54-158-59-140.compute-1.amazonaws.com] has left #wesnoth-dev []
20160919 13:31:26< vultraz> :D
20160919 13:31:31< mattsc> yay
20160919 13:31:49< mattsc> zookeeper, DeFender1031, anybody who might have an opinion: I think that vision through tunnels is taken care (except for a minor bug that I still need to take care of).
20160919 13:32:17< zookeeper> great
20160919 13:32:33< mattsc> Do you think that anything else should be added?  I could imagine zoc at most, I don’t really want to add too many options, unless there’s a real use for them.
20160919 13:32:53< zookeeper> well ZoC is one possibility, yes
20160919 13:33:00< zookeeper> but i dunno
20160919 13:33:06< vultraz> zoc through tunnels? 
20160919 13:33:16< zookeeper> ZoC is similar to ability auras and such, and we're probably not gonna have those, so...
20160919 13:33:16< DeFender1031> i'd say zoc and ability to attack through as well
20160919 13:33:26< zookeeper> ability to attack through is madness
20160919 13:33:29< vultraz> oooo
20160919 13:33:37< vultraz> wesnoth's first proper ranged attacks! :D
20160919 13:33:43< DeFender1031> vultraz, hahaha
20160919 13:34:08< DeFender1031> zookeeper, i can think of a lot of cases where people would want that
20160919 13:34:11< mattsc> DeFender1031: yeah, I’d rather not deal with attack through tunnels, although the idea certainly sounds like fun.  :)
20160919 13:34:42< mattsc> I’m also not sure whether zoc really makes all that much sense.
20160919 13:34:45< zookeeper> DeFender1031, yeah but the question of animations alone is basically unsolvable.
20160919 13:34:53< vultraz> honestly, zoc doesn't make much sense
20160919 13:35:24< vultraz> just because point a connects to point c doesn't mean point b doesn't exist.
20160919 13:35:37-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has joined #wesnoth-dev
20160919 13:35:48< zookeeper> i can imagine that someone somewhere might want ZoC, but i doubt it's _necessary_ for any usecase really
20160919 13:35:48< DeFender1031> btw, i was also thinking that just like there's [tunnel] that makes non-adjacent spaces act adjacent, it might be nice to add something that would make adjacent spaces behave non-adjacent (with various options for what it doesn't allow). Would be useful for creating things such as cliffs and fences.
20160919 13:36:12< vultraz> ah, cliffs...
20160919 13:36:22< vultraz> the eternal art nightmare
20160919 13:36:36< vultraz> for years, people have tried and failed to give us cliffs
20160919 13:36:39< vultraz> for some reason
20160919 13:36:45< vultraz> no idea why :P
20160919 13:37:20< vultraz> the solution is simply to make them unwalkable
20160919 13:37:22< mattsc> DeFender1031: yeah, fences or gates that don’t take up a full hex to actually act as such would be nice.
20160919 13:37:28< zookeeper> none of the capable artists simply never decided to do them
20160919 13:37:32< DeFender1031> vultraz, i would humbly submit that the need for cliffs to be considered impassable rather than just blocking direct movement between adjacent hexes might have something to do with it.
20160919 13:38:01< tad_> vultraz: Observation: For the past couple of days my morning sync/make dies on lobby.cpp due to memory exhaustion.
20160919 13:38:01< vultraz> not.. really?
20160919 13:38:02< vultraz> just make them unwalkable
20160919 13:38:14< DeFender1031> it's hard to make passable cliffs if they need to take up an entire hex
20160919 13:38:22< DeFender1031> much easier if it's just a border
20160919 13:38:44< zookeeper> art-wise yeah but not otherwise
20160919 13:38:53< DeFender1031> ah yes, unwalkable, not impassable. but same difference. my point is the "full hex vs. border" idea.
20160919 13:39:06< vultraz> just make them a full hex but unwalkable
20160919 13:39:10< DeFender1031> zookeeper, i'm talking art-wise
20160919 13:39:13< zookeeper> the only sane way to make cliffs is for them to take up the whole hex, the game is not equipped to handle stuff between hexes in any fashion
20160919 13:39:37< tad_> Any, yet we have tunnel which does the opposite.
20160919 13:39:46< vultraz> between hexes?
20160919 13:39:52< vultraz> why does it need to be between hexes
20160919 13:40:18< zookeeper> he just said why
20160919 13:40:24< vultraz> blah
20160919 13:40:28< vultraz> in a game of dota
20160919 13:40:33< vultraz> i shall get back to thuis
20160919 13:41:06< zookeeper> and yes north-side cliffs would be awkward to draw so they take up at least most of the hex
20160919 13:41:10< DeFender1031> okay, the issue is getting confused. let me try again to clarify. 1: the game has no way to make certain hex borders unwalkable while the hexes themselves are some type of normal walkable terrain 2: this means that any cliff art would need to be full-hex and not just on a hex border. 3: i humbly submit that it's more easy to make convincing cliffs that look good as a hex border rather than as a full hex.
20160919 13:41:12< zookeeper> but that can't really be helped
20160919 13:42:01< DeFender1031> i would also sumbit that the chasm borders would make convincing rock-faces for cliffs as well.
20160919 13:42:18< zookeeper> DeFender1031, i doubt anyone disagrees with any of your points
20160919 13:42:44< DeFender1031> (meaning, the art already exists for something else, but can't currently be used as such due to the game being all-or-nothing in terms of hex contiguity)
20160919 13:43:15< zookeeper> the chasm edges wouldn't work as-is anyway because they obviously fade to black
20160919 13:43:40-!- Bonobo [~Bonobo@2001:44b8:254:3200:d0c5:5136:c250:14b1] has quit [Quit: Leaving]
20160919 13:43:43< zookeeper> someone would have to do some serious work with them for that purpose... i know, i've tried :p
20160919 13:45:56-!- DeFender1031 [~DeFender1@46-116-114-128.bb.netvision.net.il] has quit [Read error: Connection reset by peer]
20160919 13:46:39< mattsc> Coming back to tunnels for a moment with one more question: Assume I have fog in a scenario, an allied side that does not share vision, and an allied unit on the exit hex (that I cannot see because of the fog) and passing through allied units disabled.
20160919 13:46:56-!- DeFender1031 [~DeFender1@89-139-252-125.bb.netvision.net.il] has joined #wesnoth-dev
20160919 13:47:04< mattsc> The expected behavior should be that the path through the tunnel looks possible, but then fails when I try to execute it, right?
20160919 13:47:04< DeFender1031> [9/19/2016 5:12:11 pm] <DeFender1031> zookeeper, i didn't think anyone did. It just seemed like I hd presented the point in a confusing manner, so I wanted to clarify step-by-step.
20160919 13:47:06< DeFender1031> [9/19/2016 5:12:39 pm] <DeFender1031> zookeeper, fine, the fade-to-black could be replaced with a fade-to-transparent.
20160919 13:47:08< DeFender1031> [9/19/2016 5:13:16 pm] <DeFender1031> but it's not like cliffs would need to be created from scratch
20160919 13:47:53< DeFender1031> also, the fade-to-black looks to me like a fade to shadow at the base of a cliff followed by a fade out of the shadow to whatever the terrain actually is
20160919 13:48:51< mattsc> That situation is the only I have found (so far) where the new behavior is not what it should be.
20160919 13:48:51< zookeeper> mattsc, uh...
20160919 13:49:01< tad_> DeFender1031: It may be easier to sell the idea if you could pull together some proof-of-concept artwork.
20160919 13:49:04< mattsc> And unfortunately it is not easy to fix.
20160919 13:49:34< mattsc> zookeeper: yes?
20160919 13:49:37< zookeeper> DeFender1031, fade-to-transparent would look awful
20160919 13:49:47< DeFender1031> tad_, good point. i'll do it when I next get a chance, though it'd be useless without the ability to block movement between certain adjacent hexes.
20160919 13:49:49< zookeeper> mattsc, i'm still thinking :p
20160919 13:50:25< DeFender1031> zookeeper, i don't think you understand what i meant, but tad_'s right. I'll throw together some POC.
20160919 13:51:00< DeFender1031> mattsc, there's also ambush which is a similar case, no?
20160919 13:51:04< tad_> DeFender1031: I think people can handle "Just imagine you can't cross this line." but it's hard to imagine how you'd solve the problems of fade-to-transparent or however you end up visually representing a border barrier
20160919 13:51:17< mattsc> DeFender1031: yes, but ambush works because that’s enemy units.
20160919 13:51:19< zookeeper> mattsc, vision normally extends one hex past where you can move. so if vision through tunnels worked exactly as it does elsewhere, then you should see the exit hex as long as you can move onto the enter hex.
20160919 13:51:27< zookeeper> whether tunnels should work like that in that case, i'm not sure
20160919 13:51:53< mattsc> zookeeper: oh, yes, I forgot to say that vision through tunnels must also be disabled for this.
20160919 13:51:54< tad_> Make it an option?
20160919 13:52:01< mattsc> If it is enabled, all works fine.
20160919 13:52:29< mattsc> zookeeper, tad_: it is an option now.  That was part of the point of PR 790.
20160919 13:52:51< mattsc> And it works just as one would expect, except for that one case I describe up there.
20160919 13:52:53< DeFender1031> mattsc, then i'd think movement would stop.
20160919 13:53:10< zookeeper> mattsc, humm. okay, in that case i think it just needs to fail when you try to execute it.
20160919 13:53:11< mattsc> DeFender1031: yes, so do I.
20160919 13:53:21< mattsc> Right.
20160919 13:53:25< zookeeper> i'm not sure how for example the old teleport ability implementation used to deal with that situation
20160919 13:53:42< DeFender1031> mattsc, that's the only sane behavior. the question is how much is the player told about why?
20160919 13:53:45< tad_> "badly"?
20160919 13:53:47< zookeeper> actually i guess it didn't need to since you'd see through for that you didn't own the village
20160919 13:53:52< zookeeper> s/for/fog
20160919 13:53:53< mattsc> The old implementation was much simpler in this case.  It would just fail when the exit hex was occupied.
20160919 13:55:10< mattsc> As for what the player is told, I’ll reinstate the “teleport failed” message.
20160919 13:56:05< mattsc> The problem with the new implementation is that we now need to check in the move execution whether or not the move through the unit should be possible for the tunnel.
20160919 13:56:29< mattsc> Because for some tunnels it should, for other it should not.
20160919 13:56:43< mattsc> And that is information that is currently not available in the move execution code.
20160919 13:57:00< zookeeper> i was wondering whether the move could just show as impossible, as in you can't even try (just like you can't try to move into cavewall)
20160919 13:57:02< tad_> Question: how does this effect enter_hex events?
20160919 13:57:26< zookeeper> it wouldn't
20160919 13:57:57< mattsc> zookeeper: well, that would be easy to code, but it’s inconsistent with other behavior
20160919 13:58:17< mattsc> invisble units are not taken into account when figuring out where to move to for everything else
20160919 13:58:49< zookeeper> true, but a unit in fog is a bit different
20160919 13:59:24< mattsc> how so?
20160919 13:59:34< tad_> it's visible
20160919 13:59:48< tad_> Just not until you move near it
20160919 13:59:55< zookeeper> well units in fog are in all other situations irrelevant to movement, because you can't move to or next to fog
20160919 14:00:02-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20160919 14:00:17< mattsc> Well, that’s the whole problem with not enabling vision through tunnels.
20160919 14:00:51< mattsc> It doesn’t really make sense, afaiac, but it was an option we decided to keep around because that was the old behavior.
20160919 14:01:15< tad_> So I'm a Silver Mage and I'm within range of a village. 
20160919 14:01:28< mattsc> when vision is enabled, all of this is possible.
20160919 14:01:37-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160919 14:01:41< mattsc> or in other words, none of it is a problem.
20160919 14:02:01< zookeeper> it's really tricky to come up with a good one consistent way of doing things since naturally the movement/zoc/vision rules are fundamentally incompatible with the idea of teleportation
20160919 14:02:10< mattsc> It’s just the outside case that both vision and moving through allied units on exit hexes is disabled, that it becomes an issue.
20160919 14:02:17< zookeeper> i mean, you just can't have teleport work "just like a normal move" in all cases
20160919 14:02:24< mattsc> But unfortunately, that was the old default behavior.
20160919 14:03:19< mattsc> zookeeper: agreed; but I think the behvaior needs to be consistent within teleport moves, at least
20160919 14:03:19< tad_> So step back and re-think.
20160919 14:03:55-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Read error: Connection reset by peer]
20160919 14:04:25-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160919 14:04:38< mattsc> tad_: that’s what I did (and a lot of testing).  And the question I asked at the very beginning of this was the one inconsistency I came up with.
20160919 14:05:45< zookeeper> mattsc, well, as said i think your original suggestion is the right way. in any case, the important thing is that the functionality is right; whether some rare corner case prevents your movement by making you unable to try it or by popping up a notice when you try it is something that's less crucial and more open to change if necessary.
20160919 14:07:32< tad_> So, I can see past the tunnel entrance, hence can see the exit. If the tile past the entrance would be no-fog, the exit would be. If there is a Ranger lurking, I would be stopped but I
20160919 14:07:32-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Read error: Connection reset by peer]
20160919 14:08:03< tad_> I don't see a problem. The Ranger becomes visible but can't hit me because he can't use the tunnel
20160919 14:08:15-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160919 14:08:23< mattsc> tad_: yes, all of that works like that
20160919 14:08:55< tad_> So I must not be understanding the issue.
20160919 14:08:58< mattsc> zookeeper: agreed; the problem is that in that case, movement through the tunnel is possible when it shouldn’t be.
20160919 14:09:00-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20160919 14:09:13< zookeeper> the ranger stops you only when you try to move through the tunnel, not when you step on the entrance hex, right?
20160919 14:09:15< mattsc> tad_: yeah, sorry, I was going to explain it again, I was just replying to zookeeper first
20160919 14:09:17-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160919 14:09:20-!- DeFender1031 [~DeFender1@89-139-252-125.bb.netvision.net.il] has quit [Quit: I'm not back now.]
20160919 14:09:27< mattsc> zookeeper: yes
20160919 14:09:35< mattsc> tad_: so the issue is this:
20160919 14:10:01 * tad_ peers at his coffee and decides that the lack thereof may be the true issue.
20160919 14:10:09< mattsc> The [tunnel] tag has an option that, if set, prevents you from moving through _allied_ units on the exit hex.
20160919 14:10:33< mattsc> (whether it makes sense to have that option is a different question, it needs to be there for backward consistency)
20160919 14:10:55< tad_> ok. accepted and no, it makes no sense.
20160919 14:11:04-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Read error: Connection reset by peer]
20160919 14:11:06 * mattsc is on his second cup of large coffee
20160919 14:11:07-!- fabi_ [~fabi@x2f7f108.dyn.telefonica.de] has joined #wesnoth-dev
20160919 14:11:23 * zookeeper has never had a cup of coffee
20160919 14:11:26< mattsc> or maybe large cup of coffee?  apparently I have not had enough yet either
20160919 14:11:53-!- fabi [~fabi@176.7.128.129] has quit [Ping timeout: 265 seconds]
20160919 14:11:55< zookeeper> if i make it through life without having one, i get a rare achievement
20160919 14:12:00 * mattsc has lots of vices; such as a computer game coding addiction
20160919 14:12:28< mattsc> tad_: kay, back to the case
20160919 14:12:47< mattsc> There’s also an option to disable vision through tunnels
20160919 14:12:55< tad_> good
20160919 14:13:31< mattsc> So, now if you can see what’s on the other side of the tunnel (vision enabled) or if there is an enemy unit there (whether visible or not), everything works as expected.
20160919 14:13:34-!- Shiki [~Shiki@141.57.58.245] has joined #wesnoth-dev
20160919 14:14:14< mattsc> The case I was describing is if you have fog, an allied unit that does not share vision and a unit from that ally on the exit hex (all at the same time).
20160919 14:14:17-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160919 14:14:43< mattsc> If you also have vision disabled, you cannot see that ally on the exit hex (assuming it’s far enough from any of your units)
20160919 14:15:12-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has quit [Quit: .]
20160919 14:15:20< mattsc> In that case, the pathfinder code should show you that the move is possible, but then fail when you try to execute the move.
20160919 14:15:29< mattsc> But only in that case.
20160919 14:15:57< mattsc> If you know the unit is there, the pathfinder code should not show the move as possible (and it does not)
20160919 14:16:22< mattsc> And if passing through allies is enabled, it should just move through, whether it can see the unit before moving or not (and that’s what it does).
20160919 14:16:27< tad_> I would suggest the pathfinder should take you to the entrance hex and fail at that point.
20160919 14:16:45-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Ping timeout: 265 seconds]
20160919 14:16:55< mattsc> tad_: good, that’s what I suggested to, and I just wanted to know if people agree  :)
20160919 14:17:17< mattsc> And that’s the one case where it does not do that at the moment (it moves through the unit)
20160919 14:17:19< tad_> The edge case is then if the movement points expend ON the exit hex.
20160919 14:17:29< zookeeper> i suppose none of that can make the AI explode or anything?
20160919 14:17:49< mattsc> zookeeper: not that I am aware of.  Not differently than any ambusher unit can, at least.
20160919 14:18:13< zookeeper> well that's good
20160919 14:18:17< mattsc> tad_: that’s not really handled any differently though, that’ll work as expected once the base behavior is done (I have tested that)
20160919 14:19:38< tad_> You might want to re-cycle the movement logic once you 'stop' at the entrance hex, though. In case that hex is now within normal vision range ...
20160919 14:19:50< tad_> before doing an error popup
20160919 14:20:04-!- Shiki [~Shiki@141.57.58.245] has quit [Quit: Verlassend]
20160919 14:20:32< tad_> and it might be nice, if you have to do an error popup to punt and pass the condition to the designer .. [else][message] sorta thing
20160919 14:21:14< mattsc> tad_: that last thing is something that could be done, yes.  That’s a good idea.
20160919 14:21:33< tad_> 'if [else] appears, do it, otherwise popup'
20160919 14:22:09< mattsc> tad_: that can probably be done with a normal WML event though.
20160919 14:22:22-!- DeFender1031 [~DeFender1@89-139-252-125.bb.netvision.net.il] has joined #wesnoth-dev
20160919 14:22:40< tad_> perhaps. but then we'd be asked why other places don't do that event also.
20160919 14:23:43-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20160919 14:23:47< tad_> actually, no .. do the event
20160919 14:24:28-!- DeFender1031 [~DeFender1@89-139-252-125.bb.netvision.net.il] has quit [Client Quit]
20160919 14:24:47< mattsc> tad_: I am not sure I understand what you are saying.
20160919 14:25:15< mattsc> But in any case, I have very little time left for Wesnoth before the end of the month (which is when 1.13.6 is supposed to be released)
20160919 14:25:33< mattsc> So I am going to try to fix the case I described, and that’s likely it for this release from my side.
20160919 14:25:47< tad_> I was thinking it terms of imperative WML doing the move. But this is UI action. No popup and issue an event. If the designer wants a popup they can trap the event
20160919 14:26:11< mattsc> If there’s interest in other case/options/whatever, we can pick up things later again.
20160919 14:26:17-!- Kwandulin [~Miranda@p200300760F2C712FBD836AADCF9AD8AD.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160919 14:27:02< vultraz> now zookeeper has an embargo on coffee too? D:
20160919 14:27:28< mattsc> tad_: I see
20160919 14:27:50-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has quit [Ping timeout: 265 seconds]
20160919 14:27:56< matthiaskrgr> weird
20160919 14:28:26< matthiaskrgr> when I search for memleaks in a multiplayer game of AIs, nothing at all is found
20160919 14:28:34< mattsc> tad_, zookeeper: so I am going to fix that behavior.  Which means I need to make the [tunnel] information available in move.cpp.
20160919 14:28:39< matthiaskrgr> it is like the leak sanitizer is silently disabled or something like that 
20160919 14:28:51< mattsc> Which, I think, is not really a problem, it just needs to be done.
20160919 14:28:59< vultraz> or maybe there are no leaks? :P
20160919 14:29:22 * tad_ guffaws at THAT silly idea 
20160919 14:29:35< mattsc> Everything else works either as before or as “one” (=me) would expect.
20160919 14:29:59< tad_> It's not a 'leak' if its holding onto the data and simply never using it.
20160919 14:30:07< vultraz> tad_: for the record, I am currently working with that connect engine code that caused you so much grief 
20160919 14:30:09< matthiaskrgr> vultraz: hmm. I would find that quite surprising :P 
20160919 14:30:22< matthiaskrgr> unless the game does something really funky to prevent leaks 
20160919 14:30:33< tad_> vultraz: ???
20160919 14:30:47< vultraz> remember that user_team_name thing
20160919 14:30:57< vultraz> yeah, i'm working with that mess :/
20160919 14:31:04< tad_> Oh. OK.
20160919 14:31:46< tad_> matthiaskrgr: I think the leak prevention logic is: Leak? What leak, we don't have no stinkin' leaks? (Alfred E Neuman)
20160919 14:33:19-!- DeFender1031 [~DeFender1@89-139-252-125.bb.netvision.net.il] has joined #wesnoth-dev
20160919 14:33:47< tad_> vultraz: Did you catch my comment when I came in: every morning when I do my sync-to-master and make I get a memory exhausted error from gcc on lobby.cpp and have to restart make.
20160919 14:34:07< vultraz> yeah, no idea what the problem is
20160919 14:34:19< matthiaskrgr> I can do stuff before loading a game which should create other leaks. but when I then load a MP game with AI and quit mid game, even that other leak is gone o.O
20160919 14:36:21< matthiaskrgr> oh well 
20160919 14:37:46-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: mattsc]
20160919 14:38:05< tad_> vultraz: strange thing is all I need do is re-start make. I don't need to close anything to free memory.
20160919 14:39:23< matthiaskrgr> gcc is running out of memory? 
20160919 14:40:08< tad_> zookeeper: I'm looking at THoT and the logic involving recruiting Mages. Any objection to a rewrite of the logic which takes the ability away at the point the last mage dies rather than waiting until start of next scenario?
20160919 14:40:17< tad_> matthiaskrgr: yes
20160919 14:40:52< tad_> [ 86%] Building CXX object src/CMakeFiles/wesnoth.dir/gui/dialogs/lobby/lobby.cpp.o virtual memory exhausted: Cannot allocate memory
20160919 14:47:17< matthiaskrgr> oh, interesting. the leak reporting breaks in the moment where I leave the wesnoth window unfocussed
20160919 14:47:30< matthiaskrgr> when I have it in focus / on the screen the entire time, it works 
20160919 14:59:53-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20160919 15:05:55-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has joined #wesnoth-dev
20160919 15:06:42< irker644> wesnoth: Charles Dang wesnoth:master 2dd8cfdbdcdb / data/gui/schema.cfg src/gui/widgets/control.cpp src/gui/widgets/control.hpp: GUI2: added flag to enable markup from WML https://github.com/wesnoth/wesnoth/commit/2dd8cfdbdcdb7027cb7684afcaac80b4d2b350f1
20160919 15:07:11< zookeeper> tad_, i'm not familiar with it, but by looking at the code that seems like a reasonable change
20160919 15:08:05< tad_> zookeeper: OK. I noticed it because it was missed on S11 and the inconsistency made me look deeper.
20160919 15:19:19-!- Shiki [~Shiki@141.57.58.245] has joined #wesnoth-dev
20160919 15:27:02< tad_> Can I [filter] for trait=loyal?
20160919 15:35:45-!- esr [~esr@static-71-162-243-5.phlapa.fios.verizon.net] has quit [Ping timeout: 250 seconds]
20160919 15:42:05< zookeeper> tad_, yes, with [filter_wml] [modifications] [trait] id=loyal
20160919 15:43:05< tad_> Thank you, I was slowly getting there.
20160919 15:51:52-!- Shiki [~Shiki@141.57.58.245] has quit [Remote host closed the connection]
20160919 15:52:40< zookeeper> tad_, ok so i can merge #734 now... the small_number/large_number formulas look so heavy though that i must ask whether you're really sure you tested that they work right? :>
20160919 15:56:35< zookeeper> also i'm not really sure whether the whole idea makes much sense since if you're able to kill off leaders in S1 then you don't really need any extra help and might even _prefer_ to have more enemies to harvest XP from.
20160919 15:56:36< zookeeper> but for that same reason i don't have any objections, since killing more than 1 of the leaders is pretty far-fetched.
20160919 15:59:10-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has joined #wesnoth-dev
20160919 16:06:48-!- boucman_work [~boucman@193.56.60.161] has quit [Remote host closed the connection]
20160919 16:09:17-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160919 16:09:56-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20160919 16:10:10-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Read error: Connection reset by peer]
20160919 16:10:32-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160919 16:13:05< tad_> zookeeper: Up to you. The original TODO says "two or more" and I can change it back to that pretty easy instead of "any"
20160919 16:13:57< tad_> zookeeper: I tested them for 0, 1 and 3 kills on hard and presumed they worked correctly for all other cases.
20160919 16:15:22< celticminstrel> There is only one other case, right? Or two possibly?
20160919 16:15:49< tad_> celticminstrel: '2' on hard, and [0-3] on easy and medium
20160919 16:16:03< celticminstrel> Right right, multiply by difficulties, okay.
20160919 16:17:05-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160919 16:17:28< tad_> I liked the WML because it was clearly going to work. The expression form should and is easier to read (Except that ?0| bit) but I have problems with stuff like that because it's "string replacement, then evaluation" and I get that wrong at times
20160919 16:18:50< celticminstrel> I'm trying to figure out how the map switching code was ever reached in the editor... given that it seems to be handled under HOTKEY_NULL, rather than HOTKEY_EDITOR_MAP_SWITCH.
20160919 16:18:52< tad_> I tested and am confident it does what it should, and nothing else. But the question zookeeper had was "any" vs "two or more" vs not bothering at all
20160919 16:19:47< celticminstrel> Ack, Colloquy really needs a reboot.
20160919 16:19:52< vultraz> blah, I can't get this dialog to work :/
20160919 16:19:58-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has quit [Quit: KABOOM! It seems that I have exploded. Please wait while I reinstall the universe.]
20160919 16:20:02< vultraz> the leaders don't show up
20160919 16:20:17< vultraz> and the games immediately terminate 
20160919 16:21:07-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has joined #wesnoth-dev
20160919 16:21:12< tad_> I get that when I forget to set controller=human on one 
20160919 16:21:34< celticminstrel> I didn't actually see that last thing you said, was it important?
20160919 16:21:38< vultraz> no
20160919 16:21:39-!- Kwandulin [~Miranda@p200300760F2C712FBD836AADCF9AD8AD.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160919 16:21:50< vultraz> just me lamenting i can't get the staging dialog to actually start a game :/
20160919 16:21:58< celticminstrel> Oh.
20160919 16:22:12< vultraz> i mean, it starts
20160919 16:22:14< vultraz> but with no leaders
20160919 16:22:24< celticminstrel> Now I have a clean build to debug through and see if I can locate where the control flow went wrong...
20160919 16:24:23< irker644> wesnoth: Gregory A Lundberg wesnoth:master 0493c1eba40b / data/campaigns/Heir_To_The_Throne/scenarios/ (01_The_Elves_Besieged.cfg 07_Crossroads.cfg 25_HttT_Epilogue.cfg): HttT S07 Ambushers reduced https://github.com/wesnoth/wesnoth/commit/0493c1eba40b6ea06aa077fb609e4d52f0c7c569
20160919 16:24:25< irker644> wesnoth: Lari Nieminen wesnoth:master 2970360d49e8 / data/campaigns/Heir_To_The_Throne/scenarios/ (01_The_Elves_Besieged.cfg 07_Crossroads.cfg 25_HttT_Epilogue.cfg): Merge pull request #734 from GregoryLundberg/GL_HttT_S07_ambushers_reduced https://github.com/wesnoth/wesnoth/commit/2970360d49e88b7a2892639fd382ffc3ca584062
20160919 16:25:13< vultraz> celticminstrel: btw, i added a use_markup flag to wml
20160919 16:25:22< celticminstrel> I saw that.
20160919 16:25:42< celticminstrel> Gah, did the debugger hang again. :|
20160919 16:26:48< vultraz> my eyes hurt :|
20160919 16:26:54< celticminstrel> Poor eyes.
20160919 16:27:24< celticminstrel> Wait, why do I still see the Wesnoth window when I killed it.
20160919 16:28:06< vultraz> why does the chatbox require lobby_info..
20160919 16:28:14< celticminstrel> Ah, there we go.
20160919 16:28:23< celticminstrel> Maybe for rooms?
20160919 16:28:41< vultraz>  * This class represents the collective information the client has
20160919 16:28:43< vultraz>  * about the players and games on the server
20160919 16:28:48< vultraz> then why is it called *lobby* info :|
20160919 16:29:09< celticminstrel> Don't ask me why things are named the way they are, unless you're absolutely sure I added it.
20160919 16:30:08< vultraz> and the only lobby info is in enter_lobby_mode..
20160919 16:30:20< vultraz> god, I cannot wait to clean up this file once the gui1 version is dropped
20160919 16:30:43< vultraz> ok ya know what, I'm just going to commit what I have so far
20160919 16:30:49< vultraz> and let gfgtdf play around with it
20160919 16:30:50< celticminstrel> The lobby_info can probably be passed through somehow to Connect, right?
20160919 16:30:59< celticminstrel> Uhh, hold on, you're not pushing broken code, are you?
20160919 16:31:33< vultraz> well you can't start a game 
20160919 16:32:13< celticminstrel> Exactly my point.
20160919 16:32:27< celticminstrel> If it's a branch, that's fine.
20160919 16:33:08< celticminstrel> I guess it's not exceptionally bad for master since it can be disabled in advanced prefs (provided the GUI1 connect still works), but it's not great.
20160919 16:33:08< vultraz> :/
20160919 16:33:15< vultraz> yes, that
20160919 16:33:36< celticminstrel> That's still "it's not a good idea" you know.
20160919 16:33:51< vultraz> didn't I do that with Create
20160919 16:34:10< celticminstrel> Precedent doesn't make it a good idea.
20160919 16:34:18< celticminstrel> (I don't remember if you did, possibly though.)
20160919 16:34:47< vultraz> I just want to make it easy for other people to work on it
20160919 16:35:19< celticminstrel> Stepping through code in gdb / lldb on the Mac is such a pain. It keeps jumping into std:: code even when I don't tell it to, and if I try to step out, it steps out of the function I was in instead of the std:: code.
20160919 16:35:20< tad_> They can pull your branch. Heck, they can pull from your private fork on github.
20160919 16:35:34< celticminstrel> ^
20160919 16:36:33< vultraz> would really rather have it in master
20160919 16:36:51< celticminstrel> That also doesn't make it a good idea.
20160919 16:37:52-!- ChipmunkV [~vova@static-89-94-113-91.axione.abo.bbox.fr] has joined #wesnoth-dev
20160919 16:41:54< vultraz> it's horribly inconvenient for me to work on this in a branch since then i cannot push to master
20160919 16:42:22< celticminstrel> Isn't the whole point that you shouldn't push to master?
20160919 16:42:52< vultraz> other things
20160919 16:42:56< irker644> wesnoth: Charles Dang wesnoth:master dba4bd78b3b0 / data/gui/widget/slider_default.cfg: Added short slider definition https://github.com/wesnoth/wesnoth/commit/dba4bd78b3b0d75fd89b05e3113cd68a925dc359
20160919 16:42:59< irker644> wesnoth: Charles Dang wesnoth:master 8609955b74cf / data/gui/widget/button_25x25.cfg: Refactored square/minimal button definition code https://github.com/wesnoth/wesnoth/commit/8609955b74cf0da057c7bed2e5a314467b92b3d5
20160919 16:43:16< celticminstrel> Well, you could always make a copy of your wesnoth folder if you want to work on several things at once.
20160919 16:43:42< vultraz> purge the thought!
20160919 16:43:48< celticminstrel> Why?
20160919 16:44:05-!- hk238 [~kvirc@unaffiliated/hk238] has joined #wesnoth-dev
20160919 16:44:05< vultraz> i do not want to deal with git-new-work-dir
20160919 16:44:10< celticminstrel> Huh?
20160919 16:45:04< celticminstrel> Well anyway, you can also switch away from the branch, push other stuff, and switch back.
20160919 16:45:07< vultraz> it doesn't work on windows, you have to run a special script, dunno if i have it around anymore, not to mention managing two folders and juggling compiler sessions 
20160919 16:45:20< celticminstrel> What doesn't work?
20160919 16:45:29< vultraz> git-new work dir
20160919 16:45:44< celticminstrel> Uh. What's wrong with just copying the folder in Explorer?
20160919 16:45:49< vultraz> and i dont have ccache so switching branches is a pain 
20160919 16:46:26< vultraz> so one can avoid duplicating the whole multi-gb repo? :|
20160919 16:46:53< vultraz> do you even VCS
20160919 16:47:06< celticminstrel> Well, okay, I guess that's understandable. It's also solvable though, but I don't recall exactly how to do it.
20160919 16:47:16< vultraz> git new work dir :|
20160919 16:47:28< vultraz> (or whatever the exact name is)
20160919 16:47:47< celticminstrel> I mean it's probably solvable manually as well.
20160919 16:48:51< celticminstrel> The idea would be to use a bare repository (so move wesnoth/.git to wesnoth.git or similar) and then have working copies that reference it. I'm not quite sure how that needs to be set up, though; I think you still need a .git directory...
20160919 16:51:00< celticminstrel> Ah, looks like it can even work if it's not a bare repository...
20160919 16:51:06< celticminstrel> 'git help worktree'
20160919 16:53:44< vultraz> too much work
20160919 16:53:48< celticminstrel> And doesn't rely on symbolic links like git new work dir does.
20160919 16:53:49< celticminstrel> Huh?
20160919 16:53:57< celticminstrel> It's one line. How is that too much work?
20160919 16:54:11< vultraz> I'll just... get the dialog working :|
20160919 17:03:56-!- travis-ci [~travis-ci@ec2-54-242-255-5.compute-1.amazonaws.com] has joined #wesnoth-dev
20160919 17:03:57< travis-ci> wesnoth/wesnoth#11081 (master - 2970360 : Lari Nieminen): The build passed.
20160919 17:03:57< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/161079377
20160919 17:03:57-!- travis-ci [~travis-ci@ec2-54-242-255-5.compute-1.amazonaws.com] has left #wesnoth-dev []
20160919 17:07:12-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has quit [Ping timeout: 240 seconds]
20160919 17:08:11-!- ChipmunkV [~vova@static-89-94-113-91.axione.abo.bbox.fr] has quit [Quit: ChipmunkV]
20160919 17:09:12-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160919 17:10:57-!- ChipmunkV [~vova@static-89-94-113-91.axione.abo.bbox.fr] has joined #wesnoth-dev
20160919 17:11:31-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has quit [Ping timeout: 250 seconds]
20160919 17:11:54-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has joined #wesnoth-dev
20160919 17:15:49< celticminstrel> Whee, reduced 4.1 GB to 681.3 MB thanks to git worktree. \o/
20160919 17:16:03< celticminstrel> (I had a duplicate for the 1.12 branch.)
20160919 17:18:54-!- bumbadadabum changed the topic of #wesnoth-dev to: 1.13.6 planned for late September | Wesnoth Inc. board elections: https://r.wesnoth.org/t44616 | Wesnoth Developers Channel | >>> Want to help? Go here: http://r.wesnoth.org/t42911 (and thanks!) <<< | Logs: http://irclogs.wesnoth.org | Bug tracker: http://bugs.wesnoth.org
20160919 17:19:05< bumbadadabum> corrected the typo
20160919 17:19:14< celticminstrel> Thanks.
20160919 17:19:26< celticminstrel> I'm surprised vultraz didn't do it when I first pointer it out.
20160919 17:19:45< bumbadadabum> I'm surprised you didn't do it if you're pointing it out lol
20160919 17:19:52< celticminstrel> Don't have permissions.
20160919 17:19:55< bumbadadabum> or anyone else for that matter
20160919 17:19:56< bumbadadabum> wait what
20160919 17:20:04< bumbadadabum> why am I allowed to change the topic then
20160919 17:23:07-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160919 17:26:51< vultraz> no idea
20160919 17:26:55< vultraz> what was the typo?
20160919 17:31:24< celticminstrel> Something like "Wensoth" I think.
20160919 17:31:32< celticminstrel> Don't remember exactly.
20160919 17:32:15< matthiaskrgr> wesloth  
20160919 17:40:21< DeFender1031> zookeeper, vultraz, anyone else who was involved in the cliffs discusion: https://drive.google.com/file/d/0B5F6wWfcJTDPb0RPUi1ua0pJcEE/view
20160919 17:40:39< DeFender1031> to me, the black at the bases of the cliffs just looks like shadow
20160919 17:41:45< vultraz> ah, I think i see what the problem with that model is
20160919 17:42:25< DeFender1031> but I think that being able to do stuff like this would make for a lot more interesting maps. Also note the bridge in the top right. Imagine if a tunnel were created from one side to the other, you could have an intricate system of intertwining multi-level paths
20160919 17:42:27< vultraz> because it looks like you should be able to walk to the edge
20160919 17:43:04< vultraz> but i think that could be solved with the right art
20160919 17:43:09-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160919 17:43:26< DeFender1031> vultraz, that's what I was saying before, there should be a way to make hexes which are adjacent by default behave as non-adjacent in the same ways that [tunnel] allows non-adjacent hexes to behave as adjacent.
20160919 17:44:26< DeFender1031> also note what i said about the criscrossing multilevel paths. that isn't currently possible with the "all hex boundaries are the same" model.
20160919 17:45:27-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has quit [Quit: Connection closed for inactivity]
20160919 17:46:52< zookeeper> yeah i'll give you that; if the cliffs were normal terrains, then you couldn't make bridges across the gaps without allowing units to jump down :P
20160919 17:47:19< DeFender1031> vultraz, "solving it with the right art" isn't what i meant. I meant that there should actually be the ability for hexes to only be walkable from particular adjacent hexes.
20160919 17:47:23< DeFender1031> zookeeper, exactly
20160919 17:47:52< DeFender1031> also, this would allow for stuff like units being able to attack each other across cliffs/fences, even if they can't travel that way themselves.
20160919 17:48:17< DeFender1031> I honestly think that allowing this would open up a lot of new level design possibilities.
20160919 17:49:58< DeFender1031> the idea here is that every hex in my screenshot is walkable, and intentionally so, just that the hexes on opposite sides of a cliff wall are not walkable in that particular direction.
20160919 17:50:19-!- hk238 [~kvirc@unaffiliated/hk238] has quit [Quit: http://www.kvirc.net/ 4.9.2 Aria]
20160919 17:50:39-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has joined #wesnoth-dev
20160919 17:51:47-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: So long and thanks for all the fish.]
20160919 17:52:04< celticminstrel> Is this about how a / bridge can be entered from the south even though the bridge doesn't actually connect in that direction? (For example)
20160919 17:52:16 * celticminstrel hasn't clicked the link because Firefox not open
20160919 17:54:07< DeFender1031> celticminstrel, that's somewhat tangentially related
20160919 17:54:14< DeFender1031> but it's more than that
20160919 17:55:18-!- fabi [~fabi@176.7.87.14] has joined #wesnoth-dev
20160919 17:55:34< DeFender1031> i'm saying in general, there should be a way to define boundaries between hexes so that a given hex can only be entered from specific directions but not others (or, more technically, a way to set hexes to have a different movement cost depending on what direction they're entered from).
20160919 17:55:58< tad_> Actually, it's that, but not just about bridges. What he's done is create a maze but instead of having entire tiles be unwalkable he' suggesting that each of the 6 edges be able to be marked uncross-able
20160919 17:56:21< zookeeper> well, that's so fundamentally at odds with how the game works currently that i can't begin to guess how difficult it might be to implement.
20160919 17:56:33< zookeeper> so no input from me due to insufficient data
20160919 17:56:40< DeFender1031> tad_, that's exactly it. though i'd implement it as alternate terrain types instead.
20160919 17:57:19< DeFender1031> zookeeper, it's certainly a big project, but if people like the idea, i'd be willing to start looking into the code for it.
20160919 17:57:20< tad_> The issue is that it means replacing a lot of work. Pathfinding, for insance.
20160919 17:57:24-!- gfgtdf [~chatzilla@x4e368db9.dyn.telefonica.de] has joined #wesnoth-dev
20160919 17:57:53-!- fabi_ [~fabi@x2f7f108.dyn.telefonica.de] has quit [Ping timeout: 240 seconds]
20160919 17:57:54< zookeeper> tad_, you might not like some of my comments on the thief PR :p
20160919 17:58:29< tad_> zookeeper: Actually, I'm sorta expected the general theme to be "rewrite this mess"
20160919 17:58:49< DeFender1031> anyway, what i mean by alternate terrain types is that there would be some way to say that, for example, when entering 11,7 from se, sw, and s, instead of treating the hex as flat, treat it as unwalkable.
20160919 17:59:13< DeFender1031> (but still treat it as flat if entering from elsewhere)
20160919 17:59:16< bumbadadabum> 20160919 19:56:14 error config: Unexpected characters at line start
20160919 17:59:20< bumbadadabum> what does this error mean
20160919 17:59:27< tad_> While I generally agree it would be nice to be able to do, I think it's an awfully large project and needs a lot more thinking.
20160919 17:59:29< zookeeper> tad_, well i wouldn't go that far, i think you just need to rewrite _some_ significant parts of it :>
20160919 17:59:51< DeFender1031> tad_, so let's think. no one's changing anything right now.
20160919 18:00:08< tad_> zookeeper: Only some? I'm getting off easy, then. Give me a bit and I'll look at it.
20160919 18:00:30< zookeeper> tad_, no hurry
20160919 18:00:57< tad_> DeFender1031: My knee-jerk reaction is that it's a V2 size change and not something to drop into a point release.
20160919 18:01:03< bumbadadabum> Caught a string thrown as an exception:
20160919 18:01:04< bumbadadabum> No scenariodata found
20160919 18:01:05< bumbadadabum> and this
20160919 18:01:09< bumbadadabum> what the hell
20160919 18:01:47-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160919 18:02:04-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160919 18:02:05< DeFender1031> tad_, i'm not convinced it would change anything so fundamental.
20160919 18:02:11-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160919 18:02:19< vultraz> gfgtdf: I can't seem to get the new dialog to properly launch a game. there don't seem to be any leaders :/
20160919 18:02:40< tad_> bumbadadabum: the unexpected chars I think i got one time with a missing [ .. actually I think I accidentally deleted the # 
20160919 18:02:56< DeFender1031> tad_, certainly, many algorithms such as pathfinding would need significant amounts of logic added, but i doubt it would require a core rewrite
20160919 18:03:31< DeFender1031> tad_, think of it this way, what i'm proposing is literally the exact opposite of [tunnel]
20160919 18:04:36< DeFender1031> it's saying "these hexes which were previously considered adjacent, treat them as not" rather than "these hexes which were previously considered non-adjacent, treat them as such"
20160919 18:05:05< zookeeper> ok so where and how would those rules be defined for a given terrain?
20160919 18:05:35< tad_> DeFender1031: My concern is jumping into the REALLY deep end. It touches a lot of places. Conceptually, it's not that hard to grasp. But .. you're talking about map editor, pathfinding, sight, shroud, fog, tile definitions .. and that's just off the top of my head.
20160919 18:05:52< zookeeper> are you suggesting that cliffs would be 6 terrains (one for each direction)?
20160919 18:06:13< DeFender1031> zookeeper, i would imagine there'd be some WML tag which was a SLF and also had attributes for what properties of adjacent should be dropped
20160919 18:06:32< zookeeper> or, wait, they wouldn't even be terrains so you need to first figure out _what_ they would be
20160919 18:06:44< DeFender1031> oh, and a SLF for source hex.
20160919 18:07:10< celticminstrel> zookeeper: Even 6 wouldn't be enough if you did it that way...
20160919 18:07:16< zookeeper> celticminstrel, yeah, true
20160919 18:07:21< celticminstrel> You'd need one for every possible combination of the 6 sides.
20160919 18:07:38< bumbadadabum> tad_: yeah but on the line itself it's just     {GENERIC_RUNE_ANIMATION "name=magic"}
20160919 18:07:42< bumbadadabum> and if I comment that out it works
20160919 18:07:44< DeFender1031> tad_, fair enough
20160919 18:07:48< bumbadadabum> it seems like a very weird error to get
20160919 18:07:54< celticminstrel> So uh... how many is that...
20160919 18:08:00< bumbadadabum> plus the game segfaults after I close the error message
20160919 18:08:01< celticminstrel> 2^6, right?
20160919 18:08:03< zookeeper> celticminstrel, eh actually if they were along hex borders then you'd only need 6
20160919 18:08:04< bumbadadabum> which definitely shouldn't happen
20160919 18:08:06< matthiaskrgr> hm
20160919 18:08:10< tad_> bumbadadabum: So look in the macro. Remember {..} is a text replacement 
20160919 18:08:13< DeFender1031> tad_, i'm also not suggesting htis be done now. I'm just brainstorming at the moment.
20160919 18:08:25< matthiaskrgr> I get a static_assert failure when building with clang :/ 
20160919 18:08:56< matthiaskrgr> it's inside boost :( 
20160919 18:08:57-!- mjs-de [~mjs-de@wh.Uni-Dortmund.DE] has joined #wesnoth-dev
20160919 18:08:58< bumbadadabum> tad_: if the problem was in the macro i think it'd say a line number in the macro definition
20160919 18:09:54< bumbadadabum> nvm it was in the macro
20160919 18:10:03< bumbadadabum> hmm I thought it worked differently then I guess
20160919 18:10:05< celticminstrel> matthiaskrgr: That's been known for awhile... :/
20160919 18:10:25< matthiaskrgr> http://pastebin.com/yX6AV80N 
20160919 18:10:34< celticminstrel> We can't figure out how to fix it besides editing the Boost header (which is obviously not great).
20160919 18:10:35< matthiaskrgr> celticminstrel: so the solution is to not compile with DDEBUG I guess? 
20160919 18:10:42< celticminstrel> matthiaskrgr: Maybe?
20160919 18:10:47< celticminstrel> You could try it I guess?
20160919 18:10:47< matthiaskrgr> mmh
20160919 18:11:06< tad_> DeFender1031: Brainstorming if fine and I could see 'selling' the idea. But it's a huge problem. I'd not considered the graph-theory explosion Celmin referred to .. just definingthe graph is a big problem, but you need to walk it .. A-Star should do it but you're talking changing the logic from node to edge and that's not just sit-down-and-do-it-tonight easy.
20160919 18:11:12< DeFender1031> zookeeper, yeah, i need to think more about how the terrain would be defined, that's true.
20160919 18:11:34< matthiaskrgr> ok, let's see ...
20160919 18:13:48< DeFender1031> tad_, agreed. At the moment, all I have is an idea. But I think it's a good one, and if I want to sell the idea, I have to start somewhere. Discussing it here, getting a feel for the challenges involved and brainstorming seems like the logical place to start. The idea could end up going nowhere, but if I don'tbring it up here as I have now done, there is no way it has even a chance of going anywehre.
20160919 18:15:13< zookeeper> DeFender1031, also i believe that we can't do a lot of real-time pathfinding based on SLF, it's a big performance drain and i presume tunnels only work because they're rare, and that you'd get massive slowdowns if you for example had a handful of AI units in a cliff maze where just calculating a unit's potential movement range would mean evaluating dozens of SLF's.
20160919 18:15:18< DeFender1031> zookeeper, correction, you'd only need 6 if terrain allowed for multiple overlays. Without that, you'd need a different code for each possible combination of cliffs.
20160919 18:15:33< DeFender1031> zookeeper, fair enough.
20160919 18:16:19< DeFender1031> like i said, i have to give this more thought. my question at the moment though is are people even open to this idea if I can come up with a sane way of handling it?
20160919 18:16:20< zookeeper> right, with multiple overlays you could do it with 6. painting them in the editor would be a mighty pain though :P
20160919 18:17:00< DeFender1031> zookeeper, true. ideally, i'd prefer a method that allowed painting with a single brush, or at most with three (like rails et al)
20160919 18:18:01< tad_> DeFender1031: From the comments and objections, I think the best course if you insist on pursuing this would be to do a detailed walk-through of how you'd envision editing a map, and how it would play; with an eye toward how you'd describe it to a non-programmer who just wants to create an interesting battlefield.
20160919 18:18:44< zookeeper> DeFender1031, well i think you're perhaps started to scratch the surface of potential issues and questions involved, so hard to say whether i'd be open to it at some later point... :p
20160919 18:20:12< zookeeper> i'd suggest contrasting this with cliffs-as-normal-terrains, where the only question is whether someone will draw the art.
20160919 18:20:48< tad_> Frankly, my knee-jerk reaction is the idea is good but it's not a solo project. I'd think a 4 or 5-man team and a year or two. I don't think he has any idea of the mess he's going to find just looking at the map data structuyres.  I was just there and it's not clear-cut. Look at the issues Mattsc has and 10x them.
20160919 18:20:52< DeFender1031> tad_, writing up just such a proposal would have been my logical next step if I were to pursue this. I just didn't want to put any effort into it if there wasn't any interest here.
20160919 18:22:09< DeFender1031> tad_, hey, you're the one who told me to jump into the deep end the other day :P
20160919 18:23:28< tad_> DeFender1031: I was thinking a swimming pool.  Not the Mariana Trench
20160919 18:23:44< DeFender1031> zookeeper, come to think of it, simply moving the chasm edges down by half a hex would make them full-hex cliffs. On the other hand, I really like the idea of cris-crossing paths with functional bridges and tunnels, which is simply not doable with the current terrain model.
20160919 18:24:10< gfgtdf> vultraz: gfgtdf: I can't seem to get the new dialog to properly launch a game. there don't seem to be any leaders
20160919 18:24:18< gfgtdf> vultraz: ths is about the gui2 mp connect code ?
20160919 18:24:30< vultraz> yeah
20160919 18:25:15< zookeeper> yeah, some things we just can't do. like maps with multiple layers (under and above ground, mainly).
20160919 18:25:29< gfgtdf> matthiaskrgr: its  aknow isue the recomented fix of to remove the BOOST_CONSTEXPR  marks in that file we already filed  a bugreport on boost adn it will be fixed with next boost version.
20160919 18:25:37< gfgtdf> vultraz: hmm you have smaple code ?
20160919 18:25:39< gfgtdf> sample*
20160919 18:26:03< DeFender1031> tad_, fair enough.
20160919 18:26:38< zookeeper> DeFender1031, which reminds me; if we had your little maze with a bridge, how would a unit walk under the bridge? they'd have to move onto the bridge hex from an unallowed direction, and... end up under it. what if there's a unit _on_ the bridge? etc.
20160919 18:26:56< DeFender1031> zookeeper, that's not entirely true. that at least can be simulated by splitting your map in half and using [tunnel] to transport from the top half (above ground) to the bottom half (below)
20160919 18:27:05< vultraz> gfgtdf: https://github.com/Vultraz/wesnoth/commit/81e1cf7cb2308b8fef7b4424aebc0a6498f405eb
20160919 18:27:09< tad_> zookeeper: He's thinking using [tunnel] across the sides of the bridge.
20160919 18:27:29< DeFender1031> zookeeper, yeah, what tad_ said.
20160919 18:27:37< zookeeper> urgh
20160919 18:27:47< DeFender1031> zookeeper, what's the problem?
20160919 18:27:58< tad_> it's ugly. 
20160919 18:28:15< tad_> It works but someone's going to say it's not logical. Andthey're right.
20160919 18:28:24< tad_> Which goes to layered maps.
20160919 18:28:53< zookeeper> well that would be terribly ugly and hacky? i mean it'd just look weird and wrong, there would be one hex which cannot be accessed even though it's just a bridge
20160919 18:29:50< tad_> zookeeper: He'sthinking you walk normally along the bridge and [tunnel] the sides. Still ugly and layers would be better.
20160919 18:30:38< gfgtdf> vultraz: hmm where do you cann engine->start_game() ?
20160919 18:30:52-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160919 18:31:11< DeFender1031> not sure I agree. Games like C&C timerian sun have similar terrain art and basically have tunnels work exactly like this.
20160919 18:31:17< matthiaskrgr> gfgtdf: ok.  a boost person already pointed out some commit in 1.61.0 
20160919 18:31:35< matthiaskrgr> looks like I'll have to fall back to gcc then which is a shame because it's so slow :/
20160919 18:31:42< tad_> DeFender1031: Want something less daunting, but which gets you into the code, and fixes real problems we have now: take a look at game initialization from main() to the Titlescreen where you select Campaign/MP and see if you can get it to fail gracefully when the user closes the window before the game is fully up.
20160919 18:31:43< vultraz> oh, hm...
20160919 18:32:01< gfgtdf> matthiaskrgr: well you can still update the boost header.
20160919 18:32:04< DeFender1031> tad_, that sounds less exciting :P
20160919 18:32:14< DeFender1031> heading out for a bit right now.
20160919 18:32:22< gfgtdf> matthiaskrgr: like aplying the mentioned commit manually or doaing what i said
20160919 18:32:28< matthiaskrgr> uuh 
20160919 18:32:33< vultraz> gfgtdf: oh I see
20160919 18:32:55< gfgtdf> vultraz: you are testing with local mp games ?
20160919 18:33:08< matthiaskrgr> gfgtdf: I'm good at blowing stuff up, you know ... :& 
20160919 18:33:15< vultraz> gfgtdf: yeah i haven't touched the chat box or userlist or network stuff
20160919 18:33:20< vultraz> gfgtdf: was hoping you could help with that maybe
20160919 18:33:50< vultraz> gfgtdf: start game works thanks :D
20160919 18:34:51< irker644> wesnoth: Charles Dang wesnoth:master 9721910d980b / / (12 files in 8 dirs): Initial implementation of GUI2 MP Staging dialog (Connect + Wait) https://github.com/wesnoth/wesnoth/commit/9721910d980b44d1bffae8b424b564465f1a2859
20160919 18:35:11< vultraz> it is very basically functional so celticminstrel can no longer complain
20160919 18:36:13< vultraz> gfgtdf: basically the dialog needs the lobby_info object the lobby had
20160919 18:36:19< vultraz> for the chatbox
20160919 18:36:21< vultraz> and userlist
20160919 18:37:24< vultraz> and it needs to be wired into mp wait
20160919 18:37:31< vultraz> and hook into the wesnothd connection
20160919 18:40:27< vultraz> a lot needs to be donei n general
20160919 18:41:46-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160919 18:44:32-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20160919 18:45:42< zookeeper> tad_, so how do you like the final THoT scenario?
20160919 18:45:52< zookeeper> from a gameplay point of view that is
20160919 18:47:17< vultraz> heh
20160919 18:47:27-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160919 18:47:40< tad_> zookeeper: Even running it with debug .. it's scary. After a few times through I turned off shroud and WOAH all those enemy units pre-set AND leaders making more ..
20160919 18:49:51< vultraz> i dont jthink i made any changes to that scenario
20160919 18:49:59-!- travis-ci [~travis-ci@ec2-54-242-255-5.compute-1.amazonaws.com] has joined #wesnoth-dev
20160919 18:50:00< travis-ci> wesnoth/wesnoth#11085 (master - 8609955 : Charles Dang): The build passed.
20160919 18:50:00< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/161085101
20160919 18:50:01-!- travis-ci [~travis-ci@ec2-54-242-255-5.compute-1.amazonaws.com] has left #wesnoth-dev []
20160919 18:50:28< vultraz> well, besides minor cosmetic ones
20160919 18:51:28< zookeeper> tad_, i don't even know when if ever i've looked at the map of that one. when i look at it now i have no recollection of seeing it before.
20160919 18:51:32< zookeeper> (i've never played through the campaign)
20160919 18:51:59< zookeeper> but yeah it looks about as uninviting as all those comments i've heard over the years would suggest :>
20160919 18:52:33< tad_> zookeeper: I didn't remember it either. Probably I got there, got slammed by the initial guantlet and walked away in digust.
20160919 18:53:03< vultraz> i recall i did make significant improvements to the scenario before it, though
20160919 18:53:21< zookeeper> what's the purpose of all that ugly dwarven-doors-closed.png scenery..?
20160919 18:53:29< vultraz> doors
20160919 18:53:31< vultraz> :P
20160919 18:53:35< tad_> Using debug, when it got rough I added HP or killed some and went on. I typically do left-hand rule so I ended up walkingthe ENTIRE map and going .. "When will this EVER end?"
20160919 18:53:38< zookeeper> doors to nowhere? yeah that's my point :p
20160919 18:53:53< vultraz> move in front of them and the passage opens
20160919 18:53:57< vultraz> we need better graphics
20160919 18:54:16< vultraz> someone ask doofus to draw actual doors.
20160919 18:54:22< tad_> Reading the code .. look at all those preset units. And the comments .. a door you can't open unless you did this, then that.
20160919 18:54:58< zookeeper> vultraz, there's 14 of them littered across the map... there's a few that might work as rune-activated doors, but most do nothing.
20160919 18:55:14< vultraz> tad_: and this is AFTER I cleaned the WML up :)
20160919 18:55:25 * tad_ rolls his eyes.
20160919 18:55:30< zookeeper> or do they
20160919 18:55:48< zookeeper> actually they're all positioned to look like they will actually open up passages, so nevermind, i guess there's just that many moveto events in there
20160919 18:56:05< vultraz> (last year I did significant WML cleanups of certain campaigns. never got through all of them. so you'll likely find less to do in the ones I hit)
20160919 18:56:14< vultraz> (httt was not one of them :P )
20160919 18:56:53< vultraz> i should finish northern rebirth at some point
20160919 18:57:08< tad_> vultraz: The first 6 or so levels of THoT are very clean. I have one or two fixes for each but nothing major.  If you're intersted I have a branch on my fork GL_THoT which is my WIP
20160919 18:58:26< zookeeper> i wish i could say that everyone universally hates that scenario, but it does seem like there are some people who enjoy it.
20160919 18:58:49< vultraz> tad_: :)
20160919 18:59:16< vultraz> tad_: oh, btw, if you ever get to northern rebirth, don't touch the branching path or go past scenario 9.
20160919 18:59:37< zookeeper> although i suppose that if someone came along to, ahem, completely redesign the scenario then it would not be rejected
20160919 18:59:55< vultraz> COMPLETELY REDESIGN, you say?
20160919 18:59:57< vultraz> :D
20160919 19:00:26< vultraz> this I can do
20160919 19:00:37< vultraz> well, not right now
20160919 19:00:39< vultraz> but eventually
20160919 19:00:43< tad_> vultraz: Oh? Why? Code-mess like TSG had for the path exposition? NP with stuff like that. Zookeeper said hand-off so I didn't even look at it.
20160919 19:01:08< vultraz> tad_: the branching path was never completed and isn't accessible 
20160919 19:01:19< vultraz> so unless you plan to finish it, it's not worth it :P
20160919 19:01:45< tad_> vultraz: Ah. We'll see when/if I get there. Still a lot of mainline to look at ...
20160919 19:01:56< vultraz> as for the other thing, I revamped th first 9 scenarios, but never finished the rest. so I'd likely be redoing any work you do.
20160919 19:02:29< tad_> And it's not like this is ever "done" .. I have notes for things to come back and look at again, already.
20160919 19:02:47< vultraz> NR is a huge design mess
20160919 19:02:50< vultraz> i was trying to make it less so
20160919 19:03:11< tad_> Well, more eyes never hurt.
20160919 19:03:30< tad_> But, for now .. hmm .. UI? Nose? Grindstone? /me grins
20160919 19:03:39< vultraz> :P
20160919 19:03:44< vultraz> yes, that comes first
20160919 19:03:51< vultraz> I'll finish the last 4 scenarios of nr... sometime
20160919 19:04:31< vultraz> it'd odd how my focus changed
20160919 19:04:34< tad_> Or I will or someone else will and we'll rewrite it and rewrite it again until everyone's peed in the pot and it tastes Just Right
20160919 19:04:39< vultraz> went from campaign cleanup to ui 
20160919 19:04:58< vultraz> wml to c++
20160919 19:05:08< vultraz> aka frustration by a different name :P
20160919 19:05:44< zookeeper> *cough*bootNRfrommainline*coughcough*
20160919 19:06:23< tad_> Someday, someone should check UMC for some new blood for mainline, too, no?
20160919 19:07:41< tad_> I have some ideas to work into the existing themes and timeline but got sidetracked into this cleanup and think I'm happier doing this than trying to do my own campaign.
20160919 19:08:31-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has quit [Ping timeout: 250 seconds]
20160919 19:09:03-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Quit: wedge009]
20160919 19:09:22-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20160919 19:10:07-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Client Quit]
20160919 19:10:58< zookeeper> tad_, mainlining a new campaign or two after 1.14 is certainly a possibility if you or someone else want to try to tackle that
20160919 19:13:00< tad_> The code cleanup for it, I'm up for that. "Balance" I leave to those skilled in the Black Arts. I don't grok "balance" .. never have and probably never will.
20160919 19:15:43-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160919 19:15:50-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has joined #wesnoth-dev
20160919 19:15:52< tad_> Anyway, I'm up to THoT S08 and don't like the Mage recruitment and am re-working how it goes away. Looks like fairly clear sailing after than except for "balancing" S12 so it's playable, at least on Easy, by those of us who don't do well at slug-fests
20160919 19:16:40-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20160919 19:18:40-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Client Quit]
20160919 19:18:41-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20160919 19:21:04-!- wedge010 is now known as wedge009
20160919 19:24:31-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20160919 19:26:03-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160919 19:27:23< irker644> wesnoth: Jyrki Vesterinen wesnoth:master b3528593b776 / INSTALL: Mention Boost.Thread in INSTALL https://github.com/wesnoth/wesnoth/commit/b3528593b776d40a5fb23621fa39cc3d5ffe8b28
20160919 19:28:08-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Client Quit]
20160919 19:29:04-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160919 19:36:24-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160919 19:40:43-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 244 seconds]
20160919 19:43:19-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20160919 19:43:35-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20160919 19:45:00-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160919 19:59:29-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has quit [Quit: Page closed]
20160919 20:11:21-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160919 20:14:35-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20160919 20:16:14-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160919 20:16:54-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Client Quit]
20160919 20:19:20-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160919 20:19:33-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Client Quit]
20160919 20:21:54-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160919 20:28:24-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20160919 20:30:07-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160919 20:46:28-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: So long and thanks for all the fish.]
20160919 20:47:24-!- mjs-de [~mjs-de@wh.Uni-Dortmund.DE] has quit [Remote host closed the connection]
20160919 20:48:22-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has joined #wesnoth-dev
20160919 20:48:57< tad_> zookeeper: HttT S06 Logical Thieves complete rewrite. No code motion. Whitespace changed.
20160919 20:58:20-!- ChipmunkV [~vova@static-89-94-113-91.axione.abo.bbox.fr] has quit [Quit: ChipmunkV]
20160919 21:08:39< zookeeper> tad_, oh, good that you noticed that the radius needs to be 5 and not 7 as i suggested.
20160919 21:09:12< tad_> I tested it, of course. And carefully counted the tiles.
20160919 21:11:02< tad_> I like this version better than the old. You can take all the villages in debug mode, fire turn 4, kill leader; take all, kill, fire; etc. The only way to not get thieves is to kill;kill without firing turn 4.
20160919 21:11:45< tad_> or ask for the ford and kill both leaders before the theives return
20160919 21:11:55< irker644> wesnoth: Gregory A Lundberg wesnoth:master 0594ebd1722c / data/campaigns/Heir_To_The_Throne/scenarios/06_The_Siege_of_Elensefar.cfg: HttT S06 Use capture and radius https://github.com/wesnoth/wesnoth/commit/0594ebd1722c7862b2c11348c130badf78e5e7f4
20160919 21:11:57< irker644> wesnoth: Gregory A Lundberg wesnoth:master 11d9891319a2 / data/campaigns/Heir_To_The_Throne/scenarios/06_The_Siege_of_Elensefar.cfg: HttT S06 Fix bug: Let the player know https://github.com/wesnoth/wesnoth/commit/11d9891319a21db4b94deb2d8d687b36e5888ad0
20160919 21:11:59< irker644> wesnoth: Gregory A Lundberg wesnoth:master d6f7de91acb8 / data/campaigns/Heir_To_The_Throne/scenarios/06_The_Siege_of_Elensefar.cfg: HttT S06 Thieves join if leader killed https://github.com/wesnoth/wesnoth/commit/d6f7de91acb8e99429b5e14f50811a67febcdf8b
20160919 21:12:01< irker644> wesnoth: Gregory A Lundberg wesnoth:master 9b364d25d2a5 / data/campaigns/Heir_To_The_Throne/scenarios/06_The_Siege_of_Elensefar.cfg: HttT S06 Logical Thieves https://github.com/wesnoth/wesnoth/commit/9b364d25d2a58bf2446d9396682a1e050e8d03fa
20160919 21:12:03< irker644> wesnoth: Lari Nieminen wesnoth:master d0d42236619a / data/campaigns/Heir_To_The_Throne/scenarios/06_The_Siege_of_Elensefar.cfg: Merge pull request #733 from GregoryLundberg/GL_HttT_S06_logical_thieves https://github.com/wesnoth/wesnoth/commit/d0d42236619a0e5173d099ff6fa32da2a89fe39b
20160919 21:12:36< tad_> So PR 783 should be good to go, too.
20160919 21:13:51< irker644> wesnoth: Gregory A Lundberg wesnoth:master ccce95869273 / players_changelog: HttT Player changelog https://github.com/wesnoth/wesnoth/commit/ccce95869273c550aabaa46b1bd328eb45b33961
20160919 21:13:53< irker644> wesnoth: Lari Nieminen wesnoth:master 60a6519295cc / players_changelog: Merge pull request #783 from GregoryLundberg/GL_HttT_changelog https://github.com/wesnoth/wesnoth/commit/60a6519295cca6d08de0466ff9320418ab0cc14c
20160919 21:16:45< tad_> Wow. After all those weeks with those branches in my repo it feels eerily empty now that they're gone!
20160919 21:18:34< zookeeper> now, i wonder how exactly that undead spawning when entering the cave should be done to avoid a potential landmine...
20160919 21:18:42-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Ping timeout: 265 seconds]
20160919 21:18:47< zookeeper> because it's rather silly as it is now
20160919 21:19:07< tad_> In HttT S06?
20160919 21:19:17< zookeeper> yes
20160919 21:19:23< tad_> I usually use the side entrance to avoid them
20160919 21:19:32< zookeeper> exactly :p
20160919 21:19:36-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20160919 21:19:36-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Read error: Connection reset by peer]
20160919 21:20:43< tad_> I like the effects and I don't know that it's that much of a tomato. When I do a frontal assault, I've been watching them march out every turn or two for some time so I am ready for more.
20160919 21:21:07< tad_> But, I've played it so much now I may not be the best judge.
20160919 21:23:35< tad_> How about an enter_hex instead of moveto tripwire and make the radius large .. say almost to the woods .. so they trigger when you're well away and stop the unit which triggered them?
20160919 21:23:37< zookeeper> well it's also possible, although unlikely, to just hop over the trigger tiles
20160919 21:23:42-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20160919 21:25:25-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160919 21:25:33< zookeeper> i might prefer making it an enter_hex on any cave hex, and re-position the spawn locations so that they can't the triggering unit immediately
20160919 21:25:36-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Client Quit]
20160919 21:26:20< tad_> The cave is so small, anyone at the entrance can get hit. Which is why I suggested making the radius a bit larger.
20160919 21:26:28-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160919 21:27:04< tad_> Can allow undo when they appear.
20160919 21:27:30< tad_> So you get 'em .. get scared .. and undo get run away ...
20160919 21:27:41< tad_> s/get/to/
20160919 21:27:59< zookeeper> does not compute
20160919 21:28:36< gfgtdf> tad_; allowing undo on gamechanging event will result in corrupt replays.
20160919 21:28:38< tad_> enter_hex triggers them, but allow undo. not an obvious fix, but it does allow you to get clear
20160919 21:28:51< tad_> gfgtdf: so that won't work.
20160919 21:29:21< tad_> so .. bring them in but force moves=0 so they can't move for a full turn?
20160919 21:29:23< zookeeper> yes, and it'd be cheesy to be able to undo in a situation like that
20160919 21:29:52< tad_> or moves=1 or 2 so they can't quite reach the entrance where enter_hex stopped you?
20160919 21:30:24< zookeeper> i'll see what i can do
20160919 21:31:02< tad_> I do like them, and like the walls crumbling and them appearing in former-walls.
20160919 21:32:35< gfgtdf> about whcih scneario do you talk currently ?
20160919 21:32:58< tad_> HttT S06 the cave to the far north. Tripwire when you enter from the south entrance
20160919 21:33:39< tad_> Siege of Elensefar
20160919 21:34:08< gfgtdf> zookeeper: any opinion on https://gna.org/bugs/?23841?
20160919 21:36:15-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20160919 21:37:27< zookeeper> gfgtdf, uh, looks complicated, will take a while to parse that...
20160919 21:37:40< gfgtdf> zookeeper: just read the first post of thta thread
20160919 21:37:51< gfgtdf> zookeeper: that is the one at the bottom
20160919 21:38:38-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160919 21:42:10< zookeeper> gfgtdf, in principle i don't see why not... but i can't really give a very informed opinion on something like that immediately. i mean that implementation is very simplistic and maybe it has all sorts of shortcomings WRT parameters passed to the tag or something.
20160919 21:43:04< tad_> My opinion: thought about it myself, too. didn't see the gna issue until just now. seems fraught with peril but the UMC ppl would love it since they'd avoid learning lua. the load/save issue was skipped but the concept looks sound to me.
20160919 21:44:19< tad_> It's a "true function" definition.
20160919 21:45:31< tad_> [function]name=<NAME> {other stuff to be determined} [command][/command][/function]
20160919 21:45:41< gfgtdf> zookeeper: 'WRT' ?
20160919 21:45:51< tad_> with respect to
20160919 21:46:11< gfgtdf> tad_: the load save issue is streesed in a later post in that thread (which mkaes the lua code mosre coplicated)
20160919 21:46:21 * tad_ nods.
20160919 21:46:37< tad_> That and making sure MP plays nice with it.
20160919 21:47:38< gfgtdf> tad_: there are no reasony wha it shouldn't play nice in mp.
20160919 21:47:47-!- Appleman1234 [~Appleman1@KD119104045198.au-net.ne.jp] has quit [Ping timeout: 248 seconds]
20160919 21:48:48< tad_> Still, it's arbitrary code executed out-of-line so you're want to be sure MP uses the same state and produces the same results everywhere.
20160919 21:49:22< gfgtdf> tad_: i mean i like the tag, but (like the [for] tags) i was never motivated to work on it becasue i always would write such codes in lua anyway. 
20160919 21:50:26< tad_> Which is why I didn't pursue the idea myself. But some have a hard enough time learning WML and adding Lua to the mix seems to upset them.
20160919 21:50:58< tad_> THoT does exactly that GMA issue, using Lua. I just fixed its bugs ...
20160919 21:51:13-!- Netsplit *.net <-> *.split quits: Duthlet
20160919 21:51:54< zookeeper> all i can really say right now is that the idea is good but it seems like something that could end up having all sorts of issues
20160919 21:52:08< tad_> And, to be honest, I don't think I could have fixed them, or even avoided them, using WML.
20160919 21:52:20< tad_> zookeeper: agreement here
20160919 21:54:38< tad_> It's one of those ideas which is actually pretty easy but, down the road, we're probably going to hit a snag and fixing it could mean breaking content which didn't hit that snag.
20160919 21:58:02< tad_> Here's a snag .. [function]name=SO[command][SO][/SO][/command][/function][SO][/SO]
20160919 22:01:11< tad_> And consider delayed variable replacement in the context of the inner WML
20160919 22:01:15-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20160919 22:02:15-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160919 22:04:04-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160919 22:04:11-!- Appleman1234 [~Appleman1@KD119104047049.au-net.ne.jp] has joined #wesnoth-dev
20160919 22:06:25-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20160919 22:06:49-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160919 22:07:16-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160919 22:08:12-!- tad__ [add94167@gateway/web/freenode/ip.173.217.65.103] has joined #wesnoth-dev
20160919 22:09:57-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160919 22:10:12-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has quit [Ping timeout: 240 seconds]
20160919 22:15:27< matthiaskrgr> grrr
20160919 22:15:28< matthiaskrgr> linking fails now with gcc 
20160919 22:20:36< tad__> celticminstrel: https://forums.wesnoth.org/viewtopic.php?f=21&t=44567
20160919 22:20:44< matthiaskrgr> http://pastebin.com/ywDCKkgb  
20160919 22:21:16< tad__> Looks like you're missing the pthread library.
20160919 22:21:41< tad__> "Posix Threads" --> pthread
20160919 22:21:59< matthiaskrgr> hm, what would that be 
20160919 22:22:16< matthiaskrgr> ah
20160919 22:22:25< matthiaskrgr> clthreads ? 
20160919 22:22:46< tad__> On linux it's libpthread iirc
20160919 22:23:50< matthiaskrgr> there is no such package here
20160919 22:24:13< tad__> OS/x? Let me look. Don't use it but I'll try.
20160919 22:24:34< matthiaskrgr> no, fedora 
20160919 22:24:45< tad__> OH. Hang on.
20160919 22:25:16< matthiaskrgr> the files it is complaining about are in  boost-devel but I have that installed
20160919 22:25:50< matthiaskrgr> wesnoth built for me yesterday.. :/
20160919 22:26:12< tad__> -lpthread is correct
20160919 22:27:30< matthiaskrgr> I did change CFLAGS and CXXFLAGS but that shouldn't affect linking 
20160919 22:27:41-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160919 22:28:07< tad__> Is pthread listed on LDFLAGS?
20160919 22:28:29< Aginor> it can affect linking, it depends on how the linker is invoked
20160919 22:29:27< tad__> -lpthread will look for libpthread and Fedora has it. You should have it but there should be an RPM for it. Been nearly 15 years since I last was on RHL ...
20160919 22:30:02< Aginor> no, I think pthread is just a glibc dependency nowadays
20160919 22:30:20< matthiaskrgr> which files does list the cxxflags and ldflags used for "make" ? 
20160919 22:30:32 * matthiaskrgr is not that familiar with cmake 
20160919 22:30:37< Aginor> right
20160919 22:30:48< Aginor> it's complicated
20160919 22:30:56< matthiaskrgr> for the record, I'm doing a build with asan/ubsan and debug symbols
20160919 22:31:13< Aginor> I haven't built the game for a few days, but I normally build with cmake on Fedora
20160919 22:31:24< Aginor> and I'm at work so I can't look into it right now
20160919 22:31:34< matthiaskrgr> and gcc, since clang would fail some assert inside boost
20160919 22:31:35< matthiaskrgr> ok
20160919 22:31:54< Aginor> matthiaskrgr: they're generated from the toplevel cmakelists.txt
20160919 22:31:58< tad__> I just build on Arch using gcc 6.1 and came up clean
20160919 22:32:08< Aginor> then it gets murkier
20160919 22:34:21< tad__>  Selecting the threading implementation: LD_ASSUME_KERNEL        On systems with a glibc that supports both LinuxThreads and NPTL        (i.e., glibc 2.3.x), the LD_ASSUME_KERNEL environment variable can be        used to override the dynamic linker's default choice of threading        implementation.
20160919 22:34:57< matthiaskrgr> well, I will just rebuild using systems clang, that might work 
20160919 22:34:57< tad__> man7.org/linux/man-pages/man7/pthreads.7.html
20160919 22:43:50< Aginor> matthiaskrgr: how did you invoke cmake?
20160919 22:45:46< matthiaskrgr> http://pastebin.com/4frZxfta  
20160919 22:46:35< matthiaskrgr> the CC and CXX vars are new, I the linking error was still from gcc 
20160919 22:48:31-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160919 22:48:57-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160919 22:50:25< matthiaskrgr> ok, clang-3.9 also is killed by this statis_assert *sigh*
20160919 22:52:04< matthiaskrgr> lets try building without any flags and gcc ...
20160919 22:52:28< gfgtdf> matthiaskrgr: you coudl also try to use boost 1.59 instead
20160919 22:52:54< matthiaskrgr> by compiling it locally? :S
20160919 22:53:12< gfgtdf> matthiaskrgr: you didnt compile boost 1.60 yourself ?
20160919 22:53:18< matthiaskrgr> no
20160919 22:53:26< matthiaskrgr> it's official fedora package
20160919 22:53:53< gfgtdf> matthiaskrgr: hmm on windows you usually ahve to compile boost yourself, don't know ho it works on linux.
20160919 22:54:17< matthiaskrgr> usually you just install the library package from the package maintainer and then it works
20160919 22:54:22< matthiaskrgr> ... "usually" :) 
20160919 22:55:43< Aginor> I just use the supplied packages
20160919 22:55:58< Aginor> mind you, I haven't tried to build with clang either, I just use gcc 6.1
20160919 22:57:02-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has quit [Read error: Connection reset by peer]
20160919 22:57:08-!- celmin [~celmin@unaffiliated/celticminstrel] has joined #wesnoth-dev
20160919 22:57:32< tad__> celmin: PR 792 for that bug from the forums
20160919 23:09:40< matthiaskrgr> ok, regular build succeeds
20160919 23:10:14< zookeeper> tad__, and i thought i do short and cryptic posts on the forums...
20160919 23:11:06 * tad__ chuckles
20160919 23:11:50< tad__> It is overly verbose. I s'pose I could remove 'see' and make it half that long ...
20160919 23:12:21< zookeeper> i'd like to think that if i have no idea what [role] has to do with anything there, he won't either... :p
20160919 23:13:42< tad__> OK. I'll add that I'd use [role] to record the troops at the end of one scene to recall them for the nex.t
20160919 23:25:29-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [Ping timeout: 248 seconds]
20160919 23:39:37-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20160919 23:40:19-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160919 23:41:09-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 244 seconds]
20160919 23:41:09-!- wedge010 is now known as wedge009
20160919 23:41:56< irker644> wesnoth: Gregory A Lundberg wesnoth:master 3809a69a42a2 / data/lua/wml/harm_unit.lua: Fix bug: Lua harm_unit https://github.com/wesnoth/wesnoth/commit/3809a69a42a22839eea1238ef77db1e23577a0c5
20160919 23:41:58< irker644> wesnoth: Celtic Minstrel wesnoth:master 83bd6cdfbf59 / data/lua/wml/harm_unit.lua: Merge pull request #792 from GregoryLundberg/GL_harm_unit https://github.com/wesnoth/wesnoth/commit/83bd6cdfbf59a8d4b646127ef94ce7552b7d0995
20160919 23:42:11-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Ping timeout: 244 seconds]
20160919 23:42:17-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev
20160919 23:43:53-!- Appleman1234 [~Appleman1@KD119104047049.au-net.ne.jp] has quit [Ping timeout: 244 seconds]
20160919 23:43:53-!- AI0867 [~ai@wesnoth/developer/ai0867] has quit [Ping timeout: 244 seconds]
20160919 23:44:21-!- AI0867 [~ai@wesnoth/developer/ai0867] has joined #wesnoth-dev
20160919 23:52:31-!- iwaim [~iwaim@124.146.179.10] has quit [Ping timeout: 244 seconds]
20160919 23:52:55-!- iwaim [~iwaim@rasteenie.alib.jp] has joined #wesnoth-dev
20160919 23:55:21-!- celmin is now known as celticminstrel
20160919 23:56:13< celticminstrel> Hmm, suddenly wondering if vultraz broke the MP tests with the new staging dialog?
20160919 23:57:07 * celticminstrel hasn't seen any Travis build announcements though.
20160919 23:57:51< tad__> celticminstrel: Off the top of your head, do you know where debug|kill command is located? I hit a bug with it and want to see if I can fix it.
20160919 23:58:07< celticminstrel> I think it's in synced_commands.cpp or some similar filename.
20160919 23:58:37< tad__> ty
--- Log closed Tue Sep 20 00:00:52 2016