--- Log opened Mon Mar 28 00:00:09 2016
--- Day changed Mon Mar 28 2016
20160328 00:00:09< celticminstrel> Hmm, I have an idea.
20160328 00:04:20< mattsc> Ooo, my bottleneck defense leader turned into a general on the last attack of the scenario
20160328 00:05:01< mattsc> celticminstrel: BTW, all the MAIs I have tested so far are working as they should.
20160328 00:05:14< mattsc> Granted, I haven’t tested all that many yet ...
20160328 00:05:16< celticminstrel> Yay.
20160328 00:05:45< mattsc> But I think that, except for protect_unit, this is mostly a forrmality.
20160328 00:18:34-!- travis-ci [~travis-ci@ec2-184-72-134-120.compute-1.amazonaws.com] has joined #wesnoth-dev
20160328 00:18:35< travis-ci> wesnoth/wesnoth#9075 (renderpath_redo - 2b30ffc : Andreas Löf): The build has errored.
20160328 00:18:35< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/118886429
20160328 00:18:35-!- travis-ci [~travis-ci@ec2-184-72-134-120.compute-1.amazonaws.com] has left #wesnoth-dev []
20160328 00:20:35-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [Ping timeout: 268 seconds]
20160328 00:25:23 * celticminstrel wonders what gfgtdf's C++11 status is, since the build's errors seem to only occur with C++03.
20160328 00:42:19< gfgtdf> celticminstrel: i don't think the error is related to c++03
20160328 00:43:03< celticminstrel> I don't think it's directly related to C++03, no.
20160328 00:43:42< celticminstrel> It might be something like "Travis has updated their software, and some aspect of it no longer works with C++03".
20160328 00:44:03< celticminstrel> The actual error that's occurring appears to be a Python error related to gdb.
20160328 00:46:31< gfgtdf> celticminstrel: hmm did you try to rn the mp tests locally ?
20160328 00:46:52< celticminstrel> I haven't yet.
20160328 00:50:22-!- nos__ [~nos@208.91.185.104] has joined #wesnoth-dev
20160328 01:20:24-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20160328 01:23:21< celticminstrel> This is so confusing.
20160328 01:23:24< celticminstrel> And messy.
20160328 01:23:26< celticminstrel> Hmm...
20160328 01:24:07-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 260 seconds]
20160328 01:24:07-!- wedge010 is now known as wedge009
20160328 01:24:21-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has quit [Quit: Connection closed for inactivity]
20160328 01:24:42< gfgtdf> celticminstrel: what do you mean ?
20160328 01:24:59< celticminstrel> The attacks aspect.
20160328 01:25:39< celticminstrel> I've got it working with a Lua implementation, so I could leave it like this, but I'd kinda like to do something like the advancements aspect...
20160328 01:31:07< gfgtdf> vultraz: is not the new mp lobby the default in some cases ?
20160328 01:31:13< gfgtdf> vultraz: now*
20160328 01:35:01< gfgtdf> celticminstrel: i think the c++ code shoudl be able to ask the c++ ai code about the advamcentaspect object (as opposed to the ai code passing it to the attack function which is currently the case). This woudl have the advantage that the advancement aspect coudl also be used of the unit advances as a defender
20160328 01:35:20< gfgtdf> s/of the/if the
20160328 01:35:28< gfgtdf> the unit that advances*
20160328 01:36:28< gfgtdf> not sure if this is related to what  you are currently working on though,
20160328 01:36:34< celticminstrel> Maybe... that's not really related to what I'm doing though.
20160328 01:58:50-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20160328 02:15:03-!- gfgtdf [~chatzilla@f054141215.adsl.alicedsl.de] has quit [Quit: ChatZilla 0.9.92 [Firefox 45.0.1/20160315153207]]
20160328 02:38:45< celticminstrel> I think I have it worked out. :D
20160328 02:40:57< celticminstrel> More or less.
20160328 02:43:19< mattsc> Nice.
20160328 02:44:18< celticminstrel> The aspect needs to return a table with keys "own" and "enemy".
20160328 02:45:04< celticminstrel> The value can be either a table to be converted to a unit filter, or a function that takes a unit and returns true or false.
20160328 02:45:17< celticminstrel> ^values
20160328 02:46:08< celticminstrel> It's fine to omit one of the keys, too.
20160328 02:46:09< mattsc> okay
20160328 02:46:23< mattsc> Question: If I want to switch the controller to null for the currently active human-controlled side, what’s the recommended way of doing that in master?
20160328 02:46:27< celticminstrel> At least, it should be. I should verify that I properly accounted for that case.
20160328 02:46:38< mattsc> Just doing [modify_side]controller=null doesn’t work any more.
20160328 02:46:51< mattsc> Add [end_turn] and a side_turn event?
20160328 02:47:12< celticminstrel> That's not switching the controller to null, but I suppose it would have a similar effect.
20160328 02:47:36< mattsc> … and setting the controller to null in the side turn event of the next side, is what I meant.
20160328 02:48:02< celticminstrel> I'm confused. I thought you said that didn't work.
20160328 02:48:22< mattsc> It does not work if you do it while it’s the side’s turn.
20160328 02:48:27< celticminstrel> Oh.
20160328 02:48:44-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20160328 02:49:04< celticminstrel> So if you do it while it's the side's turn and then go to the inspector, it still shows controller=human?
20160328 02:49:39-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20160328 02:50:08< mattsc> I haven’t tried that, but this:  https://github.com/wesnoth/wesnoth/blob/master/data/ai/micro_ais/scenarios/goto.cfg#L367
20160328 02:50:36< mattsc> results in the side still having human controlled when it’s its turn next time around
20160328 02:50:50< mattsc> *still being human controlled
20160328 02:50:57< celticminstrel> Oh, I experienced that, actually.
20160328 02:51:21< mattsc> This used to work in 1.12, but Wesnoth complains about it now.
20160328 02:51:35< mattsc> “warning engine: ignored attempt to change the currently playing side's controller to 'null’”
20160328 02:52:08< mattsc> I know what I suggested up there works, I was just wondering if there’s a different “recommended” way of accomplishing this.
20160328 02:52:17< celticminstrel> So the side turn event does work?
20160328 02:52:41< celticminstrel> I wonder what the reason is for disallowing it during the side's turn. Using a side turn event seems reasonable to me, though.
20160328 02:52:55< celticminstrel> I don't know if there's another recommended way.
20160328 02:54:03< celticminstrel> It crashed. :|
20160328 02:54:15< celticminstrel> It worked fine when I used get_attacks() in the console, why is it crashing now?
20160328 02:54:37< mattsc> It might have something to do with the OOS problems gfgtdf mentioned a couple days ago.
20160328 02:54:43< mattsc> :( on the crash
20160328 02:55:52< celticminstrel> And it didn't crash on what should have been the same function call just a bit earlier in the code.
20160328 02:58:41< mattsc> celticminstrel: this works: http://pastebin.com/vZj40MA3
20160328 03:02:10< celticminstrel> I can't reproduce the crash.
20160328 03:02:36-!- ancestral [~ancestral@75-168-27-21.mpls.qwest.net] has joined #wesnoth-dev
20160328 03:02:39< celticminstrel> The only thing I can think of is that it was some fluke based on me opening the console while the aspect was being calculated.
20160328 03:02:50< celticminstrel> ...which, come to think of it, is actually probably a thing that could happen.
20160328 03:03:09< celticminstrel> Hmm, should I guard against that?
20160328 03:03:54< celticminstrel> It's not like it's something that would really affect anyone.
20160328 03:10:59< celticminstrel> Okay, it works.
20160328 03:13:29< celticminstrel> Incidentally, I was testing this with a Lua facet rather than a Lua toplevel aspect.
20160328 03:15:59< celticminstrel> I still haven't updated the wiki with any of the AI changes...
20160328 03:16:09< celticminstrel> I'm not looking forward to that task. >_>
20160328 03:19:19-!- irker968 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20160328 03:20:14< mattsc> I sympathize with that
20160328 03:20:34< mattsc> I had been pushing off doing the wiki AI pages overhaul for a long time.
20160328 03:23:01-!- irker312 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20160328 03:23:01< irker312> wesnoth: Celtic Minstrel wesnoth:master 97b33cfdfca3 / / (11 files in 5 dirs): Allow the attacks aspect to be implemented with Lua https://github.com/wesnoth/wesnoth/commit/97b33cfdfca346a1ef265b7c4cf9713577eff6f3
20160328 03:23:33< celticminstrel> I've been debating whether or not to do anything towards cleanup of FormulaAI.
20160328 03:27:42< mattsc> I would not ...
20160328 03:28:09< mattsc> Do I need to recheck any of the MAIs because of that ^
20160328 03:31:35-!- travis-ci [~travis-ci@ec2-184-72-134-120.compute-1.amazonaws.com] has joined #wesnoth-dev
20160328 03:31:36< travis-ci> wesnoth/wesnoth#9076 (master - 97b33cf : Celtic Minstrel): The build failed.
20160328 03:31:36< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/118909255
20160328 03:31:36-!- travis-ci [~travis-ci@ec2-184-72-134-120.compute-1.amazonaws.com] has left #wesnoth-dev []
20160328 03:38:15< celticminstrel> Ugh.
20160328 03:39:07< celticminstrel> I don't think it should affect any of the MAIs; maybe the Fast AI since it accesses the attacks aspect, but probably not even that.
20160328 03:39:09< irker312> wesnoth: mattsc wesnoth:master 263227852a5f / data/ai/micro_ais/ (cas/ca_zone_guardian.lua micro_ai_wml_tag.lua): Micro AI CAs: use consistent indenting https://github.com/wesnoth/wesnoth/commit/263227852a5f55e1a64b5c18b02bb8a1367899c1
20160328 03:39:11< irker312> wesnoth: mattsc wesnoth:master 9ed1b39b9071 / data/ai/micro_ais/cas/ca_recruit_random.lua: Random Recruiting MAI: use require() instead of dofile() https://github.com/wesnoth/wesnoth/commit/9ed1b39b90713650f94b025104ce8bce90872c06
20160328 03:39:13< irker312> wesnoth: mattsc wesnoth:master 1e3d80df0ba4 / data/ai/micro_ais/scenarios/ (animals.cfg goto.cfg wolves.cfg): Micro AI tests: do not change controller to null for active side https://github.com/wesnoth/wesnoth/commit/1e3d80df0ba43c95efaad628c3649d245f476cdd
20160328 03:40:01< mattsc> okay; I’ll retest that one just to be sure
20160328 03:40:41< irker312> wesnoth: Celtic Minstrel wesnoth:master 5b8e53855f32 / src/ai/default/aspect_attacks.cpp: Fixup include https://github.com/wesnoth/wesnoth/commit/5b8e53855f32a63261eb4b3f026d370d0640b449
20160328 03:41:12< celticminstrel> I wonder if a forum thread should be opened on the subject of FormulaAI.
20160328 03:41:52< mattsc> For what purpose?
20160328 03:42:13< celticminstrel> To see what people think about whether or not to remove it at some point.
20160328 03:42:34< celticminstrel> There are currently 16 UMC that use it, one of which is (I assume) yours, plus one mainline campaign.
20160328 03:43:04< celticminstrel> That's from the 1.12 addons server.
20160328 03:44:38< mattsc> I don’t know …
20160328 03:45:19< celticminstrel> Don't know if a forum thread should be opened?
20160328 03:45:51< mattsc> Yes.
20160328 03:47:57-!- travis-ci [~travis-ci@ec2-54-166-17-196.compute-1.amazonaws.com] has joined #wesnoth-dev
20160328 03:47:58< travis-ci> wesnoth/wesnoth#9077 (master - 1e3d80d : mattsc): The build is still failing.
20160328 03:47:58< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/118910979
20160328 03:47:58-!- travis-ci [~travis-ci@ec2-54-166-17-196.compute-1.amazonaws.com] has left #wesnoth-dev []
20160328 03:50:21 * celticminstrel is considering duplicating the entire logic of how an aspect is calculated in the Fast MAI code...
20160328 03:51:15< mattsc> I mean, what do you expect from that?  Of those 16 UMC authors, only 3 will reply.
20160328 03:51:36< celticminstrel> Perhaps.
20160328 03:51:57< mattsc> Of those, 1 will say that they don’t care, one that they have no idea how to change things to somethign else, and one will rant that the evil Wesnoth developers never listen to the users.
20160328 03:52:06< celticminstrel> Heh...
20160328 03:52:28< mattsc> Being “slightly” cycnical here, but I expect the outcome to be something like that.  :)
20160328 03:52:35< mattsc> Then what do you do?
20160328 03:53:15< mattsc> I just managed to trigger an assert by killing the side leader in debug mode!
20160328 03:53:24-!- travis-ci [~travis-ci@ec2-184-72-134-120.compute-1.amazonaws.com] has joined #wesnoth-dev
20160328 03:53:25< travis-ci> wesnoth/wesnoth#9078 (master - 5b8e538 : Celtic Minstrel): The build is still failing.
20160328 03:53:25< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/118911119
20160328 03:53:25-!- travis-ci [~travis-ci@ec2-184-72-134-120.compute-1.amazonaws.com] has left #wesnoth-dev []
20160328 03:54:00 * pydsigner checks that he isn't using FAI
20160328 03:54:43< mattsc> pydsigner: if you do and would like to switch to Lua, both celticminstrel and myself would be happy to help
20160328 03:54:47< pydsigner> Doesn't look like I am
20160328 03:55:08< pydsigner> I knew I had some AI rules but I don't think they were FAI
20160328 03:57:52< pydsigner> * wasn't sure if they were Formula AI
20160328 03:58:18< pydsigner> Lua seems like a better way to do this though
20160328 04:00:34< aeth> The more Lua the better imo.
20160328 04:01:25< pydsigner> People are probably going to have to change things to upgrade to 1.14 anyways
20160328 04:01:39< pydsigner> May as well get the pain out of the way for them
20160328 04:04:10< mattsc> Maybe.  I’d tend to agree, in principle.  Just not 100% sure about 1.14 or 1.16.
20160328 04:04:23< mattsc> Which reminds me, do I still use FAI in my add-ons?
20160328 04:04:56< celticminstrel> ...
20160328 04:05:14< mattsc> Oh, right, I have an FAI demo scenario ...
20160328 04:05:28< celticminstrel> Why is it still failing... :(
20160328 04:06:12< celticminstrel> [Mar 27@11:53:15pm] mattsc: I just managed to trigger an assert by killing the side leader in debug mode!
20160328 04:06:13< celticminstrel> Eh? I hope that really is debug-specific...
20160328 04:07:03< mattsc> So do I.  I’ll try how to reproduce this and how specific it is tomorrow.
20160328 04:07:48< celticminstrel> FAI is more convenient for attaching AI rules to specific units.
20160328 04:08:32< celticminstrel> That aside, I don't really see any compelling reason to drop it, nor any compelling reason to keep it.
20160328 04:09:24< mattsc> yes, same here
20160328 04:09:50< celticminstrel> I consider unit formulas a mark in its favour, but a minor one.
20160328 04:10:08< celticminstrel> I know that Lua has BCAs, but from what I understand, that's less convenient than unit formulas.
20160328 04:11:15< celticminstrel> On the whole I don't think FormulaAI is better or worse than Lua. It's just a different approach.
20160328 04:12:07-!- ancestral [~ancestral@75-168-27-21.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20160328 04:15:01< celticminstrel> ...
20160328 04:15:42< celticminstrel> ......
20160328 04:15:45< celticminstrel> Oh for...
20160328 04:15:49< celticminstrel> Bah.
20160328 04:16:24< aeth> http://wiki.wesnoth.org/FormulaAI
20160328 04:16:36< celticminstrel> What about it?
20160328 04:16:36< irker312> wesnoth: Celtic Minstrel wesnoth:master 49e7e03718b3 / src/ai/lua/lua_object.cpp: Fixup include - again https://github.com/wesnoth/wesnoth/commit/49e7e03718b36fc2176eb2da01671780728f9c75
20160328 04:16:43< aeth> "Formula AI is still functional for the time being, but it is not maintained any more."
20160328 04:16:57< celticminstrel> Yes, mattsc added that header.
20160328 04:17:13< aeth> Apparently in Februray
20160328 04:17:19< SigurdFD> would that mean it should get call deprecation without any timeline for removal?
20160328 04:17:19< celticminstrel> That does not rule out the possibility of maintenance being picked up again, though.
20160328 04:17:27< celticminstrel> SigurdFD: Huh?
20160328 04:17:34< aeth> If you wanted to do things properly, I think the way to do it would be to deprecate it in 1.14 and remove it in 1.16
20160328 04:17:40< aeth> Imo
20160328 04:17:44< celticminstrel> Yes, assuming we plan to remove it.
20160328 04:17:51< SigurdFD> what aeth said
20160328 04:18:07< celticminstrel> I was also thinking it probably should not be removed for 1.14, at least - if it's removed, it should be for 1.16.
20160328 04:18:07< aeth> Well, it should be removed if no one wants to maintain it after having basically the entire life of 1.14/1.15 to step up
20160328 04:18:13< SigurdFD> or not remove it at all
20160328 04:18:22< aeth> If someone wants to maintain it, no issues, right?
20160328 04:19:23< celticminstrel> Given that I've already done a lot of work in the formula engine, I could probably maintain it if the decision is to keep it.
20160328 04:19:51< celticminstrel> Oh, mattsc, are there any other undocumented parameters to the MicroAIs?
20160328 04:20:19< mattsc> Huh?
20160328 04:20:24< celticminstrel> When going through the source I noticed that the goto MAI passes its config directly to the find_path function.
20160328 04:21:02< celticminstrel> Meaning that eg "max_cost" will likely have an effect.
20160328 04:21:14< celticminstrel> ...though it might need to be added to "optional keys" first.
20160328 04:22:38< mattsc> Hmm, I don’t remember; but if that’s sp, yes
20160328 04:22:46< celticminstrel> sp?
20160328 04:23:00< mattsc> so;  sorry
20160328 04:23:13-!- ancestral [~ancestral@75-168-27-21.mpls.qwest.net] has joined #wesnoth-dev
20160328 04:23:48< celticminstrel> It's at line 192 of ca_goto.lua
20160328 04:24:16< celticminstrel> So it does use these parameters from [args], but currently the [micro_ai} tag doesn't pass them along. Should I change that?
20160328 04:25:04< celticminstrel> The options are max_cost, ignore_units, ignore_teleport, and viewing_side.
20160328 04:25:20< celticminstrel> The latter doesn't seem very useful, mind you.
20160328 04:25:25< celticminstrel> But the other three, maybe.
20160328 04:25:32< celticminstrel> I dunno.
20160328 04:25:34< celticminstrel> What do you think?
20160328 04:26:13< aeth> why are there two empty files in the Wesnoth source code? src/hotkey/hotkey_handler_replay.cpp and src/hotkey/hotkey_handler_replay.hpp
20160328 04:26:21< mattsc> Don’t know either.
20160328 04:26:29< Aginor> hmm
20160328 04:26:32< mattsc> Let me think about it.
20160328 04:26:35< celticminstrel> Okay.
20160328 04:26:36< Aginor> probably because I never noticed
20160328 04:26:50< celticminstrel> Aginor: Shall I delete them>
20160328 04:26:51< celticminstrel> ^?
20160328 04:27:01< Aginor> checking the log first
20160328 04:27:29< celticminstrel> Wow, they're completely empty.
20160328 04:27:39< Aginor> celticminstrel: you didn't use git move
20160328 04:27:41< celticminstrel> Not even a license header or anything.
20160328 04:27:48< aeth> I only noticed them because I was playing around with this: wc -l `find | grep '\(\.hpp$\)\|\(\.cpp$\)\|\(\.c$\)\|\(\.h$\)' | grep -v '\(lua\)\|\(SDL\)'` | less
20160328 04:27:53< celticminstrel> Eh?
20160328 04:27:58< celticminstrel> So I messed up somehow?
20160328 04:28:16< Aginor> celticminstrel: there's only one commit in the log, which is you creating the empty file
20160328 04:28:30 * celticminstrel definitely used git mv for a lot of things.
20160328 04:28:34< celticminstrel> And deleting the correct file?
20160328 04:28:44< celticminstrel> I can fix this, then.
20160328 04:29:01< celticminstrel> What's the commit hash?
20160328 04:29:07< Aginor> fe0df4594724ac08abce79b011e948500dda1411
20160328 04:29:20< aeth> In case anyone is wondering, the total in wc -l in src without lua and SDL is 317494 and the total with is 397328, only counting C and C++ files.
20160328 04:29:37< Aginor> they were empty beforehand though
20160328 04:30:32< celticminstrel> Hmm.
20160328 04:30:34< aeth> I'm not sure if there are other embedded libraries other than the two SDL ones and the lua one
20160328 04:30:50< aeth> Maybe xBRZ
20160328 04:31:09< Aginor> celticminstrel: you didn't screw up
20160328 04:32:01< celticminstrel> So should I delete them?
20160328 04:32:47< Aginor> I think that's probably just as well
20160328 04:33:20< Aginor> commit 9ca66784794fa6eed0fd74558a2368f42030087e suggests they should have been deleted already
20160328 04:33:38< Aginor> (git log --follow --  ../src/hotkey/hotkey_handler_replay.hpp)
20160328 04:34:09< celticminstrel> That looks like a convenient switch.
20160328 04:35:28< irker312> wesnoth: Celtic Minstrel wesnoth:master b555bc2bb7e1 / / (7 files in 5 dirs): Delete empty source files https://github.com/wesnoth/wesnoth/commit/b555bc2bb7e1eafc4154d4b31ac355b01b7fd229
20160328 04:45:54-!- travis-ci [~travis-ci@ec2-107-22-55-162.compute-1.amazonaws.com] has joined #wesnoth-dev
20160328 04:45:55< travis-ci> wesnoth/wesnoth#9079 (master - 49e7e03 : Celtic Minstrel): The build is still failing.
20160328 04:45:55< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/118914062
20160328 04:45:55-!- travis-ci [~travis-ci@ec2-107-22-55-162.compute-1.amazonaws.com] has left #wesnoth-dev []
20160328 04:46:04< celticminstrel> :|
20160328 04:46:24< celticminstrel> ...oh. It's back to erroring now.
20160328 04:46:39< celticminstrel> Well. I guess it's X's rather than !'s though.
20160328 04:49:53< celticminstrel> So mattsc, this is what I've come up with for manually resolving the attacks aspect in the Fast MAI (and, potentially, others that want to use it) without calculating the attack analysis.
20160328 04:49:54< celticminstrel> http://pastebin.com/gJ6NNXCq
20160328 04:50:34< celticminstrel> Should I go ahead and do it, or do you think this might be a bad idea for efficiency or whatever?
20160328 04:59:37-!- SigurdFD [~SigurdFD@dynamic-acs-72-23-176-151.zoominternet.net] has quit []
20160328 05:06:55< vultraz> celticminstrel: in c++, what's the purpose of putting things in a set of plain brackets? like foo(); { bar(); } bat();
20160328 05:07:24< celticminstrel> vultraz: It introduces a new lexical scope, so any local variables declared within those braces will not be visible outside of them.
20160328 05:07:48< celticminstrel> This also means that the destructors for those variables will be called at the closing brace.
20160328 05:07:56< celticminstrel> (If they have nontrivial destructors.)
20160328 05:08:27< vultraz> ahhh
20160328 05:09:52< Aginor> very useful on occasion
20160328 05:10:04< celticminstrel> Especially in a switch case.
20160328 05:10:13< celticminstrel> When you need case-local variables.
20160328 05:10:55< vultraz> is that also why if-statements should use brackets?
20160328 05:11:15< celticminstrel> No.
20160328 05:11:40< celticminstrel> If statements without brackets still have their own lexical scope, though the only way to declare a variable in that situation is to do it within the if statement itself.
20160328 05:11:41-!- travis-ci [~travis-ci@ec2-54-166-17-196.compute-1.amazonaws.com] has joined #wesnoth-dev
20160328 05:11:42< travis-ci> wesnoth/wesnoth#9080 (master - b555bc2 : Celtic Minstrel): The build has errored.
20160328 05:11:42< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/118915770
20160328 05:11:42-!- travis-ci [~travis-ci@ec2-54-166-17-196.compute-1.amazonaws.com] has left #wesnoth-dev []
20160328 05:13:13< celticminstrel> (At least, I think that's the case. I've never actually tried declaring a local variable as the statement of a non-bracketed if statement, and finding a case where that actually does anything useful would be pretty hard.)
20160328 05:13:57< mattsc> aeth, pydsigner, SigurdFD: just one clarification, that comment was added to the wiki in February, but Formula AI had already been unmaintained for a very long time then.
20160328 05:14:28< mattsc> Not that that changes any of the arguments or conclusions, just saying that this is not an ad hoc decision we suddenly decided to take.
20160328 05:21:44-!- ancestral [~ancestral@75-168-27-21.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20160328 05:48:56-!- Kwandulin [~Miranda@p200300760F191C5C24846DE75B663772.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160328 05:55:23-!- ancestral [~ancestral@75-168-27-21.mpls.qwest.net] has joined #wesnoth-dev
20160328 06:01:45< vultraz> ah, ancestral 
20160328 06:01:58< vultraz> ancestral: I tested out the top three fonts
20160328 06:05:17-!- ancestral [~ancestral@75-168-27-21.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20160328 06:07:43< Aginor> heh
20160328 06:07:55 * Aginor is looking at video.cpp:update_rect
20160328 06:08:01< Aginor> that's some pretty dead code right there
20160328 06:08:24-!- ancestral [~ancestral@75-168-27-21.mpls.qwest.net] has joined #wesnoth-dev
20160328 06:08:28-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20160328 06:08:50< ancestral> vultraz: Yeah?
20160328 06:09:03< ancestral> Top three?
20160328 06:09:12< vultraz> ancestral: Coval, Carlito, Lato
20160328 06:09:26< vultraz> Carlito seems the best, and at least zookeeper likes it
20160328 06:09:26< ancestral> How did you make then bigger?
20160328 06:09:34< vultraz> 90 dpi
20160328 06:09:36< ancestral> *them
20160328 06:09:38< vultraz> for Carlito
20160328 06:09:52< vultraz> 84 and 88 for Lato and Coval 
20160328 06:10:25< vultraz> I tested with Coval regular not Coval Book, and I think it looked better, but honestly, though my initial impressions a few days ago were good, I think Carlito  looks better
20160328 06:10:31< vultraz> than Coval
20160328 06:10:35< vultraz> blaghs I can't type here
20160328 06:11:05< ancestral> The font scaling, does that change the dpi, or is that a different mechanism?
20160328 06:11:43< ancestral> I tested Carlito and yeah, it seemed to look good
20160328 06:11:59< ancestral> Haven’t gotten others yet
20160328 06:12:03-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20160328 06:12:22< vultraz> (actually, scratch that, I tested with Coval book)
20160328 06:12:35< vultraz> I don't know how font scaling works
20160328 06:12:37< vultraz> ask celticminstrel 
20160328 06:12:58< celticminstrel> It changes the font size.
20160328 06:13:00< vultraz> but stuff is laid out better at 100% scaling
20160328 06:13:22< vultraz> so we should tweak settings to look best at 100%
20160328 06:14:03< vultraz> I'm currently working with wesnoth using Carlito just to get a feel of it
20160328 06:15:50< celticminstrel> So if you have 150% scaling, text requested at size 12 will be drawn at size 18.
20160328 06:16:26< celticminstrel> If it was 50% scaling (only possible by editing the prefs file), it would be drawn at size 6.
20160328 06:31:35< ancestral> vultraz: I think Wesnoth sjould remain at 72 in the code; we should update the theme to compensate
20160328 06:31:59< ancestral> celticminstrel: Does it round to the nearest integer?
20160328 06:32:22< celticminstrel> I think font sizes are integers, yes.
20160328 06:33:14< ancestral> “Carlito is metric-compatible with Calibri font.”
20160328 06:33:57< ancestral> “It is a modification of popular open-source font Lato, created by Łukasz Dziedzic.”
20160328 06:34:13< vultraz> ancestral: hah XD
20160328 06:34:27< ancestral> So Lato, metrically similar to Calibri
20160328 06:37:26< vultraz> ancestral: why do you suggest keeping the 72 dpi number?
20160328 06:37:32< vultraz> it has no special significance
20160328 06:37:36< ancestral> Well yes
20160328 06:38:05< ancestral> Apps use dpi to render things to the screen
20160328 06:38:24< vultraz> I can produce about the same result as 90 dpi by using 72 dpi and 130% font scaling
20160328 06:38:26< vultraz> but..
20160328 06:38:33< ancestral> There could be a time where the system might scale or rely on the app’s dpi
20160328 06:38:43-!- nos__ [~nos@208.91.185.104] has quit [Ping timeout: 252 seconds]
20160328 06:39:06< celticminstrel> I wonder why nagas are the only race (other than the obvious ones) with both male and female units.
20160328 06:39:19< ancestral> In other words, it would be wise to have the game looking good at 72 dpi since that is what most systems expect/use for other uses
20160328 06:39:27< vultraz> ehh
20160328 06:39:29< vultraz> perhaps
20160328 06:39:42< vultraz> but then someone other than I will have to find the magic font numbers
20160328 06:39:48< ancestral> vultraz: Windows, for example
20160328 06:39:56< ancestral> Has scaling
20160328 06:40:03< ancestral> 96 dpi is an option?
20160328 06:40:23< ancestral> There could be a future where Wesnoth supports that
20160328 06:40:45< ancestral> vultraz: Let’s do the math
20160328 06:41:25< ancestral> 90/72 = 125%
20160328 06:41:47< ancestral> So multiply the font sizes by 125%
20160328 06:42:08< ancestral> So 18 becomes 23
20160328 06:43:17< ancestral> 16 becomes 20; 14 > 18; 12 > 15; 10 > 13; 9 > 11
20160328 06:43:29< ancestral> (‘>’ means “becomes”)
20160328 06:43:41< vultraz> I know that much :P
20160328 06:43:45< ancestral> :-P
20160328 06:43:57< vultraz> I could have done that myself, tbh :P
20160328 06:44:30< ancestral> Okay so
20160328 06:45:37< vultraz> values look good
20160328 06:45:43< ancestral> I need to try some of this our
20160328 06:45:46< ancestral> *out
20160328 06:46:04< vultraz> 13, 15, 18, 23, 25
20160328 06:46:10< vultraz> new GUI2 font size values
20160328 06:46:21< vultraz> (tiny, small, default, large, title)
20160328 06:47:09< vultraz> gui1 needs a similar bump
20160328 06:50:22< celticminstrel> What's with these weird numbers.
20160328 06:50:24< ancestral> vultraz: default.cfg?
20160328 06:50:28< ancestral> Anywhere else?
20160328 06:50:40< celticminstrel> Why not 12, 16, 18, 24, 28, or some such.
20160328 06:50:44< vultraz> eh?
20160328 06:50:51< vultraz> ancestral: what is defaults.cfg
20160328 06:50:54< ancestral> celticminstrel: It can be ant y number
20160328 06:51:06< ancestral> vultraz: Where the sizes are defined
20160328 06:51:11< celticminstrel> vultraz: He probably means data/gui/default.cfg?
20160328 06:51:14< ancestral> themes/default.cfg?
20160328 06:51:22< celticminstrel> Oh. Okay then.
20160328 06:51:36< vultraz> the gui2 font sizes are in data/gui/macros/_initial.cfg
20160328 06:51:54< ancestral> celticminstrel: Yes, many editors and such use numbers which relate to picas and such
20160328 06:52:10< ancestral> Typically, numbers which are multiples of 6
20160328 06:52:41< celticminstrel> 16 is not a multiple of 6.
20160328 06:52:44< vultraz> unless you mean gui1
20160328 06:52:59< ancestral> Thanks
20160328 06:53:25< ancestral> celticminstrel: 16 is a more recent addition as people want more sizes
20160328 06:54:27< celticminstrel> ...why do I have a target goal with criteria [not]owner=0...
20160328 06:54:42< celticminstrel> To start with, isn't that a location filter, not a unit filter?
20160328 06:54:58< vultraz> ancestral: I don't know if default.cfg controls all of gui1's font sizes
20160328 06:55:00< vultraz> i doubt it
20160328 06:55:04< vultraz> look in font.cpp or something
20160328 06:57:43< celticminstrel> I wonder if the AI every leaves injured units in place so that they heal 2HP.
20160328 06:57:45< celticminstrel> ^ever
20160328 06:59:12< ancestral> https://images.duckduckgo.com/iu/?u=http%3A%2F%2Fwww.jnkmail.com%2Fstuff%2Fnostalgia%2Fmacwrite.gif&f=1
20160328 06:59:47-!- aeonchild [enchilado@defocus/yummy/enchilado] has quit [Ping timeout: 248 seconds]
20160328 07:00:12< ancestral> https://en.wikipedia.org/wiki/Pica_(typography)
20160328 07:00:14< celticminstrel> Outline and Shadow are no longer standard styles, either.
20160328 07:00:28< ancestral> “The pica contains 12 point units of measure.”
20160328 07:00:48< ancestral> celticminstrel: Hey, color wasn’t invented 30 years ago :-P
20160328 07:00:51< celticminstrel> ...though TextEdit does provide a more sophisticated shadow, so maybe shadow could be considered standard.
20160328 07:01:02< celticminstrel> ancestral: Liiiies. :P
20160328 07:01:26< ancestral> In lieu of color, shadows and outlines were popular
20160328 07:01:42< celticminstrel> I guess that does make sense though.
20160328 07:03:08-!- aeonchild [enchilado@defocus/yummy/enchilado] has joined #wesnoth-dev
20160328 07:03:21< ancestral> Dithering also helped: http://macintoshgarden.org/sites/macintoshgarden.org/files/screenshots/usa.png
20160328 07:03:52< ancestral> http://macintoshgarden.org/games/where-in-the-usa-carmen-sandiego
20160328 07:04:41-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20160328 07:04:51< celticminstrel> That's not a HyperCard program, is it?
20160328 07:04:58< celticminstrel> It kinda looks like it could be.
20160328 07:05:01< ancestral> Jetrel: Imagine making Wesnoth sprites in 1-bit color!
20160328 07:05:27< ancestral> celticminstrel: It’s not afair. but it almost could have been
20160328 07:05:46< Jetrel> ancestral:  imagine The Return of the Obra Dinn!
20160328 07:05:48< celticminstrel> Myst was a HyperCard program originally.
20160328 07:06:04< celticminstrel> Huh, Jetrel was summoned.
20160328 07:06:14< irker312> wesnoth: Andreas Löf wesnoth:renderpath_redo 27b79a002219 / src/ (5 files in 2 dirs): Add a pre- and post-draw event https://github.com/wesnoth/wesnoth/commit/27b79a002219c27da7549eaf0e61b5c9225a2b46
20160328 07:06:14< ancestral> Correct
20160328 07:06:21 * Jetrel expands into a cloud of fire and smoke
20160328 07:07:13< Jetrel> celticminstrel:  yeah - I remember watching a companion video they had for Myst where they talked (briefly!) about development, and about writing code-extensions to blit the graphics off the CD.  It was really cool.
20160328 07:07:26< Jetrel> celticminstrel:  myst has a very special place in my heart as a gamer.  :)
20160328 07:07:54< celticminstrel> I never actually had the HC version myself, though.
20160328 07:08:06< Jetrel> I didn't buy the rest of the series - I was a kid with no income at the time, and it sort of "passed me by", but it was very cool.
20160328 07:08:29< Jetrel> celticminstrel:  I had that of the first game.  Many years later, I bought the iPad version of RealMyst
20160328 07:09:02< celticminstrel> I think I got that in a bundle on GoG.
20160328 07:09:11< Jetrel> The "real 3d" remake was a very, very good rendition; the sort of perfect remake someone hopes for.
20160328 07:09:14< celticminstrel> (Not the iPad version, of course.)
20160328 07:09:34< celticminstrel> I see.
20160328 07:10:43< vultraz> there's an ipad version of myst?
20160328 07:10:46< Jetrel> I mean, the early hypercard iterations were absolutely painful from a UX perspective, because the static "movement between screens" was so disorienting.
20160328 07:10:54< Jetrel> _BUT_
20160328 07:11:03< celticminstrel> vultraz: Yes, there is.
20160328 07:11:30< Jetrel> In spite of that, it was overwhelmingly worth it, because it was so insanely ahead-of-its-time in terms of what it was presenting, visually.
20160328 07:11:36< ancestral> Jetrel: I cut my teeth on HyperCard
20160328 07:11:47< Jetrel> ancestral:  believe it or not, I did too. :D
20160328 07:12:20< Jetrel> I made a really bad implementation of "nim", the math-class game where you take away 2 or 3 tokens to force the other player to take the last one.
20160328 07:12:26< vultraz> celticminstrel: huh
20160328 07:12:33< vultraz> celticminstrel: i've heard good things about the game
20160328 07:12:37< vultraz> is it worth getting
20160328 07:12:45< celticminstrel> I used to do HyperCard stuff too.
20160328 07:13:04< celticminstrel> vultraz: If you like puzzles and the point-and-click style, then I'd say it's probably worth getting.
20160328 07:13:27< Jetrel> vultraz:  it's a classic game that's ... if you're a game designer, I think it's almost "required reading" to play through it just to study the experience.
20160328 07:13:29< celticminstrel> Myst is basically one big puzzle, so if you hate puzzles, it may not be worth it.
20160328 07:13:30< ancestral> If I was older back then, I totally would have made and shipped HyperCard and World Builder games
20160328 07:14:13< ancestral> World Builder adventure games: http://macintoshgarden.org/engine/world-builder
20160328 07:14:18< vultraz> I like puzzles
20160328 07:14:29< Jetrel> vultraz:  without spoiling anything, the basic plot of the game is that you're some person (your identity as the player is deliberately left up to your imagination).  You walk into some cave and you find a book on the floor.
20160328 07:14:31< vultraz> One of my fav series is a point and click HoG puzzle-y game :P
20160328 07:15:03< celticminstrel> I'm not sure if I've encountered World Builder before.
20160328 07:15:16< celticminstrel> HoG?
20160328 07:15:47< Jetrel> vultraz:  you open this book, and read into it, and it's got this incredibly vivid description of this tiny island - you can totally picture it perfectly in your head.  You idly put your hand on one page, and then suddenly
20160328 07:15:53< Jetrel> magic happens
20160328 07:15:56< vultraz> celticminstrel: hidden object game
20160328 07:16:02< Jetrel> and you're ON the island.
20160328 07:16:06< celticminstrel> Oh...
20160328 07:17:03< Jetrel> vultraz:  so you quickly discover that this island appears to have had 4 people living on it - a man named Atrus, his wife Catherine, and his two sons Sirrus, and Achenar.  But they're gone, somehow.
20160328 07:17:56< ancestral> Kinda ashame those the two Myst supernerds never got their Myst movie done
20160328 07:18:28< Jetrel> vultraz:  and clues hint at two things - somehow, this business of the books is some serious magic, almost like the islands are being *created* by writing about them, or maybe they always existed, and the books just discover them out there in the ether...   but more importantly this family seems to have some kind of magic that lets them make these sorts of books.
20160328 07:18:54< Jetrel> vultraz:  and something terrible's happened - one of the two sons did ... something, and now something's gone horribly wrong
20160328 07:19:24< Jetrel> vultraz:  so you ... our archetypical stranger walking onto the scene ... have to figure out what happened.
20160328 07:19:49< vultraz> is it creepy?
20160328 07:19:53< Jetrel> vultraz:  is anyone still alive?  What went wrong?
20160328 07:20:02< vultraz> I recall looking at it on amazon once and getting the impression it should be creepy
20160328 07:20:06< Jetrel> vultraz:  only psychologically, it's not horror.
20160328 07:20:26< celticminstrel> I don't think it's all that creepy.
20160328 07:20:36< celticminstrel> There might be some creepy details if you look closely enough.
20160328 07:20:37< Jetrel> vultraz:  there's no gore/monsters/any of that sort of motif.
20160328 07:21:03< Jetrel> vultraz:  there is some scenery where if you stop and think about what it means, it's got some dark implications.
20160328 07:21:44< celticminstrel> That's what I meant with "if you look closely enough".
20160328 07:21:47< celticminstrel> More or less.
20160328 07:21:53< Jetrel> celticminstrel:  aye. :)
20160328 07:22:06< Jetrel> vultraz:  It is possible to lose the game, but you can't die from any traps or monsters or any of that, and there's no bogeyman coming to get you.
20160328 07:22:36< celticminstrel> So I'm using a big animals MicroAI... doesn't that mean I don't need the attacks aspect anymore...
20160328 07:22:37< Jetrel> You don't fight, and you don't kill.
20160328 07:23:56< vultraz> I was referring to creepy atmosphere
20160328 07:24:08< vultraz> sounds like that it does have
20160328 07:24:33< Jetrel_bot> vultraz:  yeah, but not like you'd usually expect
20160328 07:24:45< Jetrel_bot> vultraz:  you're not gonna get the heebie-jeebies.
20160328 07:25:08< Jetrel_bot> They're not trying to scare you, is really the thing - they are trying to tell a good story, and it has good villains.
20160328 07:31:23< Aginor> myst is a great game
20160328 07:31:32< vultraz> celticminstrel: one of my all time fav games http://www.bigfishgames.com/daily/mystery-case-files/games/diregrove
20160328 07:32:24-!- nos__ [~nos@208.91.185.104] has joined #wesnoth-dev
20160328 07:33:16< celticminstrel> ...seriously? 1 hour trial? That's ridiculous.
20160328 07:33:55< celticminstrel> That aside, I guess it sounds vaguely interesting.
20160328 07:34:25< Aginor> no linux support :/
20160328 07:38:55< ancestral> Not my cup of tea
20160328 07:38:59< vultraz> celticminstrel: the game takes maybe 8 hours or so to complete
20160328 07:39:05< vultraz> depending on your skill with puzzles
20160328 07:39:13< vultraz> I've done it in 2
20160328 07:39:25< vultraz> (that is, as a speedrun)
20160328 07:39:32< vultraz> (not a normal gameplay session)
20160328 07:40:05< celticminstrel> Then the trial should be limited by something other than time.
20160328 07:41:02< vultraz> it is, actually
20160328 07:41:08< vultraz> certain point or 1 hour
20160328 07:41:12< vultraz> whichever comes first
20160328 07:41:45< celticminstrel> Yeah, just drop the time limitation then. There's no point.
20160328 07:43:25-!- travis-ci [~travis-ci@ec2-184-72-134-120.compute-1.amazonaws.com] has joined #wesnoth-dev
20160328 07:43:26< travis-ci> wesnoth/wesnoth#9081 (renderpath_redo - 27b79a0 : Andreas Löf): The build has errored.
20160328 07:43:26< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/118927574
20160328 07:43:26-!- travis-ci [~travis-ci@ec2-184-72-134-120.compute-1.amazonaws.com] has left #wesnoth-dev []
20160328 07:43:43< ancestral> vultraz: It might make more sense to use Lato over Carlito as there are more glyphs and appears it’s more quickly updated since it’s the source
20160328 07:44:04< vultraz> ancestral: I don't think Lato looks as good, though
20160328 07:44:39< ancestral> I’m going to try it at slightly larger
20160328 07:45:07< vultraz> looking at the screenshot again, though...
20160328 07:45:09< vultraz> ehh..
20160328 07:45:10< vultraz> it could work
20160328 07:45:14< vultraz> it's hard to decide
20160328 07:45:15< ancestral> 133%
20160328 07:45:28< ancestral> In my text editor
20160328 07:45:46< ancestral> Look almost identical, except size is 1 pt difference
20160328 07:45:59< ancestral> vultraz: Also, more weights in Lato
20160328 07:46:00< vultraz> for some reason dropbox degrades my picture quality
20160328 07:46:28< ancestral> vultraz: Stick it up on vultraz.wesnoth.org
20160328 07:46:32< ancestral> or whatever
20160328 07:46:47< vultraz> I should switch to google drive
20160328 07:46:55< vultraz> I don't have a wesnoth.org domain
20160328 07:47:07< ancestral> I think you have a user acct.
20160328 07:47:22< vultraz> shadowm: do elaborate on this
20160328 07:47:53< ancestral> Jet puts art wip up occasionally?
20160328 07:48:04< vultraz> I know
20160328 07:49:03< ancestral> Otherwise, you can have ftp access on wesnoth.io if you like
20160328 07:49:36< ancestral> (I use like 1% bandwidth allotted to my server)
20160328 07:49:48< vultraz> I'll just switch to goggle drive
20160328 07:49:52< vultraz> google
20160328 07:50:08< celticminstrel> Goggle at your drive.
20160328 07:50:25< ancestral> imgur might lq too? not sure
20160328 07:51:23< ancestral> GitHub could work, but they’ll start denying requests if you hotlink too much
20160328 07:51:49< ancestral> Otherwise, turn the “project” into a web page and that’s acceotabke
20160328 07:51:57< ancestral> *acceptable
20160328 07:55:40< vultraz> ancestral: Lato https://drive.google.com/file/d/0B-mR9s8FduLLQk5qSDhEaG1uaVE/view?usp=sharing with font size values 12,13,17,20,22
20160328 07:55:52< vultraz> still needs size tweaks
20160328 07:57:59< vultraz> ancestral: more normal 12,14,18,20,22 values: https://drive.google.com/file/d/0B-mR9s8FduLLd25GV0FzYlFaajA/view?usp=sharing
20160328 08:01:13< ancestral> vultraz: Try taking a pic of the prefs dialog
20160328 08:02:51< vultraz> ancestral: https://drive.google.com/file/d/0B-mR9s8FduLLTTFpLWFtdmdhUWc/view?usp=sharing
20160328 08:03:18< vultraz> it just doesn't look as good
20160328 08:05:58< shadowm> ancestral: He doesn't have a user account and that's not how it works anyway.
20160328 08:06:01< ancestral> Honestly
20160328 08:06:06< celticminstrel> I can't tell the difference because Google is thwarting my cache.
20160328 08:06:55< ancestral> shadowm: Okay. Just trying to find solutions for our most involved developer.
20160328 08:07:09< celticminstrel> I suppose if I opened them in two separate tabs it would work...
20160328 08:07:34< ancestral> vultraz: I like 14/16/18/20/24
20160328 08:08:16< shadowm> Knowing vultraz and especially given a discussion we had earlier, I greatly doubt he'd find rsync more convenient than Google Drive/Dropbox.
20160328 08:09:31< shadowm> But in general it'd be great if he could ask the pertinent questions himself rather than leaving it to others to fill in the blanks.
20160328 08:10:16< ancestral> shadowm, vultraz: I use sftp in an ftp client for my own site. Works well for me.
20160328 08:10:37< shadowm> Same thing pretty much.
20160328 08:10:50< ancestral> Not feature rich, of course
20160328 08:11:02< shadowm> Unless you have a specific recommendation (for Windows) that will sync a local directory unattended.
20160328 08:11:25< shadowm> And which preferably has a super-shiny modern UI with no confusing options.
20160328 08:11:27< ancestral> Google Drive probably has nice things like mobile interfaces and whatnot
20160328 08:19:45< ancestral> vultraz: Okay so actually, I like 14/16/18/20/24 for Carlito. For Lato I like 12/14/16/18/22.
20160328 08:20:18< ancestral> Lato Medium might render better?
20160328 08:22:49-!- ChipmunkV [~vova@d0017-2-88-172-31-68.fbx.proxad.net] has joined #wesnoth-dev
20160328 08:23:16< ancestral> I think the line height is the only remaining issue I see
20160328 08:25:41< ancestral> Hmm, take that back
20160328 08:27:47-!- Kwandulin [~Miranda@p200300760F191C5C24846DE75B663772.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160328 08:31:04-!- mjs-de [~mjs-de@p508C8C99.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160328 08:35:40< vultraz> ancestral: I believe I did try LAto Medium during my testing yesterday
20160328 08:35:46< vultraz> it wasn't that great
20160328 08:35:54< vultraz> bit too much weight
20160328 08:36:11< ancestral> Fair enough
20160328 08:36:54< ancestral> 13/15/17/20/24 might be better
20160328 08:36:58< ancestral> vultraz: Alright here’s the deal
20160328 08:37:06< ancestral> Carlito is not being actively managed
20160328 08:37:18< ancestral> Google is to blame
20160328 08:37:20-!- aeonchild [enchilado@defocus/yummy/enchilado] has quit [Ping timeout: 250 seconds]
20160328 08:37:52< vultraz> Oh?
20160328 08:37:53< celticminstrel> You're going with weird font sizes again.
20160328 08:37:54< ancestral> https://en.wikipedia.org/wiki/Croscore_fonts#Crosextra_fonts https://bodhi.fedoraproject.org/updates/?packages=google-crosextra-carlito-fonts
20160328 08:38:46< ancestral> celticminstrel: “Weird” meaning it’s not well divisible into pica units
20160328 08:39:00< ancestral> The actual number does not matter at all
20160328 08:39:05< celticminstrel> I suppose.
20160328 08:39:37< ancestral> vultraz: Pics forthcoming
20160328 08:39:43-!- celticminstrel is now known as celmin|sleep
20160328 08:41:08< vultraz> ancestral: is active maintenance a hard selling point for us?
20160328 08:41:25< ancestral> Well
20160328 08:41:46< ancestral> What if they added Hebrew?
20160328 08:42:12< ancestral> What if they added extra characters?
20160328 08:42:38< ancestral> And what if Wesnoth wanted to support extra weights?
20160328 08:44:31< ancestral> I don’t know what conversion they did to make Carlito
20160328 08:44:42< ancestral> vultraz: http://mproud.com/wesnoth/lato-1.png http://mproud.com/wesnoth/lato-2.png
20160328 08:45:52< vultraz> Rendering is much smoother for you, I see
20160328 08:47:05< ancestral> How does 13/15/17/20/24 look with Lato for you?
20160328 08:48:00< ancestral> Lato was updated in 2013/14 with an extra few hundred glyphs
20160328 08:48:40< vultraz> pretty good, actually..
20160328 08:48:44-!- aeonchild [enchilado@defocus/yummy/enchilado] has joined #wesnoth-dev
20160328 08:51:02< vultraz> ancestral: https://drive.google.com/file/d/0B-mR9s8FduLLZHhEdkswMjB1akk/view?usp=sharing
20160328 08:52:11< vultraz> I like those numbers
20160328 08:52:17< ancestral> Yeah, the line height is the thing that needs minor adjustment, but I feel like that can be solved
20160328 08:53:53< ancestral> What did you think of Noto Sans?
20160328 08:58:29< vultraz> Nahh
20160328 08:58:33< vultraz> Too plain looking
20160328 09:01:36< ancestral> I wonder if Carlito stripped hinting
20160328 09:01:50< ancestral> vultraz: Is there a setting to turn hinting off for fonts in the Wesnoth code?
20160328 09:01:56< ancestral> Please try turning it off
20160328 09:02:31< ancestral> fonts.conf apparently
20160328 09:02:48< vultraz> that isn't used on windows
20160328 09:03:05< vultraz> do you mean cairo_font_options_set_hint_style(fo, CAIRO_HINT_STYLE_FULL);
20160328 09:03:07< vultraz> ?
20160328 09:03:13-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20160328 09:04:20< ancestral> Oh
20160328 09:04:22< ancestral> Yes
20160328 09:04:49< ancestral> vultraz: fonts.conf isn’t used at all?
20160328 09:05:00< ancestral> (I’m not talking about subpixel rendering)
20160328 09:05:04< vultraz> ancestral: https://drive.google.com/file/d/0B-mR9s8FduLLV0pEeUpkWEFOLWs/view?usp=sharing using CAIRO_HINT_STYLE_NONE and a 0.5 reduction in line height 
20160328 09:05:16< vultraz> I'm off to dinner, will check in in ~an hout
20160328 09:05:18< vultraz> hour
20160328 09:05:24< ancestral> Sure
20160328 09:05:24< vultraz> ancestral: not on windows, according to shadowm 
20160328 09:05:59< ancestral> (btw I think line height was too short, not the other way around)
20160328 09:06:30< ancestral> Or maybe it seems the baseline sits too low 
20160328 09:07:27< ancestral> It seems the hinting might not factor much in this case
20160328 09:08:22-!- Kwandulin [~Miranda@p200300760F191C5CE0E5919B26128A49.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160328 09:16:48< ancestral> I think #define GUI__TEXT_VERTICALLY_CENTRED is what needs teaking
20160328 09:16:51< ancestral> *tweaking
20160328 09:24:49< ancestral> In other news
20160328 09:24:57< ancestral> I’m getting a crash on turn 2 of the tutorial
20160328 09:24:59< ancestral> Assertion failed: (px != 0), function operator*, file Headers/boost/smart_ptr/shared_ptr.hpp, line 646.
20160328 09:26:24< ancestral> Looks like any map
20160328 09:27:12< shadowm> /fonts/fonts.conf is ignored on Windows because it defines settings that are already hardcoded on Windows.
20160328 09:27:30< ancestral> It looks like it may be ignored on OS X too
20160328 09:27:39< shadowm> (And which for whatever stupid reason are ignored on X11 if they are hardcoded.)
20160328 09:27:57< ancestral> I don’t believe I’m using X11
20160328 09:28:02< ancestral> But good to note
20160328 09:28:04< shadowm> Obviously not.
20160328 09:28:34< ancestral> It is possible to build Pango/Cairo with X11. Ah, but Wesnoth would be unaffected.
20160328 09:28:38< ancestral> Never mind then
20160328 09:29:13< ancestral> Anyway, changing the hintstyle in that file did nothing, when comparing screenshots
20160328 09:29:25< shadowm> There's really no point in building for X11 on OS X nowadays I think. I believe it doesn't even ship with the quartz-based X server nowadays?
20160328 09:29:47< ancestral> Correct, the X server isn’t installed anymore
20160328 09:30:16< ancestral> XQuartz is the open version of Apple’s X11, which I believe Apple is done supporting
20160328 09:30:31< ancestral> XQuartz is not required by most apps
20160328 09:30:42< ancestral> Except for things like GIMP
20160328 09:30:50< ancestral> UNIX ports and such
20160328 09:30:52-!- aeonchild [enchilado@defocus/yummy/enchilado] has quit [Quit: Lost terminal]
20160328 09:31:04-!- aeonchild [enchilado@defocus/yummy/enchilado] has joined #wesnoth-dev
20160328 09:31:05< ancestral> Oh, and Wine
20160328 09:31:10< shadowm> Considering that even on Linux land people want to see X11 dead...
20160328 09:31:52-!- ancestral [~ancestral@75-168-27-21.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20160328 09:51:12-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Quit: wedge009]
20160328 09:51:55-!- horrowind [~Icedove@2a02:810a:83c0:1c18:21b:fcff:fee3:c3ff] has joined #wesnoth-dev
20160328 10:02:16-!- nos__ [~nos@208.91.185.104] has quit [Quit: Leaving]
20160328 10:10:54-!- Appleman1234 [~Appleman1@KD106161147208.au-net.ne.jp] has quit [Remote host closed the connection]
20160328 10:35:36-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20160328 11:00:34-!- irker312 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20160328 11:15:02-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 260 seconds]
20160328 11:15:10-!- Appleman1234 [~Appleman1@KD106161147208.au-net.ne.jp] has joined #wesnoth-dev
20160328 11:26:24-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20160328 12:27:13-!- Kwandulin [~Miranda@p200300760F191C5CE0E5919B26128A49.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160328 12:45:39-!- Kwandulin [~Miranda@p200300760F191C1814648915AF7CF49B.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160328 13:23:58-!- irker590 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20160328 13:23:58< irker590> wesnoth: ln-zookeeper wesnoth:master 6026780539ea / data/core/terrain.cfg: Fixed wmllint mistake in 2cd4298f https://github.com/wesnoth/wesnoth/commit/6026780539eafaf2ab057103b540d1bcc34aa007
20160328 13:45:20-!- gfgtdf [~chatzilla@f054141215.adsl.alicedsl.de] has joined #wesnoth-dev
20160328 13:45:39< gfgtdf> zookeeper: you already changed the "Mdy" string?
20160328 13:45:52< zookeeper> not yet
20160328 13:46:02< zookeeper> but yeah i should get to that ASAP
20160328 13:52:58-!- travis-ci [~travis-ci@ec2-107-22-55-162.compute-1.amazonaws.com] has joined #wesnoth-dev
20160328 13:52:59< travis-ci> wesnoth/wesnoth#9082 (master - 6026780 : ln-zookeeper): The build failed.
20160328 13:52:59< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/118969451
20160328 13:52:59-!- travis-ci [~travis-ci@ec2-107-22-55-162.compute-1.amazonaws.com] has left #wesnoth-dev []
20160328 14:23:03< irker590> wesnoth: mattsc wesnoth:master c82ae94ae97e / data/ai/micro_ais/mai-defs/recruiting.lua: Random Recruiting Micro AI: fix [probability] tag https://github.com/wesnoth/wesnoth/commit/c82ae94ae97e7b08d56ff51fddb2e2ada39b6ff4
20160328 14:24:58< zookeeper> wot...
20160328 14:24:58< zookeeper> cl : Command line error D8036: '/FoReleaseDEBUG\utils' not allowed with multiple source files
20160328 14:25:04< mattsc> celmin|sleep: ^ that is the only actual bug I have found so far (I have not looked at protect_unit yet)
20160328 14:26:02< mattsc> celmin|sleep: also, the assert I wrote you about happens not only in test scenarios.  You can start HttT and it gets triggered as soon as control switches from the first to the second AI-controlled side.
20160328 14:27:43< mattsc> It’s the same one ancestral mentioned a few hours ago, btw.
20160328 14:31:27< zookeeper> wedge009, ^ any idea what's with that error?
20160328 14:45:23-!- gfgtdf [~chatzilla@f054141215.adsl.alicedsl.de] has quit [Quit: ChatZilla 0.9.92 [Firefox 45.0.1/20160315153207]]
20160328 14:54:00-!- Duthlet [~Duthlet@p4FC92119.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160328 14:59:00-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection]
20160328 14:59:06-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev
20160328 15:17:31-!- midzer [~quassel@p200300624E796C017994CA611A3D699B.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160328 15:19:48-!- Kwandulin [~Miranda@p200300760F191C1814648915AF7CF49B.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160328 15:28:33-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection]
20160328 15:28:39-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev
20160328 15:32:03-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160328 15:33:44< celmin|sleep> So I just realized that MAIs that accept tags are probably semi-broken. I think this only really affects Protect Unit and Random Recruitment.
20160328 15:34:16< mattsc> celmin|sleep: I just fixed random recruitment ^
20160328 15:34:41< celmin|sleep> I know you fixed something there, yes; but does it honour multiple probability tags?
20160328 15:35:06< mattsc> Oh, good point, I did not try that.
20160328 15:41:17< celmin|sleep> Any comments yet on the thing about pathfinder options in goto MAI?
20160328 15:41:32< mattsc> Yes; they should not be included.
20160328 15:42:05< mattsc> The way the Goto MAI works, it relies on the pathfinding to be straight pathfinding.
20160328 15:42:44< mattsc> When it says “nono-standard pathfinding” in the comment a dozen or so lines above, that means the non-standard options of the Goto MAI, not of find_path.
20160328 15:42:54< mattsc> *non-standard
20160328 15:43:19< mattsc> So, they are not hidden options and they are intentionally not added as optional keys.
20160328 15:43:32< mattsc> celmin|sleep: can you reproduce the assert?
20160328 15:43:41< celmin|sleep> Haven't tried yet.
20160328 15:44:24< mattsc> Okay; I think this is the highest priority, as it breaks almost all scenarios at the moment.
20160328 15:45:44< celmin|sleep> Trying now.
20160328 15:46:33< celmin|sleep> Also, any comments yet on that thing about Fast MAI and the attacks aspect?
20160328 15:47:03< mattsc> No, I have some question back to you about that, but don’t really have time for it.
20160328 15:47:28< mattsc> Maybe you could remind me why exactly you are trying to do that.
20160328 15:47:54< mattsc> I’m sure it’s in the logs, but we’ve talked about a lot of things in the last few days ...
20160328 15:48:30< celmin|sleep> Currently, in order to use the attacks aspect, it simply looks into the AI config and takes the first facet, if present.
20160328 15:48:45< celmin|sleep> And it relies on it to be a standard aspect.
20160328 15:49:01< celmin|sleep> So, it won't honour turns= and time_of_day=.
20160328 15:49:17< celmin|sleep> It won't work if the aspect is Lua or standard, or if the facet is Lua or standard.
20160328 15:49:46< celmin|sleep> The second standard should be composite.
20160328 15:49:57-!- celmin|sleep is now known as celticminstrel
20160328 15:50:50< celticminstrel> It's more or less the same as the case with the avoid aspect.
20160328 15:51:00-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20160328 15:51:13< mattsc> Okay, so in that case, it sounds like it is broken and, if we want it to work as the others, it should be fixed.  So what specifically are you asking me?  (sorry, not being argumentative, just (still) tired)
20160328 15:51:38< mattsc> Btw, one specific question, are turns= and time_of_day= mutually exclusive?
20160328 15:51:45< celticminstrel> They are not.
20160328 15:51:53< mattsc> Because IIRC, your code currently only deals with one or the other
20160328 15:52:07< mattsc> there’s an ‘elseif’ between the two, I think
20160328 15:52:24< celticminstrel> They aren't mutually exclusive, but if either fails to match, then the facet is not active.
20160328 15:53:01< mattsc> Oh, I see.  Right.
20160328 15:53:20< mattsc> Sorry, I’m rushing along a bit here.  I need to leave in about 30 seconds ...
20160328 15:53:46< mattsc> I’ll be back in a couple hours.
20160328 15:54:04< celticminstrel> I think I was asking because the Fast MAI presumably needs to be fast, so I'm not sure whether my function is acceptable for that. The alternative way to fix it would be somehow using ai.get_attacks().
20160328 15:56:29-!- Kwandulin [~Miranda@p200300760F191C1800D27882E14D8529.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160328 15:58:25< celticminstrel> Looks like the attacks aspect assert is a trivial error.
20160328 16:00:19< celticminstrel> Basically triggered by an attacks aspect that lacks one of the filters. Which is almost all attacks aspects, since the default lacks both of them.
20160328 16:00:32< celticminstrel> I forgot to check whether the filter exists before using it.
20160328 16:04:33< celticminstrel> I wish there was some way to step through Lua code for debugging. (unrelated to above)
20160328 16:13:31-!- mjs-de [~mjs-de@p508C8C99.dip0.t-ipconnect.de] has quit [Remote host closed the connection]
20160328 16:36:16-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20160328 16:59:19-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection]
20160328 16:59:25-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev
20160328 17:01:42< irker590> wesnoth: Celtic Minstrel wesnoth:master dd02202fc831 / src/ai/default/aspect_attacks.cpp: Fix attacks aspect assert https://github.com/wesnoth/wesnoth/commit/dd02202fc8316bc1cbe44f8958efd120d4988e4d
20160328 17:01:44< irker590> wesnoth: Celtic Minstrel wesnoth:master 32c3abbbaf3c / data/ai/micro_ais/ (10 files in 2 dirs): Micro AIs: Distinguish attribute and tag parameters and copy all of the latter https://github.com/wesnoth/wesnoth/commit/32c3abbbaf3c7bae6dcfdd3c814e0fe48aaedcf2
20160328 17:02:52< vultraz> celticminstrel: https://drive.google.com/file/d/0B-mR9s8FduLLZHhpZmtfSVlEZVk/view?usp=sharing
20160328 17:03:04< vultraz> new loadscreen design
20160328 17:03:29< celticminstrel> o.o
20160328 17:03:38< celticminstrel> Uhh.
20160328 17:03:53< celticminstrel> I can't judge it without seeing it in motion.
20160328 17:04:05< vultraz> There's no motion
20160328 17:04:14< celticminstrel> There has to be motion.
20160328 17:04:19< vultraz> Not really
20160328 17:04:27< celticminstrel> No ,really.
20160328 17:04:34< Kwandulin> looks fresh
20160328 17:04:38< vultraz> Why?
20160328 17:05:02< celticminstrel> Because otherwise there's no indication of how close it is to finishing.
20160328 17:05:59< vultraz> Yes. But there's no need for such things.
20160328 17:06:10< celticminstrel> I disagree strongly. If the load time was only a second or two, perhaps it wouldn't matter, but it's usually quite a significant delat.
20160328 17:06:15< celticminstrel> ^delay
20160328 17:08:17< vultraz> Under the current implementation, you're left staring at a static progress bar anyway. It doesn't progress until each stage is complete, which tell you nothing about how long the current stage is going to taje
20160328 17:08:19< vultraz> take
20160328 17:08:38< vultraz> 'Creating cache' is usually the longest, but there's absolutely no indication how long it may be creating the cache
20160328 17:09:01< vultraz> So really, this isn't a downgrade, IMO
20160328 17:09:50< celticminstrel> That's not really true...
20160328 17:09:52-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160328 17:10:00< celticminstrel> The bar progresses within each stage.
20160328 17:10:20< vultraz> No it doesn't
20160328 17:10:28< celticminstrel> Yes it does. I just tested now.
20160328 17:11:51< vultraz> Well, it at least does not the point where it most usually will stick with large UMC 
20160328 17:11:57< celticminstrel> Even if it only progressed when each stage is complete, though, this would still be a downgrade.
20160328 17:12:00< celticminstrel> Huh?
20160328 17:12:06< vultraz> Creating Cache
20160328 17:12:30< vultraz> If you have larrrgeee UMC it will get to that point and sit there for awhile without moving
20160328 17:12:40< celticminstrel> Ah.
20160328 17:12:52< celticminstrel> That's a minor problem that's not easily solved.
20160328 17:13:17< celticminstrel> However, even if it currently only progressed when each stage is complete, removing motion altogether is still a downgrade.
20160328 17:13:33< vultraz> Plus, knowing the stages doesn't really do anything, since any relevant info will be displayed in the error popup anyway
20160328 17:13:42< celticminstrel> Because even without progression within a stage, the position of the bar would still give you an indication of how much is left.
20160328 17:14:13< vultraz> Not really, since it's not linear.
20160328 17:14:21< vultraz> It's like the Windows Update progress bar :P
20160328 17:14:36< celticminstrel> It doesn't matter that it's not linear. It matters that it gives you a rough estimate of progress.
20160328 17:16:48< vultraz> Well, let's try this out for the next release. We will certainly get feedback, since it's a very obvious change.
20160328 17:16:55< vultraz> Just need to do some cleanup..
20160328 17:17:17< celticminstrel> Here, don't just take my word for this - here's what some profession game designer has to say on the subject: http://www.designersnotebook.com/Columns/112__BGDNT_XII/noprogre112__bgdnt_xii.htm (Scroll near bottom)
20160328 17:17:32< celticminstrel> ^professional
20160328 17:17:42< celticminstrel> This is a very bad change.
20160328 17:18:17< celticminstrel> The loadscreen looks nice, I'll give you that much; but a lack of motion is completely unacceptable.
20160328 17:18:32< vultraz> Dota 2 has a loading screen with no motion or even a 'loading' label :p
20160328 17:18:35-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160328 17:18:51< celticminstrel> That doesn't change the argument.
20160328 17:18:57< vultraz> I figure if Valve does it, it must not be all bad
20160328 17:19:01< celticminstrel> Just marks it as one of the ones that does it wrong. :P
20160328 17:19:20< vultraz> But again, we cannot decide this with two people
20160328 17:19:25< vultraz> We'll need community feedback
20160328 17:21:26< celticminstrel> You could wait for a bit more dev feedback too.
20160328 17:22:34< celticminstrel> Did you read the relevant section in the link?
20160328 17:24:32< vultraz> yes
20160328 17:28:25< vultraz> I agree with the 'spamming progressbar' bit
20160328 17:28:43< celticminstrel> Which bit was that, the IE comment?
20160328 17:31:42< vultraz> the bit about progressbars that fill up again and again and again
20160328 17:31:55< celticminstrel> Ah, yeah.
20160328 17:32:08< celticminstrel> I really think it's important to see movement as well, though.
20160328 17:41:12-!- travis-ci [~travis-ci@ec2-184-72-134-120.compute-1.amazonaws.com] has joined #wesnoth-dev
20160328 17:41:13< travis-ci> wesnoth/wesnoth#9086 (master - 32c3abb : Celtic Minstrel): The build has errored.
20160328 17:41:13< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/119014275
20160328 17:41:13-!- travis-ci [~travis-ci@ec2-184-72-134-120.compute-1.amazonaws.com] has left #wesnoth-dev []
20160328 17:44:13-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160328 17:46:17< mattsc> celticminstrel: Yes, the Fast MAI needs to be fast, and I cannot tell you right now whether that code will do that or not.
20160328 17:46:29< mattsc> Also, rats, now I need to test all the MAIs again.  :|
20160328 17:47:14< celticminstrel> Hmm. Maybe I'll do some comparisons.
20160328 17:48:12< mattsc> I just don’t have time to do any more than some very cursory things right now; maybe tonight, but I can’t promise that
20160328 17:50:04-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160328 17:52:10< mattsc> celticminstrel: I assume that you don’t need a confirmation of that, but the assert is fixed for me too; thanks
20160328 17:55:03< mattsc> celticminstrel: do we really need to allow the MAIs to accept several [filter] tags?  (and the like)
20160328 17:55:32< celticminstrel> No, and in fact they don't right now; the [micro_ai] tag will copy spares into the CA, but they won't be used.
20160328 17:56:04< celticminstrel> Except for [unit] in Protect Unit and [probability] in Random Recruiting, the CAs only take the first tag.
20160328 17:56:15< mattsc> but it will for [probability] and … whatever the recruiting one is called?
20160328 17:56:20< celticminstrel> Yes.
20160328 17:56:30< mattsc> okay; you answered my question before I asked it  :)
20160328 17:56:44< celticminstrel> I could've added a case to MAIH to avoid copying spare tags when they won't be used, but I figured it wasn't worth the trouble.
20160328 17:58:37< mattsc> agreed
20160328 18:43:24< mattsc> celticminstrel: simple_attack appears broken now
20160328 18:43:31< celticminstrel> :(
20160328 18:43:51< mattsc> The AI is not attacking with the correct units first.
20160328 18:44:16< mattsc> And I am quite certain that when I tested that either earlier today or yesterday, it was working.
20160328 18:44:45< celticminstrel> So, I think the chance of an elf having a given trait is 0.2*1 + 0.8*0.25 = 0.4. Does this math make sense?
20160328 18:45:46 * celticminstrel is assuming that all traits are equally likely.
20160328 18:45:59 * celticminstrel is also currently booting up the simple_attack test scenario.
20160328 18:46:38< celticminstrel> The only MAI I actually tested after that tag change was random recruitment.
20160328 18:47:50< mattsc> Yes, that math should be correct, I think.
20160328 18:51:36< celticminstrel> Whoops, looks like I overlooked the Simple Attack MAI when updating the required/optional keys.
20160328 18:52:40< celticminstrel> Oh, by the way, the rusher recruit engine has an option to minimize the recruitment on the first turn - should that be exposed in the MAI?
20160328 18:53:35< mattsc> I’d say let’s not do that.
20160328 18:54:26< celticminstrel> Okay... it also ignores the ca_score given in [micro_ai], I think.
20160328 18:54:34< celticminstrel> Unless I missed something.
20160328 18:56:00< mattsc> That’s possible.  That one really wasn’t written with MAI in mind, it was just added because it was easy.
20160328 18:56:14< celticminstrel> Looks easy to fix though.
20160328 18:56:47< mattsc> But make sure you don’t break the ExpAI if you do so.
20160328 18:57:02< celticminstrel> The fix is just in the wrapper CA, so it wouldn't affect the ExpAI.
20160328 18:57:10< mattsc> Oh, okay.
20160328 18:58:24< mattsc> Yeah, I see.  That could be fixed quite easily.
20160328 19:09:10< celticminstrel> For some reason the rush recruit leader died the last two times I tried that test scenario. I hope this was just random chance and not a bug.
20160328 19:09:59< mattsc> You did not watch how it happened?
20160328 19:10:20< mattsc> celticminstrel: what other work are you going to do on the MAIs?
20160328 19:10:24< celticminstrel> I was distracted the second time. I don't remember if I was watching the first time.
20160328 19:10:45< celticminstrel> I'm probably going to do something in the Fast MAI.
20160328 19:10:51< celticminstrel> Other than that, no plans at the moment.
20160328 19:10:57< mattsc> Okay; that scenario does not have a “planned” outcome, so him dying is not at all impossible.
20160328 19:11:04< celticminstrel> (Something about how it handles the attacks aspect.)
20160328 19:11:29< celticminstrel> I did change the scenario so that the random recruiter also recruits mages with weight 4.
20160328 19:11:50< celticminstrel> (That was for testing that multiple tags worked.)
20160328 19:11:51< mattsc> okay, sounds like a useful test
20160328 19:11:54< mattsc> right
20160328 19:12:05< celticminstrel> Useful enough to commit?
20160328 19:12:17< mattsc> Either way.
20160328 19:12:25< celticminstrel> ?
20160328 19:12:42< mattsc> I don’t care whether you commit that or not.  Either way is fine by me.
20160328 19:12:47< celticminstrel> Ah.
20160328 19:13:15< mattsc> I’m going to hold off testing the Fast AI then.
20160328 19:13:34< mattsc> And I am going to wait with simple_attack until after you fix that.
20160328 19:13:40< mattsc> But I’ll continue with the others.
20160328 19:13:46< celticminstrel> Oh right, I didn't push that yet.
20160328 19:13:57< mattsc> I’m almost ready to check out what’s going on with protect_unit.
20160328 19:14:11< mattsc> Unfortunately, duty calls again …
20160328 19:14:18< irker590> wesnoth: Celtic Minstrel wesnoth:master d4104e01c5f0 / data/ai/micro_ais/cas/ca_recruit_rushers.lua: Rush Recruit MAI: Pass requested CA score to recruit engine https://github.com/wesnoth/wesnoth/commit/d4104e01c5f05df98b8ca33b28471ad817ad651d
20160328 19:14:20< irker590> wesnoth: Celtic Minstrel wesnoth:master c8b47e48a167 / data/ai/micro_ais/mai-defs/misc.lua: Simple Attack MAI: Fix filters being ignored https://github.com/wesnoth/wesnoth/commit/c8b47e48a167c2ffe9839806aa75e32be9c2e92e
20160328 19:15:32-!- Duthlet [~Duthlet@p4FC92119.dip0.t-ipconnect.de] has quit [Remote host closed the connection]
20160328 19:29:36< mattsc> celticminstrel: FYI, simple_attack is working again and both recruiting engines are fine.
20160328 19:34:26< celticminstrel> Yay.
20160328 19:34:47< celticminstrel> Right now I'm analyzing the distribution of special initiate abilities in my campaign.
20160328 19:35:04< mattsc> I have a couple things I want to check out and (probably) fix in the Goto AI, but those don’t have anything to do with your changes.
20160328 19:35:14< celticminstrel> The bats, right?
20160328 19:35:25< mattsc> yes
20160328 19:35:28< mattsc> and ...
20160328 19:36:01< mattsc> oh, check the behavior of the “guardian thieves”.  I think they do what they are supposed to do, just want to be sure.
20160328 19:36:44< mattsc> And then I’ll check what protect_unit is doing.
20160328 19:37:08< mattsc> unless you have already fixed it in the meantime?
20160328 19:37:16< celticminstrel> I haven't looked at it.
20160328 19:37:44< mattsc> Okay; I did look at it briefly to confirm what you said (that it’s broken)
20160328 19:37:52< celticminstrel> Last I checked the mage was stepping into danger and dying.
20160328 19:38:15< mattsc> Yes, same here; and same in The_Elves_Besieged also.
20160328 19:38:22< celticminstrel> Yeah.
20160328 19:38:42< mattsc> I’m sure the fix will be easy, we just have to find it.  :(
20160328 19:38:51< mattsc> Umm, I meant :)
20160328 19:49:55-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160328 19:50:46-!- travis-ci [~travis-ci@ec2-23-23-47-74.compute-1.amazonaws.com] has joined #wesnoth-dev
20160328 19:50:47< travis-ci> wesnoth/wesnoth#9087 (master - c8b47e4 : Celtic Minstrel): The build has errored.
20160328 19:50:47< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/119052950
20160328 19:50:47-!- travis-ci [~travis-ci@ec2-23-23-47-74.compute-1.amazonaws.com] has left #wesnoth-dev []
20160328 19:57:41-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160328 20:19:49-!- mjs-de [~mjs-de@x4db68d6b.dyn.telefonica.de] has joined #wesnoth-dev
20160328 20:32:32-!- Kwandulin [~Miranda@p200300760F191C1800D27882E14D8529.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160328 20:42:33-!- mjs-de [~mjs-de@x4db68d6b.dyn.telefonica.de] has quit [Remote host closed the connection]
20160328 21:11:54-!- horrowind [~Icedove@2a02:810a:83c0:1c18:21b:fcff:fee3:c3ff] has quit [Quit: horrowind]
20160328 21:23:38-!- boucman [~rosen@wesnoth/developer/boucman] has left #wesnoth-dev []
20160328 21:28:19-!- ChipmunkV [~vova@d0017-2-88-172-31-68.fbx.proxad.net] has quit [Quit: ChipmunkV]
20160328 21:30:53-!- prkc [~prkc@108.61.123.83] has quit [Ping timeout: 244 seconds]
20160328 21:39:08-!- Appleman1234 [~Appleman1@KD106161147208.au-net.ne.jp] has quit [Ping timeout: 244 seconds]
20160328 21:39:37-!- nos__ [~nos@208.91.185.104] has joined #wesnoth-dev
20160328 21:45:46-!- prkc [~prkc@catv-80-98-243-98.catv.broadband.hu] has joined #wesnoth-dev
20160328 22:00:18-!- mjs-de [~mjs-de@x4db68d6b.dyn.telefonica.de] has joined #wesnoth-dev
20160328 22:08:48< mattsc> celticminstrel: quick comment in passing: when I reload an intermediate save of any of the test scenarios, I get
20160328 22:08:50< mattsc> “warning ai/config: side 2: aspect with id=[version] lacks default config facet!”
20160328 22:08:53< mattsc> for each side
20160328 22:09:29< celticminstrel> Not actually a problem, but a little weird... will look into it.
20160328 22:11:30< celticminstrel> It's probably something like "version" being written to save files.
20160328 22:14:33-!- irker590 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20160328 22:35:50-!- Appleman1234 [~Appleman1@KD106161136087.au-net.ne.jp] has joined #wesnoth-dev
20160328 22:36:15-!- mjs-de [~mjs-de@x4db68d6b.dyn.telefonica.de] has quit [Remote host closed the connection]
20160328 23:07:49< celticminstrel> The behaviour of return_guardian is perfect for my scenario, but the configurability of it is terrible.
20160328 23:08:03< celticminstrel> IOW I would need a separate [micro_ai] tag for each guardian.
20160328 23:10:56< celticminstrel> A BCA or unit formula might be a bit better, though still not quite ideal... since then the code is duplicated in every unit.
20160328 23:11:23< celticminstrel> Really, ai_special=guardian is the optimal solution, except for that minor detail of not returning to the guard post IIRC.
20160328 23:18:43-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20160328 23:23:04-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [Ping timeout: 250 seconds]
20160328 23:58:13-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
--- Log closed Tue Mar 29 00:00:40 2016