--- Log opened Sun May 25 00:00:13 2014
--- Day changed Sun May 25 2014
20140525 00:00:13-!- tomreyn [~tomreyn@megaglest/team/tomreyn] has quit [Quit: tomreyn]
20140525 00:02:53-!- ancestral [~ancestral@12.23.74.29] has quit [Quit: i go nstuf kthxbai]
20140525 00:06:05< irker584> wesnoth: ancestral wesnoth:master 29490feef346 / data/core/units/khalifate/Hakim.cfg: Added [portrait] to Hakim http://git.io/KAxhmA
20140525 00:06:07< irker584> wesnoth: Charles Dang wesnoth:master a43bc197e320 / data/core/units/khalifate/Hakim.cfg: Merge pull request #168 from ancestral/master http://git.io/MIRmHw
20140525 00:06:37< shadowm> vultraz, ancestral: No profile=?
20140525 00:07:10< shadowm> Oh, it already is there. Then in what circumstances could the lack of [portrait] blocks pose a problem?
20140525 00:07:23-!- ancestral [~ancestral@mobile-166-137-184-181.mycingular.net] has joined #wesnoth-dev
20140525 00:08:03< shadowm> ancestral: See the logs.
20140525 00:08:39-!- trewe [~trewe@2001:8a0:d139:9701:626c:66ff:fe92:9b7c] has quit [Quit: quit]
20140525 00:10:37< shadowm> gfgtdf, iceiceice : The announcement draft is up, including the sections that concern you two: http://forums.wesnoth.org/viewtopic.php?f=16&t=40487
20140525 00:11:11< shadowm> vultraz, _8680_ , c74d : Review plz.
20140525 00:11:20< ancestral> Every unit uses [portrait] to identify its portrait image
20140525 00:11:29< ancestral> Except Hakim
20140525 00:11:47< ancestral> Which means, in my bestiary project, I don’t see a portrait for him
20140525 00:12:31< shadowm> Unit portraits are actually defined by the `profile` attribute.
20140525 00:12:55< ancestral> What is portrait used for then?
20140525 00:13:03< shadowm> The [portrait] tag is optional and part of an (IIRC deliberately on mordante's part) undocumented hack that only concerns the [message] dialog.
20140525 00:13:48< shadowm> It was kind of more important and essential during the portrait format transition in 1.5.x.
20140525 00:14:34< shadowm> Nowadays the engine knows that it should look under a transparent/ subdir (in both directions) to resolve portrait paths.
20140525 00:14:55< mattsc> shadowm: are you planning to announce 1.11.15 today still?  (sorry for asking if you said it already, I have no time to read the logs atm)
20140525 00:15:01< shadowm> (Or maybe it's only one direction.)
20140525 00:15:02< iceiceice> shadowm: I edited part of the phrasing
20140525 00:15:18< shadowm> iceiceice: Hm, I'd appreciate it if I was informed of what the edits entailed.
20140525 00:15:41< shadowm> mattsc: I never plannd to announce 1.11.16 today, only to finish up the announcement for tomorrow.
20140525 00:16:07< ancestral> shadowm: What do you mean optional? Is portrait still being used with message?
20140525 00:16:12< shadowm> The minimum requirement for announcing is 24 hours after all tier-1 packages are done, IIRC.
20140525 00:16:16< iceiceice> "none of this has a compatibility path for code written before 1.11.14 -> ... for code written between versions 1.11.7 and 1.11.14
20140525 00:16:34< shadowm> ancestral: Uh yeah, portraits are still shown in [message] dialogs.
20140525 00:16:39< mattsc> shadowm: okay, thanks.  Just asking because I don’t know yet whether I’ll be able to get the OS X package done today.
20140525 00:16:57< shadowm> iceiceice: I don't really see how we need to specify a lower boundary but okay.
20140525 00:17:07< ancestral> Then files should have [portrait] there, right?
20140525 00:17:08< mattsc> shadowm: oh, and the ‘still’ was meant as ‘still today, not ‘still planning’.  :)
20140525 00:17:44< shadowm> iceiceice: Also, I said "for 1.11.14 and earlier".
20140525 00:17:58< iceiceice> y but we didnt break compatibility with anything for 1.11.6 and earlier
20140525 00:18:16< shadowm> Grr I still read that as 'why' D:
20140525 00:18:18< iceiceice> this sounds better, and it lets someone who is reading quickly, just the bold parts, know that if they havent ported over yet this didnt afect them
20140525 00:18:31< iceiceice> shadowm: idk man i learned y = yes from wesnoth
20140525 00:18:58< shadowm> Must be the crazy MP people's fault.
20140525 00:19:09< iceiceice> is this really something none of the devs do but me?
20140525 00:19:09< iceiceice> y is much faster than yes, and most of the time i agree with people anyways :D
20140525 00:19:22< shadowm> mattsc: Ah, okay, no need to worry then.
20140525 00:19:25< vultraz> shadowm: no mention about the wmltoold gui?
20140525 00:19:28< vultraz> tools*
20140525 00:19:48< shadowm> mattsc: http://wiki.wesnoth.org/ReleasingWesnoth#The_real_annoucement says I am supposed to wait 72 hours at most.
20140525 00:20:06< shadowm> ancestral: [message] will guess from profile=, is what I said.
20140525 00:20:35< shadowm> It'll also guess from its own image= attribute, which is still used.
20140525 00:20:46< vultraz> Besides that, looks fine
20140525 00:20:47-!- ancestral_ [~ancestral@mobile-166-137-184-181.mycingular.net] has joined #wesnoth-dev
20140525 00:20:54< ancestral_> shadowm: So there’s no point for 100 other unit files to have [portrait]s?
20140525 00:21:02< shadowm> vultraz: There is no mention of whatever you are talking about in either the changelogs or R_N.
20140525 00:21:12< shadowm> I am abiding by what the changelogs and R_N say.
20140525 00:21:25< vultraz> Well that sucks
20140525 00:21:28< shadowm> ancestral_: Not really.
20140525 00:21:32< vultraz> Who forgot the changelog entry?
20140525 00:21:41< shadowm> As I said, it was kind of required during the 1.5.x portrait format transition.
20140525 00:21:46-!- gfgtdf [~chatzilla@e177118194.adsl.alicedsl.de] has joined #wesnoth-dev
20140525 00:21:46< iceiceice> just put it in the release notes
20140525 00:21:49< iceiceice> why not
20140525 00:22:01< iceiceice> we should tout Elvish_Hunter's hard work
20140525 00:22:09< shadowm> Later a more generic solution was implemented for UMC.
20140525 00:22:31< iceiceice> not everything in the release notes always corresponds to something in a changelog anyways
20140525 00:22:39< shadowm> There's a fair chance I'm wrong, of course, but the easiest way to check is the following:
20140525 00:22:59< mattsc> shadowm: I’ll definitely try to get it done tonight, just not sure yet that I’ll get it all done.  Otherwise first thing tomorrow morning.
20140525 00:23:06< shadowm> ancestral_: :lua wesnoth.wml_actions.message { image = "path/path/path", speaker = "narrator", message = "testage" }
20140525 00:23:34< ancestral_> You didn't answer my question
20140525 00:23:48< ancestral_> Oh wait you did
20140525 00:24:14< ancestral_> Alright, I need to parse profile
20140525 00:24:37-!- ancestral [~ancestral@mobile-166-137-184-181.mycingular.net] has quit [Ping timeout: 240 seconds]
20140525 00:24:38-!- ancestral_ is now known as ancestral
20140525 00:24:42< shadowm> I also believe the [portrait] tags convey more information needed for RTL languages that profile= alone doesn't.
20140525 00:24:59< shadowm> But again, it's all undocumented.
20140525 00:26:19< shadowm> ancestral: In any case, if you really need an authoritative source with regards to how [portrait] and profile=/image= interact with each other, I suggest asking mordante himself. I can only give you my word as a UMC coder that [portrait] is not needed in practice.
20140525 00:26:53< shadowm> iceiceice: I cannot write of what I don't know about.
20140525 00:27:10< shadowm> But vultraz seems overly interested on the subject, so why not let him write an entry?
20140525 00:27:11-!- ancestral [~ancestral@mobile-166-137-184-181.mycingular.net] has quit [Quit: Smell ya later!]
20140525 00:27:26-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [Ping timeout: 258 seconds]
20140525 00:27:32< SigurdFD> vultraz: It looks like the gui is only in master
20140525 00:28:09< shadowm> Well, seems my work here is complete, then.
20140525 00:28:32< vultraz> Can't it go to 1.12 too?
20140525 00:28:44< shadowm> FEATURE FREEZE.
20140525 00:28:51< shadowm> Come on people.
20140525 00:28:53< vultraz> It's not a feature
20140525 00:28:58< shadowm> Yes, it is.
20140525 00:29:01< vultraz> It's a wmltools extension
20140525 00:29:25< shadowm> At any rate, if SigurdFD is correct, it's not in 1.11.15.
20140525 00:29:35-!- Octalot [~noct@132.53.125.91.dyn.plus.net] has quit []
20140525 00:29:48< shadowm> I won't announce the addition of something that doesn't physically exist in the release tarballs.
20140525 00:30:14< vultraz> Well I believe it should be added
20140525 00:30:26< vultraz> It's a major help to UMC authors
20140525 00:30:32< vultraz> And it doesn't break anything
20140525 00:30:39< vultraz> It doesn't change wmltools
20140525 00:30:50< shadowm> You should either prod the author of take full responsibility for it then.
20140525 00:30:57< shadowm> *or
20140525 00:32:00< shadowm> But if you ask me, people have been stretching the freeze policy so far this cycle, it's basically turned into stretchy goo at this point.
20140525 00:32:48< vultraz> Do you want just the file or all its history
20140525 00:35:42< shadowm> Is that a question?
20140525 00:35:47< vultraz> ?
20140525 00:35:51< shadowm> I don't really know what to make of it.
20140525 00:36:21< vultraz> Do you want me to just copy the file over to the 1.12 branch or do you expect a cherry pick of all the commits around that file
20140525 00:36:23< vultraz> ?
20140525 00:36:26< shadowm> "Do you want" -- at what point did I imply I wanted something? I just listed your options.
20140525 00:36:37< iceiceice> i dont see any problem of tacking on at the bottom of the release notes "by the way, a feature not part of 1.11.15 but which may be helpful to you is ..."
20140525 00:37:01< shadowm> Deciding how to go about integrating it into the 1.12 tree isn't my business.
20140525 00:37:03< iceiceice> if its not important enoguh to announce the gui, then i dont see that its very important for me to update wmllint for minute features that only exist in subversions
20140525 00:37:18-!- mjs-de [~mjs-de@f053055092.adsl.alicedsl.de] has quit [Remote host closed the connection]
20140525 00:37:32< shadowm> Wow.
20140525 00:37:43< shadowm> Yeah we are surely going to get somewhere like that.
20140525 00:37:56< iceiceice> just saying
20140525 00:38:04< shadowm> wmllint substitution rules count as bug fixes for obvious reasons.
20140525 00:38:30< shadowm> Adding a new GUI? By all means it's a feature. Sure, it doesn't affect the game, so whatever. But it's not my responsibility to backport it.
20140525 00:38:34< iceiceice> y but the fact is almost no users tested 1.11.7
20140525 00:38:48< iceiceice> and even the most knowledgeable umc developers were oblivious to "remove_carryover_when_leader_lost"
20140525 00:38:50< shadowm> "1.11.7".
20140525 00:39:25< shadowm> I think you mean the set of versions wherein { version >= 1.11.7 && version < 1.11.16 }.
20140525 00:39:25< iceiceice> you are asking me to learn python to address this change
20140525 00:39:51< shadowm> The alternative is that you had made a compatibility path available.
20140525 00:39:56-!- ancestral [~ancestral@mobile-198-228-215-011.mycingular.net] has joined #wesnoth-dev
20140525 00:40:02< shadowm> This did not happen, so you get stuck with the wmllint alternative.
20140525 00:40:16< iceiceice> well this wasnt my commit anyways
20140525 00:41:02< iceiceice> idk i'm going to try to fix bugs that seem more consequential instead i think
20140525 00:41:05< shadowm> Then why did you take it upon yourself to update wmllint?
20140525 00:41:17< iceiceice> to appease you
20140525 00:41:34< shadowm> I asked a question and you responded with the question about how to update wmllint.
20140525 00:41:50< shadowm> You could have told me "I don't think so, but it's not my change anyway" back then.
20140525 00:42:26< iceiceice> its not really about whose commit it is, i discussed the idea and i helped test it
20140525 00:42:33< iceiceice> and it fixed up an earlier bad commit of mine
20140525 00:43:01< iceiceice> and i do agree it would be ideal if wmllint fixed these two flags
20140525 00:43:03< shadowm> Good, then I see there's really no reason to complain about "having to" dea with wmllint.
20140525 00:43:27< iceiceice> but it might be better for someone else to address that if it's considered a "bug" as i don't know python and my time is better spent elsewhere
20140525 00:43:51< shadowm> It'd not be the first or the last syntax change that happens that doesn't get a wmllint conversion rule.
20140525 00:43:52< iceiceice> it seems liek a very low priority "bug"
20140525 00:43:52< iceiceice> for the reasons i described
20140525 00:44:00< iceiceice> ok
20140525 00:44:08< shadowm> It's however still one of the scarcingly few syntax changes taking place mid-feature-freeze.
20140525 00:44:31< irker584> wesnoth: Charles Dang wesnoth:1.12 0fcbb0a1639d / data/tools/GUI.pyw: Backported Elvish_Hunter's wmltools GUI from master http://git.io/NmVVxQ
20140525 00:44:54-!- ancestral [~ancestral@mobile-198-228-215-011.mycingular.net] has quit [Quit: Smell ya later!]
20140525 00:45:07< shadowm> I personally don't care about whether it was all fully planned or not, but I don't like having to include trollbait in announcements.
20140525 00:45:53< shadowm> Someone later could point to those (or the changelogs, which in fact happens) to blame us for every single bad thing that happens to them.
20140525 00:46:37< shadowm> "wesnoth developers dont really care about compatibility or addons and break the syntax all the time"
20140525 00:47:13-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20140525 00:52:16-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20140525 00:52:28< iceiceice> i'm only concerned about criticism if its actually legitimate
20140525 00:53:01< iceiceice> it was bad not to include a backwards path but unfortunately we just didnt think of it and no one who was more experienced caught the mistake...
20140525 00:54:04< shadowm> I kind of assumed everyone here was familiarized with wmllint to some extent (its existence being the bare minimum).
20140525 00:54:09-!- kex [~kex@89.205.75.19] has joined #wesnoth-dev
20140525 00:55:02< shadowm> Still, I believe the wmltools GUI thing was a subject that was being discussed between vultraz and I specifically.
20140525 00:55:48< iceiceice> shadowm: there's no documentation that i am aware of about how to add conversion rules to wmllint
20140525 00:56:00< iceiceice> i would be happier even if i could just pass it a sed expression
20140525 00:56:13< shadowm> Of course there isn't, hence in the past we've followed either of the following two approaches:
20140525 00:56:37< shadowm> 1) Filing a feature request (yes) for the wmllint maintainer
20140525 00:56:53< shadowm> 2) The DIY, which is what I and several others have been forced to do for years
20140525 00:57:29< shadowm> There is no permanent wmllint maintainer atm, so (1) is kind of impossible.
20140525 00:58:00-!- aboutGod [~aboutGod@static-72-66-66-50.washdc.fios.verizon.net] has joined #wesnoth-dev
20140525 00:58:33< irker584> wesnoth: Charles Dang wesnoth:1.12 97fd0f794be7 / changelog: Changelog entry for 0fcbb0a1639d http://git.io/OLppMw
20140525 00:58:41-!- kex [~kex@89.205.75.19] has quit [Ping timeout: 252 seconds]
20140525 00:58:57< shadowm> All that said, given enough time and prodding I may occasionally take care of rules myself, assuming I'm informed of a need for them in advance.
20140525 00:59:30-!- noy [~Noy@wesnoth/developer/noy] has quit [Read error: Connection reset by peer]
20140525 00:59:55-!- happygrue [~happygrue@wesnoth/developer/wintermute] has quit [Ping timeout: 240 seconds]
20140525 00:59:59< iceiceice> gfgtdf: i'm looking at wml_exception.?pp,
20140525 01:00:07-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20140525 01:00:10< iceiceice> i think maybe we should add something that wraps around the enum parser
20140525 01:00:40-!- noy [~Noy@wesnoth/developer/noy] has quit [Client Quit]
20140525 01:01:08< iceiceice> idk... i guess most of the time if some config from the user has a bad value we just stick a default in an run with it
20140525 01:01:14< iceiceice> idk if that is considered a feature or a bug
20140525 01:01:31< shadowm> That's how it usually works.
20140525 01:01:33< iceiceice> i'm inclined to give them an error window, at least in debug mode or something, and then run with it
20140525 01:02:10< shadowm> If we had a strict WML mode things would be generally nicer for non-lazy UMC creators.
20140525 01:02:53< shadowm> It's not a rare thing for me to get stuck inspecting misbehaving WML code only because I didn't spot an obvious typo.
20140525 01:03:14< iceiceice> i think its one of the most frustrating things about umc development tbh
20140525 01:04:20< shadowm> The introduction of the Lua-based event processing in... 1.7.x?... at least introduced validation of action tag names, and some action tags may validate attribute values if they are supposed to be keywords from a list, but it's not guaranteed to be the case.
20140525 01:04:48< shadowm> The downside is that one failed action aborts the whole event handler.
20140525 01:05:22< shadowm> Well, that's not a bad thing per se -- unless that happens right after a [kill][/kill]-type situation. :p
20140525 01:06:00< gfgtdf> iceiceice: so you want something like a lexical_cast_default_only_in_debug_mode ?
20140525 01:06:15< shadowm> Since I work a lot with cutscenes, my constant fear is that such a thing may happen and the player might not realize that the moment a "Lua traceback" appears on the top left corner, they are supposed to quit the game and report back to the author.
20140525 01:06:23< iceiceice> i want like "lexical_cast_wml"
20140525 01:06:28< iceiceice> and "lexical_cast_wml_default"
20140525 01:06:55< iceiceice> i guess i would set it up so that it would check for game_config::debug_mode
20140525 01:07:04< iceiceice> but if we make a strict mode or smth we could change it later
20140525 01:07:07< iceiceice> or a preference
20140525 01:07:20< iceiceice> that doesn't seem too important,
20140525 01:07:24< shadowm> When I designed the new WML loading error report I made it a goal to emphasize the need for feedback (when the error comes from an add-on) for a similar reason.
20140525 01:08:17< iceiceice> the point is that if its bad and the condition is on, it would show an error message like when you get OOS
20140525 01:08:27< shadowm> Dunno, perhaps it's possible to trap Lua errors and I should look into it at a later point.
20140525 01:08:46< shadowm> *trap and ascertain whether they come from the console or the add-on
20140525 01:09:13< iceiceice> idk it just occurs to me that if i'm going to start putting the smart enum code everywhere then maybe i should do this
20140525 01:10:22< iceiceice> save some typing later
20140525 01:10:31< iceiceice> i guess i could just make lexical_cast_wml an alias for lexical cast for now
20140525 01:12:37-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20140525 01:13:07-!- aboutGod [~aboutGod@static-72-66-66-50.washdc.fios.verizon.net] has left #wesnoth-dev []
20140525 01:18:04< gfgtdf> iceiceice: so teh main featore is that lexical_cast_wml shodul give a more user foendly error message ?
20140525 01:18:11< gfgtdf> the main feature
20140525 01:18:31< iceiceice> the main feature is that it gives an error message at all
20140525 01:18:59< iceiceice> right now if we use lexical_cast_default(cfg["controller"], team::AI)
20140525 01:19:01< iceiceice> or something liek this
20140525 01:19:25< iceiceice> choosing the default because of a malformed string doesnt resultin any error log
20140525 01:19:37< iceiceice> or giving the wrong value of "campaign_type" in a campaign
20140525 01:19:43< iceiceice> or the wrong value [scenario] in a top level tag
20140525 01:19:56< iceiceice> or in any other place where theres only a few possibilities
20140525 01:21:14< shadowm> Anyone thinks the coloring here is too in-your-face about the contents? http://forums.wesnoth.org/viewtopic.php?f=16&t=40487
20140525 01:21:21-!- Necrosporus_ [~Necrospor@unaffiliated/necrosporus] has joined #wesnoth-dev
20140525 01:24:19-!- Necrosporus [~Necrospor@unaffiliated/necrosporus] has quit [Ping timeout: 240 seconds]
20140525 01:29:25< mattsc> Can’t be in your face enough for this.  IMO.
20140525 01:30:41< iceiceice> did anyone "deem" that "all of this is necessary for the goal of making wesnoth reliable and internaly consistent" ?
20140525 01:30:52< iceiceice> you make it sound like there is an argument among the devs that you dont agree with 
20140525 01:31:02< iceiceice> all that happened is that we made a mistake and didnt include a backwards compatibility path
20140525 01:31:16< iceiceice> and it was released shortly after
20140525 01:31:37< iceiceice> i dont know that anyone who is arguing that the lack of a backwards compatibility path is necessary for any goal
20140525 01:31:55< iceiceice> the colors are fine though
20140525 01:36:52< iceiceice> are you going to revise or shall i?
20140525 01:39:48< iceiceice> shadowm: ^^
20140525 01:40:17-!- SigurdFD [~SigurdFD@24.154.98.89] has quit [Ping timeout: 276 seconds]
20140525 01:40:28< mattsc> Just to clarify: my comment was about coloring only, not content  
20140525 01:41:51< mattsc> That’s how I understood the question (didn’t really have time to think about content.  Note to self: don’t comment if you don’t have time to think about conent …  :P )
20140525 01:45:41< shadowm> iceiceice: Revise what?
20140525 01:46:03< shadowm> Oh, uh.
20140525 01:46:22< iceiceice> i reread what you wrote
20140525 01:46:24< shadowm> iceiceice: What makes it seem like there's an argument?
20140525 01:47:17< iceiceice> shadowm: this is what you wrote. gfgtdf and iceiceice have been fixing bugs... unfortunately this has inevitably resulted in breaking compatibility. <-- this is illogical
20140525 01:47:22< shadowm> I generally favor the passive voice in announcement, but that's not because I agree or disagree with what I'm writing, more like the contrary.
20140525 01:47:34< iceiceice> "all of this has been deemed necessary to support... " <-- implies by someone other than you
20140525 01:47:39< iceiceice> do you not want us to fix bugs in 1.12?
20140525 01:47:41< iceiceice> or do you?
20140525 01:47:49< shadowm> I just addressed that part.
20140525 01:48:09< shadowm> I'm not supposed to provide opinions (positive or negtive
20140525 01:48:12< shadowm> ffff
20140525 01:48:27< shadowm> I'm not supposed to provide opinions (positive or negative) in the middle of an announcement, hence my use of the passive voice.
20140525 01:48:34< iceiceice> the 3 sentences that are written are acutally full of inuendo
20140525 01:48:43< iceiceice> the passive voice comes off as passive aggressive imo
20140525 01:48:57< iceiceice> i am willing to accept criticism for my glaring msitakes,
20140525 01:49:06< iceiceice> but i dont think the effort to fix bugs in 1.12 should be criticized...
20140525 01:49:17< iceiceice> because no one is actually questioning whether we should be fixing bugs in 1.12 or not?
20140525 01:49:21< shadowm> I don't see how I'm critizing the effort there.
20140525 01:49:29< iceiceice> unless you would like to actually have an argument about that
20140525 01:49:35< shadowm> It sounds to me like I'm actually restating that the intention is to fix bugs.
20140525 01:49:43< iceiceice> you write that "unfortunately, it inevitably leads to backwards compatiblitiy problems"
20140525 01:49:52< iceiceice> i dont agree with that at all
20140525 01:50:09< shadowm> Well, then "happily, this breaks compatibility"?
20140525 01:50:28< iceiceice> unfortunately we made a mistake and did not include backwards compatilbity paths for these syntax changes
20140525 01:51:36< iceiceice> not, "the unfortunate decision of gfgtdf and iceiceice to fix bugs has inevitably broken backwards compatibility... but it was `deemed necessary.` so there."
20140525 01:52:16< shadowm> In view of how there doesn't seem to be a chance for us to make amends within the necessary timeframe, I've published my final draft.
20140525 01:52:53< shadowm> This is also the last announcement I write for Wesnoth.
20140525 01:54:06< shadowm> I hereby apologize for all the trouble I have caused during all these years and promise to not inconvenience you any further.
20140525 02:01:49-!- gfgtdf [~chatzilla@e177118194.adsl.alicedsl.de] has quit [Quit: ChatZilla 0.9.90.1 [Firefox 29.0.1/20140506152807]]
20140525 02:03:25-!- bumbadadabum [~bumbadada@d155109.upc-d.chello.nl] has quit [Ping timeout: 240 seconds]
20140525 02:04:40-!- RiftWalker [~nathan@ip24-252-126-205.no.no.cox.net] has quit [Remote host closed the connection]
20140525 02:09:43< mattsc> iceiceice, shadowm: I am not a good mediator and, as I said before, don’t have time to stick around right now, so I won’t even try that, however …
20140525 02:10:30< mattsc> iceiceice: I did not interpret shadowm’s text the way you did, although I can see how that could happen.  shadowm: I don’t think we need to mentione iceiceice and gfgtdf by name to get the message across.
20140525 02:11:28< mattsc> I do however think that incompatabilies with previous version should be mentioned in bold, highlighted, colorful text, esp. given how many people got tripped up by the intorduction of fight_on_… already.
20140525 02:12:22< shadowm>  Yeah, I agree that there isn't really a need to mention them by name, which is why I removed that portion.
20140525 02:12:41< iceiceice> y i think it should be in colorful highlighted text, and its even fine to say "iceiceice and gfgtdf made a mistake"
20140525 02:12:42< iceiceice> its true of course
20140525 02:12:52< mattsc> Well, I do think the portion should be kept (and in in-your-face colors)
20140525 02:12:54< shadowm> But iceiceice also raised objections about the wording of the justification later on, which is why I ended up dropping the whole block.
20140525 02:12:57< iceiceice> i dont like it if the release note conveys infighting among the devs thoguh, we are generally pretty good about not exposing stuff like that
20140525 02:13:12< mattsc> iceiceice: as I said, I did not interpret it that wat.
20140525 02:13:12-!- sachith500 [~kvirc@112.134.162.118] has joined #wesnoth-dev
20140525 02:13:15< mattsc> *way
20140525 02:13:23< shadowm> I still don't see how what I wrote could be interpreted as infighting without reading today's log.
20140525 02:13:44< iceiceice> its hard for me to argue now because what was written is deleted and i didnt backup anything
20140525 02:13:54< mattsc> Well, the people involved in this discussion *did* real today’s log.
20140525 02:13:55< shadowm> And I admit it's my fault and that I have been in a rather bad mood as of late, which is why I have decided to stop doing PR for Wesnoth.
20140525 02:14:10< shadowm> s/And/But/
20140525 02:14:20< mattsc> shadowm, iceiceice: I volunteer to rewrite the text, trying to give it as neutral a spin as I can handle.
20140525 02:14:33< mattsc> … but I cannot do that right now.
20140525 02:14:45< mattsc> I can do it either later tonight or tomorrow morning.
20140525 02:15:21< shadowm> In the interest of transparency: http://pastebin.com/c6WbBRW6
20140525 02:15:46< iceiceice> i apologize if you feel i was overly critical
20140525 02:16:03-!- _8680_ [~8680@2002:4404:712c:0:9838:9696:b9b7:a0c1] has quit [Ping timeout: 252 seconds]
20140525 02:16:24< iceiceice> i realize that it is very difficult to write large volumes of good clean text
20140525 02:16:28-!- _8680_ [~8680@2002:4404:712c:0:1d71:cc44:5deb:afb6] has joined #wesnoth-dev
20140525 02:18:46< shadowm> It's still best that I step away to prevent my own problems from tainting my perception of your work, and by proxy, the image of the project that we present to the audience every time a release takes place.
20140525 02:19:34-!- Ard0nik [~user@adsl-75-28-103-232.dsl.irvnca.sbcglobal.net] has joined #wesnoth-dev
20140525 02:19:46< shadowm> It's a troublesome pattern and I agree that it's detrimental to Wesnoth.
20140525 02:20:53< mattsc> shadowm, iceiceice: well, I’ve been in a really sour mood the last couple days (for reasons both work and Wesnoth), but as I said, if it’s just that paragraph that needs (minor IMO) rewording, I can do that a little later.
20140525 02:20:55< shadowm> I believe vultraz might be interested in writing words the next time, since he's always seemed to enjoy reviewing everything I write (for Wesnoth or my add-ons).
20140525 02:21:11-!- Ardonik [~user@adsl-75-28-99-95.dsl.irvnca.sbcglobal.net] has quit [Ping timeout: 252 seconds]
20140525 02:21:31< mattsc> Well, if one of the native-English speakers volunteers, that’s much better than I, of course.  :P
20140525 02:21:49-!- SigurdFD [~SigurdFD@24.154.98.89] has joined #wesnoth-dev
20140525 02:22:09< shadowm> I'd suggest a second alternative, but I have concerns that the alternative might publish complete whitepapers on every release's technical specifications instead of accessible announcements.
20140525 02:22:13< mattsc> (and btw, the sour mood about Wesnoth has nothing to do with any other person involved; it’s all my own doing)
20140525 02:22:26< shadowm> Yeah, that happens. :\
20140525 02:27:34< vultraz> I'm not very good with official-sounding announcements
20140525 02:27:50< vultraz> Plus I'm busy researching the possibility of putting Wesnoth on Steam
20140525 02:28:10< vultraz> I'm reading the Steamworks docs, and it looks like there's a bit of technical setup involved
20140525 02:28:13< vultraz> shadowm would have to do shit
20140525 02:30:44-!- sachith500 [~kvirc@112.134.162.118] has quit [Read error: Connection reset by peer]
20140525 02:32:09-!- Ivanovic_ [~ivanovic@frnk-5f74d9a4.pool.mediaWays.net] has joined #wesnoth-dev
20140525 02:32:46< vultraz> And I don't understand half of this >_>
20140525 02:33:16< iceiceice> i committed the yellow warning back
20140525 02:34:37-!- Ivanovic [~ivanovic@wesnoth/developer/ivanovic] has quit [Ping timeout: 240 seconds]
20140525 02:35:01-!- Ivanovic_ [~ivanovic@frnk-5f74d9a4.pool.mediaWays.net] has quit [Changing host]
20140525 02:35:01-!- Ivanovic_ [~ivanovic@wesnoth/developer/ivanovic] has joined #wesnoth-dev
20140525 02:35:36< iceiceice> the changes were, i replaced the wording "inevitable breaking of compatibility", with "due to oversight" and replaced the "all this has been deemed necessary" with "we apologize"
20140525 02:36:18-!- Ivanovic_ is now known as Ivanovic
20140525 02:37:22< iceiceice> anyone else, feel free to take it from here, its not particularly reasonable of me to ask to be allowed to write the criticisms of myself
20140525 02:37:45-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20140525 02:38:20< iceiceice> i feel that this wording criticizes me more directly without suggesting that we argued about whether the fix we committed was a good idea at all
20140525 02:38:36< iceiceice> but feel free to rewrite as appropriate
20140525 02:39:17< iceiceice> i just felt obligated to write a baseline since it got removed based on my objections earlier
20140525 02:42:18-!- kex [~kex@89.205.75.19] has joined #wesnoth-dev
20140525 02:43:16< iceiceice> thanks for posting back the old version, shadowm
20140525 02:47:02-!- kex [~kex@89.205.75.19] has quit [Ping timeout: 265 seconds]
20140525 03:02:07-!- sachith500 [~kvirc@112.134.162.118] has joined #wesnoth-dev
20140525 03:08:13< Coffee_irc> sorry to just chime in, but didn't we always have problems with start of scenario saves and such in multiplayer :P
20140525 03:08:43< Coffee_irc> see http://forums.wesnoth.org/viewtopic.php?f=15&t=34363
20140525 03:09:50< Coffee_irc> it's just we have completely new bugs to replace the old bugs now ;)
20140525 03:10:52< Coffee_irc> it would be good if we could have a mechanism for reproducing mp bugs
20140525 03:11:45< Coffee_irc> but I don't see some of the more annoying bugs being fixed for 1.12
20140525 03:13:38< shadowm> vultraz: Why me?
20140525 03:14:12< vultraz> because you know how to deal with shit
20140525 03:14:34< shadowm> Just to make things clear, we (which one of us, I don't remember) already said we would still use our own MP servers, not... whatever the other option was.
20140525 03:14:51< shadowm> Keeping that in mind, I still don't see how I'd have to be involved.
20140525 03:14:59< vultraz> shadowm: if you have time, log into the Steamworks site with your Steam account and take a look at the SDK and Steampipe documentations https://partner.steamgames.com/
20140525 03:16:03< vultraz> I'm drafting a mail to the ML right now
20140525 03:16:55< shadowm> Hang on, I'm busy reading the mesmerizing teasers there.
20140525 03:19:59< shadowm> Let's see if I can still remember my steam password.
20140525 03:20:19< shadowm> I don't like this, you know.
20140525 03:20:24< shadowm> Hiding documentation behind logins.
20140525 03:21:04< vultraz> It's propitiatory, they can't just give it to anyone. They need to know people looking at it are actual Steam users. 
20140525 03:21:16< shadowm> "Propitiatory".
20140525 03:21:25< shadowm> Trully the word of the day.
20140525 03:21:35< shadowm> *truly
20140525 03:22:15< vultraz> proprietary
20140525 03:22:17< vultraz> sorry
20140525 03:22:28< shadowm> It takes 0 cents to make an account, so that seems kind of pointless.
20140525 03:22:39< shadowm> Unless I need to have purchased stuff like Greenlight?
20140525 03:22:46< shadowm> In that case no way I'll get in.
20140525 03:23:04< shadowm> ... yeaaaaaaaaaah, vultraz, this isn't going to work.
20140525 03:23:11< shadowm> "You can access the Steamworks SDK and documentation by visiting https://partner.steamgames.com/, signing in, and accepting the SDK Access Agreement. The account you do this from must not be limited."
20140525 03:23:21< shadowm> "not be limited"
20140525 03:23:27< vultraz> Well I'm looking at it
20140525 03:23:31< vultraz> Do you not own a steam game
20140525 03:24:02< shadowm> How many times have I told you (in my channel) that I don't.
20140525 03:24:13< vultraz> Oh right, just Portal
20140525 03:24:24< shadowm> It was a free giveaway, so it doesn't count.
20140525 03:24:39< Coffee_irc> shadowm: just post your credit card details here and someone will make you a full account :P
20140525 03:24:49< vultraz> xD
20140525 03:25:09< vultraz> (no but really, this is unfortunate. You're the one who can best understand half of this)
20140525 03:25:11< shadowm> I think I'd rather reply to one of those guys offering me a thousand pounds on the forum support address.
20140525 03:25:44< shadowm> *million
20140525 03:26:56< shadowm> vultraz: You can't even drop me a few hints without violating that ridiculous NDA?
20140525 03:27:20< shadowm> Because I gather that there's pretty much an NDA involved in the agreement.
20140525 03:28:05< shadowm> See, this is why I've always had misgivings about the whole thing, even though I'm not an active part of the Free software movement.
20140525 03:29:27< shadowm> vultraz: Are you 100% sure you aren't mistaking me for someone else?
20140525 03:29:39< vultraz> WHo else would you be?
20140525 03:29:41< shadowm> I mean, what could they possibly have to do with the forums or add-ons server or MP server.
20140525 03:30:12< shadowm> Those are all ours and I can't imagine how  I'd have to change anything for Steam.
20140525 03:30:30< vultraz> There's a lot of technical stuff about steampipe, etc
20140525 03:30:56< shadowm> There are plenty more actual capable coders here that aren't me.
20140525 03:31:14< iceiceice> Coffee_irc: 
20140525 03:31:17< iceiceice> we should update that forum post
20140525 03:31:22< iceiceice> some of those are fixed now
20140525 03:31:51< iceiceice> its not necessary for the original host to reload, *i believe*
20140525 03:31:58< Coffee_irc> iceiceice: I thought you were volunteering me for this steam thing for a sec :P
20140525 03:32:11< iceiceice> lol i havent read since your refernce to that opst
20140525 03:32:14< iceiceice> i have to catch up
20140525 03:32:41< Coffee_irc> iceiceice: there have been lots of commits recently
20140525 03:32:53< Coffee_irc> is that something you could summarize for irc?
20140525 03:33:26< shadowm> vultraz: I went and stole the documentation and I seriously don't see what I have to do with this.
20140525 03:33:33< iceiceice> ok i can give a status update line by line to jb's post
20140525 03:33:39< iceiceice> -Make sure the original host reloads the game.
20140525 03:33:55< shadowm> Assuming the pastebin here isn't forged, anyway.
20140525 03:33:56< iceiceice> the reason for this was that any ai controlled side which was marked hidden by the wml maintainer,
20140525 03:34:03< iceiceice> would appear as network
20140525 03:34:07< iceiceice> for any observer or client other than the host
20140525 03:34:14< iceiceice> and not be fixable when you reload
20140525 03:34:26< iceiceice> you can manually reset using :control but that is confusing for many players
20140525 03:34:31< Coffee_irc> iceiceice: oh, so these bugs are mostly fixed now?
20140525 03:34:35< iceiceice> this was fixed by me with soliton's help
20140525 03:34:39< iceiceice> *great help
20140525 03:34:43< shadowm> vultraz: It kind of sounds like we'll have to invest on our own hosting for this, though.
20140525 03:34:49< iceiceice> uhm many of them are still broken i guess
20140525 03:34:53< shadowm> CDN hosting, to be specific, which we don't have.
20140525 03:34:53< iceiceice> but that one is fixed
20140525 03:35:01< iceiceice> -Never use a start of scenario save. Always use a manual save, created by the original host.
20140525 03:35:11< iceiceice> this is i guess true still, but its worse now, in that its not just hat the save is corrupted,
20140525 03:35:17< iceiceice> but you get an "invalid scenairo id number" when you try to load
20140525 03:35:30< iceiceice> idk maybe its actually an imprvoment if you just got a corrupted game before :O
20140525 03:35:35< Coffee_irc> iceiceice: I figure that some of these bugs can be solved in 1.13
20140525 03:35:43< iceiceice> -Never use an end of scenario save (while the End Scenario button is visible)
20140525 03:35:46< iceiceice> i dont know about this issue
20140525 03:35:53< Coffee_irc> I have 2 computers here I can use for testing
20140525 03:35:54< iceiceice> -Never use the Back to Turn function. Ever. Not even when playing in SP.
20140525 03:35:56< iceiceice> i knew about this, but now we give an actual warning
20140525 03:36:12< Coffee_irc> I posted a while back all the bugs I found in MP
20140525 03:36:14< iceiceice> -When loading a local game from the main menu, use multiplayer -> local game -> load game
20140525 03:36:14< iceiceice> the warning tells you to do this
20140525 03:36:23< iceiceice> Do not delete any units in your recall list. Ever. Even if the unit is useless, deleting a unit in your recall list will cause OOS for other players.
20140525 03:36:26< iceiceice> i guess that was fixed
20140525 03:36:29< iceiceice> where's your list?
20140525 03:36:32< iceiceice> are they on the bug tracker?
20140525 03:36:43< Coffee_irc> it was a forum post
20140525 03:36:53< iceiceice> i think i don't know about it
20140525 03:36:56< shadowm> vultraz: To be fair, CDN hosting is (AFAIK, I'm not an authoritative source) cheaper than full web hosting since you just throw crap at the host and watch it come at the other end for your users, but still.
20140525 03:37:08< iceiceice> most of the bugs i worked on fixing were allt he bugs reporte by slowthinker abut mp
20140525 03:37:20< Coffee_irc> iceiceice: http://forums.wesnoth.org/viewtopic.php?f=10&t=38487
20140525 03:38:04< iceiceice> ok the first two oos i think are fixed by gfgtdf
20140525 03:38:07< shadowm> vultraz: I don't know, perhaps I'm jumping to conclusions since I'm averting my eyes from the rest in case they decide to send drones to my house to shoot me in the head for reading this.[3~
20140525 03:38:08< iceiceice> should confirm though
20140525 03:38:21< iceiceice> also random start time is mysteriously working apparently
20140525 03:38:30< vultraz> shadowm: go on steam and buy something for 99 cents and then you can read it
20140525 03:38:31< iceiceice> i wrote this ont he bug tracker,
20140525 03:38:38< iceiceice> no one knows how it got fixed :)
20140525 03:38:39< Coffee_irc> iceiceice: it might be "working" by chance :P
20140525 03:38:45< shadowm> vultraz: I don't know how to do that, I don't have a credit card.
20140525 03:38:48< iceiceice> idk no one reported that they can still reproduce it
20140525 03:38:54< vultraz> ..you what?
20140525 03:39:07< Coffee_irc> iceiceice: it could generate the same time of day when testing so you have to test a couple of times
20140525 03:39:19< iceiceice> so i thoguht about how to fix it once
20140525 03:39:22< iceiceice> and then i looked at the code
20140525 03:39:30< iceiceice> and i saw that it was already doing more or less what i was going to do
20140525 03:39:33< vultraz> you don't have a credit card, paypal, or any online or non-cash method of payment belonging to you at all?
20140525 03:39:38< Coffee_irc> iceiceice: actually I might have tried to fix this sometime myself
20140525 03:39:40< iceiceice> it now randomizes the time of day in the mp_connect_engine
20140525 03:39:46< vultraz> How the hell do you buy anything
20140525 03:39:47< iceiceice> before transmitting to the server
20140525 03:39:48< Coffee_irc> I can't remember
20140525 03:40:00< shadowm> vultraz: Magnets.
20140525 03:40:00< iceiceice> idk if you can reproduce it would be good, but i ahve failed so far
20140525 03:40:15< shadowm> An infinite array of magnets.
20140525 03:40:21< iceiceice> i think thunderstruck fixed it inadvertantly althoguh he denies it :p
20140525 03:40:32< vultraz> I'm also scared of the drones, since I showed the video to both you and Ivanovic
20140525 03:40:36< Coffee_irc> iceiceice: I seem to remember working with thunderstruck on it
20140525 03:40:53< vultraz> But come on, we're the dev team, how else are we supposed to work on this
20140525 03:41:27< iceiceice> Coffee_irc: https://github.com/wesnoth/wesnoth/blob/master/src/multiplayer_connect_engine.cpp#L327
20140525 03:41:29< shadowm> Well, my principles dictate that I'm not directly participating on anything that involves an icky "developer agreement".
20140525 03:41:31< iceiceice> this is where it happens now
20140525 03:41:45< iceiceice> that function is called with true in exactly one place,
20140525 03:41:46< iceiceice> https://github.com/wesnoth/wesnoth/blob/master/src/multiplayer_connect_engine.cpp#L467
20140525 03:41:52< iceiceice> just after shuffle sides is applied
20140525 03:41:55< shadowm> vultraz: As I said you have plenty more people here who are better able toh elp.
20140525 03:42:00< iceiceice> which is more or less what my planned fix was
20140525 03:42:02< iceiceice> so it looks pretty much fixed to me
20140525 03:42:13< Coffee_irc> iceiceice: ok
20140525 03:42:14< iceiceice> althoguh i havent tested thoroughly
20140525 03:42:23< iceiceice> maybe i didnt undrestand what the real problem is
20140525 03:42:43< iceiceice> it hoguht you get oos because each client generates it separately
20140525 03:42:45< vultraz> shadowm: I trust your judgement
20140525 03:42:49< vultraz> and skilll
20140525 03:43:10< vultraz> You made codeoxes scrollable on mobile
20140525 03:43:28< Coffee_irc> iceiceice: that's it, it used to generate a random time of day on each client which could differ
20140525 03:45:10< iceiceice> y i've seen that on 1.10
20140525 03:45:25< iceiceice> ok let me look at the rest of your list again
20140525 03:46:02< iceiceice> ask gfgtdf about the randomization issue,
20140525 03:46:05< iceiceice> i think that is probably a nonissue now
20140525 03:46:25< iceiceice> iiuc the [start] message from the server now contains a random seed, so i think even prestart is properly seeded...
20140525 03:46:32< iceiceice> but take with a grain of salt, until you hear it from gfgtdf :)
20140525 03:46:52< iceiceice> idk about carryover gold errors
20140525 03:46:59< Coffee_irc> iceiceice: many problems would be solved by that
20140525 03:47:20< iceiceice> i think the online replay errors you mention are fixed
20140525 03:47:30< iceiceice> prestart and start are now synchronzied,
20140525 03:47:33< iceiceice> so changes to variables should be committed to replay
20140525 03:47:46< iceiceice> "do not work if side has a dwarf leader" i think that will be fine also
20140525 03:48:32< iceiceice> yeah so the SOS saves being out of sync because of prestart is one thing
20140525 03:48:34< Coffee_irc> really?
20140525 03:48:48< Coffee_irc> so there are no bugs in mp for 1.12? :P
20140525 03:48:55< iceiceice> the bugs that i know of are
20140525 03:48:58< iceiceice> we cannot load SOS saves at all
20140525 03:49:12< Coffee_irc> SOS?
20140525 03:49:15< iceiceice> start of scenario
20140525 03:49:22< iceiceice> i invented a new wesnoth acroynm :)
20140525 03:49:26< iceiceice> its the least i could do
20140525 03:49:44< Coffee_irc> fair enough
20140525 03:49:46< iceiceice> maybe should be SoS
20140525 03:49:54< Coffee_irc> we always had problems in the past with them anyway
20140525 03:50:28< iceiceice> so there are a bunch of bugs that i fixed that aren't mentioned here
20140525 03:50:29< iceiceice> one is like,
20140525 03:50:33< iceiceice> if you play a 2 v 2 game
20140525 03:50:39< iceiceice> where there is only one player for each side
20140525 03:50:42< iceiceice> *each team
20140525 03:50:52< iceiceice> and the host dc's,
20140525 03:51:00< iceiceice> the engine gets confused,
20140525 03:51:08< iceiceice> even if the 2nd guy reassigns the team to an observer,
20140525 03:51:15< iceiceice> he gets vision of the other team and loses vision of his own side
20140525 03:51:35< iceiceice> i never really figured out what exactly was going wrong there,
20140525 03:51:51< iceiceice> but when i made the "idle" controller type, i made dropped sides automatically go to that
20140525 03:51:54< iceiceice> and that basically fxied the problem
20140525 03:52:47< Coffee_irc> ok, so it sounds like 1.12 will actually be somewhat polished in terms of mp gameplay
20140525 03:54:10< Coffee_irc> the major 2 bugs that annoy mp players I think are not on my list :P
20140525 03:54:24< Coffee_irc> they are being disconnected and now knowing it
20140525 03:54:44< Coffee_irc> and returning from disconnect with no movement points
20140525 03:54:50< iceiceice> y 
20140525 03:55:20< iceiceice> i'm not sure exactly what should be done about that...
20140525 03:55:31< iceiceice> there's two options i guess
20140525 03:55:35< Coffee_irc> I volunteer to help debug that
20140525 03:55:48< Coffee_irc> I have a laptop and desktop that I can use to test
20140525 03:55:50< iceiceice> the isseu i think is that sides only do the "upkeep phase" when [init_side] is sent over the network
20140525 03:56:02< iceiceice> so if someone dc's, it never gets sent
20140525 03:56:12< iceiceice> but when you reload, it doesnt get added if it is missing
20140525 03:56:18< iceiceice> so the upkeep phase doesn't happen
20140525 03:56:29< iceiceice> the two things you could do are, (1) make the server send [init_side] instead
20140525 03:56:42< iceiceice> (2) make the replay viewer keep track of whether upkeep happened as part of the game state
20140525 03:56:45< iceiceice> and do it if it didn't happen
20140525 03:56:52< iceiceice> (2) is probably the way to go i guess
20140525 03:57:02< iceiceice> also (3) the brilliant idea i didnt think of :)
20140525 03:58:03< Coffee_irc> iceiceice: I'm happy to volunteer at some point in the future to test this
20140525 03:58:31< Coffee_irc> I'm thinking that I can use iptables and introduces artifical lag/disconnects on my local network to test
20140525 03:58:59< iceiceice> thats a pretty good way to test
20140525 03:59:05< iceiceice> esp for the not sure if i dc'd thing
20140525 03:59:11< iceiceice> for the other one, i think it would be sufficient just to like
20140525 03:59:18< iceiceice> start two clients and a local wesnothd instance on one machine,
20140525 03:59:33< iceiceice> and kill one of them with KILL signal
20140525 03:59:39< iceiceice> when it isn't its turn
20140525 03:59:39< iceiceice> or like,
20140525 03:59:42< iceiceice> i guess the way it actually ahppens,
20140525 03:59:45< iceiceice> you bring up a menu
20140525 03:59:49< iceiceice> so that you block network processing
20140525 03:59:53< iceiceice> during the other guys turn
20140525 03:59:55< iceiceice> then he ends turn
20140525 04:00:00< iceiceice> so its your turn but you are gone
20140525 04:00:07< iceiceice> then you kill with TERM / "disconnect"
20140525 04:00:27< Coffee_irc> iceiceice: laptops have a wireless toggle that can be used
20140525 04:01:08< iceiceice> ok well i actually added this bug to the wiki under "not so easy coding"
20140525 04:01:14< iceiceice> i cant remember what i was trying to fix instead
20140525 04:01:24< Coffee_irc> anyway, you're saying the forum announcement should be updated?
20140525 04:01:30< iceiceice> until recently we had legitimate blockers about getting stuck in a campaign in the connect menu
20140525 04:01:45< iceiceice> idk i think most of these fixes have been in previous changelogs
20140525 04:02:17< iceiceice> i dont know if they made it into release notes, but they aren't like "new in 1.11.15"
20140525 04:03:11< iceiceice> do you think that that bug of not getting moves is more significant than start of scenario saves not working?
20140525 04:03:39< Coffee_irc> iceiceice: IMO yes
20140525 04:04:00< iceiceice> alright well i think i could probably fix it
20140525 04:04:01< vultraz> Sent out an ML related to this steam discussion
20140525 04:04:08< iceiceice> or figure it out eventually
20140525 04:04:09< Coffee_irc> iceiceice: if a start of scenario save doesn't work it is just ignored and another save loaded from
20140525 04:04:19< iceiceice> or honestly i bet gfgtdf coudl figure it out in like 5 minutes
20140525 04:04:21< Coffee_irc> people figure this stuff out quick ;)
20140525 04:04:57< iceiceice> so there is a known work around fo rthis bug also
20140525 04:05:04< iceiceice> its documented in all of the tournaments
20140525 04:05:20< iceiceice> you like, take the broken save and watch it in the replay viewer or something
20140525 04:05:22< Coffee_irc> yeah, but that's (not mincing words) crap
20140525 04:05:22< iceiceice> and then save it again
20140525 04:05:33< iceiceice> idk at least there's a workaround
20140525 04:05:35-!- ancestral [~ancestral@12.23.74.29] has joined #wesnoth-dev
20140525 04:05:36-!- ancestral [~ancestral@12.23.74.29] has quit [Client Quit]
20140525 04:05:41< iceiceice> the start of scenario saves thing has no workaround,
20140525 04:05:45< shadowm> vultraz: I don't know know how I made my scrollboxes scrollable on mobile, really. All testing I did was with Firefox's "responsive design" view (allows constraining viewport size) and Chrome's mobile emulation mode (which is a little more complete than that).
20140525 04:05:50< iceiceice> apparently the guy who posted ont he bug report thinks thats very important,
20140525 04:05:58< iceiceice> because if his add-on campaign is borked,
20140525 04:06:00< Coffee_irc> iceiceice: the start of scenario save thing is not a big deal IMO
20140525 04:06:01< iceiceice> he can release a patch,
20140525 04:06:09< iceiceice> and people can use the start of scenario saves rather than restart often
20140525 04:06:19< shadowm> vultraz: If I had to guess it's just that previously, codebox scrollbars were too small because my site didn't declare a viewport policy, thus the whole thing was zoomed out.
20140525 04:06:26< Coffee_irc> when people choose to continue a mp campaign I see that they use a saved turn from one of the players
20140525 04:08:07< iceiceice> ok, well if you want to make an effort to fix that one i think we should talk to gfgtdf 
20140525 04:08:15< iceiceice> and we can probably put together a patch and test it
20140525 04:08:36< Coffee_irc> I'm happy to help where I can, although it is not an area I am familiar with
20140525 04:08:40< Coffee_irc> codewise
20140525 04:08:40< iceiceice> i think i know what the cause is but he also changed the replay viewer backlog recently
20140525 04:08:50< iceiceice> he might say that ther's like an elegant fix 
20140525 04:09:14< iceiceice> it happens quite oftne that i think of a patch and get it to work and then there's like a 2-3 line gfgtdf solution
20140525 04:09:59< Coffee_irc> iceiceice: I'm not complaining if all these bugs are finally getting fixed
20140525 04:10:34< SigurdFD> iceiceice: fwiw, I think the zero move thing is more important. Also, start of scenario do have a workaround...At least for one use case.
20140525 04:11:06 * shadowm eats vultraz's HTML email.
20140525 04:11:31< shadowm> Oh well, the quote formatting is also broken in the HTML part.
20140525 04:11:53< shadowm> Oh wait, "our logo".
20140525 04:12:00< shadowm> Never mind, that's you talking.
20140525 04:13:58< iceiceice> SigurdFD: whats the workaround
20140525 04:14:00< iceiceice> ?
20140525 04:14:55< SigurdFD> load the SoS in SP, then use the resulting turn 1 save. But I've only tested it in a Local 1-player MP game.
20140525 04:15:20< iceiceice> i see
20140525 04:15:20< shadowm> vultraz: The devs ML is public.
20140525 04:15:29< SigurdFD> ^ load the resulting turn 1 save in MP, that is.
20140525 04:16:03< iceiceice> y it makes sense
20140525 04:16:27< iceiceice> it might be good to post that on the bug report also
20140525 04:17:36< vultraz> shadowm: what's wrong with my HTML email
20140525 04:17:39< shadowm> vultraz: Well, I don't think I have anything to contribute there at this point other than continuing the drone joke.
20140525 04:17:59< shadowm> vultraz: Nothing, I just dislike HTML email.
20140525 04:18:22< shadowm> I think the use of rich markup in emails should be limited to sappy love letters.
20140525 04:18:42< vultraz> If I wanted to send a sappy love letter it wouldn't be via email
20140525 04:18:50< shadowm> I know over 80% of the Internet disagrees with me. *shrugs*
20140525 04:19:01< SigurdFD> iceiceice: ok
20140525 04:19:02< iceiceice> fwiw i blame microsoft for this
20140525 04:20:10< iceiceice> Coffee_irc: 
20140525 04:20:10< iceiceice> Carryover gold errors:
20140525 04:20:10< iceiceice> - not calculated correctly if reloaded or if, in offline play, a "back to turn" is used.
20140525 04:20:10< iceiceice> i guess that this one is caused because you reload mp through the sp path?
20140525 04:20:30< iceiceice> it probably doesnt affect if you use mp -> load game i would guess
20140525 04:20:52< Coffee_irc> iceiceice: I think so
20140525 04:21:09< iceiceice> i dont know about this one: - experience modifier gets reset from scenraio 2 to default of 100
20140525 04:21:23< iceiceice> i never noticed or tested that
20140525 04:21:40< Coffee_irc> iceiceice: I had to work around that by comparing to xp of elvish sorceress
20140525 04:21:44< Coffee_irc> which is 100 xp
20140525 04:21:50< Coffee_irc> but that is a bad workaround
20140525 04:21:55< iceiceice> i see
20140525 04:22:24< iceiceice> i really should change the unit test system
20140525 04:22:28< iceiceice> so that you can give it an mp campaign
20140525 04:22:37< iceiceice> or a "test campaign" i guess
20140525 04:22:53< Coffee_irc> I'm actually looking at the test scenario now
20140525 04:22:55< iceiceice> so we can write some tests for stuff like this
20140525 04:23:00< Coffee_irc> where are the files for it?
20140525 04:23:26< iceiceice> so the wml unit tests are in src/tests/scenarios
20140525 04:23:36< iceiceice> err thats wrong
20140525 04:23:47< iceiceice> data/test/scenarios
20140525 04:24:03< iceiceice> the schedule of the tests is just in root, its called wml_test_schedule
20140525 04:24:18< iceiceice> the number indicates what is supposed to happen
20140525 04:24:46< Coffee_irc> when you type "./wesnoth --test" where is the data file?
20140525 04:25:40< iceiceice> when you type ./wesnoth --test id
20140525 04:25:46< iceiceice> it will try to load scenario named id
20140525 04:25:52< iceiceice> when it goes to data/_main.cfg,
20140525 04:25:57< iceiceice> #TEST will be defined int he preprocessor
20140525 04:26:03< iceiceice> which will lead you to data/test/_main.cfg
20140525 04:26:10< iceiceice> so test is layed out more or less like an add-on
20140525 04:26:55< iceiceice> some of the stuff there is used by the C++ unit tests though
20140525 04:27:41< iceiceice> ./wesnoth --unit runs it as a unit test, so it is assuming it is running from command line and supposed to be passing an exit code reflecting what happened
20140525 04:28:05< iceiceice> but its not significantly different from --test other than that
20140525 04:28:43< iceiceice> wesnoth --unit scen --showgui should be sort of the same as --test in terms of what it looks like from the gui
20140525 04:30:28-!- kex [~kex@89.205.75.19] has joined #wesnoth-dev
20140525 04:32:06< Coffee_irc> iceiceice: thanks, I also found data/scenario-test.cfg
20140525 04:32:47< iceiceice> y i guess thats true
20140525 04:32:50< iceiceice> maybe some of those shold be moved around idk
20140525 04:33:19< iceiceice> until recently data/test/ was not loaded even when you did --test
20140525 04:33:22< iceiceice> i enabled that
20140525 04:35:01-!- kex [~kex@89.205.75.19] has quit [Ping timeout: 240 seconds]
20140525 04:38:23-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20140525 04:40:52-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20140525 04:41:27-!- ancestral [~ancestral@12.23.74.29] has joined #wesnoth-dev
20140525 04:42:25-!- noy [~Noy@wesnoth/developer/noy] has quit [Client Quit]
20140525 04:46:03< mattsc> shadowm, Ivanovic: 1.11.15 OS X package is up on SF
20140525 04:48:27< SigurdFD> iceiceice, Coffee_irc: posted the workaround to #22068. Tested with 3-player local game of The Great Quest as well. afaict, it worked.
20140525 04:50:10< ancestral> shadowm: Not to go OT on the portrait subject, but the “profile” is not for transparent images, where [portrait] is. 
20140525 04:50:42< shadowm> ancestral: profile= works for transparent images if you don't mind getting them displayed in the help screen too, although there's also small_profile= since 1.9.x to help with that.
20140525 04:50:58< ancestral> shadowm: Profile does not point to the transparent png
20140525 04:51:03< ancestral> I need transparent pngs
20140525 04:51:06< shadowm> It does if you want it to.
20140525 04:51:20< ancestral> Not without parsing the string and adding “transparent” into it
20140525 04:51:27< shadowm> Sorry, I'm still speaking from the point of view of a UMC coder, not a mainline code consumer.
20140525 04:51:37< ancestral> shadowm: I’m processing existing units, not creating UMC
20140525 04:51:52< shadowm> I know, I keep forgetting that's your goal.
20140525 04:52:04< ancestral> Anyway, the patch got merged
20140525 04:52:08< shadowm> Consuming WML as data is kind of a different subject.
20140525 04:52:49< shadowm> You need to take into account everything the engine does or might do with the data if your intention is to support UMC at a later point.
20140525 04:53:15< ancestral> wmlparser2.py helps with that regard
20140525 04:53:50< shadowm> Isn't that for reading a WML tree?
20140525 04:54:10< ancestral> It’s for reading WML in .cfg files for units, races, etc.
20140525 04:54:17< shadowm> Yeah, but not interpreting it.
20140525 04:54:23< ancestral> What do you mean?
20140525 04:54:32< ancestral> It actually runs the Wesnoth engine
20140525 04:54:36< shadowm> Interpreting the possible meanings for profile=.
20140525 04:55:03< ancestral> like “profile": "portraits/humans/cavalier.png"
20140525 04:55:25< shadowm> For example, wmlparser2.py can't possibly know that profile=unit_image means something other than "a path to a file named 'unit_image' (no extension)".
20140525 04:56:13< ancestral> Would someone put something other than a path?
20140525 04:56:30< shadowm> Or that in the absence of a [portrait] tag, profile= points to the portrait used both in help and in-game unless small_profile= is specified (in which case small_profile is used in help).
20140525 04:56:31< ancestral> Are you talking about using macros? wmlparser2.py handles that too
20140525 04:57:42< shadowm> Or that if you let P be the path given in profile= and B its filename, and P/transparent/B exists, the game will use P/transparent/B in messages.
20140525 04:58:17< shadowm> No, I'm not talking about macros, I'm talking about the way values for some attributes are interpreted.
20140525 04:58:32< shadowm> I don't think I can be any clearer about it.
20140525 05:03:42< ancestral> Of course the logic isn’t hardcoded in the data files
20140525 05:04:12< ancestral> You’re right in that profile is what I should be looking at first
20140525 05:04:29-!- lord-knyghtmare [~tahsinjah@203.188.163.173] has joined #wesnoth-dev
20140525 05:04:39< ancestral> “The engine first looks for the image in the transparent subdirectory and if found that image is used. If not found it will use the image as-is.”
20140525 05:04:50< ancestral> So, yes, I can parse it
20140525 05:05:24< iceiceice> SigurdFD: thanks for reporting 
20140525 05:06:53< lord-knyghtmare> Does the micro ai tag support the a list for side?
20140525 05:07:15< lord-knyghtmare> Example: side=2,3,4,5,6,7,8,9,10,11,12,13,14,15,16
20140525 05:07:37< lord-knyghtmare> Or, do I have to make an individual one for every side?
20140525 05:15:57-!- sachith500|2 [~kvirc@112.134.162.118] has joined #wesnoth-dev
20140525 05:21:24-!- lord-knyghtmare [~tahsinjah@203.188.163.173] has quit [Quit: lord-knyghtmare]
20140525 05:34:46-!- SigurdFD [~SigurdFD@24.154.98.89] has quit []
20140525 05:49:39-!- Sulfur [~Miranda@p5B327137.dip0.t-ipconnect.de] has joined #wesnoth-dev
20140525 05:55:52-!- sachith500|3 [~kvirc@112.134.162.118] has joined #wesnoth-dev
20140525 05:56:13-!- sachith500|4 [~kvirc@112.134.162.118] has joined #wesnoth-dev
20140525 05:56:15-!- sachith6 [~kvirc@112.134.162.118] has joined #wesnoth-dev
20140525 05:56:15-!- sachith1 [~kvirc@112.134.162.118] has joined #wesnoth-dev
20140525 06:00:51-!- sachith6 [~kvirc@112.134.162.118] has quit [Client Quit]
20140525 06:00:54-!- sachith1 [~kvirc@112.134.162.118] has quit [Client Quit]
20140525 06:00:57-!- sachith500|4 [~kvirc@112.134.162.118] has quit [Client Quit]
20140525 06:00:59-!- sachith500|3 [~kvirc@112.134.162.118] has quit [Quit: KVIrc 4.2.0 Equilibrium http://www.kvirc.net/]
20140525 06:01:01-!- sachith500|2 [~kvirc@112.134.162.118] has quit [Quit: KVIrc 4.2.0 Equilibrium http://www.kvirc.net/]
20140525 06:18:39-!- kex [~kex@89.205.75.19] has joined #wesnoth-dev
20140525 06:19:31-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20140525 06:23:05-!- kex [~kex@89.205.75.19] has quit [Ping timeout: 252 seconds]
20140525 06:25:29-!- shadowm_desktop [ignacio@wesnoth/developer/shadowmaster] has quit [Ping timeout: 264 seconds]
20140525 06:25:55-!- sachith500|2 [~kvirc@112.134.205.211] has joined #wesnoth-dev
20140525 06:27:11-!- shadowm_desktop [ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20140525 06:28:55-!- sachith500 [~kvirc@112.134.162.118] has quit [Ping timeout: 258 seconds]
20140525 06:33:59-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20140525 06:34:21-!- noy [~Noy@S01067cb21b205894.vs.shawcable.net] has joined #wesnoth-dev
20140525 06:34:27-!- noy [~Noy@S01067cb21b205894.vs.shawcable.net] has quit [Changing host]
20140525 06:34:27-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20140525 06:34:29-!- Guest36501 [~cib@p508BC650.dip0.t-ipconnect.de] has joined #wesnoth-dev
20140525 06:41:26-!- Necrosporus_ is now known as Necrosporus
20140525 06:47:35-!- irker584 [~irker@fehu.ai0867.net] has quit [Quit: transmission timeout]
20140525 06:53:22-!- [Relic] [~relic@99-58-54-211.lightspeed.milwwi.sbcglobal.net] has quit [Quit: Leaving]
20140525 06:56:41-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20140525 06:57:55-!- Sulfur [~Miranda@p5B327137.dip0.t-ipconnect.de] has quit [Ping timeout: 240 seconds]
20140525 06:58:43-!- Sulfur [~Miranda@p5B327137.dip0.t-ipconnect.de] has joined #wesnoth-dev
20140525 07:04:33< Necrosporus> I get "error display: could not open image 'misc/ellipse-nozoc-nozoc-bottom.png'" having misc/ellipse-nozoc in my unit description
20140525 07:04:48-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20140525 07:10:18-!- shadowm_desktop [ignacio@wesnoth/developer/shadowmaster] has quit [Read error: error:1408F10B:SSL routines:SSL3_GET_RECORD:wrong version number]
20140525 07:10:25< Necrosporus> a mistype in wiki:     “01~100” would become “01,02,...,10,11,...,999,100”. 
20140525 07:10:58-!- shadowm_desktop [ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20140525 07:13:40-!- aquileia [2edf524c@gateway/web/freenode/ip.46.223.82.76] has joined #wesnoth-dev
20140525 07:17:27-!- irker476 [~irker@fehu.ai0867.net] has joined #wesnoth-dev
20140525 07:17:27< irker476> wesnoth: Chris Beck wesnoth:master 986a87228786 / src/ (make_enum.hpp team.cpp team.hpp): failed conversion of enums results in error dialog, in debug mode http://git.io/I-uUqA
20140525 07:17:27< irker476> wesnoth: Chris Beck wesnoth:master d146d9cc98e4 / changelog: update changelog http://git.io/Za8YYg
20140525 07:17:28< irker476> wesnoth: Chris Beck wesnoth:master d77f0f9c864e / data/core/units/khalifate/Hakim.cfg: Merge branch 'master' of git://github.com/wesnoth/wesnoth http://git.io/Zp2BaQ
20140525 07:17:48< iceiceice> gfgtdf: ^
20140525 07:18:25-!- iceiceice [~chris@cpe-66-108-20-80.nyc.res.rr.com] has quit [Quit: Leaving]
20140525 07:20:14< Necrosporus> Is this error a bug or not? I know ellipse behavior is changed, but I guess, -nozoc should not be automatically appended as unit may use custom ellipse, having no zoc
20140525 07:35:16-!- prophile [~alynn@oftn/member/prophile] has joined #wesnoth-dev
20140525 07:38:19-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20140525 07:53:25-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20140525 08:02:26-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20140525 08:06:53-!- kex [~kex@89.205.75.19] has joined #wesnoth-dev
20140525 08:11:30-!- boucman_work [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20140525 08:11:35-!- boucman_work [~rosen@wesnoth/developer/boucman] has quit [Client Quit]
20140525 08:11:39-!- kex [~kex@89.205.75.19] has quit [Ping timeout: 258 seconds]
20140525 08:12:11< Necrosporus> Seems like not quite a bug
20140525 08:12:25< Necrosporus> nevertheless I'm not sure if this behavior is ok
20140525 08:13:25< Necrosporus> zookeeper, what do you think, is it (-nozoc being automatically appended to ellipse even if it's already defined as ellipse-nozoc) OK or not?
20140525 08:13:47< zookeeper> probably ok
20140525 08:14:32-!- iceiceice [~chris@cpe-66-108-20-80.nyc.res.rr.com] has joined #wesnoth-dev
20140525 08:14:46< iceiceice> necrosporus: if you see typos on the wiki you can just edit
20140525 08:14:50< iceiceice> it is a wiki after all :)
20140525 08:14:56< Necrosporus> but I might have a custom ellipse for specific unit, not necessarily wanting it to be changed automatically when the unit becames a leader or loses zoc
20140525 08:15:01< iceiceice> what's the bug with the image file names?
20140525 08:16:07< zookeeper> then you'll just have to duplicate the images. inconvenient, but not a big deal.
20140525 08:16:09< Necrosporus> http://wiki.wesnoth.org/AnimationWML there's no 'edit' link beside one in template
20140525 08:16:41< Necrosporus> zookeeper, and if you introduce other type of ellipse, triplicate it?
20140525 08:17:19< Necrosporus> or even tetraplicate 
20140525 08:17:35< Necrosporus> I guess there must be a key to revert to old behavior
20140525 08:17:53< iceiceice> Necrosporus: the relevant C++ is here: https://github.com/wesnoth/wesnoth/blob/master/src/unit.cpp#L2042
20140525 08:18:24-!- thunderstruck [~zaibotren@cpc13-sgyl31-2-0-cust696.18-2.cable.virginm.net] has joined #wesnoth-dev
20140525 08:19:08-!- sachith500|2 [~kvirc@112.134.205.211] has quit [Read error: Connection reset by peer]
20140525 08:19:25-!- sachith500|2 [~kvirc@112.134.205.211] has joined #wesnoth-dev
20140525 08:19:31< Necrosporus> even worse
20140525 08:19:49< Necrosporus> What if unit has no zoc and is leader?
20140525 08:20:15< Necrosporus> I guess I should check in dead water with enemy leader being stunned
20140525 08:21:58< iceiceice> y it looks like its not intended that you will assign the ellipse to be a string with -nozoc included
20140525 08:22:11< iceiceice> if you want to use that file i guess you should just copy it and rename it withotut he -nozoc
20140525 08:23:55< Necrosporus> Actually in my case I do not have to define ellipse at all as I do not use custom one and unit is lvl0
20140525 08:28:12< thunderstruck> 20140524 17:39:46< gfgtdf> iceiceice, thunderstruck: do you knwo about https://gna.org/bugs/?21801 ?
20140525 08:28:50< thunderstruck> I know nothing about preview-maps, but I now about how game config is generally loaded
20140525 08:28:56< iceiceice> i dont really know anything about it
20140525 08:29:12< iceiceice> i speculated that it mgiht have to do with the new minimap code?
20140525 08:29:13< iceiceice> its not clear when this became an issue i think
20140525 08:29:19< iceiceice> gfgtdf wrote that he didnt think that was the issue
20140525 08:29:41-!- shadowm_desktop [ignacio@wesnoth/developer/shadowmaster] has quit [Ping timeout: 252 seconds]
20140525 08:30:06< iceiceice> i assume that gfgtdf profiled it in VC
20140525 08:32:29< Necrosporus> Is there some debug command to 'unvictory?'
20140525 08:32:50< Necrosporus> I wanted to move enemy leader to other place and got victory immediately I changed its side
20140525 08:34:18< iceiceice> no
20140525 08:34:24< thunderstruck> iceiceice: But did he manage to reproduce it?
20140525 08:34:37< thunderstruck> I don't see any such delay on my machine.
20140525 08:34:39< iceiceice> thunderstruck: you'll have to ask him
20140525 08:35:53< iceiceice> i get a small delay but not noticeably slow
20140525 08:39:53< thunderstruck> iceiceice: Could you try to remove a call "load_game_config_for_game()" in create_engine constructor inside multiplayer_create_engine.cpp file?
20140525 08:40:05< thunderstruck> And see if you still get a delay.
20140525 08:40:47< thunderstruck> iceiceice: line number if ~350
20140525 08:40:52< thunderstruck> s/if/is
20140525 08:41:07< thunderstruck> I would do it myself, but I don't see any delay
20140525 08:41:33< Necrosporus> I get 3 or 4 seconds delay
20140525 08:41:56< Necrosporus> Also game is slow for me in other aspects
20140525 08:42:02< iceiceice> ok, i'm trying it
20140525 08:42:29< Necrosporus> for example when I start a scenario scrolling to other part is slow
20140525 08:42:45< Necrosporus> But when I already scrolled it becomes normal 
20140525 08:43:20< iceiceice> it seems pretty much the same
20140525 08:43:27< iceiceice> hard to say
20140525 08:43:59< thunderstruck> iceiceice: Thanks for doing it. In that case it's probably not related to config loading.
20140525 08:44:21< iceiceice> y np
20140525 08:50:51-!- mjs-de [~mjs-de@f053055092.adsl.alicedsl.de] has joined #wesnoth-dev
20140525 08:52:23-!- Gambit [~derek@wesnoth/developer/grickit] has quit [Remote host closed the connection]
20140525 09:10:24< Necrosporus> If I upload an addon to server, will wesnoth upload hidden files like .git too?
20140525 09:15:06< vultraz> you need to exclude those with a _server.ign file
20140525 09:15:07-!- tomreyn [~tomreyn@megaglest/team/tomreyn] has joined #wesnoth-dev
20140525 09:17:51< AI0867> iceiceice: wesnoth.org:8080/job/Wesnoth/155/branch=master,compiler=4.6,label=Gentoo-64/console
20140525 09:19:02< Necrosporus> vultraz, do I need to exclude *pbl and _info.cfg manually? like in https://github.com/shikadilord/After_the_Storm/blob/master/_server.ign
20140525 09:19:50< Necrosporus> Also here's no .git
20140525 09:20:49-!- sachith500|2 [~kvirc@112.134.205.211] has quit [Read error: Connection reset by peer]
20140525 09:21:15-!- sachith500|2 [~kvirc@112.134.205.211] has joined #wesnoth-dev
20140525 09:25:57< vultraz> Necrosporus: if you want
20140525 09:26:58< Necrosporus> _server.pbl containing password should be excluded automatically?
20140525 09:29:19< vultraz> the pbl is never uploaded to the server
20140525 09:29:33< vultraz> But some of us like to exclude it to be through
20140525 09:30:14-!- bumbadadabum [~bumbadada@d155109.upc-d.chello.nl] has joined #wesnoth-dev
20140525 09:43:27< AI0867> loonycyborg: any idea what to do about those warnings in 4.4/4.5? Just disable them for those builds?
20140525 09:47:35-!- EdB [~edb@85.69.242.6] has joined #wesnoth-dev
20140525 09:51:34-!- mordante [~mordante@roadie.xs4all.nl] has joined #wesnoth-dev
20140525 09:51:34-!- mordante [~mordante@roadie.xs4all.nl] has quit [Changing host]
20140525 09:51:34-!- mordante [~mordante@wesnoth/developer/mordante] has joined #wesnoth-dev
20140525 09:51:48< mordante> servus
20140525 09:53:23< mordante> Ivanovic, the other libs I assume you mean regarding the random?
20140525 09:55:06-!- kex [~kex@89.205.75.19] has joined #wesnoth-dev
20140525 09:59:27-!- kex [~kex@89.205.75.19] has quit [Ping timeout: 245 seconds]
20140525 10:02:30-!- adityapande [75cd8003@gateway/web/freenode/ip.117.205.128.3] has joined #wesnoth-dev
20140525 10:02:30< mordante> Necrosporus, I can't reproduce the bug, please file a report with the exact steps how to reproduce
20140525 10:05:01< adityapande> any projects for me to get started ?
20140525 10:08:19-!- bumbadadabum [~bumbadada@d155109.upc-d.chello.nl] has quit [Read error: Connection reset by peer]
20140525 10:15:43-!- Guest36501 [~cib@p508BC650.dip0.t-ipconnect.de] has quit [Ping timeout: 276 seconds]
20140525 10:16:14< mordante> adityapande, what kind of projects are you looking for?
20140525 10:17:28-!- irker476 [~irker@fehu.ai0867.net] has quit [Quit: transmission timeout]
20140525 10:23:16< adityapande> mordante: i want to work on some easier projects scheduled for the newer releases
20140525 10:23:48-!- aquileia [2edf524c@gateway/web/freenode/ip.46.223.82.76] has quit [Ping timeout: 240 seconds]
20140525 10:23:51< adityapande> mainly coding projects
20140525 10:24:49< mordante> any area of coding that you are interested in?
20140525 10:25:38< adityapande> ya mainly ai 
20140525 10:26:42< adityapande> but other areas might also interest me
20140525 10:28:43-!- Guest36501 [~cib@p5DC7514B.dip0.t-ipconnect.de] has joined #wesnoth-dev
20140525 10:29:25-!- prophile [~alynn@oftn/member/prophile] has quit [Quit: The Game]
20140525 10:30:34< mordante> you can look at a bug https://gna.org/bugs/index.php?go_report=Apply&group=wesnoth&func=browse&set=custom&msort=0&report_id=101&advsrch=0&status_id=1&resolution_id=0&submitted_by=0&assigned_to=0&category_id=101&bug_group_id=104&severity=0&priority=0&summary=&details=&sumORdet=0&history_search=0&history_field=0&history_event=modified&history_date_dayfd=25&history_date_monthfd=5&history_date_yearfd=2014&chunksz=150&spamscore=5&boxoptionw
20140525 10:30:34< mordante> anted=1#options
20140525 10:30:56< mordante> or at and easy coding project http://wiki.wesnoth.org/EasyCoding#Improvements_to_AI
20140525 10:31:31< adityapande> thanks
20140525 10:31:33< mordante> Crab_ and mattsc know most about the AI, but Crab_ is not that often around
20140525 10:31:53< mordante> if you have any questions just ask
20140525 10:32:21< adityapande> which version of wesnoth to clone in git?
20140525 10:33:16< mordante> not sure what you exactly mean, with Git you clone the entire repository do you mean which branch to work on?
20140525 10:41:29< adityapande> ya which branch
20140525 10:42:11< mordante> best work on master
20140525 10:42:48< mordante> we're working on finishing the stable release 1.12, and fixes should be in both 1.12 and master
20140525 10:43:22< mordante> but the person committing your patches can backport them to 1.12 if they are needed there
20140525 10:43:38< adityapande> ok
20140525 10:44:02< adityapande> thanks
20140525 10:45:25< mordante> you're welcome
20140525 10:48:48-!- happygrue [~happygrue@wesnoth/developer/wintermute] has joined #wesnoth-dev
20140525 10:48:52-!- EdB [~edb@85.69.242.6] has quit [Quit: Konversation terminated!]
20140525 10:50:22-!- adityapande [75cd8003@gateway/web/freenode/ip.117.205.128.3] has quit [Quit: Page closed]
20140525 10:53:29< c74d> 2014-05-25 02:22:13 <shadowm> I'd suggest a second alternative, but I have concerns that the alternative might publish complete whitepapers on every release's technical specifications instead of accessible announcements.
20140525 10:54:03< c74d> I doubt I would, unless people put technical whitepapers in the release notes, which I quite doubt.
20140525 10:54:43< c74d> I may be unsuitable for other reasons though, like not being a developer.
20140525 10:55:21-!- wesbot changed the topic of #wesnoth-dev to: tagged 1.11.15 (1.12 beta5), announcing "soon" | string+feature freeze active on 1.12 | 240 bugs, 344 feature requests, 29 patches | Logs: http://irclogs.wesnoth.org | Alternate logs: http://wesnoth.debian.net | Don't paste on IRC! Use a pastebin: http://pastebin.com | http://imagebin.org
20140525 10:55:57< c74d> Though if it’s viewed as a public relations job, I suppose filling it with someone external wouldn’t be all that unusual.
20140525 11:03:43< AI0867> Necrosporus: ellipse-leader-nozoc-selected-bottom.png
20140525 11:16:08-!- ancestral [~ancestral@12.23.74.29] has quit [Quit: ancestral]
20140525 11:20:14-!- Sulfur [~Miranda@p5B327137.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20140525 11:43:14-!- kex [~kex@89.205.75.19] has joined #wesnoth-dev
20140525 11:48:40-!- kex [~kex@89.205.75.19] has quit [Ping timeout: 276 seconds]
20140525 11:55:56< mordante> I'm off bye
20140525 11:56:05-!- irker881 [~irker@fehu.ai0867.net] has joined #wesnoth-dev
20140525 11:56:05< irker881> wesnoth: Mark de Wever wesnoth:master b81b66f9b99a / src/map_location.cpp: Fix compilation. http://git.io/ZeWitw
20140525 11:56:21-!- mordante [~mordante@wesnoth/developer/mordante] has quit [Quit: Leaving]
20140525 11:57:42-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20140525 11:58:27-!- Sulfur [~Miranda@p5B327137.dip0.t-ipconnect.de] has joined #wesnoth-dev
20140525 12:05:23< c74d> (Oh, to be clear, I was acting on the assumption that I’m the “second alternative”. I’m fairly certain I am, but I suppose it might be someone else.)
20140525 12:10:21< AI0867> I don't think anyone else would be accused of writing whitepapers =P
20140525 12:39:38-!- aquileia [2edf524c@gateway/web/freenode/ip.46.223.82.76] has joined #wesnoth-dev
20140525 12:40:30-!- Gambit [~derek@wesnoth/developer/grickit] has joined #wesnoth-dev
20140525 12:40:54-!- markus_ [~mjs-de@g228128119.adsl.alicedsl.de] has joined #wesnoth-dev
20140525 12:44:31-!- mjs-de [~mjs-de@f053055092.adsl.alicedsl.de] has quit [Ping timeout: 252 seconds]
20140525 12:50:30-!- gfgtdf [~chatzilla@f054141213.adsl.alicedsl.de] has joined #wesnoth-dev
20140525 12:53:24< gfgtdf> thunderstruck: when i profiled that issue, my compiler told me that create_terrain_maps called by gamemap s consstructor takes >25% of the time from the time that i profiled  (i profiled often back/next) clicking but i dont remember exactly
20140525 12:55:01< gfgtdf> thunderstruck: and to me it seems like we o that for every of the preview maps but we nornly only have to do i once (reading the terrains) but i dont know that code very much
20140525 13:01:07-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Ping timeout: 245 seconds]
20140525 13:01:15-!- stikonas_ [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20140525 13:05:34-!- kex [~kex@89.205.75.19] has joined #wesnoth-dev
20140525 13:07:08< gfgtdf> iceiceice: about units loosing movements: maybe we can jut store the "init_side_done_" in savegames and then call maybe_do_init_side in playsingle_controller::play_side ?
20140525 13:07:47< iceiceice> y that probably works
20140525 13:08:44< iceiceice> it can probably just be a boolean in "game_data"?
20140525 13:11:03< gfgtdf> iceiceice: hm i why not keeping it private in play_controller, read in in its contructor and save it here: https://github.com/wesnoth/wesnoth/blob/master/src/play_controller.cpp#L737
20140525 13:12:58< gfgtdf> iceiceice: note that it defaults to true
20140525 13:13:22< iceiceice> i didnt realize that the playcontroller object was written in save games :X
20140525 13:15:45< iceiceice> y i think this is the right idea
20140525 13:16:15< iceiceice> i guess also could look at the replay controller to see hwo its done there/
20140525 13:16:15< iceiceice> since apparently it works there?
20140525 13:17:00< iceiceice> hmm actually i guess thats not quite right,
20140525 13:17:09< iceiceice> the part thats broken is getting from snapshot
20140525 13:17:46< iceiceice> ok i will try
20140525 13:20:04-!- Samual [~dioteckte@xonotic/core-team/Samual] has joined #wesnoth-dev
20140525 13:21:43-!- markus_ is now known as mjs-de
20140525 13:31:43< gfgtdf> thunderstruck: s/i profiled often back/next/i profiled often clicking the back/next buttons of the mp gui
20140525 13:51:55< Necrosporus> does anyone want to try reproducing my bug?
20140525 13:57:38< irker881> wesnoth: Chris Beck wesnoth:master e0b0716d2a79 / src/make_enum.hpp: fixup enum cast error reporting http://git.io/hIR3aA
20140525 14:03:59< iceiceice> gfgtdf: it looks like that worked 
20140525 14:04:13< iceiceice> im going to test on previous version to make sure my test made sense
20140525 14:05:53< gfgtdf> Necrosporus: i currenty have o wesnoth on this pc
20140525 14:06:27< gfgtdf> iceiceice: the init_side_done_ thing ?
20140525 14:07:20< iceiceice> y
20140525 14:07:27< iceiceice> i am still testin, i made a pull request thoguh
20140525 14:07:28< Necrosporus> gfgtdf, no wesnoth?
20140525 14:07:44< Necrosporus> iceiceice, regarding my bug, I can't reproduce it on 1.11.15
20140525 14:07:48< iceiceice> Necrosporus: which bug?
20140525 14:07:55< Necrosporus> assertion failure
20140525 14:08:00< iceiceice> oh
20140525 14:08:03< iceiceice> the inspect thing?
20140525 14:08:06< Necrosporus> yes
20140525 14:08:41< Necrosporus> inspect works fine on 1.11.15 even though && was not replaced 
20140525 14:08:42< iceiceice> does it affect master?
20140525 14:08:47< Necrosporus> I do not know
20140525 14:09:00< iceiceice> which version doe sit affect?
20140525 14:09:10< iceiceice> like 1.11.14+ or something?
20140525 14:09:26< Necrosporus> iceiceice, fc1ca38
20140525 14:09:57< gfgtdf> iceiceice: hmm that pr wokred ? do you know when maybe_do_init side is called ?
20140525 14:10:21< iceiceice> so it might be something wrong with my testing method
20140525 14:10:30< iceiceice> what i did was
20140525 14:10:37< iceiceice> start networked 2player game on the free lands
20140525 14:10:44< iceiceice> bring up side 2 preference menu
20140525 14:10:47< iceiceice> side 1 ends turn
20140525 14:10:55< iceiceice> kill side 2 client from terminal with ^C
20140525 14:11:08< iceiceice> at that point side 1 should be in the corrupted state
20140525 14:11:16< iceiceice> having not recieved init side
20140525 14:11:19< iceiceice> they select the "save and abort" option
20140525 14:11:33< iceiceice> then i try to reload in mp -> load game
20140525 14:11:41< iceiceice> im going to test to make sure that this *didn't* work before my commit
20140525 14:12:17< gfgtdf> iceiceice: hm i tihnk savig while the other cient still ahs teh enu open
20140525 14:12:22< gfgtdf> the menu
20140525 14:12:43< iceiceice> hmm ok that might be right
20140525 14:15:07< iceiceice> hmm ok
20140525 14:15:39< iceiceice> so the method you just described, doesnt seem to cause the bug in current master
20140525 14:15:40< iceiceice> e0b0716d2a79485f0c422b0f1aa6aa340990f91a
20140525 14:16:35-!- lord-knyghtmare [~tahsinjah@203.188.163.173] has joined #wesnoth-dev
20140525 14:16:51< iceiceice> so that's rather confusing actually
20140525 14:22:12< loonycyborg> AI0867: I'm wondering why they're on 4.4, 4.5 only :/
20140525 14:23:06< iceiceice> loonycyborg, AI0867: i looked at the log, it looks like it is stalling out during config loading maybe?
20140525 14:23:53< loonycyborg> I mean the warnings
20140525 14:24:30-!- [Relic] [~relic@99-58-54-211.lightspeed.milwwi.sbcglobal.net] has joined #wesnoth-dev
20140525 14:25:34< iceiceice> oh maybe idk what you are talking about
20140525 14:27:22< AI0867> loonycyborg: broken compiler? It's at least missing a warning option
20140525 14:28:24< loonycyborg> -Wno-unused-local-typedefs was added later..
20140525 14:28:30< loonycyborg> hmm
20140525 14:28:41< AI0867> loonycyborg: it seems you're using jobs=4 on the gentoo VMs. Are you sure that's a good idea?
20140525 14:29:01< AI0867> s/VMs/VM/
20140525 14:29:17< loonycyborg> two builds won't be started in parallel on the same vm
20140525 14:29:23< loonycyborg> so it makes sense to use all cores
20140525 14:30:18< gfgtdf> iceiceice: so you did it how it is desccribed here : https://gna.org/bugs/?21397
20140525 14:30:54< iceiceice> i did exactly like that
20140525 14:30:58< iceiceice> i find that the leader has gold
20140525 14:30:58< iceiceice> *sry
20140525 14:30:59< loonycyborg> ok, disabled strict compliation for those two compilers
20140525 14:31:00< iceiceice> the leader has moves
20140525 14:31:59< iceiceice> anyone on channel: can someone take a minute to try to reproduce this bug report on master?
20140525 14:31:59< iceiceice> https://gna.org/bugs/?21397
20140525 14:32:05< gfgtdf> iceiceice: hamm maybe you can mage a turn 1 bweween where you move both leader so that they have no moves left ? and them obsrver the gold too ?
20140525 14:32:25< gfgtdf> make
20140525 14:32:35< iceiceice> ok
20140525 14:33:13< iceiceice> oh y i think you are right
20140525 14:34:01< gfgtdf> iceiceice: so you did reproduce ?
20140525 14:34:11< iceiceice> trying tnow
20140525 14:34:13< Necrosporus> http://pastebin.com/061b4M5L does it say anything to you?
20140525 14:36:01< iceiceice> ok i reproduced
20140525 14:41:53< gfgtdf> iceiceice: and did teh fix work ?
20140525 14:42:01< iceiceice> i jumped to the wrong branch...
20140525 14:42:08< iceiceice> i need to clean out old topic branches :(
20140525 14:42:17< iceiceice> fixup_maybe_do_init_side != fixup_init_side_done
20140525 14:44:12< gfgtdf> lol
20140525 14:44:17< iceiceice> ok the commit i pushed didnt fix it
20140525 14:44:21< gfgtdf> iceiceice: what does teh fixup_maybe_do_init_side  done
20140525 14:44:29< iceiceice> idk something from months ago
20140525 14:44:30< gfgtdf> s/done/fix
20140525 14:45:04< gfgtdf> iceiceice: ok did you tra adding a call do maybe_do_init_side into https://github.com/wesnoth/wesnoth/blob/master/src/playsingle_controller.cpp#L682
20140525 14:45:30< iceiceice> i will try
20140525 14:47:11< iceiceice> does maybe_do_init_side need to be in that do while loop?
20140525 14:47:16< iceiceice> i guess probably?
20140525 14:48:12< iceiceice> gfgtdf: here was fixup_maybe_do_init_side
20140525 14:48:13< iceiceice> https://github.com/cbeck88/wesnoth/compare/wesnoth:1.12...fixup_maybe_do_init_side
20140525 14:48:33< gfgtdf> iceiceice: hmm i thought ti wouldnt since it onyl matters on relading 
20140525 14:48:51< iceiceice> here was my comments about it
20140525 14:48:51< gfgtdf> iceiceice: but mabe if we put it inside teh loop we can remove those calls from playturn.cpp ...
20140525 14:48:51< iceiceice> https://github.com/wesnoth/wesnoth/pull/138
20140525 14:49:15< iceiceice> ok maybe i'll make two branches
20140525 14:54:13< iceiceice> its pretty wierd to me the way that maybe_do_init_side takes a side number
20140525 14:54:21< iceiceice> would we ever want to kill this funciton for other than the current side??
20140525 14:54:30< iceiceice> s/kill/call
20140525 14:56:16< gfgtdf> iceiceice: hmm ye i dont know a reason neigher, same is true for do_replay
20140525 14:57:19-!- Sulfur [~Miranda@p5B327137.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20140525 14:57:59< iceiceice> idk maybe someone had grand plans in the past...
20140525 14:58:07< iceiceice> i think it might be a good idea to look at all the places we call functions like that
20140525 14:58:22< iceiceice> and use like play_controller::current_side or wahtever the right accessor is
20140525 14:58:28< iceiceice> if that's actually what we always do
20140525 14:58:43< iceiceice> if that interferes with the grand plans they can refactor later
20140525 14:59:27< iceiceice> ok putting it outsid ethe do_while loop works to fix the basic problem
20140525 15:00:30< AI0867> loonycyborg: well, there are two other build VMs, and building wesnoth isn't the only thing the server is expected to do
20140525 15:01:45< gfgtdf> iceiceice: i think passing that value is there for very ling already so most likley not grand plans
20140525 15:02:00< loonycyborg> AI0867: maybe nice those processes then?
20140525 15:02:00< gfgtdf> long
20140525 15:02:38< loonycyborg> So it'll both utilize all cpus always and not interfere with other processes too much.
20140525 15:03:57< gfgtdf> iceiceice: https://github.com/wesnoth/wesnoth/blob/f0f3a5b873c2c6640b334d120dbe123a4bda5333/src/play_controller.cpp#L292
20140525 15:04:19< gfgtdf> iceiceice: that is from 2006
20140525 15:04:55-!- gfgtdf [~chatzilla@f054141213.adsl.alicedsl.de] has quit [Quit: ChatZilla 0.9.90.1 [Firefox 29.0.1/20140506152807]]
20140525 15:05:36-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20140525 15:05:52< aquileia> Hi noy
20140525 15:06:48-!- thunderstruck [~zaibotren@cpc13-sgyl31-2-0-cust696.18-2.cable.virginm.net] has quit [Quit: leaving]
20140525 15:07:45-!- thunderstruck [~zaibotren@cpc13-sgyl31-2-0-cust696.18-2.cable.virginm.net] has joined #wesnoth-dev
20140525 15:08:53-!- lord-knyghtmare [~tahsinjah@203.188.163.173] has quit [Quit: lord-knyghtmare]
20140525 15:10:11< AI0867> loonycyborg: there's also memory usage (particularly at link-time), but I suppose we have enough of that
20140525 15:10:51< loonycyborg> linking may eat a lot of memory for long time, but only with debug builds
20140525 15:13:47< AI0867> what did you change for #157?
20140525 15:22:31< aquileia> noy, would you mind discussing the position in timeline for a khalifate campaign?
20140525 15:23:37< noy> unfortunately, I've got limited time right now....I should be out working. 
20140525 15:23:46< aquileia> ok, no problem
20140525 15:23:47< noy> what do you mean by a"position in timeline"
20140525 15:23:53< aquileia>  Captain_Wrathbow and I planned for 501 YW, zookeeper recommended 417 YW
20140525 15:24:16< aquileia> and as your draft tells us relations begin around 500...
20140525 15:26:05< noy> uh, I don't know what the dates correspond to in the regular timeline anymore
20140525 15:26:32< aquileia> 417 is before the Garardine dynasty
20140525 15:26:51< aquileia> 501 is the battle between Eldred and Delfador
20140525 15:27:34< iceiceice> Coffee_irc, SigurdFD, anyone else with an interest: I think gfgtdf and I worked out a simple fix to the bug where mp disconnections could cause you to lose all your moves in the reloaded game
20140525 15:28:44< iceiceice> if anyone could help to test PR 169 for possible merge it would be appreciated
20140525 15:29:01< iceiceice> also PR 170 is an alternate version, but only to consider if 169 is bad
20140525 15:29:14< noy> I didn't think first contact between the later colonizers and the khalif ate were a major event
20140525 15:29:25< noy> according to the draft notes.
20140525 15:29:27< aquileia> noy: Someone (Eldred in 501, a contestant of Garard) would hire a khalifate army
20140525 15:29:36< iceiceice> the bug report is here: https://gna.org/bugs/?21397
20140525 15:29:50< noy> ...I also don't think that they would be very much up to being hired either
20140525 15:29:53< aquileia> It wouldn't be the first contact
20140525 15:30:23< aquileia> The issue is rather: Do they know of each other in 417?
20140525 15:30:48< noy> I'm sure that they may have known each other early on... 
20140525 15:31:42< noy> I think there was knowledge before 417
20140525 15:32:07< noy> think marco polo and the chinese
20140525 15:32:24< aquileia> Is it viable to have 10.000 khalifate soldiers... let's put is as "come to the aid of a well paying friend"?
20140525 15:32:26< noy> no
20140525 15:32:32< noy> absolutely not
20140525 15:32:50< aquileia> damn
20140525 15:33:12< noy> because raising an army like that would be very difficult. 
20140525 15:33:49< aquileia> At what scale are wesnothian armies at all?
20140525 15:33:51< noy> and if they had limited contact with other empires, who would they trust?
20140525 15:33:56< zookeeper> aquileia, unspecified, really
20140525 15:34:24< zookeeper> i don't recall there being much concrete numbers given anywhere
20140525 15:34:34< noy> I'm just thinking sociologically, rather than an exact wesnoth example
20140525 15:34:47< zookeeper> for all we know, konrad's HttT army could be anything from a few thousand to tens of thousands
20140525 15:35:40< aquileia> in that case, what about around 1000 mercenaries?
20140525 15:36:18< noy> that's more possible, but I'd be hesitant. 
20140525 15:36:18< aquileia> The kingdom is destabilized, thus they could turn the tide of a war easily
20140525 15:36:38< aquileia> That's valid for both possible dates
20140525 15:36:57< noy> mercenaries in previous eras were heavily rooted in a culture
20140525 15:37:24< noy> you didn't see, anatolian mercenaries operating in Europe
20140525 15:37:39< aquileia> But our 10.000 are based on a RL example: Greek mercenaries in Persia
20140525 15:37:41< noy> or berbers, or whatever
20140525 15:39:07< zookeeper> noy, so what _is_ the intended way for the khalifate to interact with wesnoth? if they're to wesnoth what chine was to europe then they'd pretty much just exist in isolation.
20140525 15:39:17< zookeeper> china, even
20140525 15:39:39< noy> that's probably a good example Zoo
20140525 15:39:49< noy> They trade, there is some contact
20140525 15:40:01< noy> maybe there is a skirmishes 
20140525 15:40:48< noy> the ten thousand seems interesting, but not quite apt aquileia 
20140525 15:41:15< aquileia> And if our story is the reason they have no further contact? Not being paid for their services is pretty bad
20140525 15:41:17< noy> remember that even though the khalif ate have built large cities in the fringes of the desert
20140525 15:41:46< noy> their society still operates in semi-marginal terrain
20140525 15:42:09< noy> drawing up 10k men for someone else's fight would be difficult at best
20140525 15:42:43< aquileia> Eldred/his replacement in 417 has lots of money
20140525 15:42:53< aquileia> just to say...
20140525 15:43:12< noy> yes, but they and the wesnothians are extremely different societies
20140525 15:43:22< noy> persians and greeks knew each other for quite some time
20140525 15:43:24< aquileia> and as I said, we can reduce them by a factor of 10
20140525 15:43:39< zookeeper> aquileia, switch to dwarves? :P then everything could work out just fine
20140525 15:43:41< noy> Wesnothians and the people of the desert, did not
20140525 15:44:05< noy> I can see a smaller group could, and might happen
20140525 15:47:21< noy> think about the logistics for example
20140525 15:47:37< noy> after a few scrapes and battles, who would repair their armour and weapons?
20140525 15:47:49< iceiceice> AI0867, loonycyborg: how did you guys fix this btw? out of curiosity. http://wesnoth.org:8080/job/Wesnoth/155/branch=master,compiler=4.6,label=Gentoo-64/console
20140525 15:48:01< noy> nobody would have ever seen their equipment before... let alone know how to repair it
20140525 15:49:00< loonycyborg> iceiceice: I didn't do it :P
20140525 15:49:07-!- Guest36501 [~cib@p5DC7514B.dip0.t-ipconnect.de] has quit [Ping timeout: 240 seconds]
20140525 15:49:28< aquileia> noy, that's one of their problems on the way back
20140525 15:49:51< aquileia> They were hired for one single battle
20140525 15:50:29< aquileia> but it turned out differently
20140525 15:50:31< zookeeper> logistics have never made much sense in this game
20140525 15:50:46< zookeeper> it's not like a detail like that breaks anyone's suspension of disbelief
20140525 15:51:17< iceiceice> how would the khalifate even send an army across the desert?
20140525 15:51:18< iceiceice> do they ship water on camels or smth?
20140525 15:52:06< iceiceice> i'm supposed to think of them as like, ~500 CE arabic nation, and not like the people in Dune?
20140525 15:52:20< aquileia> iceiceice: Their employer provides provisions and they exhaust the local resources
20140525 15:52:33< iceiceice> i mean.. theres not alot of local resources in the desert
20140525 15:52:41< aquileia> on they way back, they thus have to take a different route in order not to starve
20140525 15:52:43< iceiceice> presumably they work quite hard to establish whatever oasis they have
20140525 15:53:18< zookeeper> considering how difficult it is, i don't know why people don't rather just make campaigns about any of the other races :p
20140525 15:53:53< iceiceice> i guess you can make a story about getting water from cactuses at the perfect season or something :)
20140525 15:53:54< aquileia> because the khalifate are ripe, and the idea is in fact quite old (two years, I think)
20140525 15:54:29< zookeeper> iceiceice, cacti could be enemies you need to hit (blade weapons i guess) but they'd sting you back
20140525 15:54:52< iceiceice> hehe
20140525 15:55:01< iceiceice> the dreaded cursed cactus quintain
20140525 15:55:16< zookeeper> here lies Abdul-Raʿuf, killed by a cactus
20140525 15:55:29< noy> iceiceice: they would travel between fortified oasis 
20140525 15:55:42< noy> or other water sources
20140525 15:56:05< AI0867> iceiceice: neither did I. I suspect it's intermittent whenever not enough CPU/IO bandwidth is available to actually run the test to completion
20140525 15:56:22< noy> that's how arab societies operated, even great one
20140525 15:56:25< aquileia> zookeeper, iceiceice: We didn't plan to make scenarios of the journey towards wesnoth... with that gimmick though... ;)
20140525 15:56:40< iceiceice> hmm so the idea is not like, they march from khalifate capital with an army of 10,000
20140525 15:56:59< iceiceice> but perhaps they travel trhoguh the desert along many paths?
20140525 15:57:01-!- gfgtdf [~chatzilla@f054141213.adsl.alicedsl.de] has joined #wesnoth-dev
20140525 15:57:03< iceiceice> and merge at the destination?
20140525 15:57:12< aquileia> exactly
20140525 15:57:41< gfgtdf> iceiceice: do you think there s a good reason to have differet classes playsingle_controller and playmp-controler ?
20140525 15:58:18< iceiceice> only that it is what we have and it works
20140525 15:58:34< gfgtdf> hm ok
20140525 15:58:35< iceiceice> idk, is the scope of RiftWalker's project to merge them?
20140525 15:58:57-!- Sulfur [~Miranda@p5B327137.dip0.t-ipconnect.de] has joined #wesnoth-dev
20140525 15:58:59< gfgtdf> iceiceice: idk he wasnt online the lasg time i asked him
20140525 15:59:02< gfgtdf> last
20140525 15:59:15< iceiceice> thunderstruck: do you know? ^
20140525 15:59:34< gfgtdf> iceiceice: did you fix 169 wrk for you ?
20140525 15:59:36< gfgtdf> work
20140525 15:59:45< iceiceice> y it worked for me but only in testing the basic variation
20140525 15:59:49< aquileia> noy: So... is the cultural issue a blocker?
20140525 16:00:37< aquileia> Given enough gold, you can persuade pretty much everyone you're his friend
20140525 16:00:45< iceiceice> AI0867, loonycyborg: are all the jenkins builds with build=release?
20140525 16:00:58< loonycyborg> yes atm
20140525 16:01:07< aquileia> but if it doesn't work for you...
20140525 16:01:23< iceiceice> i think it might be worthwhile to add a debian debug build, i encountered an issue a while ago where the test executable would segfault but only in debug mode
20140525 16:01:39< iceiceice> i think beacuse it was concealed by optimizations
20140525 16:01:57< AI0867> 17:50 < zookeeper> logistics have never made much sense in this game ← the war will feed the war
20140525 16:02:08< iceiceice> probly not appropriate to add to travis but maybe jenkins
20140525 16:02:10< iceiceice> idk
20140525 16:03:18< aquileia> AI0867, iceiceice: We'll, Wesnoth models upkeep & provisions & bones (for undead) with gold
20140525 16:03:29< aquileia> *well
20140525 16:03:42< aquileia> it's easy and understandable
20140525 16:04:09< iceiceice> y but its *very* far from realistic
20140525 16:04:38< iceiceice> you can be a very wealthy empire but it doesnt mean you can conquer russia in the winter...
20140525 16:04:52< iceiceice> or send a navy to antarctica
20140525 16:05:30< iceiceice> idk maybe im just supposed to suspend belief and assume that my spearman unit might actualy just represent one man :p
20140525 16:05:52< aquileia> Climate isn't really respected, right. The worst offenders are caves full of villages - explanation? balancing...
20140525 16:06:07< aquileia> But more complexity is no solution
20140525 16:06:22< iceiceice> idk its only dwarves and trolls that have cave villages, no?
20140525 16:06:25< iceiceice> maybe some orcs?
20140525 16:06:42< iceiceice> its not clear to me that trolls eat anything (maybe rocks?)
20140525 16:06:56< iceiceice> i thought the dwarves have mushroom farms in NR, similar to dwarf fortress :)
20140525 16:07:18< iceiceice> i didnt feel there was a lore gap there
20140525 16:07:35< aquileia> I mean those uninhabited caves
20140525 16:07:36-!- cib [~cib@p5DC7514B.dip0.t-ipconnect.de] has joined #wesnoth-dev
20140525 16:07:58< thunderstruck> iceiceice: Maybe. Some logic will be different for MP and SP, though.
20140525 16:08:00-!- cib is now known as Guest47335
20140525 16:08:40< thunderstruck> iceiceice: So, maybe that separation should still exist.
20140525 16:08:59< AI0867> iceiceice: to be fair, the last two attempts to invade russia were made with no preparation for a winter war in russia
20140525 16:09:50< iceiceice> there's also just some places where the logistics make it infeasible to properly invade no matter how much money and power you have
20140525 16:09:58< iceiceice> like, Switzerland, or Afghanistan
20140525 16:10:13< AI0867> well, napoleon never intended to invade russia, he just wanted a decisive victory against the russian army
20140525 16:10:23< iceiceice> Russians tried to invade Afghanistan in 1970... that didnt go so well
20140525 16:10:36< AI0867> but they wouldn't commit to a battle, then destroyed everything that could be of use to the french in their retreat
20140525 16:10:42< iceiceice> US moved in in 2000s... 
20140525 16:10:56< iceiceice> helicopters made it possibel to take capital but its hard to say they controlled the country
20140525 16:11:24< iceiceice> i think afghanistan is like one of the classic historical places that empires continually try and fail to seize
20140525 16:11:25< AI0867> germany might have pulled it off, but hitler explicitly forbade any preparations for combat during the winter, as the war had to have been long finished by then
20140525 16:11:55< AI0867> they don't even try with switzerland
20140525 16:12:23< noy> aquileia: why would you spend money on them, when you could buy someone else?
20140525 16:13:00< AI0867> because we want kalifa in a fight somewhere ;)
20140525 16:13:14< noy> heh, of course
20140525 16:13:24< aquileia> Because the other options don't want to take a side or are in fact against you
20140525 16:13:51< iceiceice> gfgtdf: idk maybe we should just merge 169 to master
20140525 16:14:04< iceiceice> and test periodically for a few days / see if others report bugs
20140525 16:14:08< iceiceice> its hard to see that it could break anything
20140525 16:14:11< aquileia> the orcs already get tribute for not attacking, the elves are with Delfador, ...
20140525 16:14:15< noy> well, I think its also careful not to make it too obvious
20140525 16:14:20< gfgtdf> iceiceice: i have no probem with that
20140525 16:14:32< noy> and its really not that plausible, as if you're trying to shoehorn them in
20140525 16:14:43< Necrosporus> iceiceice, do you think I need to run git bisect?
20140525 16:14:50< noy> if you want to do it, I think a few hundred or so might be believable.... maybe mercenary travellers. 
20140525 16:15:02< iceiceice> Necrosporus: for the inspect assertion?
20140525 16:15:02< iceiceice> or new bug
20140525 16:15:14< Necrosporus> yes
20140525 16:15:24< aquileia> noy: That would pretty much wipe out the story
20140525 16:15:29< noy> but I'd just see that being a threat to local populaces and quickly be attacked as a source of fiction
20140525 16:15:29< Necrosporus> I can't reproduce it with 1.11.15
20140525 16:15:29< iceiceice> thats not an answer, the former or the latter?
20140525 16:15:37< iceiceice> ok
20140525 16:15:37< noy> *friction
20140525 16:15:46< noy> Whats wrong with dwarves?
20140525 16:15:52< Necrosporus> iceiceice, no new bugs so far
20140525 16:16:01< iceiceice> i think the most important thing is to get it reproduceable
20140525 16:16:10< iceiceice> you cant bisect until you've doen that pretty much
20140525 16:16:23< iceiceice> if its not reproduceable then theres not much we can do
20140525 16:16:34< aquileia> A few hundred khalifate can't retreat through half a continent of enemy territory
20140525 16:16:41< iceiceice> once its reproduceable i would say best thing si to get a core dump / backtrace etc.
20140525 16:17:35< iceiceice> can there be like a khalifate warlord or bandit or something, with a few followers?
20140525 16:17:36< aquileia> noy: Around 500, the dwarves are occupied with themselves
20140525 16:17:51< noy> what a few hundred cannot do, it would be even more difficult for a couple of thousand
20140525 16:17:55< aquileia> remember "Dwarves doors" and the following scenarios?
20140525 16:18:43< aquileia> The dwarves shut down most relations and aren't really active until NR brings them back from their hiding places
20140525 16:19:04< aquileia> noy: So whom would you hire?
20140525 16:19:58< noy> See, I'd suggest dwarves who didn't want to stay underground. 
20140525 16:20:05< aquileia> The orcs are persuaded later on by Asheviere, but Eldred already pays them off for a cease-fire
20140525 16:20:38< aquileia> are there any dwarves available in force?
20140525 16:20:40< noy> I think that's more plausible than the Dues ex machina of a group that had no providence in the area showing up
20140525 16:21:40< noy> who knows, I can't see that ALL dwarves would listen to their king's edict
20140525 16:22:35-!- gfgtdf [~chatzilla@f054141213.adsl.alicedsl.de] has quit [Quit: ChatZilla 0.9.90.1 [Firefox 29.0.1/20140506152807]]
20140525 16:23:59< irker881> wesnoth: Chris Beck wesnoth:master 239cb5fdc622 / src/play_controller.cpp: save play_controller.init_side_done as a toplevel member of snapshot http://git.io/S0-SwA
20140525 16:24:01< irker881> wesnoth: Chris Beck wesnoth:master 77ead906927b / src/playsingle_controller.cpp: call maybe_do_init_side when the engine starts a sides turn http://git.io/IiaOdQ
20140525 16:24:03< irker881> wesnoth: Chris Beck wesnoth:master 28d0e2df0e8d / src/ (play_controller.cpp playsingle_controller.cpp): Merge pull request #169 from cbeck88/fixup_init_side_done http://git.io/A9SR7A
20140525 16:25:43-!- sachith500 [~kvirc@112.134.55.85] has joined #wesnoth-dev
20140525 16:26:23< happygrue> It could be more possible in a larger band of mercinaries to have a group of Khalifate.  A small group of mounted archers or something that play a key role in a battle under the banner of a larger outfit
20140525 16:27:17< noy> sure, that might work
20140525 16:28:23< irker881> wesnoth: Chris Beck wesnoth:master 7b876194bd79 / changelog: update changelog http://git.io/kyXYVQ
20140525 16:29:40-!- sachith500|2 [~kvirc@112.134.205.211] has quit [Ping timeout: 265 seconds]
20140525 16:31:23< happygrue> perhaps an engament between two mounted detachments is turned into a rout because one side has Khalifate archers along and the other has no archers at all, so when the main armies meet one side has no cavalry left and is defeated... etc.
20140525 16:32:39< aquileia> noy: "Cut off from contact with the surface, the dwarves of Knalga fought on, but knew their utter subjugation could not be averted forever. " Doesn't sound as if they would be available for hire
20140525 16:32:48< happygrue> specificallt Rami I was thinking:  http://units.wesnoth.org/trunk/mainline/en_US/Rami.html
20140525 16:33:12< happygrue> or higher level elite forces
20140525 16:33:14< aquileia> Except if that was shortly after 501, that would change things
20140525 16:34:58< aquileia> Nope, it was at least a human generation before 534
20140525 16:37:17< aquileia> happygrue, noy: A small khalifate detachment wouldn't make it back, I fear
20140525 16:37:32< noy> shrug
20140525 16:38:02< aquileia> We can reduce the number up to the point where on nightmare you'll end the campaign with 5% of your troops
20140525 16:38:05< happygrue> I missed the first part of the setup, but I was talking about a small detachment playing a key role in a much larger army
20140525 16:38:20< aquileia> but... you can't play without units
20140525 16:38:21< happygrue> take your 10,000, and have a few hundred khalifate horse archers in there
20140525 16:39:12< happygrue> is this backstory only or is the idea to recruit them?
20140525 16:39:17< aquileia> happygrue: It's the story of the khalifate troops on their way home... the rest of the army wouldn't come with them
20140525 16:39:56< aquileia> and thus you have only khalifate units
20140525 16:40:17< happygrue> Ah, I see
20140525 16:41:11< aquileia> As I said it doesn't have to be 10,000 - but a few hundred + an aspect ratio of 10 persons per unit means it'll be the hardest campaign ever created
20140525 16:41:39< happygrue> what is the scale of UtBS?
20140525 16:41:50< happygrue> that has a similar "just a few left" feel to it last I played it
20140525 16:41:57< happygrue> often it's just one person per unit
20140525 16:41:58< aquileia> "Take these 20 units... you'll get no reinforcements for the rest of the campaign, and you'll have to cross half of Wesnoth"
20140525 16:42:03< happygrue> scale is totally relative IMO
20140525 16:42:24< aquileia> sure
20140525 16:42:29< happygrue> so the idea is no recruiting ever?
20140525 16:42:49< aquileia> you recruit from your population pool
20140525 16:43:47< aquileia> and fighting soldiers can't provision themselves (explanation for the upkeep limit)
20140525 16:44:05-!- Netsplit *.net <-> *.split quits: Necrosporus, boucman
20140525 16:44:53< aquileia> I have a basic draft for the first 5 scenarios (not sufficiently converted to Wesnoth though) if you are interested
20140525 16:45:08< aquileia> But I guess you don't want to spoil the fun
20140525 16:45:17< happygrue> hehe
20140525 16:45:37< happygrue> if it could be playable at some point then sure I'll wait
20140525 16:45:54< aquileia> at least that draft is a lot shorter than the first 3 chapters of the book
20140525 16:45:58-!- Netsplit over, joins: Necrosporus
20140525 16:46:08< happygrue> about to be afk a bit now.  IMO the fluff is what justifies what we choose to do, not something that tells us what we CAN'T do
20140525 16:46:11< happygrue> WINR
20140525 16:46:33< happygrue> so do something that's fun to play and the fluff can be made to work around it.  :D
20140525 16:47:04-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20140525 16:55:21-!- wesbot changed the topic of #wesnoth-dev to: tagged 1.11.15 (1.12 beta5), announcing "soon" | string+feature freeze active on 1.12 | 242 bugs, 344 feature requests, 29 patches | Logs: http://irclogs.wesnoth.org | Alternate logs: http://wesnoth.debian.net | Don't paste on IRC! Use a pastebin: http://pastebin.com | http://imagebin.org
20140525 16:58:14< aquileia> noy, zookeeper, happygrue, iceiceice: Thanks for the discussion, it really helps to find weak spots in the story and clean them up
20140525 17:00:15< zookeeper> aquileia, a detachment can make it back... by sea
20140525 17:01:19< aquileia> Yes, but first they have to arrive there from Weldyn
20140525 17:02:09< aquileia> As soon as they fought their way out of Wesnoth and are in khalifate territory, the worst part is over
20140525 17:02:16-!- RiftWalker [~nathan@ip24-252-126-205.no.no.cox.net] has joined #wesnoth-dev
20140525 17:02:51< aquileia> but this will be a campaign where you can't prevent some losses
20140525 17:03:22< zookeeper> well i don't know what the idea is WRT the population limit
20140525 17:03:45< aquileia> shikadibot: WRT
20140525 17:03:51< zookeeper> with regards to
20140525 17:03:57< aquileia> ah, ok
20140525 17:04:01< zookeeper> if you mean you can't recruit at all, then i think that's probably not going to be very fun because you'll need to avoid losses like the plague
20140525 17:04:23< aquileia> You can recruit in the limits of your available troops
20140525 17:04:54< zookeeper> as pointed out, UtBS is conceptually a similar thing with your entire people travelling with you, and there's no limitations to recruitment there, and that doesn't make affect the feel really.
20140525 17:05:07< aquileia> on easy the limit is no issue, on nightmare... well, it warrants the name
20140525 17:05:12-!- travis-ci [~travis-ci@ec2-54-198-70-55.compute-1.amazonaws.com] has joined #wesnoth-dev
20140525 17:05:12< travis-ci> [travis-ci] wesnoth/wesnoth#2788 (master - 7b87619 : Chris Beck): The build has errored.
20140525 17:05:12< travis-ci> [travis-ci] Build details : http://travis-ci.org/wesnoth/wesnoth/builds/25997447
20140525 17:05:12-!- travis-ci [~travis-ci@ec2-54-198-70-55.compute-1.amazonaws.com] has left #wesnoth-dev []
20140525 17:06:05< zookeeper> anyway, this is wesnoth, having a group of a few hundred people fight their way through half the continent doesn't sound extraordinary at all
20140525 17:07:38< iceiceice> i guess i suspended disbelief when i played UtBS
20140525 17:07:52< iceiceice> but i also didnt think too hard about it, the story drew my attention elsewhere :)
20140525 17:08:12< aquileia> sure, a few hundred soldiers escape the wrath of Asheviere...
20140525 17:09:01< zookeeper> and i suggested the sea route because that way you at least get to see and interact with more races and whatnot, if half the campaign is just a trudge through a big desert then you'll have to be a genius to make it be fun :p
20140525 17:09:14< aquileia> What is less realistic, 1000 mercenaries being hired or 100 defeating Wesnoth?
20140525 17:10:02< zookeeper> and WRT dwarves; there's at least those "surghan mercenaries" in SoF which don't seem very nice folks and seem like they hire their muscle to the highest bidder and don't mind attacking even other dwarves
20140525 17:10:59< aquileia> zookeeper: The exact route is the next thing we'll need. But in any case, they can't go back the way they came, so the sea is their best bet
20140525 17:11:27< zookeeper> they're hired by xenophobic elves to kill other dwarves... so i doubt they care what some dwarvish king somewhere says
20140525 17:12:09< aquileia> It's not a kings edict, they can't get out of their tunnels
20140525 17:12:40< zookeeper> not all dwarves live(d) in knalga
20140525 17:12:58< zookeeper> it was just presumably their largest city/empire/whatever
20140525 17:14:17< aquileia> My take of wesnothian history is that all of the heart mountains are occupied with orcs. There's not too much mountains other than that
20140525 17:15:53< zookeeper> i... don't really see it like that. do you risk running into orcs anywhere in the heart mountains? sure... but that doesn't mean there's not many other dwarvish settlements which just haven't been overrun for one reason or another
20140525 17:15:56< aquileia> And the real reason behind dwarves not being a possibility for me: They wouldn't struggle in the mountains, and I have 2 nice scenarios designed...
20140525 17:16:26< aquileia> Yes, I know, deus ex machina...
20140525 17:16:47< iceiceice> idk theres two ways to go about it right?
20140525 17:16:49-!- kex [~kex@89.205.75.19] has quit [Remote host closed the connection]
20140525 17:16:54< iceiceice> you can make the scenarios and then invent the fluff after
20140525 17:17:00< iceiceice> or do the ESR method
20140525 17:17:00< iceiceice> and write the story
20140525 17:17:02< aquileia> but I want to crush the players army under rocks... heavy rocks
20140525 17:17:09< iceiceice> then apparently just scribble in the map editor what you need
20140525 17:17:22< iceiceice> and balance the 5% of hexes where battle occurs, after playtesting it
20140525 17:17:24-!- kex [~kex@89.205.75.19] has joined #wesnoth-dev
20140525 17:18:06< zookeeper> i thought you were simply discussing dwarves in the context of alternate mercenaries eldred could have hired... not in the context of who are the protagonists of the campaign
20140525 17:18:28< aquileia> iceiceice: In the beginning was the word, and then they built the world around it
20140525 17:19:13< aquileia> zookeeper: Ah, I'd have no problem to mix in some dwarves in the first scenario
20140525 17:20:02< aquileia> That'd even be of advantage for the multiplayer branch
20140525 17:20:41< aquileia> currently the second player would have to wait through most of the first scenario, guarding the camp
20140525 17:21:07< aquileia> If he controls the dwarves, that's nice
20140525 17:21:32-!- kex [~kex@89.205.75.19] has quit [Ping timeout: 245 seconds]
20140525 17:22:26< zookeeper> anyway, if we're talking about dwarves as possible other protagonists, then i think that'd be a _lot_ easier to do and there'd be even more of plausible story possibilities
20140525 17:22:29< aquileia> We can't promise that we'll do MP supprt, though
20140525 17:24:45< zookeeper> or you could do it with undead! an orphaned army of skeletons and ghosts and whatnot, desperately trying to get back home :~(
20140525 17:24:57< aquileia> Wrathbow's first criterion was "khalifate campaign", so a full conversion is unlikely
20140525 17:25:25< aquileia> Home to whom?
20140525 17:25:43< aquileia> Wait, I better shouldn't have asked...
20140525 17:26:59< zookeeper> well that's the twist!
20140525 17:27:21< zookeeper> they'd be trying to get back home to friends and family... but in the end they find out that they're all dead and they've been with them all along
20140525 17:27:23< zookeeper> :troll smiley:
20140525 17:28:12< aquileia> If you create that campaign, I'll play it
20140525 17:29:22< aquileia> but for now I'll go back to the heap of work waiting for me...
20140525 17:30:01-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20140525 17:30:31-!- stikonas_ [~gentoo@wesnoth/translator/stikonas] has quit [Ping timeout: 252 seconds]
20140525 17:30:41-!- stikonas__ [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20140525 17:34:29< iceiceice> quick question:
20140525 17:34:36< iceiceice> does gettext run before macros?
20140525 17:34:37< iceiceice> or after?
20140525 17:34:52< iceiceice> if I use N_("lawful") as a string argument to a macro,
20140525 17:35:09< iceiceice> will that cause like a bajillion instances of N_("lawful") in the pot file?
20140525 17:35:36< aquileia> Doesn't the pot file boil them down to one instance?
20140525 17:35:46< iceiceice> i have *no* idea!
20140525 17:36:15< aquileia> I thought it only lists unique strings (per namespace, that is)
20140525 17:38:26< iceiceice> ok thx
20140525 17:39:25< iceiceice> and what are the rules about manipulating translatable strings ?
20140525 17:39:34< iceiceice> i cant do something like concatenate two strings and expect it to be translatable,
20140525 17:39:43< iceiceice> but if I preview all the strings that could occur that way then that is ok?
20140525 17:39:49< iceiceice> can i do that in a code comment?
20140525 17:40:03< iceiceice> if i knew where to look this stuff up i would...
20140525 17:41:25< Necrosporus> aquileia, wesnothian armies usually below 20 soldiers :)
20140525 17:44:29< aquileia> happygrue, zookeeper: Uhm, is there an analogon to "a dwarf" for the khalifate? A khalifa? Or do I have to write khalifate soldier all the time?
20140525 17:45:00< Necrosporus> > for all we know, konrad's HttT army could be anything from a few thousand to tens of thousands // in my case it's about 80 
20140525 17:45:41< zookeeper> aquileia, well do you need to identify them like that everywhere? they're humans so "men" would probably fit a lot of the time
20140525 17:46:30< zookeeper> "i saw two hundred <analogon>s behind that hill" would sound pretty weird if it was commonly used
20140525 17:46:32< aquileia> I'm not thinking too much about the story but rather the discussions to come
20140525 17:47:02< zookeeper> well, i don't know of such a word
20140525 17:47:30< aquileia> E.g. I'm summarizing today's discussion and it's a bit weird to write "khalifate man/soldier/troops/..." all the time
20140525 17:48:53< aquileia> But I'll have to cope with it, I guess
20140525 17:59:00-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20140525 18:15:25-!- Sulfur [~Miranda@p5B327137.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20140525 18:17:16< happygrue> aquileia: no help from me regarding khalifate language there, but I can see your point.  Perhaps noy has something he's been using in his writeup?
20140525 18:17:40< happygrue> also, with regard to "defeating wesnoth" you're meaning they are able to evade the might of Wesnoth, right?
20140525 18:18:17< happygrue> because Khalifate are about as good a people to evade major amry conflict as we have (other than drakes)
20140525 18:18:57< happygrue> they are very mobile and could probably outrun orcs and trolls in the open, but even their mounted units can move quickly over the hilly terrian, outrunning almost everyone else there
20140525 18:19:12< happygrue> they aren't going to get away from elves in a forest... ;)
20140525 18:20:11< happygrue> but some hundreds could make a fighting escape for quite a long way by being able to get to the terrian at the end of the scenario that would allow them to outrun whomever they were currently fighting against, to throw out a possible plot point
20140525 18:24:16-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20140525 18:33:21< aquileia> happygrue: Sorry, I was afk
20140525 18:33:41< happygrue> as was I for a while, no worries. :D
20140525 18:34:45< aquileia> They play a major part in Eldred's victory against Delfador, but it's rather because Delfador's troops aren't full soldiers
20140525 18:35:44< aquileia> As the Wesnothian army was severely defeated at the ford of Abez, both sides take what troops they can get
20140525 18:36:39< aquileia> And that mercenary detachment of khalifate veterans (+ optionally dwarves) turnes the tide of the  battle quite massively
20140525 18:39:03< aquileia> The opposing army flees and takes considerable losses, but later on the khalifa(te) will be repaid...
20140525 18:45:39< irker881> wesnoth: Chris Beck wesnoth:master 684d1c35c64e / src/ (callable_objects.cpp reports.cpp unit_types.cpp unit_types.hpp): unit_type::alignment to use MAKE_ENUM macro http://git.io/Yd8_Wg
20140525 18:48:09< aquileia> happygrue: Would you like to playtest the campaign when it's finished? We're surrently a team of two, and I'll have to play it on easy myself...
20140525 18:48:18< aquileia> *currently
20140525 18:49:28< happygrue> aquileia: is it a mp campaign or sp?
20140525 18:49:31< happygrue> but sure in either case
20140525 18:50:36< aquileia> We will finish the SP mode, and if we still have time, we'll transfer most AI allies to a second player
20140525 18:51:31< aquileia> The army has two generals which usually have different tasks
20140525 18:52:16< aquileia> like: A defends the main road while B launches an attack via a hidden path
20140525 18:53:41< aquileia> and of course we could go crazy and add the enemy as a third side (victory by killing at least n units)
20140525 18:53:58< aquileia> It'd be hard to balance the latter, though
20140525 18:54:29< aquileia> After all, all players should advance to the next scenario
20140525 18:54:43< happygrue> Indeed
20140525 18:54:59-!- kex [~kex@89.205.75.19] has joined #wesnoth-dev
20140525 18:55:17< happygrue> Perhaps the way to do it is to say you need to kill an average of x units per scenario after the last one to win, so that it's not over until the end
20140525 18:55:20< aquileia> And if Rigor takes the opposing side...
20140525 18:55:30< happygrue> though you really could just win by killing everyone.  Tricky stuff :D
20140525 18:56:05< happygrue> anyway, afk again but I'm happy to look at it once it's ready for a testing runthrough
20140525 18:56:11< aquileia> I think both sides should be able to win in the end, so I'd refrain from the ultimate goal "kill everyone"
20140525 18:56:32< aquileia> but playtesting will show what's feasible
20140525 18:56:40< happygrue> yes, but in practice with two humans it could happen that one side just does :D
20140525 18:57:02< aquileia> Yes, sure
20140525 18:57:24< aquileia> If e.g. you played against me, I'm as good as dead =P
20140525 18:57:58< aquileia> That means for MP testing we'll need three players of approximately equal level
20140525 18:58:55< aquileia> Or... we design it as two noobs against a ladder professional!
20140525 18:59:29-!- kex [~kex@89.205.75.19] has quit [Ping timeout: 258 seconds]
20140525 19:05:31-!- Guest47335 [~cib@p5DC7514B.dip0.t-ipconnect.de] has quit [Ping timeout: 240 seconds]
20140525 19:26:33-!- shadowm_desktop [ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20140525 19:42:30< Necrosporus> where's 1.10? http://www.wesnoth.org/gettext/
20140525 19:43:58-!- aquileia [2edf524c@gateway/web/freenode/ip.46.223.82.76] has quit [Quit: Page closed]
20140525 19:44:46< shadowm> 1.10 had a long and productive life and it's our hope that it'll finally find peace and happiness in heaven.
20140525 19:50:16< irker881> wesnoth: gfgtdf wesnoth:1.12 29f8834ab72d / src/ (ai/actions.cpp playsingle_controller.cpp): cleanup check_victory in ai actions http://git.io/ZFJT9Q
20140525 19:50:18< irker881> wesnoth: gfgtdf wesnoth:1.12 65c3c1c18dc0 / src/ (playmp_controller.cpp playsingle_controller.cpp): remove check_victory in play_human_turn http://git.io/VZ2B8w
20140525 19:55:17< Necrosporus> shadowm, but I would like to see statistics and possibly fetch translation files
20140525 19:55:44< Necrosporus> also translators might need it in case if same addon exists for 1.10 and 1.11
20140525 19:56:27< shadowm> Translators should be working on mainline 1.12 at this time, not 1.10 mainline or UMC.
20140525 19:57:03< irker881> wesnoth: gfgtdf wesnoth:master 9f399d89797b / src/ (ai/actions.cpp playsingle_controller.cpp): cleanup check_victory in ai actions http://git.io/05C3cQ
20140525 19:57:06< irker881> wesnoth: gfgtdf wesnoth:master d1d70cb7b1ed / src/ (playmp_controller.cpp playsingle_controller.cpp): remove check_victory in play_human_turn http://git.io/sEa4Kg
20140525 19:59:47< Necrosporus> shadowm, translators should be working on whatever they want to work, also they may need to refer to old translations in order to make new ones
20140525 20:12:30-!- gfgtdf [~chatzilla@f054141213.adsl.alicedsl.de] has joined #wesnoth-dev
20140525 20:13:10< gfgtdf> iceiceice: in the make enum macro after assert(...) my compiler complains because we dotn return a value after that
20140525 20:13:16< gfgtdf> con't
20140525 20:13:18< gfgtdf> don't
20140525 20:13:31< iceiceice> oh
20140525 20:13:52< iceiceice> can you push whatever is the right hting to do?
20140525 20:14:23< iceiceice> i changed from `throw "something"` b/c i thought assert is prefereable to that but maybe i'm wrong
20140525 20:14:53< iceiceice> there are a lot of `throw 42`'s int he code also ...
20140525 20:15:04< iceiceice> idk if that makes the compiler try to search for catch 42 blocks
20140525 20:16:08-!- travis-ci [~travis-ci@ec2-54-197-185-19.compute-1.amazonaws.com] has joined #wesnoth-dev
20140525 20:16:08< travis-ci> [travis-ci] gfgtdf/wesnoth-old#110 (luaslpit1 - 142388b : gfgtdf): The build passed.
20140525 20:16:08< travis-ci> [travis-ci] Build details : http://travis-ci.org/gfgtdf/wesnoth-old/builds/26006023
20140525 20:16:08-!- travis-ci [~travis-ci@ec2-54-197-185-19.compute-1.amazonaws.com] has left #wesnoth-dev []
20140525 20:17:44< gfgtdf> iceiceice: hmm i dnt realy know a value that makes sense t return there
20140525 20:18:38< iceiceice> its for the corrupted enum, right?
20140525 20:18:42< gfgtdf> iceiceice: ye
20140525 20:18:46< iceiceice> coudl just return "deadbeef" or "bad enum" or something
20140525 20:19:12< gfgtdf> mh ye why not
20140525 20:19:16< iceiceice> or empty string
20140525 20:23:53< irker881> wesnoth: Chris Beck wesnoth:master 9ac310e1916b / src/ (7 files in 3 dirs): team::controller enum to use MAKE_ENUM macro http://git.io/UXrZvQ
20140525 20:23:55< irker881> wesnoth: Chris Beck wesnoth:master b2ac9c5b3140 / src/ (ai/actions.cpp playmp_controller.cpp playsingle_controller.cpp): Merge branch 'master' of git://github.com/wesnoth/wesnoth http://git.io/ieWb3w
20140525 20:27:24-!- travis-ci [~travis-ci@ec2-54-221-172-150.compute-1.amazonaws.com] has joined #wesnoth-dev
20140525 20:27:24< travis-ci> [travis-ci] wesnoth/wesnoth#2791 (master - d1d70cb : gfgtdf): The build was broken.
20140525 20:27:24< travis-ci> [travis-ci] Build details : http://travis-ci.org/wesnoth/wesnoth/builds/26006494
20140525 20:27:24-!- travis-ci [~travis-ci@ec2-54-221-172-150.compute-1.amazonaws.com] has left #wesnoth-dev []
20140525 20:30:10< gfgtdf> iceiceice: what does check_victory_always_one  test ?
20140525 20:30:28< iceiceice> i dont know what that is, let me look
20140525 20:30:52< iceiceice> ohhh
20140525 20:30:52< iceiceice> the unit test
20140525 20:31:13-!- travis-ci [~travis-ci@ec2-54-221-172-150.compute-1.amazonaws.com] has joined #wesnoth-dev
20140525 20:31:13< travis-ci> [travis-ci] gfgtdf/wesnoth-old#111 (master - 684d1c3 : Chris Beck): The build was broken.
20140525 20:31:13< travis-ci> [travis-ci] Build details : http://travis-ci.org/gfgtdf/wesnoth-old/builds/26006340
20140525 20:31:13-!- travis-ci [~travis-ci@ec2-54-221-172-150.compute-1.amazonaws.com] has left #wesnoth-dev []
20140525 20:31:55< gfgtdf> 	iceiceice: my last commit removes a possible check_victory after turn refresh events
20140525 20:32:00< iceiceice> it checks that if you start a scenario
20140525 20:32:05< iceiceice> with two always defeated sides
20140525 20:32:07< iceiceice> like
20140525 20:32:22< iceiceice> so it starts a 2 v2 scenario
20140525 20:32:25< iceiceice> in which one team is normal
20140525 20:32:27< iceiceice> and one team is both always defeated
20140525 20:32:37< iceiceice> it checks that the scenario immediately ends
20140525 20:33:34< gfgtdf> iceiceice: hm ok that ist not cofusing that this fails after taht commit
20140525 20:33:36< gfgtdf> that
20140525 20:34:13< gfgtdf> iceiceice: do you think we shoudl put a check after side initilisation (side .. turn events) back in ?
20140525 20:34:37< iceiceice> imo y i think so
20140525 20:35:44< iceiceice> gfgtdf: i looked at oyur commit message
20140525 20:35:52< gfgtdf> iceiceice: hm i dont get ewhta is wrong here: https://s3.amazonaws.com/archive.travis-ci.org/jobs/26006341/log.txt
20140525 20:36:02< gfgtdf> what*
20140525 20:36:05< iceiceice> i think the expensive thing about check_victory is that it invalidates the map and forces redraw?
20140525 20:36:18< iceiceice> but will check_victory really be called often enoguh for that o matter?
20140525 20:36:19< gfgtdf> iceiceice: ye there is a comment about that in check_victory
20140525 20:37:00< gfgtdf> ye the play human turn loops rns permanenty durign teh human turn even if teh user does nothing
20140525 20:37:09< gfgtdf> loop runs
20140525 20:37:16< iceiceice> oh i dint know that
20140525 20:37:23< iceiceice> the log you linke dmeans, the C++ unit tests gave message 200
20140525 20:37:26< iceiceice> i have no idea what it means
20140525 20:38:11-!- trewe [~trewe@2001:8a0:d139:9701:626c:66ff:fe92:9b7c] has joined #wesnoth-dev
20140525 20:38:36< iceiceice> hmm
20140525 20:38:42< iceiceice> no
20140525 20:38:49< iceiceice> ok i guess the log you linked isnt the same as the test that ran?
20140525 20:38:49< iceiceice> https://travis-ci.org/wesnoth/wesnoth/jobs/26006495
20140525 20:38:54< gfgtdf> iceiceice: if we want that check_victory back in we could put it in do_init_side at teh end
20140525 20:39:06< iceiceice> y ok
20140525 20:39:21< iceiceice> i mean its up to you sort of, we can make it do whatever 
20140525 20:39:27< iceiceice> i woudl just like to have tests sot hat we know when things are changing
20140525 20:39:36< gfgtdf> iceiceice: no my link came from http://travis-ci.org/gfgtdf/wesnoth-old/builds/26006340
20140525 20:39:46< iceiceice> oh its a different build
20140525 20:39:49< gfgtdf> iceiceice: which is from current master with no chenges by me
20140525 20:40:07< gfgtdf> iceiceice:  just me pushing ti to gfg... affter pulling
20140525 20:40:09< gfgtdf> it
20140525 20:40:56< iceiceice> idk i have never seen 200 exit code before
20140525 20:41:05< iceiceice> it mgiht be some random fault
20140525 20:41:40< iceiceice> i will say, it makes sense to me that if you start the game and one team is defeated, it should end immediately
20140525 20:42:09< iceiceice> wesnoth normally ends right now if one side doesnt have a leader or smth iirc
20140525 20:42:16< irker881> wesnoth: mattsc wesnoth:1.12 db45be4330b4 / projectfiles/Xcode/ (3 files in 2 dirs): Xcode project update for 1.11.15 http://git.io/biKsmA
20140525 20:42:29< gfgtdf> iceiceice: hm ye but teh previous behaviour was different
20140525 20:43:11-!- kex [~kex@89.205.75.19] has joined #wesnoth-dev
20140525 20:43:13< iceiceice> still, whats the motivation to change it?
20140525 20:44:35< gfgtdf> iceiceice: none i said the main reason for that cahnge in the commit message
20140525 20:48:17-!- kex [~kex@89.205.75.19] has quit [Ping timeout: 264 seconds]
20140525 20:49:19< iceiceice> hm ok 
20140525 20:49:26< iceiceice> well i would prefer to keep the test if possible
20140525 20:49:54< iceiceice> i guess if you want to remove it, then maybe can fix it or add a test for a similar kind of scenario?
20140525 20:51:27< gfgtdf> iceiceice: no i think we can try to keep it
20140525 20:52:04< RiftWalker> iceiceice:  In response to your earlier question, I don't plan on merging the playsingle- and playmp-controller classes
20140525 20:52:08< gfgtdf> iceiceice: in teh test teh sides whetre ai sides ?
20140525 20:52:18< iceiceice> no all human
20140525 20:52:21< RiftWalker> at least, not at first
20140525 20:52:27< iceiceice> i have osme with ai sides
20140525 20:52:50< iceiceice> RiftWalker: ok, good to know
20140525 20:52:52< gfgtdf> iceiceice: i think it woudl be interesting f in the previousl iplementation it works with ai sides
20140525 20:53:25< iceiceice> what does "works" mean?
20140525 20:53:29< iceiceice> it means the scenario continues?
20140525 20:53:58< gfgtdf> iceiceice: wel whethe rteh test doesnt fail
20140525 20:54:03< gfgtdf> well whether
20140525 20:54:08< iceiceice> oh you want to add a test?
20140525 20:54:15< iceiceice> y that's easy
20140525 20:54:52-!- travis-ci [~travis-ci@ec2-54-221-172-150.compute-1.amazonaws.com] has joined #wesnoth-dev
20140525 20:54:52< travis-ci> [travis-ci] wesnoth/wesnoth#2792 (master - b2ac9c5 : Chris Beck): The build was broken.
20140525 20:54:52< travis-ci> [travis-ci] Build details : http://travis-ci.org/wesnoth/wesnoth/builds/26007560
20140525 20:54:52-!- travis-ci [~travis-ci@ec2-54-221-172-150.compute-1.amazonaws.com] has left #wesnoth-dev []
20140525 20:56:01< iceiceice> the file is in data/test/scenarios/test_check_victory.cfg
20140525 20:56:17< iceiceice> the new test is just one line below the always case
20140525 20:56:32< iceiceice> you can mark with ai instead of human
20140525 20:57:17< gfgtdf> iceiceice: hm i just looked at te test an i think it's confusing: FAIL_IF_NO_END  uses the turn1 side 2 event to test whether the game continued, but even thought i deledted the test after the side/turn events. The check after the side/turn end shoudl still prevent it from reacing the side 2 turn
20140525 20:58:10< gfgtdf> iceiceice: this one https://github.com/wesnoth/wesnoth/blob/master/src/playsingle_controller.cpp#L657
20140525 20:58:22< iceiceice> hmm then maybe i should refacotr the test if its confusing
20140525 20:58:28< iceiceice> it is fairly convoluted
20140525 20:59:11< gfgtdf> iceiceice: no not teh test, the result is
20140525 20:59:57< iceiceice> hmm maybe it has to do with [end_turn] implementation?
20140525 21:03:11< iceiceice> hmm so oine thing thats kind of funny
20140525 21:03:16< iceiceice> i'm adding MAKE_ENUM to places
20140525 21:03:22< iceiceice> i realized that in play_turn.cpp
20140525 21:03:28< iceiceice> when it processes netowrk data to change_controller,
20140525 21:03:37< iceiceice> it just ignores signals to make a side empty
20140525 21:03:45< iceiceice> apparently
20140525 21:03:53< gfgtdf> iceiceice: where do you throw end turn exception ?
20140525 21:04:23< iceiceice> idk i dont throw it, i just used [end_turn] tag
20140525 21:04:35< iceiceice> so i guess in game_events/wml_actions.cpp
20140525 21:05:29< gfgtdf> iceiceice: hm ok you used it inside the start event it seems
20140525 21:05:52< iceiceice> is that bad?
20140525 21:07:34< gfgtdf> iceiceice: hm not obviously but i still wouldnt use it becasue id say tecnicly we are not inside a "turn" in a start event
20140525 21:08:08< iceiceice> hmm ok
20140525 21:08:09< gfgtdf> iceiceice: this is not based on my information about teh code
20140525 21:08:32< iceiceice> this is subtlety i have never fully grasped, the distinction between "side 1 turn 1" and "start"
20140525 21:08:43< iceiceice> i can change the unit tests to do that instead 
20140525 21:09:10< gfgtdf> iceiceice: hm i dont think its that important
20140525 21:11:57< gfgtdf> iceiceice: ok now i realy dont understand it: i onyl changed https://github.com/wesnoth/wesnoth/blob/master/src/playsingle_controller.cpp#L809 but i think that code souldnt even be called in this case because  end_turn_ is true before we enter that function ?
20140525 21:14:08< iceiceice> hmm
20140525 21:14:33< iceiceice> what else has changed recently?
20140525 21:14:39< iceiceice> does maybe_do_init_side call check_vicotry?
20140525 21:15:39< iceiceice> is there any way that my commit is involved?
20140525 21:15:41< iceiceice> i guess not
20140525 21:16:16< iceiceice> hmm i guess i could compile the debug version
20140525 21:16:18< iceiceice> and run in gdb
20140525 21:16:23< iceiceice> and set breakpoints in that function
20140525 21:17:24-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20140525 21:19:52-!- gfgtdf_ [~chatzilla@f054168225.adsl.alicedsl.de] has joined #wesnoth-dev
20140525 21:21:53< RiftWalker> gfgtdf: I'm looking at the code you asked about. It isn't necessary to my project to remove playsingle_scenario, but it depends mainly on whether we're keeping playsingle_controller. From what I can see, little if anything would be broken by using playmp_controller (and therefore playmp_scenario), but I need to take a closer look. For now I'm going to keep it to avoid any breakage, since it doesn't really affect my project goals.
20140525 21:22:05-!- gfgtdf [~chatzilla@f054141213.adsl.alicedsl.de] has quit [Ping timeout: 258 seconds]
20140525 21:22:17-!- gfgtdf_ is now known as gfgtdf
20140525 21:23:05< iceiceice> gfgtdf: do you know what "handle_generic_event" is called?
20140525 21:23:08< iceiceice> in play_single_controller?
20140525 21:23:25< gfgtdf> iceiceice: hm my ai manager i think
20140525 21:23:29< gfgtdf> by
20140525 21:24:03< iceiceice> hmm so the only hits i get for "throw end_turn_exception"
20140525 21:24:07< iceiceice> are 4 in src/menu_events
20140525 21:24:14< iceiceice> releated to :droid and :idle
20140525 21:24:17< iceiceice> 4 in plaympcontroller
20140525 21:24:23< iceiceice> and 1 in playsingle controller, which is right there
20140525 21:24:28< iceiceice> in handle_generic_event
20140525 21:24:42< gfgtdf> iceiceice: synced_context::pull_remote_user_input calso calls handle generic event but you dont use mp sync in taht sxenario plus it not a mp 
20140525 21:24:53< iceiceice> so i think that function must be throwing all the end turn exceptions in sp games
20140525 21:27:06< iceiceice> hmm ok i have now built most of wesnoth in debug mode
20140525 21:27:28< iceiceice> got a big slowdown when i had to link the first big hunk...
20140525 21:28:37-!- mjs-de [~mjs-de@g228128119.adsl.alicedsl.de] has quit [Remote host closed the connection]
20140525 21:29:39< gfgtdf> RiftWalker: i think it woudl be nice if we could merge playsinge/playmp controller first there is little reason to keep then especialy sinc ethere is already mp cod ein plamy controller and also it coudl be nice to be able to connect to a mp server during teh actual game
20140525 21:30:58< gfgtdf> since there is*, mp code in playsingle_controller*, the*
20140525 21:31:23< gfgtdf> iceiceice: im current compiling master agin
20140525 21:31:25< iceiceice> hmm i am finding that the units tests re all good with the build on master
20140525 21:32:06< iceiceice> that doesnt make sense
20140525 21:35:09-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Remote host closed the connection]
20140525 21:35:28< iceiceice> i dont understand
20140525 21:36:01< iceiceice> why would the unit test pass on me and fail on traivs?
20140525 21:36:51< iceiceice> i think it might be spurious timeout, maybe we should rebuild?
20140525 21:37:07< iceiceice> for some reason i cant rerun the test on travis
20140525 21:38:43< iceiceice> i think i'm going to push an empty commit to make travis rebuild
20140525 21:39:18< iceiceice> do you think thats a bad idea?
20140525 21:40:05< iceiceice> hmm maybe i shold just try to refresh github credentialy
20140525 21:40:32< gfgtdf> iceiceice: hm not teh best idea i think, you can also just rerung teh test un your fork maybe ?
20140525 21:40:32< iceiceice> ok that was it
20140525 21:41:05< gfgtdf> rerun
20140525 21:41:29< iceiceice> its ok its rebuilding now
20140525 21:41:32< iceiceice> on travis
20140525 21:41:49< iceiceice> i had to log out and log back in again... ig uess the gui got confused and i was in a half logged in state
20140525 21:41:49-!- stikonas__ [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20140525 21:42:39< iceiceice> idk my assumption may be wrong, but i assume that when i build on my machine with clang it should give the same output as travis everytime
20140525 21:42:42< iceiceice> for clang
20140525 21:42:51< iceiceice> its never been wrong before
20140525 21:43:13< iceiceice> other than random travis timeouts
20140525 21:45:22-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20140525 21:47:38< gfgtdf> iceiceice: hm about which if teh failed test are ou talking ?
20140525 21:47:40< gfgtdf> od
20140525 21:47:42< gfgtdf> of
20140525 21:48:24< iceiceice> the one you mentioned
20140525 21:48:35< iceiceice> check_victory_always_one
20140525 21:48:36< iceiceice> it didn't timeout for me
20140525 21:48:39< iceiceice> it passed
20140525 21:48:53< iceiceice> the build was a success
20140525 21:48:56< iceiceice> on my machine
20140525 21:54:06< iceiceice> gfgtdf: do you know where [modify_side] is synced?
20140525 21:54:11< iceiceice> or how it comes to be synced?
20140525 21:54:34< gfgtdf> iceiceice: hm modify_side is a normal action i think ?
20140525 21:54:39< iceiceice> im trying to figure out how it works if the host says [modify_side] controller = null
20140525 21:54:56< iceiceice> i know that mattsc currently uses this in ai test scenarios
20140525 21:54:59< iceiceice> or did at some point
20140525 21:55:19< gfgtdf> well actions are usualy executed on all clients simulaniously
20140525 21:55:20< mattsc> what? what?
20140525 21:55:21< iceiceice> because i was looking at the play_turn network handling code, and it looks like clients will ignore such a message
20140525 21:55:27< mattsc> :)
20140525 21:55:35< iceiceice> so i'm trying to figure out if that is mp safe or not
20140525 21:55:45< gfgtdf> iceiceice: hm i dont think modify_side will invoke any network traffic
20140525 21:56:00< iceiceice> so wait if wml changes a side to ai, it will cause out of sync?
20140525 21:57:20< iceiceice> ok that is fine i was just curious
20140525 21:57:31< iceiceice> itloks like that is probably the case
20140525 21:57:39< gfgtdf> iceiceice: hm why shodul it ? teh wml is usualy executed on all clients
20140525 21:57:41< gfgtdf> should
20140525 21:57:54< gfgtdf> so all client will know that that is nt an ai side
20140525 21:57:57< gfgtdf> now
20140525 21:58:17< iceiceice> hmm
20140525 21:58:31< iceiceice> y ok i see
20140525 21:59:27-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [Ping timeout: 245 seconds]
20140525 22:00:19-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20140525 22:03:15< Necrosporus> Can I do something like [ai] [avoid] something but for single unit instead of whole side?
20140525 22:03:34< mattsc> No
20140525 22:05:17< mattsc> Well, yes, but only by writing your own Formula or Lua AI.  Might not be worth the effort …
20140525 22:07:14< iceiceice> !!! it happened again :(
20140525 22:07:51< iceiceice> gfgtdf:
20140525 22:07:54< iceiceice> i think i'm going to commit some debugging output to the test
20140525 22:08:07< iceiceice> like, [wml_message] or smth
20140525 22:08:21< iceiceice> so we can see if it is timing out in wml or before it gets there
20140525 22:11:22-!- sachith500 [~kvirc@112.134.55.85] has quit [Ping timeout: 276 seconds]
20140525 22:11:42-!- travis-ci [~travis-ci@ec2-54-197-185-19.compute-1.amazonaws.com] has joined #wesnoth-dev
20140525 22:11:42< travis-ci> [travis-ci] wesnoth/wesnoth#2792 (master - b2ac9c5 : Chris Beck): The build was broken.
20140525 22:11:42< travis-ci> [travis-ci] Build details : http://travis-ci.org/wesnoth/wesnoth/builds/26007560
20140525 22:11:42-!- travis-ci [~travis-ci@ec2-54-197-185-19.compute-1.amazonaws.com] has left #wesnoth-dev []
20140525 22:21:47< iceiceice> hmm maybe i shoul djust do a clean build here
20140525 22:23:48< gfgtdf> iceiceice: hm when i tested en_turn inside a start event i got teh expected behaviour which is https://github.com/wesnoth/wesnoth/blob/master/src/playsingle_controller.cpp#L657 beeing called.
20140525 22:24:45< iceiceice> i think it might just be travis being stupid
20140525 22:25:32< iceiceice> hmm... althoguh...
20140525 22:26:40< gfgtdf> iceiceice: i thougth abotu the [option]type=dialog and i think its possible. But the problem is form where to get data about the game to the lua content since we cannot use teh function that ar eingame available
20140525 22:27:20< iceiceice> so i can reproduce the timeout now from that commit
20140525 22:27:23< iceiceice> but only when i build in debug mode
20140525 22:28:17< gfgtdf> so you think the check_victory_always_one fail is caused by timeout ?
20140525 22:29:31< iceiceice> i dont know
20140525 22:30:29< iceiceice> i will run with show gui
20140525 22:31:22-!- kex [~kex@89.205.75.19] has joined #wesnoth-dev
20140525 22:31:34< iceiceice> y the scenario isn't ending :(
20140525 22:34:34< iceiceice> i wonder what was wrong with my tests before
20140525 22:34:39< iceiceice> i guess i need to always build clean or something
20140525 22:36:08< iceiceice> hmm theres something really wierd going on
20140525 22:36:43-!- kex [~kex@89.205.75.19] has quit [Ping timeout: 276 seconds]
20140525 22:40:51< gfgtdf> iceiceice: my i tested normal scenario mode (replaced [test] with [scenario]) and i woin immideately
20140525 22:41:13< iceiceice> that's wierd
20140525 22:41:37< gfgtdf> iceiceice: i still had to click the "show objectives" aways though
20140525 22:41:42< gfgtdf> away
20140525 22:42:06< iceiceice> y that happens in test also
20140525 22:42:08< gfgtdf> iceiceice:  so i won after 1 click :)
20140525 22:42:13< iceiceice> i think it is disabled in noninteracive
20140525 22:44:22< gfgtdf> iceiceice: no wait
20140525 22:44:34< iceiceice> i am trying to repeat what you said
20140525 22:44:35< gfgtdf> iceiceice: i still had a wesnoth version from before my commit
20140525 22:44:42< iceiceice> oh
20140525 22:46:48< iceiceice> it doesnt make a difference for me when i load as multipalyer
20140525 22:46:50< iceiceice> instead of test
20140525 22:47:50< iceiceice> ok so what did we learn
20140525 22:47:54< iceiceice> the test is legitimately timing out
20140525 22:47:57< gfgtdf> iceiceice: hm ok i think i found teh reason why i was so confused
20140525 22:48:48< gfgtdf> iceiceice: i mistook check_victory_always_one with check_victory_always_two and wondered why the [end_turn] didint work
20140525 22:49:20< iceiceice> i think i was confused becuase i thought it failing when it was supposed to be looping
20140525 22:49:26< iceiceice> and actually its looping when it was suppsoed to failing
20140525 22:49:32< iceiceice> er, ending
20140525 22:50:26< iceiceice> ok so it hink... we understand now what's happening?
20140525 22:50:30< iceiceice> like you said, 
20140525 22:50:39< iceiceice> it never enters that loop when end turn is true
20140525 22:51:54< gfgtdf> iceiceice:, mattsc: is there an ai that does absolutely nothing ? not even pressing end turn ?
20140525 22:52:09< iceiceice> there is idle mode
20140525 22:52:09< iceiceice> *idle controller
20140525 22:52:15< iceiceice> i made that to get that behavior
20140525 22:52:41< gfgtdf> hm  no i want to test teh ai codepath
20140525 22:52:50< iceiceice> oh
20140525 22:52:57< iceiceice> then i think the answer is no
20140525 22:53:03< gfgtdf> or is there an ai that does nothing but presses end turn ?
20140525 22:53:10< iceiceice> i think idle ai does that
20140525 22:53:36< Necrosporus> mattsc, can you make it possible somehow?
20140525 22:53:55< Necrosporus> at least in 1.13, if it won't make it into 1.12
20140525 22:54:33< iceiceice> hmm so gfgtdf: how should we fix master?
20140525 22:55:12< gfgtdf> iceiceice: i need to do some testing with ai first
20140525 22:55:26< gfgtdf> iceiceice: becaue i dotn want different behaviour ai <-> human controller 
20140525 22:55:28< gfgtdf> don't
20140525 22:55:50< iceiceice> ok
20140525 22:55:59< Necrosporus> iceiceice, do you think it's ok to have separate side for each special unit?
20140525 22:56:11< iceiceice> what does the special unit do?
20140525 22:56:17< gfgtdf> iceiceice: so teh question im investigatin is: did teh test before my commit ran the same events when exceuting ai or human controller
20140525 22:56:44< iceiceice> hmm so i can make duplicates of the tests that have ai controllers
20140525 22:56:58< Necrosporus> iceiceice, guard certain location
20140525 22:57:12< Necrosporus> though ai_special=guardian is not enough
20140525 22:57:13< iceiceice> built into the test_check_victory.cfg i have an argument to set sides 2,3 to an argument
20140525 22:57:17< iceiceice> controler
20140525 22:57:28< gfgtdf> iceiceice: ok do teh following ad an a side turn end event in tha scenario, make side 1 ai aith idle ai
20140525 22:57:30< iceiceice> do you want to check what happens when all are ai or just some are ai?
20140525 22:57:47< gfgtdf> in that scenario onyl side 1 actualy does anything
20140525 22:58:01< iceiceice> ok
20140525 22:58:15< iceiceice> so its like test_end_turn
20140525 22:58:26< iceiceice> but where ai presses the button
20140525 22:58:59< gfgtdf> iceiceice: add a "side turn end" event and remember whether it is fired, in the case of check_victory_always_one and check whether that is consitent when making side 1 a ai controlled with idle_ai
20140525 22:59:36< gfgtdf> iceiceice: my guiess is that this in unconsitent without my commit
20140525 22:59:40< iceiceice> ok
20140525 23:00:11< iceiceice> is it actually important that the controller types are always defate?
20140525 23:00:40< gfgtdf> iceiceice: sry i dont know defate
20140525 23:00:45< iceiceice> *defeated
20140525 23:01:00< gfgtdf> ok i'll look that word up :)
20140525 23:02:23< gfgtdf> hmm idk why i didnt recongize that word
20140525 23:04:09< gfgtdf> iceiceice: i erpsonyl believe that teh scenario end conition shoudl be the same for ai and human, otherwise one migth for exampe diod a side an invoke a differnt victory_condition with that, or casue defeat on one side or not on another side
20140525 23:04:11< gfgtdf> personly
20140525 23:04:21< gfgtdf> driod
20140525 23:04:32< gfgtdf> on another client
20140525 23:04:53< iceiceice> y i agree 
20140525 23:04:55< iceiceice> i thinkt hat's important
20140525 23:05:42< gfgtdf> my proposes solution is to leave my last commit but maybe add a check_victory at the end of do_init_side
20140525 23:05:46< gfgtdf> proposed
20140525 23:05:59< iceiceice> ok
20140525 23:06:07< iceiceice> i need to make sure i have the idle ai syntax right, im almost readyt or un
20140525 23:06:31< gfgtdf>             [ai]
20140525 23:06:33< gfgtdf>                 ai_algorithm=idle_ai
20140525 23:06:34< gfgtdf>             [/ai]
20140525 23:06:36< gfgtdf> did it for me.
20140525 23:06:55-!- grzywacz [~grzywacz@wesnoth/developer/grzywacz] has joined #wesnoth-dev
20140525 23:07:22-!- ancestral [~ancestral@12.23.74.29] has joined #wesnoth-dev
20140525 23:07:37< mattsc> Necrosporus: and keep the rest of the AI behavior.  No, that’d mean essentially rewriting most of the AI.
20140525 23:08:01< mattsc> Necrosporus: have you checked out the Micro AI guardians.  There are a whole bunch of them, maybe one of the will do the job?
20140525 23:08:22< mattsc> http://wiki.wesnoth.org/Micro_AIs#Guardian_Micro_AIs
20140525 23:10:09< iceiceice> y
20140525 23:10:09< iceiceice> ok
20140525 23:10:20< iceiceice> so i am finding that in normal, i do fire side turn end event before vicotry
20140525 23:10:37< iceiceice> and with idle ai i dont
20140525 23:11:14< gfgtdf> iceiceice: :o i gto teh opposite bahaviour 
20140525 23:11:18< gfgtdf> got
20140525 23:11:55< iceiceice> so i think i'm not getting idle ai actually
20140525 23:11:59< iceiceice> because my side one player is moving
20140525 23:12:04< iceiceice> but it still ends turn
20140525 23:12:17< gfgtdf> iceiceice: no ne need im to do nothing
20140525 23:12:39< iceiceice> well either way
20140525 23:12:40< gfgtdf> iceiceice: otherwise his "user action" will casue a check_victory
20140525 23:12:50< iceiceice> in one case "side turn end" results in a chat
20140525 23:12:53< iceiceice> and in the other it doesnt
20140525 23:13:13< gfgtdf> iceiceice: did you chack with or without my comit ?
20140525 23:13:32< gfgtdf> iceiceice: that migth be just becasue you as human didi  move a unit but teh ai did
20140525 23:13:36< gfgtdf> did
20140525 23:13:38< iceiceice> oh
20140525 23:13:39< gfgtdf> didn't
20140525 23:13:39< Necrosporus> mattsc, I guess it will. Do you think this is better than just modify movement_costs?
20140525 23:13:42< iceiceice> i think i'm checking on master
20140525 23:13:53< iceiceice> (b2ac9c5-Clean)
20140525 23:14:03< mattsc> Necrosporus: Yes.  No.  Depends.  ;)
20140525 23:14:20< mattsc> Depends very much on what you want to do.
20140525 23:14:44< mattsc> Also, my general advice to everybody writing UMC content is: do it the way _you_ want to do it.
20140525 23:14:46< gfgtdf> iceiceice: i tihnk the fist move of the leader will end you scenario in that cse
20140525 23:14:55< gfgtdf> or endtuen whatever comes first
20140525 23:14:56< mattsc> Don’t worry about what other people might think.
20140525 23:14:57< gfgtdf> endturn
20140525 23:15:20< Necrosporus> mattsc, make unit avoid terrain where it has poor defense
20140525 23:15:22< iceiceice> y i think its the move of the leader
20140525 23:15:26< iceiceice> ok so it makes sense then
20140525 23:15:31< iceiceice> the you dont get the chat
20140525 23:15:48< iceiceice> heres my scenario
20140525 23:15:53< iceiceice> i'm going to recompile with your commit
20140525 23:16:02< iceiceice> err
20140525 23:16:02< iceiceice> wait do i have your commit or not?
20140525 23:16:19< Necrosporus> Like make AI wait player at water edge instead of trying to cross water and die on opposite edge
20140525 23:17:18< Necrosporus> Is it possible to make AI also rotate units at water edge and use healers like a player would do?
20140525 23:17:18< gfgtdf> iceiceice: yes you do
20140525 23:17:36< gfgtdf> iceiceice: that one that break teh travis test
20140525 23:17:45< iceiceice> i wonder why i'm not getting idle ai
20140525 23:18:05< iceiceice> http://pastebin.com/vCA2KQrP
20140525 23:19:02< iceiceice> oh
20140525 23:19:05< iceiceice> i forgot to paste it to the bottom :X
20140525 23:19:45< iceiceice> ok
20140525 23:19:53< iceiceice> i'm actually getting that the event fires in both cases
20140525 23:19:54< iceiceice> side turn end
20140525 23:20:43< gfgtdf> iceiceice: ye but without my commit that broke the travis tet youl woudl only get it on ai sid ei think
20140525 23:20:58< iceiceice> i see
20140525 23:21:10< iceiceice> do you want me to revert and confirm?
20140525 23:21:13< iceiceice> that would be very easy, lots of ccache now :)
20140525 23:21:30< gfgtdf> iceiceice: hm aiready rtested but you can test again
20140525 23:21:34< gfgtdf> already
20140525 23:21:35< iceiceice> ok its not important
20140525 23:22:00< gfgtdf> iceiceice: shall i commit an untested fi to master ?
20140525 23:22:11< iceiceice> sure
20140525 23:23:02< mattsc> Necrosporus: all of that is possible in principle, but far from easy
20140525 23:23:38< irker881> wesnoth: gfgtdf wesnoth:master 387017d6d2f6 / src/play_controller.cpp: add a check_victory after init side http://git.io/_WhlKg
20140525 23:23:42< iceiceice> Necrosporus: maybe you can have a special lua ai stage
20140525 23:23:44< iceiceice> that runs first
20140525 23:23:53< iceiceice> moves the special unit
20140525 23:23:55< iceiceice> and sets it moves to zero
20140525 23:23:59< iceiceice> and then you have the other stages run?
20140525 23:24:03< Necrosporus> mattsc, it's easy to make AI not move units into water by making water movement cost > their max_moves
20140525 23:24:04< gfgtdf> iceiceice: ^
20140525 23:24:09< mattsc> Necrosporus: check out the bottleneck Micro AI test scenario - that’s probably the closest to what you are describing (but might not be applicable to your scenario anyway)
20140525 23:24:23< iceiceice> y so my view is,
20140525 23:24:28< iceiceice> now that we ahve defeat condition
20140525 23:24:40< iceiceice> it is okay to run check_vicotry as many times as we like
20140525 23:24:42< mattsc> Necrosporus: well, if it’s easy, then just do it  :D
20140525 23:24:57< iceiceice> until we get performance problems 
20140525 23:25:02< Necrosporus> though that's a dirty hack
20140525 23:25:04< gfgtdf> iceiceice: well there is still some case that defeat_condition cannot cover: when all sides are allied
20140525 23:25:17< iceiceice> hnmm
20140525 23:25:33< Necrosporus> mattsc, so what about making [avoid] also recognize unit type or unit ids?
20140525 23:25:39< iceiceice> can they not just declare an additional side that is never defeated and is enemies?
20140525 23:25:46< iceiceice> and has no units?
20140525 23:26:02< iceiceice> thats not so much work
20140525 23:26:03< Necrosporus> like [avoid] terrain=smth ... [filter] id=unit_id [/filter] [/avoid]
20140525 23:26:13< gfgtdf> iceiceice: well he umber of sides are limited
20140525 23:26:15< gfgtdf> number
20140525 23:26:17< gfgtdf> is
20140525 23:26:29< Necrosporus> mattsc, would it be as hard?
20140525 23:26:41< iceiceice> y i guess thats the onyl case that may struggle
20140525 23:26:49< iceiceice> and they can victory when enemies defeated anyways
20140525 23:26:57< iceiceice> as long as there i a human
20140525 23:27:00< gfgtdf> iceiceice: assume someone wants to makean 'alliance' mod where sides can make themself allies but still onyl win when tehy are the onyl one left
20140525 23:27:01< mattsc> Necrosporus: [avoid] is a SLF.  It does that already.
20140525 23:27:05< iceiceice> and code up vectory check themself
20140525 23:27:09< iceiceice> in wml
20140525 23:27:15< Necrosporus> mattsc, but does it have standard unit filter?
20140525 23:27:31< iceiceice> sorry, i meant to say, they can set "victory_when_enemies_defeated=no"
20140525 23:27:35< Necrosporus> I mean what if it would also have SUF in a special tag like [filter]?
20140525 23:27:42< mattsc> Necrosporus: it’s a SLF.  It takes everything any Standard Location Filter takes.  Not more, not less.
20140525 23:27:58< gfgtdf> iceiceice: i think victory_when_enemies_defeated=no is crap in mp, 
20140525 23:28:24< gfgtdf> iceiceice: hmm but maybe it fixes exactly this case hmm
20140525 23:28:41< Necrosporus> mattsc, then it must work with several avoid tags
20140525 23:29:00< Necrosporus> make units of one type avoid one terrain, of other type other terrains and so on
20140525 23:29:04< mattsc> AFAIK, there can be only one.
20140525 23:30:06< iceiceice> the thunderdome!
20140525 23:30:14< gfgtdf> iceiceice:  hm ye i think victory_when_ememied_defeated works in this casr
20140525 23:30:36-!- yann [~dwitch@nan92-1-81-57-214-146.fbx.proxad.net] has joined #wesnoth-dev
20140525 23:33:12< mattsc> I seem to remember a bug report that multiple [avoid] tags don’t work, but I can’t find it right now.
20140525 23:41:50-!- tomreyn [~tomreyn@megaglest/team/tomreyn] has quit [Quit: tomreyn]
20140525 23:41:54-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20140525 23:43:10< iceiceice> gfgtdf: 
20140525 23:43:15< iceiceice> its actually pretty hard to write a unit test for this
20140525 23:43:18< iceiceice> the thing is that
20140525 23:43:25< iceiceice> once end level has occurred and the scenario is ending
20140525 23:43:33< iceiceice> the test is sort of over
20140525 23:43:46< iceiceice> and checking if side end turn will now happen is sort of meaningless
20140525 23:44:01< iceiceice> i guess in principle it could fire a global varible change or something??
20140525 23:44:17< iceiceice> that is persistent, right?
20140525 23:44:27< gfgtdf> iceiceice: hm butwe want to know is whether side turn end is calles before check_victory
20140525 23:44:33< gfgtdf> iceiceice: hm ye it is
20140525 23:44:43< iceiceice> ohh
20140525 23:44:46< iceiceice> so you think that 
20140525 23:44:47< gfgtdf> you can read you persistent dat in your userdata dir
20140525 23:44:53< iceiceice> if the team is both sides always defeated,
20140525 23:45:02< iceiceice> but the side turn end event declares victory
20140525 23:45:05< iceiceice> it should be victory?
20140525 23:45:40< gfgtdf> iceiceice: :/ this is complicated
20140525 23:46:07< iceiceice> ok i'm not going to write a test that tests which one wins
20140525 23:46:13< gfgtdf> iceiceice: i just notices that :/ and :s look the same on my client. Becasue teh client does some changes
20140525 23:46:23< iceiceice> i'm just going to check that the defeat happens
20140525 23:46:28< iceiceice> and that the global variable is indeed set
20140525 23:46:30< iceiceice> in a subsequent test i guess
20140525 23:51:05< Necrosporus> I found another probably-bug
20140525 23:51:40< Necrosporus> Insert this into [unit_type] http://pastebin.com/Qbmmh8aj and try to hover cursor over new trait
20140525 23:52:18< Necrosporus> a space missing
20140525 23:53:25-!- travis-ci [~travis-ci@ec2-54-198-70-55.compute-1.amazonaws.com] has joined #wesnoth-dev
20140525 23:53:25< travis-ci> [travis-ci] wesnoth/wesnoth#2794 (master - 387017d : gfgtdf): The build was fixed.
20140525 23:53:25< travis-ci> [travis-ci] Build details : http://travis-ci.org/wesnoth/wesnoth/builds/26014240
20140525 23:53:25-!- travis-ci [~travis-ci@ec2-54-198-70-55.compute-1.amazonaws.com] has left #wesnoth-dev []
20140525 23:53:31< gfgtdf> iceiceice: ^
20140525 23:56:25< iceiceice> gfgtdf: do you know what "immediate=no" does in set_global_variable?
20140525 23:56:25< iceiceice> does it just mean that it waits for a synced context?
20140525 23:56:29< iceiceice> or smth?
20140525 23:56:46< gfgtdf> iceiceice: no i dont implement it
20140525 23:56:57< gfgtdf> iceiceice: didn't
20140525 23:57:06< gfgtdf> iceiceice:  i always wonders what it does
20140525 23:57:09< iceiceice> ok
20140525 23:59:05-!- Jetrel_ [~Jetrel@c-75-73-180-126.hsd1.mn.comcast.net] has joined #wesnoth-dev
--- Log closed Mon May 26 00:00:45 2014