--- Log opened Sun Mar 02 00:00:24 2014
20140302 00:00:50< Aishiko_laptop> well that was 7 seconds of exciting
20140302 00:02:04< Aishiko_laptop> I opened codeblocks loaded the workspace for BFW and then 7 seconds in it blew up looking for boost... =P
20140302 00:02:47< shadowm> Blew up?
20140302 00:03:31< Aishiko_laptop> errored on me... but "blew up" sounds more exciting
20140302 00:04:03< shadowm> Are you on Windows?
20140302 00:05:40< Aishiko_laptop> linux, I think I might not have the dev files I need
20140302 00:06:53< fabi> Aishiko_laptop: apt-get build-dep wesnoth
20140302 00:07:20-!- cib0 [~cib@p20030067CE0E5A01267703FFFEE75B84.dip0.t-ipconnect.de] has quit [Remote host closed the connection]
20140302 00:07:35< fabi> And then install one of the boost libs which is missing on the build dependencies of the older version.
20140302 00:07:36< shadowm> Yeah, well, that command fabi provides will work on Debian and derivatives.
20140302 00:07:52< Aishiko_laptop> fabi, no such command
20140302 00:08:03< fabi> rpm based distro?
20140302 00:08:26< Aishiko_laptop> CentOS on this machine and Gentoo on my main workhorse
20140302 00:08:32< fabi> Ah yes.
20140302 00:08:59< Aishiko_laptop> I tired yum instead of apt-get but sadly didn't work =P
20140302 00:09:01< shadowm> Oh my, centos. ;(
20140302 00:10:23< Aishiko_laptop> I needed something that would work after a ubuntu once again screwed the pooch on me so I grabbed the Centos disc (it was the first one I found)
20140302 00:11:54< irker924> wesnoth: ln-zookeeper wesnoth:master d05cc8b4c3c0 / data/campaigns/Eastern_Invasion/maps/ (08_Training_the_Ogres.map 11_Captured.map): Revert "Updates to the new EI maps" http://git.io/_DYCSg
20140302 00:12:27< shadowm> Aishiko_laptop: Well, the INSTALL file lists our dependencies. I have no idea what package names would provide them on Cent OS.
20140302 00:12:27< zookeeper> bumbadadabum, ^ i'll deal with re-applying some of the stuff tomorrow
20140302 00:12:33< bumbadadabum> ok
20140302 00:12:35< zookeeper> just did a simple revert for now
20140302 00:12:37< Aishiko_laptop> this laptop doesn't have alot of power. I doubt it would be able to compile wesnoth without taking all day
20140302 00:12:47< bumbadadabum> apologies for the inconvenience
20140302 00:13:17< Aishiko_laptop> shadowm, it actually doesn't appear to have dev files for boost, so I might just go and do them all manually
20140302 00:14:13< shadowm> Oh, building boost is such a chore, but fortunately we don't depend on too many binary boost libraries.
20140302 00:14:17< zookeeper> no problem, but when doing map changes you might want to check the scenario files for "terrain" or something to catch possible breakages.
20140302 00:14:54< zookeeper> bumbadadabum, i take it that you have ended up doing scenario-side updates in some cases when you've updated maps, right?
20140302 00:14:54-!- thunderstruck [~zaibotren@cpc13-sgyl31-2-0-cust696.18-2.cable.virginm.net] has quit [Quit: leaving]
20140302 00:15:06< Aishiko_laptop> shadowm, they had the boost binaries, just no dev files
20140302 00:15:28< bumbadadabum> zookeeper: I did some things in the earlier scenarios
20140302 00:15:49< bumbadadabum> ATM I need to do a full playthrough to iron stuff out
20140302 00:15:58< bumbadadabum> but I can't really due to exams
20140302 00:18:43< Aishiko_laptop> shadowm, ohh wait I found it they have it postfixed with -devel
20140302 00:18:43< zookeeper> okay
20140302 00:19:18< shadowm> Aishiko_laptop: Right.
20140302 00:19:59-!- sachith500 [~kvirc@112.134.125.130] has joined #wesnoth-dev
20140302 00:21:39< sachith500> hello everyone :D
20140302 00:23:44< Aishiko_laptop> wb sachith500 
20140302 00:23:59< sachith500> :D
20140302 00:29:28< shadowm> "Null Stone"...
20140302 00:29:48< shadowm> Well, I can't say it's less creative than "Sceptre of Fire".
20140302 00:31:21< zookeeper> uh-huh
20140302 00:31:29< zookeeper> shadowm, did you replay or skip captured?
20140302 00:31:47< shadowm> zookeeper: I finished it.
20140302 00:32:12< shadowm> As I said before, I used :lua to set the terrains according to the WML's expectations, before triggering the events again.
20140302 00:32:16< zookeeper> oh, right, with that terrai... yeah
20140302 00:32:22< zookeeper> how did it play out?
20140302 00:32:58< shadowm> Well, I guess, mostly because I managed to crush the orcs and trolls quickly enough.
20140302 00:33:23< shadowm> And I didn't encounter any resistance on my way to the western exit.
20140302 00:33:51< zookeeper> right, for some reason i didn't even think of putting any resistance there. i guess there should be (some light one)
20140302 00:34:38< shadowm> I can't really compare it with the previous version because I don't remember the gameplay that well.
20140302 00:36:07< anonymissimus> mordante: doesn't work, looks the same: http://pastebin.com/XvahdUB2
20140302 00:39:25< zookeeper> shadowm, well you don't have to compare to the old one, so if it didn't annoy your or feel stupid or way too easy/hard then i guess i'm happy with it
20140302 00:40:33< anonymissimus> jamit: ping ?
20140302 00:40:50< jamit> anonymissimus: Just around briefly at the moment
20140302 00:41:41< anonymissimus> okay, I could test still atm but then no more
20140302 00:43:45< jamit> I still don't understand the error. I'll be looking at it in the evening my time, so I'd expect you to be asleep. :)
20140302 00:46:34< anonymissimus> okay, I'll read logs
20140302 00:46:45-!- anonymissimus [~chatzilla@HSI-KBW-149-172-228-192.hsi13.kabel-badenwuerttemberg.de] has quit [Quit: ChatZilla 0.9.90.1 [Firefox 11.0/20120312181643]]
20140302 00:48:02-!- trewe [~trewe@124.164.54.77.rev.vodafone.pt] has quit [Quit: quit]
20140302 00:51:12< jamit> anonymissimus: Try changing the line with the error (750) to: inline typename smart_list<Data>::template iterator_base<Value, Reversed> &
20140302 00:53:37< jamit> Adding both "typename" and "template" to that line does not break my compile with gcc, and it might help VC figure out what that line means.
20140302 00:53:37< jamit> (And if anyone using clang wants to make sure this does not break clang, please do so.)
20140302 00:53:38< jamit> anonymissimus: By the way, if this test is successful, you'll probably still get similar errors, just pushed down to line 768 or 769.
20140302 00:55:39< jamit> And I'm off for a few hours.
20140302 00:59:17-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [Ping timeout: 252 seconds]
20140302 01:01:25< Aishiko_laptop> http://wiki.wesnoth.org/CompilingWesnoth#Prerequisites  updated to include CentOS 6.4/Red Hat 6 build options 
20140302 01:02:16< Aishiko_laptop> small thing but it might help the next person
20140302 01:04:50< _8680_> “build system is not yet final -- SCons and CMake are both in contention” — Is that still true?
20140302 01:06:35< Aishiko_laptop> _8680_, I personally like cmake but thats just because I use it more
20140302 01:08:37-!- jamit [~jamit@wesnoth/developer/jamit] has quit [Ping timeout: 264 seconds]
20140302 01:14:01< shadowm> _8680_: We never made a decision, and probably never will do so anyway.
20140302 01:14:45< shadowm> Both are equally maintained and used by mainline developers, so it's unlikely either will ever be dropped.
20140302 01:15:26-!- kex [~kex@89.205.75.19] has joined #wesnoth-dev
20140302 01:17:02< Aishiko_laptop> I just added scons to the list of packages on the CentOS entry so that the end user can make whatever choice they want (scons in CentOS is 2.0x)
20140302 01:17:19< Aishiko_laptop> and now I'm off, to die a horrible death
20140302 01:20:37-!- kex [~kex@89.205.75.19] has quit [Ping timeout: 264 seconds]
20140302 01:21:55< shadowm> Yay, undead ambushing me in the swamps! I can tell I'll enjoy this scenario... not.
20140302 01:27:13-!- wesbot changed the topic of #wesnoth-dev to: release of 1.11.11 maybe on Sunday | string+feature freeze active on master | 212 bugs, 350 feature requests, 28 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
20140302 01:29:37-!- spoffy [~spoffy@152.78.175.8] has quit [Ping timeout: 264 seconds]
20140302 02:19:31-!- sachith500 [~kvirc@112.134.125.130] has quit [Ping timeout: 252 seconds]
20140302 02:22:08-!- sachith500 [~kvirc@112.134.7.80] has joined #wesnoth-dev
20140302 02:24:52< sachith500> hello, is there anyone here who can answer a few questions about gsoc proposals for me? :D
20140302 02:27:16< happygrue> maybe
20140302 02:27:38< sachith500> hey happygrue :D
20140302 02:27:44< happygrue> hello
20140302 02:27:47< sachith500> I was just curious about what experience we should list
20140302 02:27:59< sachith500> I was thinking project relevant experience only maybe?
20140302 02:28:08< sachith500> otherwise it would start looking more like a resume
20140302 02:28:12< sachith500> :P
20140302 02:28:13< happygrue> I agree
20140302 02:28:19< sachith500> all right :D
20140302 02:28:59< sachith500> is there any specific experience you would like mentioned for the data analysis proposal?
20140302 02:29:01< happygrue> Ultimately, we're not looking for the most experienced person, we're looking for someone who looks like they can do a good job on the project
20140302 02:29:10< sachith500> oh all right :D
20140302 02:29:46< sachith500> I will ask someone to review it when i have a reasonable draft
20140302 02:30:00< sachith500> I'll probably have gone overboard with it :D
20140302 02:31:33< happygrue> I don't have anything in particular that I'm looking for, though I would favor a proposal that has some hope of getting something going with the presentation side (step 3) over one that wasn't going to attempt that.
20140302 02:32:29< sachith500> oh right
20140302 02:32:34< sachith500> like my idea for unit stats?
20140302 02:32:56< sachith500> on the units.wesnoth page?
20140302 02:33:01< sachith500> only it would be more accessible
20140302 02:37:12-!- sachith500 [~kvirc@112.134.7.80] has quit [Ping timeout: 240 seconds]
20140302 02:38:12< happygrue> well, more that the first two parts are critical, and the third part is somewhat optional, in that it's not *required*, but more as time permits
20140302 02:38:55< happygrue> ideally a proposal would should a good plan that looks like it could get to at least part of step 3, some presentation or data that is useful and easy to use for UMC authors (and me!) to look at the data
20140302 02:40:09< happygrue> that could mean cross linking it to other places, like units.wesnoth, but I'm not acually sure how hard that is to get working in a manner that is somewhat clean and useful... I just don't know.  :D
20140302 02:40:16-!- sachith500 [~kvirc@112.135.85.50] has joined #wesnoth-dev
20140302 02:40:22< happygrue> but sure, it could be an idea in there.
20140302 02:40:34< sachith500> happygrue, so sorry I dced for a bit
20140302 02:41:36< sachith500> yeah I agree
20140302 02:41:43< sachith500> the first 2 parts are what's paramount :)
20140302 02:42:13< happygrue> yes, I would rather they be solid rather than rushed trying to get the 3rd part "done"
20140302 02:42:37< sachith500> true
20140302 02:42:53< happygrue> it would be great to get something working for the last bit though.  ;)
20140302 02:43:07< sachith500> I think the first part shouldn't be too difficult
20140302 02:43:12< sachith500> step 1 I mean
20140302 02:43:24< sachith500> in terms of implementation
20140302 02:43:59< sachith500> I'll try to have a basic skeletal structure for the plan by the end of today
20140302 02:45:34< sachith500> hehe mediawiki is interesting, I'm enjoying myself editing the proposal
20140302 02:45:54< happygrue> hehe
20140302 02:46:41< sachith500> I added a small brainstorm section for now
20140302 02:46:49< sachith500> with issues I'm currently considering
20140302 02:46:59< happygrue> Seems like a good idea.
20140302 02:47:18< sachith500> I have 2 PR already
20140302 02:47:24< sachith500> should I do some more?
20140302 02:47:36< sachith500> or should I work harder on my proposal?
20140302 02:47:51< sachith500> *merged PR
20140302 02:48:47< sachith500> just added that to the page as well :)
20140302 02:49:03-!- werlley [~werlley@187-41-159-188.user.veloxzone.com.br] has joined #wesnoth-dev
20140302 02:51:15< happygrue> both are good.
20140302 02:51:45< happygrue> Have you been looking at fixing bugs or the easy coding projects on the wiki?
20140302 02:51:47< sachith500> OK :D
20140302 02:51:52< sachith500> fixing bugs
20140302 02:51:56< sachith500> easy ones were outdated
20140302 02:51:59< sachith500> when I looked at them
20140302 02:52:03< happygrue> yeah, could be
20140302 02:52:04< sachith500> don't know how they are now
20140302 02:52:05< sachith500> :)
20140302 02:52:11< sachith500> I started about a month back :P
20140302 02:52:44< sachith500> first bug I started work on was regarding undead, cause I really love them ^^
20140302 02:55:30< happygrue> hehe
20140302 02:55:39< happygrue> they are fun!
20140302 02:55:45< sachith500> indeed :D
20140302 02:57:48< Aishiko_laptop> PR?
20140302 02:59:16< sachith500> pull request
20140302 02:59:19< sachith500> github
20140302 02:59:26< sachith500> :D
20140302 03:00:36< Aishiko_laptop> ahhh I have none as git aggrevates me beyond "git clone xxxx dir"
20140302 03:01:15< Aishiko_laptop> you may thank the linux kernel for that, I got more sent back to me because git didn't format it right then anything else
20140302 03:02:31< sachith500> :D
20140302 03:03:14< sachith500> Aishiko_laptop: have you noticed any other gsoc lurkers around here?
20140302 03:03:19< Aishiko_laptop> interesting error c++: unrecognized option '-Qunused-arguments'
20140302 03:03:19< Aishiko_laptop>  followed by can't open output file it is a directory
20140302 03:03:39-!- kex [~kex@89.205.75.19] has joined #wesnoth-dev
20140302 03:03:55< sachith500> I haven't registered anyone else other than you :D
20140302 03:03:56< _8680_> Aishiko_laptop: Are you using Clang or GCC?
20140302 03:03:57< Aishiko_laptop> sachith500, 2 or 3 have come in and spent 5 to 60 minutes and then left
20140302 03:04:06< sachith500> ah :D
20140302 03:04:15< Aishiko_laptop> _8680_, gcc 4.4.8 I think
20140302 03:04:33< shadowm> Aishiko_laptop: scons or cmake?
20140302 03:04:40< Aishiko_laptop> cmake
20140302 03:04:44< _8680_> 4.4? That’s rather old.
20140302 03:04:48< shadowm> Uhhhhh...
20140302 03:04:57< _8680_> I believe that error would be caused by Clang arguments being used for GCC.
20140302 03:05:24< shadowm>   if(UNIX AND NOT CMAKE_COMPILER_IS_GNUCXX)   # Assume the compiler is the clang compiler.
20140302 03:05:30-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20140302 03:05:42< shadowm> cmake is doing something wrong.
20140302 03:06:03< shadowm> What happens if you explicitly set the compiler to g++ ?
20140302 03:06:06< Aishiko_laptop> it got to 98% then tried to link ../wesnoth and couldn't as it wasn't a fill
20140302 03:06:43< Aishiko_laptop> ok its 4.4.7-4 64bit version
20140302 03:06:58< sachith500> Aishiko_laptop: was it building properly earlier?
20140302 03:08:20< Aishiko_laptop> sachith500, no clue, this install is a week old
20140302 03:08:29-!- kex [~kex@89.205.75.19] has quit [Ping timeout: 252 seconds]
20140302 03:10:53< Aishiko_laptop> I had to pull the drive for recovery and slapped in an old harddrive quickly slapped Centos with the webdev stuff on it, and my VM of windows for the c# class and didn't bother trying out anything else because at the time I pressed for time and I needed something to work for my classes right then and there
20140302 03:12:10-!- iceiceice [~chris@cpe-66-108-20-80.nyc.res.rr.com] has joined #wesnoth-dev
20140302 03:12:14-!- bumbadadabum [~bumbadada@d155109.upc-d.chello.nl] has quit [Read error: Operation timed out]
20140302 03:14:00< sachith500> ahh :\
20140302 03:17:16< Aishiko_laptop> if it does this on my desktop, then I'll be in trouble =)
20140302 03:17:27-!- Ivanovic_ [~ivanovic@x2f3c987.dyn.telefonica.de] has joined #wesnoth-dev
20140302 03:17:27-!- Ivanovic_ [~ivanovic@x2f3c987.dyn.telefonica.de] has quit [Changing host]
20140302 03:17:27-!- Ivanovic_ [~ivanovic@wesnoth/developer/ivanovic] has joined #wesnoth-dev
20140302 03:17:53< Aishiko_laptop> my desktop is where I'll do most of my dev work, its got a bit more power under the chassis 
20140302 03:17:55< shadowm> Aishiko_laptop: Could you pastebin the whole cmake output? http://pastebin.com
20140302 03:18:11< Aishiko_laptop> shadowm, I can post what's left
20140302 03:19:32< irker924> wesnoth: Ignacio R. Morelle wesnoth:master 140bb6f19fe8 / data/campaigns/Sceptre_of_Fire/scenarios/2_Closing_the_Gates.cfg: SoF S2: Redraw terrains when closing the gates so the gate animation works http://git.io/KLbpmw
20140302 03:19:56< Aishiko_laptop> shadowm, http://pastebin.com/EgKFYAz0
20140302 03:20:38-!- Ivanovic [~ivanovic@wesnoth/developer/ivanovic] has quit [Ping timeout: 252 seconds]
20140302 03:20:53< shadowm> Did you make a subdir to run cmake there, or you ran cmake directly on the top dir?
20140302 03:21:27< Aishiko_laptop> its only covering from 76% until 98% where it tries to link to the directory, appearance expecting a file in a subdirectory
20140302 03:21:27-!- Ivanovic_ is now known as Ivanovic
20140302 03:22:48-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Ping timeout: 240 seconds]
20140302 03:23:23< Aishiko_laptop> err that should read expecting a directory
20140302 03:23:30< Aishiko_laptop> and it should have been in a subdirectory
20140302 03:26:29< Aishiko_laptop> I'm willing to bet its something I did or I din't set up the build environment right
20140302 03:28:09-!- tomreyn [~tomreyn@megaglest/team/tomreyn] has quit [Quit: tomreyn]
20140302 03:29:22< shadowm> Aishiko_laptop: That's why I'm asking where you ran cmake.
20140302 03:29:38< shadowm> Although I guess the correct question is where you pointed cmake to, plus that.
20140302 03:31:19< Aishiko_laptop> in a folder called wesnoth in a folder named gsoc while the git code is in the folder called wesnoth in a different location (my home folder ie /home/username/wesnoth)
20140302 03:31:40< Aishiko_laptop> the other should have pointed to /home/username/gsoc/wesnoth
20140302 03:32:30< shadowm> cmake should be run in a subdir of the top source dir. For example, my top source dir is ~/src/wesnoth, so I'd make a subdir called "cmake-build", change into it, and run "cmake .." from there.
20140302 03:33:40< shadowm> If the top source dir is called wesnoth and you run cmake from there, I wouldn't be too surprised if it runs into problems when trying to write a file [path]/wesnoth/../wesnoth.
20140302 03:34:41< Aishiko_laptop> see I told you it was most likely something I did!
20140302 03:35:58< Aishiko_laptop> though I could have sworn you could do out of directory builds with cmake
20140302 03:35:59< shadowm> Also, to convince cmake that you are running gcc so it stops trying to pass it clang arguments, I'd try giving it "g++" as the compiler.
20140302 03:36:52< shadowm> Although, hm.
20140302 03:37:14< shadowm> I don't know, maybe. I don't really know what's going on there.
20140302 03:37:57< shadowm> (Mostly because I don't use cmake.)
20140302 03:38:44< Aishiko_laptop> well I'll try that and see what happens
20140302 03:39:09< iceiceice> quick question about 1.11.11 again:
20140302 03:39:17< shadowm> If the problem persists, I'd try looking at things from cmake's point of view: what is ../wesnoth and what does it contain?
20140302 03:39:34< iceiceice> so i guess it was going to happen today but now its "not today", are we going to do it tomorrow does that mean?
20140302 03:39:37< shadowm> Using the same working directory as cmake as base for .., that is.
20140302 03:39:53< shadowm> iceiceice: The topic says "maybe on Sunday" and it's Sunday now.
20140302 03:40:03< iceiceice> oh sorry for my blindness ><
20140302 03:40:05< fabi> Aishiko_laptop: Maybe it is a good idea to first build wesnoth with scons. If that runs you can see if you get cmake to work. Maybe because you need it for your favorite ide.
20140302 03:40:06< shadowm> Well, Sunday, 00:40 for me.
20140302 03:41:07< fabi> Oh, a release.
20140302 03:41:20< shadowm> I'm sure this won't be a problem with anyone because everyone understands that we are in feature freeze and master should be release-quality at all times, right?
20140302 03:41:55< fabi> "master should be in release-quality at all times" 
20140302 03:42:01< shadowm> Or 1.12 instead of master as soon  as the branch comes into existence, which is rumored to happen right after the 1.11.11 tag.
20140302 03:42:02< fabi> Your wet dream :-)
20140302 03:42:33< shadowm> This should really be a common sense thing, alas.
20140302 03:43:35< shadowm> But after we have a branch, you people really won't have any excuses.
20140302 03:44:08< fabi> Excuses? That sounds if someone did something wrong.
20140302 03:44:41< shadowm> Not so far, it's mostly a cautionary thing.
20140302 03:44:44< iceiceice> i'm just trying to figure out when i should be sprinting to fix bugs, when there will be a release about which we might get some testing feedback etc.
20140302 03:45:20< iceiceice> ofc i am not pushing anything that will break master
20140302 03:45:28< fabi> Well, that single new string is still in the codebase. 
20140302 03:45:38< Aishiko_laptop> shadowm, its still doing the c++: unrecognized option '-Qunused-arguments'  only it now says g++, I'll try scons and see what happens, I'm glad I'm hitting this now instead of later, I've got time to find out what is going on
20140302 03:45:49< iceiceice> what the uppercase -> small caps change?
20140302 03:45:53< shadowm> iceiceice: Suuuure. ;)
20140302 03:46:08< iceiceice> does that even translate?
20140302 03:46:25< shadowm> What caps change?
20140302 03:46:39< iceiceice> there was a string chagne "Save and Abort -> Save and abort"
20140302 03:46:57< iceiceice> we talked on irc, you told me about "sentence case" as i recall
20140302 03:47:03< shadowm> I already did the pofix stuff to keep the translation crowd happy, so you need not worry about it anymore.
20140302 03:47:20< iceiceice> oh, what string change are you talking about fabi?
20140302 03:47:43< fabi> iceiceice: You are talking about a string that changed after the freeze?
20140302 03:47:45< shadowm> His own, the addition of "VJ/JP" to some textdomain.
20140302 03:47:52< iceiceice> oh
20140302 03:48:20< fabi> iceiceice: "Save and Abort -> Save and abort" happened after the freeze?
20140302 03:48:55< iceiceice> yes, shadowm and Espreon complained about improper case in it
20140302 03:49:15< iceiceice> then i committed a change and learned i shouldnt have done that but i guess it has worked out somehow
20140302 03:49:22< fabi> Well, if the freeze is already broken, why do I need to revert my stuff then?
20140302 03:49:27< shadowm> No, your change was okay.
20140302 03:49:45< shadowm> It counts as a typo, which means you just were supposed to add it to pofix -- which I did for you.
20140302 03:49:55< iceiceice> i see, thanks
20140302 03:50:21< shadowm> fabi: Are you familiarized with string freeze policy?
20140302 03:50:32< fabi> It seems not.
20140302 03:51:01< shadowm> fabi: https://mail.gna.org/public/wesnoth-dev/2014-02/msg00107.html -- second paragraph after the bullet/asterisk list.
20140302 03:53:05< shadowm> iceiceice: The only "not okay" thing about it is that we are apparently supposed to cram all the pofix stuff in the same commit; that is, the original source change, the pofix addition, and the pofix pass results.
20140302 03:53:55< iceiceice> i really should understand the whole gettext thing better...
20140302 03:53:58< shadowm> But meh, no-one here would be eligible for a "perfect committer" award, so whatever.
20140302 03:54:01< iceiceice> so it is sensitive to the commit history?
20140302 03:54:09< shadowm> Not even me.
20140302 03:55:49< Aishiko_laptop> shadowm, and that alone makes those us new here feel that our mistakes are not going to end our involvement
20140302 03:56:03< shadowm> iceiceice: Depends on what that is supposed to mean.
20140302 03:56:56< shadowm> I mean, it's always nice when a commit is sufficiently self-contained that it can be reverted and reapplied as many times as you want without leaving any related changes behind, but most of the time that doesn't happen.
20140302 03:57:04< iceiceice> i see 
20140302 03:57:25< iceiceice> i think i didn't get the memo about pofix files anyways, is there a note about this on the wiki that i missed?
20140302 03:58:17< shadowm> iceiceice: Hm, the email I linked above -- to figure out the rest I asked Ivanovic and Espreon directly because I found the documentation n the pofix script itself a tad confusing.
20140302 03:58:38< iceiceice> ok
20140302 03:58:49< shadowm> The script is in utils/pofix.py. It's Python.
20140302 03:59:09< shadowm> Badly indented Python.
20140302 03:59:45< irker924> wesnoth: Ignacio R. Morelle wesnoth:master 62d3783f414d / utils/pofix.py: pofix: I forgot to add a conversion annotation http://git.io/uaPdDQ
20140302 03:59:54< shadowm> ^ See? Not an atomic commit.
20140302 04:00:33< shadowm> It's not a problem because it's an honest mistake, not a habit!
20140302 04:03:07-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20140302 04:07:01< sachith500> hello can someone give me feedback on the format of http://wiki.wesnoth.org/GSoC_sachith500_Proposal
20140302 04:07:17< sachith500> should I just stick to using the questionnaire?
20140302 04:12:55< Aishiko_laptop> I did but thats just me
20140302 04:14:01-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Ping timeout: 264 seconds]
20140302 04:15:11< Aishiko_laptop> shadowm, this time it was successful
20140302 04:15:33< shadowm> Yay.
20140302 04:15:51< shadowm> What was, cmake or scons=
20140302 04:16:06< shadowm> s/=/?/
20140302 04:16:40< Aishiko_laptop> cmake -D CMAKE_CXX_COMPILER=g++ 
20140302 04:16:52< shadowm> Ah, cmake, okay.
20140302 04:17:38< Aishiko_laptop> well there was a directory after it but basically that was it, I just need to see if there is a way to set gcc as the default compiler
20140302 04:18:37< shadowm> Aishiko_laptop: But you said it was still complaining about an unrecognized option. In that case, changing the default compiler to g++ (which was already the case to begin with) won't really make a difference.
20140302 04:18:41< Aishiko_laptop> it installed but still says 1.10.5
20140302 04:19:10< shadowm> Developers generally don't install the game, it's a waste of time.
20140302 04:19:43< shadowm> You can run it right from the source directory, although it's easier to do so with scons than cmake. For cmake builds you have to manually point Wesnoth to the top directory.
20140302 04:19:45< Aishiko_laptop> I figured I had made it might as well see if it was a higher version then the one I had already =P
20140302 04:20:29< shadowm> For example, if you built it under ./cmake-build in the topdir, the wesnoth binary will be in cmake-build, and you'll run it as `./wesnoth ..` from cmake-build.
20140302 04:20:49< Coffee_irc> just want to say another thing is that you can parallerl build with scons with "scons-j <num_processors>"
20140302 04:20:59< Coffee_irc> this is much faster than single threaded compilation
20140302 04:21:10< shadowm> Coffee_irc: With cmake/make you can also do that:  make -j<num>
20140302 04:21:45< Aishiko_laptop> actually it didn't install it I ran it from the build directory and it loaded 1.11.10
20140302 04:21:53< shadowm> You missed the whitespace between 'scons' and '-j'.
20140302 04:22:01< Coffee_irc> yes, oops
20140302 04:22:18< Aishiko_laptop> Coffee_irc, really I think scons might become my default then =)
20140302 04:22:32-!- Gambit [~derek@wesnoth/developer/grickit] has quit [Remote host closed the connection]
20140302 04:23:32< Aishiko_laptop> though it is an interesting change from the tapestry map to a paper map on a wooden table
20140302 04:23:48< shadowm> scons leaves the binaries (but only the binaries) in the top dir, so it's possible to take advantage of automatic data dir detection: http://shadowm.rewound.net/blog/?p=212
20140302 04:24:15< shadowm> Arguably, someone should include .. and ../.. in the automatic detection so it also works for the cmake crowd.
20140302 04:27:01< shadowm> That someone could even be a first-time contributor!
20140302 04:28:37< fabi> shadowm: Hell no. Putting in ../.. in the automatic dir detection is not a good idea.
20140302 04:28:47< Aishiko_laptop> ?
20140302 04:29:00< shadowm> fabi: Why not?
20140302 04:29:35< fabi> It will cause people with several Wesnoth versions in the same directory in trouble.
20140302 04:29:42< fabi> be in trouble
20140302 04:29:49< Aishiko_laptop> ohhh 
20140302 04:29:51< shadowm> Eh... how?
20140302 04:30:18< shadowm> Unless you dropped the Wesnoth data dir _inside_ another Wesnoth data dir, I mean...
20140302 04:30:21< Aishiko_laptop> because it might detect the wrong version's directory? maybe?
20140302 04:30:47< fabi> My binary is in wesnoth-master
20140302 04:32:25< fabi> Well, if you check for the data dir in the same directory first it should work.
20140302 04:32:55< shadowm> Do wesnoth-master/../data/_main.cfg or wesnoth-master/../../data/_main.cfg exist for you?
20140302 04:33:45< Aishiko_laptop> shadowm, quick question, in upgrading versions it created a new directory for everything, would a version upgrade, check for old add-ons compare against server and download those that match?  for example if you had After_the_Storm in 1.10 and its available for 1.12 it would go get it and install it?
20140302 04:34:22< fabi> The later is available.
20140302 04:34:23< shadowm> Aishiko_laptop: No, Wesnoth currently doesn't try to do anything with past versions' preferences or user data directories.
20140302 04:34:54< Aishiko_laptop> I found that out what I'm asking is would that be a untility that might be wanted?
20140302 04:35:21< Aishiko_laptop> or even possible, I guess would be a better question
20140302 04:36:15< shadowm> Hm, yeah, not sure how wanted a general "(try to) migrate preferences/add-ons from previous version" function would be.
20140302 04:37:00< shadowm> Doable? Since the core WML syntax (what is a tag, what is an attribute, and so on) hasn't changed since 0.5.x or so, yeah, it is doable in principle.
20140302 04:37:41< shadowm> The problem is that we do a terrible job at ensuring backwards compatibility between any given stable series.
20140302 04:38:23< shadowm> So it may not necessarily be a matter of reading the old version's preferences file in and writing it back to the new version's config dir. Or it _might_ be, but it's not guaranteed to be the case in the future.
20140302 04:38:32< Aishiko_laptop> well in this case its merely a check old add-ons and compare to current addons and addon server
20140302 04:39:00< Aishiko_laptop> so no trying to convert anything but download the addons that have been updated to work with the new version
20140302 04:39:33< shadowm> Every directory or *.cfg file name in data/add-ons is for all intents and purposes an add-on id, so you have a way to generate a list of add-ons to check from there.
20140302 04:41:42< AI0867> fabi: I wonder what your directory structure looks like then
20140302 04:43:29< shadowm> As a side note: I dream of a Wesnoth where add-ons are add-ons and not a bunch of assorted files/directories that get loaded unconditionally together.
20140302 04:43:53< Aishiko_laptop> so in theory I could take the _info.cfg files in the old directory and use them to compare against the server.
20140302 04:44:08< shadowm> I mean, it would immediately solve this annoying issue I have with having to remove add-ons that define terrains I also define in my campaigns, whenever I want to use the editor, and fabi wouldn't have to lift a finger.
20140302 04:44:41< fabi> ?
20140302 04:45:07< shadowm> All add-ons that are designed to work in the editor are loaded at the same time unconditionally.
20140302 04:45:13< Aishiko_laptop> you could view your add-ons and say load 1 and 2 automatically but not 3, unless I tell you to?
20140302 04:45:30< shadowm> Yeah, something like that.
20140302 04:47:12< shadowm> Aishiko_laptop: _info.cfg files are currently optional and only used: 1) for cosmetic purposes (the add-on name/title is obtained from there) by two or three dialogs; 2) for tracking versions.
20140302 04:47:56< Aishiko_laptop> hmm, it would need an auto-genterated cfg file, a dialog to show installed add-ons, a way to designate which ones to load/auto-load/unload and a way to tell the loading part of the program to not load certain add-ons
20140302 04:48:11< shadowm> The only thing we can really trust about an add-on at this time is its id (directory/file name).
20140302 04:48:32< AI0867> the current preprocessor-based system could probably handle it, but it requires the authors to cooperate
20140302 04:48:42< shadowm> That said, I would like to think of _info.cfg as a first step towards containerized add-ons.
20140302 04:48:55< AI0867> if the add-on author forgets his ifdef guards, it gets included everywhere
20140302 04:49:33< Aishiko_laptop> AI0867, could we have that, added into the a cfg file when its loaded?
20140302 04:49:52< Aishiko_laptop> as in uploaded to the addon server as part of the _info.cfg file?
20140302 04:50:15< AI0867> Aishiko_laptop: current system works for campaigns only
20140302 04:50:25< AI0867> a [campaign] tag has some defines
20140302 04:50:35< shadowm> AI0867: The thing is that people have figured out new ways to tamper with other add-ons ([lua] at runtime), we might eventually _need_ containerized add-ons.
20140302 04:50:47< AI0867> and when the campaign is started, the entire WML tree is reread with those defines defined
20140302 04:51:09< Aishiko_laptop> ohh lovely
20140302 04:51:26 * Aishiko_laptop asks you to note the sarcasm
20140302 04:51:35< shadowm> Aishiko_laptop: Remember someone proposed glorifying SP tampering in the form of "SP modifications"?
20140302 04:51:39< fabi> shadowm: Let's do that first when 1.12 is branched away. It is also a precondition for the "enable real mods" feature request I have already worked on.
20140302 04:51:42< shadowm> Er, not Aishiko_laptop, AI0867 .
20140302 04:51:49< AI0867> however, multiplayer is guarded (as a whole) by MULTIPLAYER and the editor by EDITOR
20140302 04:51:55-!- kex [~kex@89.205.75.19] has joined #wesnoth-dev
20140302 04:52:25-!- happygrue [~happygrue@wesnoth/developer/wintermute] has quit [Ping timeout: 264 seconds]
20140302 04:52:27< shadowm> fabi: Wesnoth 2.0 coming this 2015?
20140302 04:52:52< AI0867> I'm not sure why we would _need_ it though, as installing this stuff is voluntary
20140302 04:52:53< fabi> Well, I don't care much about the version naming. 
20140302 04:53:27< shadowm> AI0867: Oh, people don't know what they install nowadays. That's why antivirus vendors still make a profit.
20140302 04:54:07< shadowm> You could have a SuperCool Fun MP Unit Pack somewhere, injecting WML events even on singleplayer, and people wouldn't know.
20140302 04:54:56< Aishiko_laptop> shadowm, well a stop-gap to the containerized add-ons which is a good idea, is having the addons stored in say /add-ons instead of /data/add-ons and just create a symbolic for each individual add-on and if you don't want it loaded it just deletes the link, until you want to load it
20140302 04:55:02< AI0867> and the point of it would be?
20140302 04:55:14< shadowm> There may also be legitimate reasons for me to not want someone to, say, play AtS with a Deterministic Gameplay SP modification installed.
20140302 04:55:40< AI0867> Aishiko_laptop: ~add-ons doesn't get included as a whole
20140302 04:56:01< shadowm> Okay, perhaps I'm future-proofing too much. My main concern right now is add-on conflicts in the editor.
20140302 04:56:13< AI0867> individual ~add-ons/Foo.cfg and ~add-ons/Foo/_main.cfg get loaded
20140302 04:56:19< AI0867> in separate preprocessors
20140302 04:56:34< AI0867> though the resulting configs get merged
20140302 04:56:45< shadowm> Aishiko_laptop: Well, when I said I removed add-ons, I actually meant moving them around.
20140302 04:56:51-!- kex [~kex@89.205.75.19] has quit [Ping timeout: 252 seconds]
20140302 04:57:25< AI0867> so you could add a 'disabled' field to _info.cfg that stops it from getting loaded at this point
20140302 04:57:27< shadowm> It starts to look a bit messy after a while: http://pastebin.com/8zszCeQY
20140302 04:57:42< AI0867> though other add-ons could still include stuff from it directly
20140302 04:58:17< Aishiko_laptop> shadowm, I figured as much
20140302 04:58:35< AI0867> shadowm: think a 'disable' button would fit in the 1.13 add-on manager?
20140302 04:58:40< Aishiko_laptop> for the moving them around
20140302 04:58:42< shadowm> AI0867: Yeah, it's just that the config manager doesn't know about _info.cfg.
20140302 04:59:15< shadowm> AI0867: Hm, the add-ons manager... well.
20140302 04:59:41< shadowm> It'd require yet another re-engineering since I obviously don't want to connect to add-ons.wesnoth.org every time I want to enable/disable already installed add-ons.
20140302 05:00:04-!- mattsc [~mattsc@154.20.32.245] has quit [Quit: Ciao]
20140302 05:00:54< Aishiko_laptop> this is beginning to sound like another gsoc level project, something will either take a long time to get done or never get done otherwise
20140302 05:02:42< fabi> shadowm: Add-ons and the editor is indeed a problem.
20140302 05:03:32-!- vorobeez [~quassel@85.142.148.12] has joined #wesnoth-dev
20140302 05:05:24-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20140302 05:11:13-!- werlley [~werlley@187-41-159-188.user.veloxzone.com.br] has quit [Ping timeout: 240 seconds]
20140302 05:14:38< Aishiko_laptop> shadowm, do you mean something like http://pastebin.com/Ni53U1eJ
20140302 05:17:24< Aishiko_laptop> or it might be better to have the cfg for each senerio/campaign read only when you're playing it and it goes back to vanilla defaults afterwards until you load the next customized campaign...
20140302 05:20:08< shadowm> Well, the [campaign] node(s) has/have to be read from the add-on, so...
20140302 05:21:02< shadowm> Leveraging the _info.cfg file sounds like the simplest way to do this, though.
20140302 05:21:12< Aishiko_laptop> I aggree
20140302 05:22:13< shadowm> As for the Add-ons Manager, probably start it displaying currently installed add-ons, allow to disable and remove them, and have an option there to connect to the add-ons server for installing or upgrading add-ons.
20140302 05:22:25< Aishiko_laptop> however is there a way to tell the WML to ignore the other .cfgs if it finds a Disable Flag in the info?
20140302 05:23:21< Aishiko_laptop> shadowm, I would include enabling them, if they are locally installed and you don't want to remove them when disabled you should have a way to re-enable them without having to download them again
20140302 05:24:04< shadowm> Well, yes, disabling and enabling add-ons is of course part of the same functionality, separate from removing them.
20140302 05:24:52< vorobeez> hello everybody
20140302 05:25:18< vorobeez> i am watching at /src/team.cpp and in method void team::team_info::read(const config &cfg) use a some type "config". Where can i see code of this type? 
20140302 05:25:44< shadowm> vorobeez: src/config.hpp has the interface to the config class, src/config.cpp has the implementation code.
20140302 05:26:04< Aishiko_laptop> shadowm, OK, just saying I've seen people go ohh disable and remove and suddenly its one option and you have to redownload to enable!
20140302 05:26:13< vorobeez> shadown: thank you
20140302 05:26:30< shadowm> Aishiko_laptop: The config manager atm does this: for every file f in { data/add-ons/*/_main.cfg, data/add-ons/*.cfg }: preprocess-and-parse f
20140302 05:26:51< shadowm> So adding the _info.cfg disable flag check would be trivial.
20140302 05:27:49< Aishiko_laptop> OK
20140302 05:29:04< shadowm> (That's a rather oversimplified pseudocode. In reality it goes more like for every file f in { ... }: do-a-bunch-of-confusing-checks-then-preprocess-and-parse-or-add-to-twml_error-report f)
20140302 05:30:12< shadowm> I remember saying "no, let's keep the legacy data/add-ons/*.cfg structure for single-file add-ons" some time ago, but after the WML error reports improvement projects I actually regret that.
20140302 05:30:23< shadowm> AI0867: ^
20140302 05:30:33< shadowm> *project, singular
20140302 05:30:33< Aishiko_laptop> but adding a check at the begining to see if you should go further is easily done
20140302 05:31:24< shadowm> Aishiko_laptop: Well, the accurate part of my oversimplification is that the caller only has to call this magic preprocess-and-parse pseudofunction on a single path and the rest is done automagically from there.
20140302 05:32:05< shadowm> The single path is, for example, data/add-ons/My_addon/_main.cfg. This is congruent with how WML authors are supposed to structure their add-ons.
20140302 05:33:06< shadowm> Wesnoth literally does nothing more than load that _main.cfg. If the author wants, that _main.cfg may load more files from the add-on (like units/My_Unit_Type.cfg, scenarios/Scenario1.cfg, etc.), but it's up to the _main.cfg to define which files and in what order.
20140302 05:33:23< shadowm> So if you skip the _main.cfg, you skip the whole add-on.
20140302 05:35:32< Aishiko_laptop> ohh that's nice
20140302 05:35:47< Aishiko_laptop> for our purposes at any rate
20140302 05:37:17< shadowm> But then there's the matter that another add-on may manually include stuff from a disabled add-on. Two possible options: 1) Allow that to happen, let the user disable that dependant add-on too if they want; 2) overcomplicate the WML preprocessor's implementation details with extra checks for disabled add-ons and any paths that they might provide.
20140302 05:38:06< shadowm> Having been to the WML preprocessor's implementation site rather recently and having witnessed the insanity that happens within, I'd opt for 1, myself.
20140302 05:39:06< shadowm> The preprocessor is the part of the WML reading pipeline that does things like substituting {file/inclusions/like_this.cfg} when reading WML files.
20140302 05:40:32< Aishiko_laptop> right, at this point I would agree that let it go at first and then check later to see if that extra step is needed
20140302 05:47:43-!- ahi [~ahi@41.34.240.50] has joined #wesnoth-dev
20140302 06:14:35-!- ancestral [~ancestral@72.21.225.66] has joined #wesnoth-dev
20140302 06:35:48< sachith500> so the replays at http://replays.wesnoth.org/
20140302 06:35:56< sachith500> all have difficulty listed
20140302 06:36:04< sachith500> but so far everything i found has only "normal"
20140302 06:40:03-!- kex [~kex@89.205.75.19] has joined #wesnoth-dev
20140302 06:44:10-!- ancestral [~ancestral@72.21.225.66] has quit [Quit: i go nstuf kthxbai]
20140302 06:44:37-!- kex [~kex@89.205.75.19] has quit [Ping timeout: 264 seconds]
20140302 06:44:38-!- Gallaecio_ [~quassel@84.120.219.139.dyn.user.ono.com] has joined #wesnoth-dev
20140302 06:46:25-!- Gallaecio [~quassel@84.120.219.139.dyn.user.ono.com] has quit [Ping timeout: 264 seconds]
20140302 07:10:27< AI0867> shadowm: repoke me after 1.12 is branched off
20140302 07:10:41< AI0867> and I'll get rid of it
20140302 07:12:23< fabi> sachith500: The MP part of the game supports difficult levels since 1.11.something. Most people play the stable 1.10 series, thus there is no difficult different than normal on the archive.
20140302 07:13:22< sachith500> oh all right fabi
20140302 07:13:23< sachith500> thanks
20140302 07:13:42< sachith500> I checked the 1.11 too, should check a few more games maybe
20140302 07:13:53< sachith500> do you know if they're correctly recording the difficulty now
20140302 07:14:16< fabi> sachith500: Only MP Campaigns make use of the difficulty system.
20140302 07:14:24< sachith500> yeah i know
20140302 07:14:28< sachith500> for example in WC
20140302 07:14:34< sachith500> would the nightmare be recorded
20140302 07:14:37< sachith500> as the difficulty?
20140302 07:14:45< sachith500> *would "nightmare"
20140302 07:15:02< fabi> Only if it uses the new mechanism already.
20140302 07:15:19< sachith500> ah so the guy who made it would have to use it too?
20140302 07:15:34< fabi> Sure
20140302 07:16:01< fabi> The designer of every campaign must define the difficult levels manually.
20140302 07:16:05< shadowm> AI0867: But then everybody will hate me.
20140302 07:16:44< fabi> shadowm, AI0867: What are you after?
20140302 07:17:53< sachith500> fabi do you know where exactly the savegamewml is generated for replays?
20140302 07:18:06< sachith500> for multiplayer games?
20140302 07:18:47< fabi> No, sorry. I don't know it.
20140302 07:18:55< shadowm> Hm, does [variation] work correctly with gendered units?
20140302 07:18:56< sachith500> ok :)
20140302 07:19:05< fabi> I would assume it is a game_state thing.
20140302 07:19:15< fabi> Thus somewhere in game_state.cpp maybe.
20140302 07:20:58< sachith500> ok I'll take a look
20140302 07:23:02< shadowm> Maybe combining [variation] inherit=yes with [male] and [female] wasn't my brightest idea ever.
20140302 07:24:15< shadowm> Okay, no, there's something broken here.
20140302 07:24:48< shadowm> It seems as though animations/etc. from the [female] definition for a [variation] that is inherit=yes are being ignored.
20140302 07:27:09-!- stevedes [0e613edb@gateway/web/freenode/ip.14.97.62.219] has joined #wesnoth-dev
20140302 07:28:14< shadowm> Hopefully it isn't that I should be defining the [variation] in the base unit's [female] definition?
20140302 07:28:30< shadowm> This is starting to make my head spin.
20140302 07:28:58< shadowm> It doesn't help that I've seen some of the ugly implementation details and IIRC variations are internally separate unit types and argh.
20140302 07:30:35< shadowm> But no, tying variations to genders rather than the other way around doesn't fix it.
20140302 07:31:48< shadowm> Oh, maybe it's because I'm using the old variation_name instead of variation_id?
20140302 07:32:00< Aishiko_laptop> what exactly is it doing and what should it be doing? as its soundling like its using the male sprites for female sprites
20140302 07:32:01< shadowm> (And Wesnoth doesn't even warn me about it, for some reason.)
20140302 07:32:17< shadowm> Yes, but it's another layer more complicated than that.
20140302 07:33:13< Aishiko_laptop> well I'm willing to learn if you have the time and more importantly the energy to explain it
20140302 07:33:32< shadowm> Unit types may also have general variations and this is the first time I use that feature with a unit that has more than one gender, so I'm not too sure of what I'm doing here.
20140302 07:34:13< shadowm> Both genders and variations involve inheritance of definition bits that aren't overridden, so for the variation's female variation (uh...) there are two layers of inheritance.
20140302 07:36:31< shadowm> It seems like only one layer works, so I only get the base -> gender -> variation inheritance for an instance of the unit variation with gender=female, rather than base -> gender -> variation -> gender.
20140302 07:37:23< shadowm> Wait.
20140302 07:38:15< shadowm> Argh, it looks like it has the labels from the base -> female unit type, but the animations are from the base -> variation (male) unit type?
20140302 07:38:50< Aishiko_laptop> shadowm, I found a minor bug and have fixed it, now I need to figure out how to properly send it up
20140302 07:39:16< shadowm> Is it an existing reported bug?
20140302 07:39:46-!- albince [~albince@117.221.145.23] has joined #wesnoth-dev
20140302 07:40:14< Aishiko_laptop> shadowm, would a minor reworking the flow help?  Have it choose the gender, and apply that variation and then do the other variation on that object once its had the first wave of inheritance done?
20140302 07:40:32< Aishiko_laptop> shadowm, I don't think its been reported
20140302 07:40:43< Aishiko_laptop> I guess that would be the first step
20140302 07:41:01< shadowm> I'm not in control of the inheritance flow since that's all done in the C++.
20140302 07:41:16< Aishiko_laptop> ohhh
20140302 07:41:32< shadowm> And for all intents and purposes, when I'm working on my add-ons, I'm not a mainline developer, so...
20140302 07:44:12< Aishiko_laptop> of course, I thought you were doing something in mainline.  Would it be possible to let the mainline create the unit and then once its been created use the addon to modify it?
20140302 07:45:26< shadowm> Well, by necessity the mainline engine manages unit types defined by add-ons, there's not much of a choice there.
20140302 07:47:12< shadowm> Could I fix up this broken border case at runtime and patch animations on top? ... Maybe. Kind of. It would be hacky and I'd rather fall back to the easiest and most obvious method to accomplish this.
20140302 07:47:25< shadowm> Which is [unit_type] coupled with [base_unit].
20140302 07:48:33< shadowm> Using variations of a unit_type instead of another unit_type has increased benefits for unit filtering operations in WML, though, because you know it's the same unit type with a variation.
20140302 07:49:00< Aishiko_laptop> of course
20140302 07:49:03< shadowm> Rather than a different unit type that's an edited clone of the original.
20140302 07:49:30< Aishiko_laptop> right
20140302 07:51:01< Coffee_irc> is it too late to sneak in a quick bug fix before the release of 1.11.11?
20140302 07:51:11< fabi> Coffee_irc: no
20140302 07:51:17< fabi> Coffee_irc: Not too late.
20140302 07:51:25< Coffee_irc> fabi: cheers
20140302 07:51:30< shadowm> Coffee_irc: Does it break other things?
20140302 07:51:43< Coffee_irc> shadowm: it is a simple fix
20140302 07:51:46< Coffee_irc> one line
20140302 07:52:01< Coffee_irc> the mp create screen defaults always to the top, instead of the last game you played
20140302 07:52:07< shadowm> Because that's really the only criterion to take into account when the topic doesn't say "releasing".
20140302 07:52:57< shadowm> Coffee_irc: The map selection screen? I was going to ask the other day about that and then forgot.
20140302 07:54:55< shadowm> OOOOOOOOO
20140302 07:55:07< irker924> wesnoth: David Mikos wesnoth:master dee9868ac82a / changelog src/multiplayer_create_engine.cpp: fix bug with MP create defaulting to top due to incorrect variable initializatio http://git.io/4lYGrg
20140302 07:55:10< shadowm> I missed an opportunity to add another Copy-to-clipboard button.
20140302 07:55:36< Aishiko_laptop> Coffee_irc, I was wondering the same thing I fixed a minor but and just sent the pull request.
20140302 07:56:02< shadowm> It's now staring me in the face. The :inspect dialog. Although... first it'd be a good idea to have that WML it prints be written through the regular WML writer into a stringstream, since what we have now in there doesn't look like valid WML.
20140302 07:56:23< Coffee_irc> Aishiko_laptop: was it for the mp create screen bug?
20140302 07:56:33< Coffee_irc> I belive that should now be fixed
20140302 07:56:57< Aishiko_laptop> Coffee_irc, no I don't do MP it was to "Removed the image of Delfador while Kaylan is speaking so the right image is shown."
20140302 07:57:22-!- shadowm_desktop is now known as evilshadowm
20140302 07:57:28< Aishiko_laptop> yeah removing one line of code in a lua file isn't what I'd call a code fix.
20140302 07:57:49-!- albince [~albince@117.221.145.23] has quit [Ping timeout: 264 seconds]
20140302 07:58:27< Coffee_irc> Aishiko_laptop: once you do the hard hards and do enough quality pull requests or get accepted into GSoC you get commit access and can push changes
20140302 07:58:28< Aishiko_laptop> wait it didn't fix it
20140302 07:58:51< Coffee_irc> *yards
20140302 07:59:21< Aishiko_laptop> Coffee_irc, I know that is why I sent a pull request (basically I took that to mean, hey look I did a fix can you look at it?)
20140302 08:00:09< Aishiko_laptop> I basically said the wrong thing
20140302 08:01:09< Aishiko_laptop> it did I just had to redo the cache for it to take effect
20140302 08:01:20< Aishiko_laptop> I'm willing to start small
20140302 08:03:06< irker924> wesnoth: David Mikos wesnoth:master c625af6b13ef / changelog: fix spelling error in last commit changelog entry http://git.io/VzCLJw
20140302 08:04:35< Coffee_irc> Aishiko_laptop: that's about the only way to begin to tackle code from a large project like wesnoth
20140302 08:07:11< Coffee_irc> the good thing about wesnoth is that it's been around for a while and is mature enough that most things have already been started by the time you want to look at them
20140302 08:10:39< Coffee_irc> of course there is still plenty to do
20140302 08:11:17< Aishiko_laptop> Coffee_irc, for me its the other way around. if I've got a blank slate as it were, I'm able to better attack the problem,  but figuring out other's code solo is very difficult for me
20140302 08:12:35-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20140302 08:12:58< Coffee_irc> Aishiko_laptop: it is always difficult to figure out other people's code, but here you'll find many of the original authors willing to help you in general terms
20140302 08:13:44< sachith500> coffee_irc, do you know where the savegamewml generation is handled?
20140302 08:13:51< Coffee_irc> boucman, who incidentally just joined the channel, helped me a great deal by explaining why he made the animation system as it is for certain areas
20140302 08:14:24< boucman> morning Coffee_irc :)
20140302 08:14:31< Coffee_irc> afternoon boucman
20140302 08:14:51< irker924> wesnoth: Ignacio R. Morelle wesnoth:master 220f86036ea8 / data/gui/default/window/gamestate_inspector.cfg: gui2/tgamestate_inspector: Reindent http://git.io/2DN-HA
20140302 08:16:36< Aishiko_laptop> morning boucman, I've been wanting to ask you a question or two about one of the coding tasks and it says to talk to you
20140302 08:16:57< boucman> there are still coding tasks mentionning me ? wow :)
20140302 08:17:05< boucman> but sure, ask off :)
20140302 08:18:01< Aishiko_laptop> yes, lol.  This one is the one about doing customized recall costs and it says to talk to you for "the syntax needs to be discussed and refined, but the overall idea is good" and to make sure the whiteboard is updated
20140302 08:18:17< shadowm> Wheee https://dl.dropboxusercontent.com/u/21371130/screenshots/gui2-tooltip-glitch.png
20140302 08:19:03< boucman> ah... well i'm really a retired dev. i'd gladly discuss and explain the code I did, but i'm too far away from what's going on to discuss future syntax
20140302 08:19:29-!- evilshadowm is now known as shadowm_desktop
20140302 08:19:35< boucman> and this task is not related to my code, so I can provide little insight, sorry
20140302 08:19:48< Aishiko_laptop> ohh, wow maybe we should check to see if its been done and left like the last task I tried to do
20140302 08:20:24< Coffee_irc> Aishiko_laptop: boucman did the animation engine and reviewed incoming patches IIRC
20140302 08:20:25< boucman> hehe, yes 
20140302 08:21:04< Aishiko_laptop> which is the section I'm applying to work on in the end (moving from many small images to spritesheets)
20140302 08:21:24< Coffee_irc> boucman is your man for that ;)
20140302 08:21:46< Ivanovic> shadowm: using pofix.py is very simple
20140302 08:21:51< Ivanovic> go to the wesnoth root
20140302 08:22:11< Ivanovic> call: ./pofix.py po/wesnoth*/wesnoth*pot
20140302 08:22:25< Coffee_irc> sachith500: I don't know where that is handled
20140302 08:22:33< sachith500> nvm I think i got it
20140302 08:22:33< sachith500> :D
20140302 08:22:43< Ivanovic> manual action: check how many .bak files were created (so how many files were changes), you can do so by calling a diff between .bak and normal files
20140302 08:23:03< shadowm> Ivanovic: Why not git diff?
20140302 08:23:04< Ivanovic> if only the intended changes happened, call: ./utils/pofix.py po/wesnoth*/*.po
20140302 08:23:15< Coffee_irc> Aishiko_laptop: for 1.11.2 a whole new syntax was developed for series of images between myself and boucman
20140302 08:23:16< shadowm> Or is there a gettext catalogue diff tool I'm currently unaware of?
20140302 08:23:17< Ivanovic> git diff should also work
20140302 08:23:27< Ivanovic> there is no need for such a tool!
20140302 08:23:35< Ivanovic> just use *ANY* diff tool you like
20140302 08:23:35< irker924> wesnoth: Aishiko wesnoth:master c14537cf0083 / data/campaigns/Heir_To_The_Throne/scenarios/02_Blackwater_Port.cfg: Update 02_Blackwater_Port.cfg http://git.io/jviemw
20140302 08:23:37< irker924> wesnoth: Ignacio R. Morelle wesnoth:master 5594aeec32d8 / data/campaigns/Heir_To_The_Throne/scenarios/02_Blackwater_Port.cfg: Merge pull request #114 from Aishiko/master http://git.io/hdouSw
20140302 08:23:38< Aishiko_laptop> right now it seems like a simple fix to do the move to the sheets
20140302 08:23:40< Ivanovic> and check what pofix.py did
20140302 08:23:45< Ivanovic> so do NOT run a pot-update
20140302 08:23:50< Coffee_irc> Aishiko_laptop: the square bracket syntax for animations
20140302 08:23:50< Ivanovic> first apply pofix.py
20140302 08:24:13< Coffee_irc> Aishiko_laptop: you could extend this for spritesheets or create a new syntax
20140302 08:24:17< Aishiko_laptop> ohh wow thats alot
20140302 08:24:19< shadowm> Ivanovic: Yes... You already explained all this to me the other day.
20140302 08:24:44< shadowm> Ivanovic: On the other hand, I'd like to know why my pot-updates (don't ask) are so noisy.
20140302 08:25:07< shadowm> Is there something more to running make pot-update you didn't mention in ReleasingWesnoth?
20140302 08:25:21< shadowm> Aishiko_laptop: Merged your PR.
20140302 08:25:56< Aishiko_laptop> Coffee what I was thinking was just expand it so the locator can include or not include a string of comma seperated numbers and when it gets to the image.cpp it then either pulls the individual small image or ... the image off the sprite sheet, processes as normal, stores in mem as normal and volia!  
20140302 08:26:02< Aishiko_laptop> shadowm, I saw thank you
20140302 08:26:21< Coffee_irc> Aishiko_laptop: the reasoning at the time for the syntax was here: http://forums.wesnoth.org/viewtopic.php?f=10&t=38198
20140302 08:26:22< Ivanovic> no, nothing
20140302 08:26:32< Ivanovic> this all depends on your local system setup and does not matter at all
20140302 08:26:49< Ivanovic> you know, it depends on stuff like when it thinks it should do linebreaks
20140302 08:26:52< Coffee_irc> Aishiko_laptop: yeah, extending the syntax would be simple, but the image loading would need to change I think
20140302 08:26:59< Aishiko_laptop> shadowm, that is alot of floating text on that screenshot
20140302 08:27:06< Coffee_irc> Aishiko_laptop: that's the part that I am not familiar with
20140302 08:27:11< Ivanovic> the only thing you can do to check if a pot-update introduced new strings is to compare the stats of a complete lang
20140302 08:27:18< shadowm> Ivanovic: But yes, git diff and git diff --stat are much easier to use than the thing you suggested to do above when working with a git working tree, you should try them some day.
20140302 08:27:35< Aishiko_laptop> Coffee_irc, I think so too, it would need a function to get the area of the spritesheet and pass that back for processing
20140302 08:27:39-!- albince [~albince@117.221.145.23] has joined #wesnoth-dev
20140302 08:27:47< Ivanovic> shadowm: what matter is that you should check if .bak files are created
20140302 08:27:56< Ivanovic> if they are: make sure only the intended strings are changed
20140302 08:28:09< shadowm> Those would appear in git status.
20140302 08:28:12< Ivanovic> if not: don't bother, nothing changed, your fix/change to pofix.py is useless
20140302 08:28:16-!- kex [~kex@89.205.75.19] has joined #wesnoth-dev
20140302 08:28:25< shadowm> Back to pot-update, somehow it tried to cram strings from other textdomains into wesnoth-ai here.
20140302 08:28:27< Ivanovic> how you achieve this: does not matter
20140302 08:28:34< Aishiko_laptop> boucman, thank you, sorry that your still being referenced, but if you know where I could look to see if they  have done that yet (the custom recall costs) I'd take a look and see if they have or not
20140302 08:28:47< Ivanovic> that is strange, because your tools should work exactly like mine
20140302 08:29:06< irker924> wesnoth: fendrin wesnoth:master 50c3d000361a / src/reports.cpp: Revert a string change. http://git.io/2xWoOA
20140302 08:29:10< Ivanovic> are you sure this is not a case of some newly introduced strings?
20140302 08:29:19< boucman> no, i'd grep the source for "recall" I guess, maybe check the WML syntax reference, which should be mostly reliable
20140302 08:29:28< shadowm> They looked like strings from core textdomains and wesnoth-ai is a WML textdomain.
20140302 08:30:03< Ivanovic> shadowm: gettext should always tell you the references where the "new" strings come from
20140302 08:30:24< shadowm> Also, am I the only one who gets wmlxgettext warnings from several assorted files?
20140302 08:30:46< Aishiko_laptop> thank you I'll do that but, with so little to go on, I'll have to find the maintainer for that subsystem =)  assuming its not been done and they still want to do it
20140302 08:31:05< shadowm> Ivanovic: Ohhh, maybe it's because I'm using make -j<N>.
20140302 08:31:07< Ivanovic> there are a lot of warnings about the creation of the depths of some of the entries
20140302 08:31:43< Aishiko_laptop> Coffee_irc, my time line currently has the first 3 weeks to do the syntax and code changes and then 9 weeks to fix every thing that breaks because of it! lol
20140302 08:32:20< Ivanovic> make update-po4a-man && make update-po4a-manual --jobs=6 && make pot-update --jobs=6 && make mo-update
20140302 08:32:29< Ivanovic> this is what i use from a cmake generated tree
20140302 08:32:48-!- kex [~kex@89.205.75.19] has quit [Ping timeout: 240 seconds]
20140302 08:33:08< Coffee_irc> Aishiko_laptop: if you have enough free time I think that would be about right, comparing to how long it took to implement the square bracket syntax
20140302 08:33:09< Ivanovic> so if you are also using the cmake build system you should be fine
20140302 08:33:19< Ivanovic> if you are relying on other commands: no idea what you do!
20140302 08:33:38< Coffee_irc> Aishiko_laptop: but the bug fixing at the end and integration always seems to take a little longer than expected ;)
20140302 08:33:45< shadowm> I _am_ using cmake, there's no other way to get a Makefile from wesnoth's source tree here.
20140302 08:33:53< Aishiko_laptop> Coffee_irc, really? wow
20140302 08:34:19< Coffee_irc> Aishiko_laptop: I never rewrote any of the image loading though, just changed the syntax
20140302 08:34:21< Aishiko_laptop> and I thought I was way off base thinking there has to be more to it then this!!
20140302 08:35:08< Coffee_irc> it could be harder than it looks... or not... :P
20140302 08:35:18< Ivanovic> shadowm: and are you using the same steps?
20140302 08:35:34< Ivanovic> "-j 6" should be identical to what i do for the two steps
20140302 08:36:06< shadowm> Ivanovic: Surely I don't need to update-po4a-man and update-po4a-manual if I'm only interested in the game textdomains?
20140302 08:36:20< shadowm> That is, everything from data/ and src/.
20140302 08:36:34< Aishiko_laptop> well the getting of the images parts is easy, making it work with the masks and modifiers that come after it (like day and night, or status effects like poisoned or slowed... has me concerned
20140302 08:37:16< shadowm> Ivanovic: This is the diff for po/wesnoth-ai/POTFILES.in I get, which looks suspiciously like the entire set of C++ source files: http://pastebin.com/E1MHKDwJ
20140302 08:37:55< Coffee_irc> Aishiko_laptop: I think you'll find that the masks and such are already implemented, but if you change the way images load you will possibly run into this problem
20140302 08:37:57< shadowm> I just tried without -j, and I still got that.
20140302 08:38:03< Ivanovic> shadowm: strange one, for me that file is empty
20140302 08:38:30< Aishiko_laptop> I mean I can see reusing the existing syntax entirely as is, but the [n~m] being the frames not the image numbers
20140302 08:38:40< shadowm> And wesnoth-ai/wesnoth-ai.pot of course contains a whole barrage of strings from the C++.
20140302 08:39:04< Coffee_irc> Aishiko_laptop: there are many ways to tackle the WML implementation of spritesheets
20140302 08:39:19< shadowm> And this is why I guess I'm never performing a pot-update test run again. <.<
20140302 08:39:29< Aishiko_laptop> Coffee_irc, that is exactly what I'm thinking since its not an image but a frame but if its passed in after the frame is gotten, I wonder if it would just work
20140302 08:39:58< Coffee_irc> Aishiko_laptop: you could use the existing expansion and create a new ~SpriteSheet() image modifier, or create a new variable instead of image=""
20140302 08:40:09< Ivanovic> shadowm: which version of cmake are you using?
20140302 08:40:22< Ivanovic> over here i got cmake version 2.8.12.2
20140302 08:40:34< shadowm> fabi: Re http://git.io/2xWoOA, how does adding a nondescript TODO line help you remember to revert the revert?
20140302 08:40:34< Ivanovic> might be something strange going on in the "collect files to use" function
20140302 08:40:48< shadowm> 2.8.12.1.
20140302 08:41:27< Aishiko_laptop> The goal at the moment is to make the change over invisible to everyone but those that work on the animation subsystem and those that make the sprites
20140302 08:41:44< Aishiko_laptop> including the WML and everything else
20140302 08:43:59< Ivanovic> okay, that should not make a difference
20140302 08:44:16< Coffee_irc> Aishiko_laptop: you'll probably find an existing tool to convert individual pngs into spritesheets
20140302 08:44:17< Aishiko_laptop> but having said that somewhere there is going to need to be a decision made on whether it is a sheet or not.
20140302 08:44:50< Coffee_irc> Aishiko_laptop: if you haven't spoken with Jetrel I would
20140302 08:44:57< Coffee_irc> he is the art director here
20140302 08:45:37< Aishiko_laptop> Coffee_irc, I have not but that's because he's not been mentioned before now to me (or while I was around)  I'll have to do that tomorrow
20140302 08:45:50< Ivanovic> okay, someone added something to utils/pofix.py but did not apply the change to the pot file
20140302 08:45:53< Ivanovic> *not* good!
20140302 08:46:02< Aishiko_laptop> OK technically later today
20140302 08:46:10< Coffee_irc> Aishiko_laptop: it is an idea that has come up many times in the past
20140302 08:46:31< Coffee_irc> Aishiko_laptop: you'll find talks about this on the forums if you do a search
20140302 08:46:58< Coffee_irc> some are some years old though
20140302 08:47:20< Ivanovic> i am not applying pofix on all existing po and pot files in po/
20140302 08:48:12< Coffee_irc> Aishiko_laptop: http://forums.wesnoth.org/viewtopic.php?f=2&t=29311&p=417382&hilit=sprite+sheet+spritesheet#p417469
20140302 08:48:39< Gallaecio_> Ivanovic: do you know how to take this screenshot: http://www.wesnoth.org/manual/dev/images/game-screen-1.11.9.jpg ?
20140302 08:48:56< Ivanovic> Gallaecio_: it is all described in the README.txt
20140302 08:49:04< Gallaecio_> Ivanovic: Thanks!
20140302 08:49:07< Coffee_irc> Aishiko_laptop: nothing happened from that, but it could be a useful read
20140302 08:52:20< Aishiko_laptop> thank you I'm reading some of the forums now
20140302 08:52:34< sachith500> anyone here familiar with multiplayer server communication with the game? specially with regards to replay archive generation?
20140302 08:53:12< Coffee_irc> sachith500: Soliton possibly, although I am not sure?
20140302 08:53:29< sachith500> I thought he was away :D
20140302 08:53:35< sachith500> yeah he would know
20140302 08:55:26< irker924> wesnoth: Nils Kneuper wesnoth:master aa5615066f1b / po/wesnoth-ai/ (58 files): applied utils/pofix.py on all of po/ http://git.io/4qmOAQ
20140302 08:55:28< irker924> wesnoth: Nils Kneuper wesnoth:master e73bd5e5368e / utils/pofix.py: add steps for running pofix.py http://git.io/vF3RIg
20140302 08:56:29< irker924> wesnoth: Nils Kneuper wesnoth:master 5c3a3e46a831 / po/ (5 files in 5 dirs): updated Slovak translation http://git.io/SLARpA
20140302 08:57:13-!- albince [~albince@117.221.145.23] has quit [Ping timeout: 240 seconds]
20140302 09:02:46-!- thunderstruck [~zaibotren@cpc13-sgyl31-2-0-cust696.18-2.cable.virginm.net] has joined #wesnoth-dev
20140302 09:03:08< fabi> shadowm: It was a pure revert, the TODO line was in the code already.
20140302 09:08:34< Ivanovic> lovely...
20140302 09:08:52< Ivanovic> the pofix.py run actually creates 2 fuzzy strings in the AI domain when doing a pot-update
20140302 09:11:38< shadowm> fabi: That's not what I see in the web diff, but okay...
20140302 09:12:31< shadowm> Ivanovic: The pofix line from http://git.io/4qmOAQ is not from me, it was already there.
20140302 09:12:56< shadowm> I think I complained about it the other day, but since nobody gave me a definitive answer I didn't touch it.
20140302 09:13:09< shadowm>     # conversion added in 1.11.0-dev
20140302 09:13:11< shadowm>     ("$unit.type", "$unit.language_name"),
20140302 09:13:39< shadowm> I mean, it's pretty obvious that everything that doesn't say 1.11.10+dev was already there, right?
20140302 09:14:50< Ivanovic> you know your wml, does changing data/ai/micro_ais/scenarios/goto.cfg to show unit.language_name sound like a good idea?
20140302 09:15:10< Ivanovic>  #: data/ai/micro_ais/scenarios/goto.cfg:603
20140302 09:15:19< Ivanovic>  #: data/ai/micro_ais/scenarios/goto.cfg:631
20140302 09:15:29< shadowm> It's in translatable [message] message= attributes, so it's good.
20140302 09:15:47< Ivanovic> so either we alter goto.cfg or remove the replacement
20140302 09:18:11< shadowm> That is, $unit_store.type gives you the untranslatable unit type _id_ (e.g. "Elvish Fighter") instead of the translatable unit type name that $unit_store.language_name gives ( _ "Elvish Fighter"), so of course you want the latter in a translatable message.
20140302 09:18:16< Ivanovic> fabi: several strings from data/campaigns/Legend_of_Wesmere/scenarios/chapter2/07_Elves_Last_Stand.cfg seem to have been removed
20140302 09:18:31< fabi> Ivanovic: Yes, is removal a problem?
20140302 09:18:32< Ivanovic> fabi: is this intentional?
20140302 09:18:41< Ivanovic> i just want to know if this is intentional
20140302 09:18:52< fabi> Yes, it is.
20140302 09:18:56< Ivanovic> removal is always acceptable as long as you don't say shortly after the pot-update " we need the strings back"
20140302 09:18:58< Ivanovic> ;)
20140302 09:20:05< shadowm> Now, for all I care that pofix entry should actually be a wmllint check.
20140302 09:20:29< Ivanovic> shadowm: sounds sane
20140302 09:20:45< Ivanovic> but the pofix.py entry still makes sense to fix the issue of the "Broken" string which i will now fix
20140302 09:21:14-!- shadowm_desktop [ignacio@wesnoth/developer/shadowmaster] has quit [Ping timeout: 240 seconds]
20140302 09:22:03< shadowm> > Author: Ignacio R. Morelle <shadowm2006@gmail.com>
20140302 09:22:10< irker924> wesnoth: Nils Kneuper wesnoth:master 0be71a927687 / data/ai/micro_ais/scenarios/goto.cfg: made string translateable and match pofix.py replacement http://git.io/pDp7Hw
20140302 09:22:23< shadowm> AI0867: Remind me what I need to chnge in my github account configuration to have web merges use the correct email address?
20140302 09:22:57 * shadowm hits himself with the squeaky mallet.
20140302 09:23:13< Ivanovic> shadowm: does the added explaination in pofix.py help you and is it clear enough?
20140302 09:23:21-!- shadowm_desktop [ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20140302 09:24:06< shadowm> "Your Primary GitHub Email will be used for Account related notifications (account changes, billing receipts, etc) as well as any web based GitHub operations, such as edits and merges made via the web." <-- I guess it's this, too bad I don't get more fine-grained control over it...
20140302 09:24:33-!- trademark_ [~trademark@nsg93-8-88-175-59-164.fbx.proxad.net] has joined #wesnoth-dev
20140302 09:24:48< fabi> hi trademark_
20140302 09:26:25< shadowm> Ivanovic: Yes, looks good to me.
20140302 09:27:59< shadowm> Hm, these tasks seem mostly disk-bound to me, I wonder who would need parallelization for pofix.
20140302 09:28:12< irker924> wesnoth: Nils Kneuper wesnoth:master 7bab3cccee85 / / (624 files in 29 dirs): pot-update and regenerated doc files http://git.io/iGbjeQ
20140302 09:28:46< Ivanovic> shadowm: it was useful when there were about 500 replacements in the 1.8 pofix.py file
20140302 09:29:02< shadowm> # find data/campaigns/ -name '*.cfg' -print0 | xargs -0 utils/pofix.py
20140302 09:29:11< shadowm> How relevant is this from the "(old) example usage" section?
20140302 09:29:34< Ivanovic> might be interesting to see if the change was actually completely applied to to cfg files
20140302 09:29:45< Ivanovic> but honestly i think the last time i used this is like 500 years ago
20140302 09:31:43< shadowm> Ah, okay, it iterates over the substitution pairs and for every single one of them iterates over every line in the file... yeah, it's very CPU-bound, all right.
20140302 09:37:05< shadowm> I'd propose swapping the iteration subjects since the file is generally more likely to be longer than the substitution table, but there might be a (causality-based?) reason for the current design I don't see.
20140302 09:38:43-!- mordante [~mordante@wesnoth/developer/mordante] has joined #wesnoth-dev
20140302 09:38:58< Aishiko_laptop>          morning mordante 
20140302 09:38:59< mordante> servus
20140302 09:39:03< mordante> hi Aishiko_laptop 
20140302 09:39:45< Aishiko_laptop> your up early for the weekend!
20140302 09:39:59< sachith500> :P
20140302 09:40:19< mordante> it's not that early anymore ;-) 10:40 local time
20140302 09:40:36< Aishiko_laptop> ohh well its about 6 hours earlier here =P
20140302 09:40:55< Ivanovic> shadowm: no idea
20140302 09:41:37< Ivanovic> but yeah, the assumption that the po files got more lines than pofix.py tends to be correct
20140302 09:44:22< mordante> anonymissimus I have an alternative patch, for when you're around
20140302 09:45:30< shadowm> Who is "res"? (developer)
20140302 09:45:39< mordante> Aishiko_laptop, -Qunused-arguments is a clang warning, are you using ccache?
20140302 09:46:24< Aishiko_laptop> mordante, I was wondering about the const display::tdrawing_layer display::drawing_buffer_key::layer_groups[] = {     right before that there is a FIXME to merge that function with t_drawing_layer.  Was wondering how I'd go about doing that.
20140302 09:46:49< Aishiko_laptop> I'm not sure
20140302 09:49:49< Aishiko_laptop> I'm using the default but telling it to use g++  and I stil get those warnings
20140302 09:51:20-!- albince [~albince@117.221.145.23] has joined #wesnoth-dev
20140302 09:52:45< irker924> wesnoth: Ignacio R. Morelle wesnoth:master 7d445702a088 / data/core/about.cfg: Make a credits entry conform to the "Name (<nicks>)" convention http://git.io/ogrYrQ
20140302 09:52:48< irker924> wesnoth: Ignacio R. Morelle wesnoth:master 7c61120c5613 / data/core/about.cfg: Relocate a misplaced credits entry from patch #1139 http://git.io/5OuaJA
20140302 09:52:51< irker924> wesnoth: Ignacio R. Morelle wesnoth:master c706117fbb97 / data/core/about.cfg: Sort a couple of credits entries by name http://git.io/U9f8VQ
20140302 09:52:54< irker924> wesnoth: Ignacio R. Morelle wesnoth:master c796334fad1d / data/core/about.cfg: Relocate misplaced credits entry from PR #94 http://git.io/anHzGQ
20140302 09:52:57< shadowm> Answer: not a developer.
20140302 09:53:39-!- SpoOkyMagician [~chatzilla@74-136-58-34.dhcp.insightbb.com] has joined #wesnoth-dev
20140302 09:55:08< mordante> Aishiko_laptop, no idea, it was a patch submitted by somebody and obviously never finished
20140302 09:55:24< mordante> are you using CMake or SCons?
20140302 09:57:14-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20140302 09:58:32< shadowm> ../../data/campaigns/Eastern_Invasion/scenarios/11_Captured.cfg, line 960: {MODIFY_UNIT side,canrecruit=1,yes canrecruit no} -> {MODIFY_UNIT side,canrecruit=yes,yes canrecruit no}
20140302 09:58:46< shadowm> Yay. A spurious wmllint warning.
20140302 09:59:27-!- SpoOkyMagician [~chatzilla@74-136-58-34.dhcp.insightbb.com] has quit [Quit: just checking something. later.]
20140302 09:59:29< shadowm> Wait.
20140302 09:59:51< shadowm> Oh right, the filter, the attribute, and the value, okay.
20140302 10:00:51-!- ALourenco [0252303e@gateway/web/freenode/ip.2.82.48.62] has joined #wesnoth-dev
20140302 10:03:02-!- mjs-de [~mjs-de@f048145047.adsl.alicedsl.de] has quit [Remote host closed the connection]
20140302 10:03:08< irker924> wesnoth: Ignacio R. Morelle wesnoth:master 38e6ee682276 / data/campaigns/Eastern_Invasion/scenarios/11_Captured.cfg: EI S11: Swap an attribute value list so as to not trigger a wmllint conversion http://git.io/2CDdxQ
20140302 10:03:59< zookeeper> shadowm, huh, what? ambushing scenarios are the best :>
20140302 10:06:50< Gallaecio_> Why does http://www.wesnoth.org/manual/dev/manual.en.html#_abilities list both “Cures” (3rd) and “Unpoison” (last) with practically identical descriptions? Should I report this as a bug, or is this on purpose?
20140302 10:06:53-!- stevedes [0e613edb@gateway/web/freenode/ip.14.97.62.219] has quit [Ping timeout: 245 seconds]
20140302 10:08:59-!- Yasin [~Yasin@41.129.34.167] has joined #wesnoth-dev
20140302 10:09:35< Gallaecio_> Is Charles Dang <exodia339@gmail.com> here?
20140302 10:09:41< shadowm> That's vultraz.
20140302 10:09:53< Gallaecio_> vultraz: ↑
20140302 10:28:12-!- spoffy [~spoffy@152.78.175.8] has joined #wesnoth-dev
20140302 10:45:13< shadowm> Okay...
20140302 10:45:51< shadowm> I don't think gender variations of a unit type's variation have the means to be constructed from the unit type variation as opposed to the unit type.
20140302 10:46:17< shadowm> Now that I'm reading the C++, that is.
20140302 10:47:12-!- Octalot [~noct@31.185.149.167] has joined #wesnoth-dev
20140302 10:47:48< shadowm> I expect a similar issue with a unit type variation's variation as well... assuming that is even allowed.
20140302 10:49:35< shadowm> I don't even see a way to tell during the variation or gender variation's construction that it's not the base unit type since the variation's unit_type id is the unit_type id.
20140302 10:50:49-!- albince [~albince@117.221.145.23] has quit [Ping timeout: 240 seconds]
20140302 11:02:44-!- Matei [~razvan@p16.eregie.pub.ro] has joined #wesnoth-dev
20140302 11:05:38< Yasin> Good morning. I have a question regarding building. I get 22 errors, starting with a syntax errors(missing semi-colons). All are in smart_list.hpp A warning before the errors says: dependent name is not a type.
20140302 11:06:19< Ivanovic> Yasin: which operating system, which compiler, which build system and which boost version are you using?
20140302 11:07:46< Yasin> Windows, visual studio 2008, and using the dependency package from the compiling on windows page.
20140302 11:08:02< Ivanovic> ooookay, no idea how to fix this
20140302 11:08:18< Ivanovic> that is: the number of people using windows to compile is freaking low
20140302 11:08:37< Ivanovic> my guess this is compiler and/or dependency related
20140302 11:09:36-!- Aishiko_laptop [~unknown@cpe-065-191-176-226.nc.res.rr.com] has quit [Ping timeout: 240 seconds]
20140302 11:10:09< Ivanovic> anyone around who is using windows to compile?
20140302 11:10:10< Yasin> Ok. There is still another method(Crab's) I can try.
20140302 11:10:53< Ivanovic> i know that our windows builds are created with a linux crosscompiler
20140302 11:11:08< Ivanovic> so it *might* be that building under windows is currently broken, but i don't know for sure
20140302 11:11:22< loonycyborg> Ivanovic: Only 'daily' builds
20140302 11:11:45< Ivanovic> loonycyborg: but you use mingw to build the windows package, right?
20140302 11:11:48< loonycyborg> I build releases with mingw on windows
20140302 11:11:56< loonycyborg> windows XP
20140302 11:12:05< loonycyborg> I multiboot to it
20140302 11:13:24< loonycyborg> I think the easiest way of building on windows currently is codeblocks project
20140302 11:13:35< loonycyborg> or msvc
20140302 11:14:37< loonycyborg> ask shadowm, I think he has a setup with easiest way in his windows vms.
20140302 11:15:50< Yasin> msvc9?
20140302 11:15:56< fabi> Ivanovic: So you release today, branching away 1.12?
20140302 11:17:46< loonycyborg> I don't know details of codeblocks and msvc. My own setup is using scons+mingw and might be hard to reproduce without knowing what you're doing.
20140302 11:19:45-!- Gallaecio_ is now known as Gallaecio
20140302 11:21:45-!- bumbadadabum [~bumbadada@d155109.upc-d.chello.nl] has joined #wesnoth-dev
20140302 11:22:07-!- grzywacz [~grzywacz@wesnoth/developer/grzywacz] has joined #wesnoth-dev
20140302 11:23:02< Ivanovic> fabi: some time later today
20140302 11:23:06< Ivanovic> like in about 4h
20140302 11:29:37< fabi> Ivanovic: Hurry up, before Russia starts the next World War.
20140302 11:30:40< Ivanovic> fabi: that was planned for tomorrow, right?
20140302 11:30:45< Ivanovic> like when we announce they start...
20140302 11:30:46< Ivanovic> ;)
20140302 11:31:28< shadowm> Yasin: projectfiles/CodeBlocks/README.txt in the Wesnoth repo has up-to-date instructions and files for building Wesnoth using CodeBlocks. If you want/need to use MSVC++ 9 you might want to take a look at http://wiki.wesnoth.org/CompilingWesnothOnWindows , but I haven't tried that myself.
20140302 11:35:34< Yasin> shadowm, that's the page I'm following. At the bottom it says the last time this method was successfully followed was more than two years ago though.
20140302 11:35:45< Yasin> I'll try the codeblocks method if I fail with msvc.
20140302 11:36:27< shadowm> I think I read something about builds with MSVC++ having been broken recently, but no idea if anyone has figured out a solution yet.
20140302 11:37:27< shadowm> For CB we recommend using gcc, which tends to work since most of the people using Linux here also use gcc.
20140302 11:39:13-!- Matei [~razvan@p16.eregie.pub.ro] has quit [Ping timeout: 264 seconds]
20140302 11:39:15< Yasin> There were about four errors where a typename was missing in smart_list.hpp. Now there are only two errors: LINK : fatal error LNK1104: cannot open file 'VCOMPD.lib'
20140302 11:39:24< Yasin> and Cannot open include file: 'vld.h':
20140302 11:43:32< mordante> Yasin, are you using msvc?
20140302 11:43:57< Yasin> Yes.
20140302 11:44:59< mordante> Yasin, please test with this patch http://paste.debian.net/84868/ and please pastebin the remaining errors 
20140302 11:45:24-!- prkc [~negusnyul@catv-188-142-184-108.catv.broadband.hu] has joined #wesnoth-dev
20140302 11:45:31-!- prkc [~negusnyul@catv-188-142-184-108.catv.broadband.hu] has left #wesnoth-dev []
20140302 11:46:20< Yasin> I already fixed that error.
20140302 11:46:40< Yasin> The two remaining errors are the ones I posted above.
20140302 11:50:50< mordante> Yasin, have you define HAVE_VISUAL_LEAK_DETECTOR for your compiler?
20140302 11:50:58< mordante> defined*
20140302 11:52:11-!- tomreyn [~tomreyn@megaglest/team/tomreyn] has joined #wesnoth-dev
20140302 11:53:07< Yasin> I see what you mean. I haven't. I just install it from here right? https://vld.codeplex.com/releases/view/82311
20140302 11:55:20< mordante> I'm not sure, but it looks like it
20140302 11:55:43< mordante> I still expect your compiler command contains /DHAVE_VISUAL_LEAK_DETECTOR, which causes the issue
20140302 11:56:24-!- spoffy [~spoffy@152.78.175.8] has quit [Ping timeout: 240 seconds]
20140302 11:57:11< ALourenco> Hi
20140302 11:57:20< ALourenco> I already applied for the SpriteSheet project
20140302 11:57:22< ALourenco> http://wiki.wesnoth.org/GSoCSpriteSheetsALourenco
20140302 11:57:41< ALourenco> here's the lik, I don't know if you need it to add to the list...
20140302 11:57:43< Yasin> mordante, where can I find that?
20140302 12:01:27< mordante> Yasin, in the settings under compiler there is some field called output or command (don't have access to MSVC here)
20140302 12:01:42< shadowm> ALourenco: Pages created from the source of http://wiki.wesnoth.org/SoC2014_Template_of_Student_page will be automatically listed in SummerOfCodeIdeas -- yours is lacking the relevant category inclusions.
20140302 12:03:24< mordante> ALourenco, also please avoid the horizontal scrollbars in the wiki page, it makes reading annoying
20140302 12:04:19< mordante> ALourenco, also best use »foo at bar dot org« for your e-mail address
20140302 12:04:19-!- _8680_ [~8680@2002:4404:712c:0:76de:2bff:fed4:2766] has quit [Ping timeout: 245 seconds]
20140302 12:04:42-!- _8680_ [~8680@2002:4404:712c:0:76de:2bff:fed4:2766] has joined #wesnoth-dev
20140302 12:04:42-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20140302 12:09:00< Yasin> I think I got rid of that one. I was building with Deubg_with_VLD. Rebuilding.
20140302 12:09:02< vultraz> Gallaecio: around?
20140302 12:09:13< Gallaecio> vultraz: yes!
20140302 12:10:47< mordante> ah yeah that probably was the issue, when it works could you pastebin your fix, then I'll commit it
20140302 12:10:48< shadowm> FYI: I have changed the markup and text in http://wiki.wesnoth.org/SoC2014_Template_of_Student_page a bit, mostly so it looks more proper.
20140302 12:11:22< shadowm> I won't edit existing student pages, though, that doesn't seem proper (beyond the edit I did yesterday for lipkab's page, which was breaking everything).
20140302 12:12:54< shadowm> And pages with both the previous and current format can coexist from a technical standpoint.
20140302 12:13:27< shadowm> Not that there is much format to speak of beyond the existence of a Description anchor.
20140302 12:14:59< ALourenco> ok, i'll take a look to that. The horizontal bars weren't on purpose. Sorry about that
20140302 12:15:14-!- ALourenco [0252303e@gateway/web/freenode/ip.2.82.48.62] has quit [Quit: Page closed]
20140302 12:15:15-!- Matei [~razvan@82.76.24.247] has joined #wesnoth-dev
20140302 12:16:34< vultraz> Gallaecio: so it seems both definitions should be clarified more. Unpoison only unpoisons, and will not heal, HOWEVER it *prevents* additional healing on that unit this turn. Cures will either heal you, or cure you of poison if you have it that turn (the Heals abilities only stop poison from taking effect that turn)
20140302 12:17:14-!- _8680_ [~8680@2002:4404:712c:0:76de:2bff:fed4:2766] has quit [Ping timeout: 245 seconds]
20140302 12:17:38< vultraz> I was copying the Unpoison text from the definition given in macros/abilities.cfg , so that might warrant clarification as well
20140302 12:18:20-!- _8680_ [~8680@2002:4404:712c:0:76de:2bff:fed4:2766] has joined #wesnoth-dev
20140302 12:18:56< Gallaecio> What is the trait for unpoison? It does not seem to be in the «wesnoth-help» translation domain. And your definition of unpoison is the same as the definition of Curation in wesnoth-help, so I don’t really see the difference :S
20140302 12:19:46< shadowm> Ability, not trait.
20140302 12:20:18< vultraz> Internally, Cures is Unpoison + Hea8
20140302 12:20:22< shadowm> For all intents, Cures = Heals +8 + Unpoison
20140302 12:20:24< vultraz> Heal8*
20140302 12:20:56< shadowm> Meaning Unpoison is not necessarily paired with a healing ability, and it doesn't have to be Heals +8.
20140302 12:21:04< vultraz> ^ 
20140302 12:21:27< vultraz> That should be clarified in the description but...unfortunately...string freeze :/
20140302 12:21:57< shadowm> The Dwarvish Annalist and Loremaster in THoT are examples of a unit with Unpoison and Unpoison + Heals +4, respectively.
20140302 12:23:10< Gallaecio> The game help only shows “cures”.
20140302 12:23:24< Gallaecio> In the list of abilities, that is.
20140302 12:23:27< shadowm> The game help by default only shows abilities for units you have encountered.
20140302 12:23:55< Gallaecio> And heals+8 has unpoisoning implicit.
20140302 12:24:26< shadowm> It may only show abilities for units it knows it exists, in any case, and there are no core units that have the ability, so you'd need to load THoT or an add-on that uses the ability.
20140302 12:24:37< Gallaecio> “cures” does not heal-
20140302 12:24:46< shadowm> No, heals +8 does not have implicit unpoisoning.
20140302 12:25:02< Gallaecio> It does according to the help :S
20140302 12:25:04< shadowm> Cures, however, is accompanied by heals +8.
20140302 12:25:12< Gallaecio> oO
20140302 12:25:23< shadowm> Let me put it in WML terms:
20140302 12:25:31< vultraz> "A poisoned unit cannot be cured of its poison by a healer, and must seek the care of a village or a unit that can cure.""
20140302 12:25:39< vultraz> That's what heals4 and 8 say
20140302 12:25:49< shadowm> heals+4: ABILITY_HEALS; heals+8: ABILITY_EXTRA_HEAL; unpoison (really "curing"): ABILITY_UNPOISON
20140302 12:26:12< shadowm> cures (really heals+8,cures): ABILITY_CURES = ABILITY_EXTRA_HEAL + ABILITY_UNPOISON
20140302 12:26:41< shadowm> (Furthermore, heals+4 and heals+8 are both "healing" internally).
20140302 12:27:06< shadowm> If you aren't confused then that means I didn't accomplish my goal.
20140302 12:27:21< vultraz> The ability description for Unpoison perhaps isn't clear enough, and she be clarified to say it does not include healing on its own
20140302 12:27:35< vultraz> As does healing clarify it does not cure poison
20140302 12:27:41< vultraz> the ability that does both is Cures
20140302 12:27:48< vultraz> That's what we're saying
20140302 12:28:32< shadowm> This mess is really a direct result of the ability reimplementation in 1.11.x. Originally, all abilities were hardcoded by id in the C++, and we had a single atomic 'curing' ability that did both the healing and unpoisoning parts.
20140302 12:28:39< shadowm> Er, 1.1.x.
20140302 12:28:45< shadowm> Muscle memory.
20140302 12:29:48< shadowm> Then abilities became modular and cures became confusing. Later someone found a use case for unpoisoning without healing and it became even more confusing.
20140302 12:30:39< Gallaecio> I did not know «cures» could heal. If that is so, its current definition (A unit which can cure an ally of poison, although the ally will receive no additional healing on the turn it is cured of the poison.) is incomplete.
20140302 12:31:02< Gallaecio> Ivanovic: is this important enough to skip the string freeze?
20140302 12:32:19< shadowm> Now, why does the manual explain campaign or UMC-specific abilities anyway?
20140302 12:32:50< vultraz> shadowm: I just added all the abilities that were in macros/abilities.cfg
20140302 12:32:53< shadowm> I mean, as I said unpoison is only used in THoT. It may be a mainline campaign, but it's still a single campaign.
20140302 12:33:10< shadowm> Oh wait.
20140302 12:33:15< vultraz> ...
20140302 12:33:31< shadowm> Well, Khalifate has a heals+8 unit with no cures.
20140302 12:33:42< shadowm> But no unpoisoning unit, so my argument still applies.
20140302 12:34:06< Gallaecio> There is also the issue that «unpoison» does not appear in the F1 help (checked in the translation file).
20140302 12:34:25< vultraz> [23:23:16]	shadowm	The game help by default only shows abilities for units you have encountered.
20140302 12:34:27< shadowm> Because it's called cures.
20140302 12:34:55< shadowm> ABILITY_UNPOISON defines an ability whose internal id is "curing", and is named "cures".
20140302 12:35:01< Gallaecio> Isn’t cures the one that heals as well?
20140302 12:35:16< shadowm> ... :)
20140302 12:35:29< vultraz> It's...confusing :P
20140302 12:35:32< shadowm> Yes and no, I already explained this above.
20140302 12:35:41< Gallaecio> Oh, wait…
20140302 12:36:35< vultraz> Gallaecio: so what do you intend to do
20140302 12:36:37< Gallaecio> So lets say a unit has abilities «heal+4» and «cures», and another unit has only «cures». They are called «curing» and «unpoisoning» respectively in the help?!
20140302 12:36:43< shadowm> Even though cures and heals +8 are usually found coupled together, the help system should still display them separately.
20140302 12:36:48-!- sachith500|2 [~kvirc@112.135.85.77] has joined #wesnoth-dev
20140302 12:36:56< shadowm> Gallaecio: Internal ids are internal ids, you don't get to see them as a user.
20140302 12:37:15< shadowm> But to answer your question, both are "curing".
20140302 12:37:34< shadowm> Their ids, that is. The name you see is "cures".
20140302 12:38:02< shadowm> So let me put it like this: when you check the White Mage's description, you should see separate links for heals +8 and cures.
20140302 12:38:26< Gallaecio> vultraz: first I intend to understand why there are 3 habilities (heal+4, heal+8 and cures) but 3 unmatching names for them in the manual: Cures, Heal (heal+4, heal+8), Unpoison.
20140302 12:38:32< shadowm> If the manual refers to "cures" as the whole deal (heals +8 plus cures), then it's doing it wrong.
20140302 12:39:14< Gallaecio> shadowm: the manual refers to both “Cures” and “Unpoison” as the same thing. That is the issue.
20140302 12:39:35< shadowm> It should only refer to cures.
20140302 12:40:25-!- sachith500 [~kvirc@112.135.85.50] has quit [Ping timeout: 264 seconds]
20140302 12:40:32< shadowm> Healing (heal+4, heal+8) should then be explained separately from cures because they _are_ separate, from the player's point of view.
20140302 12:40:40< vultraz> shadowm: so I should not have added an entry about Unpoison?
20140302 12:41:06< shadowm> I don't know what you did, vultraz, I haven't read the Wesnoth manual in a decade.
20140302 12:41:24< shadowm> Decade minus one year and a half.
20140302 12:41:43< vultraz> I added Ability entries for Feeding and Unpoison
20140302 12:41:55< vultraz> Why Feeding wasn't there already I have no idea
20140302 12:42:16< shadowm> vultraz: You are a WML coder. To you and me, Unpoison and Cures are things. To everyone else, Unpoison does not exist and Cures is a single thing and it's separate from heals +8.
20140302 12:42:38< vultraz> Except in THoT
20140302 12:42:44< shadowm> Even in THoT.
20140302 12:43:11< shadowm> Start THoT, use :debug, create an Annalist and a Loremaster and tell me what you see.
20140302 12:43:46< shadowm> (And at this point I should apologize for not remembering earlier that the WML is not the user's point of view.)
20140302 12:44:14< shadowm> (And actually, the WML macros in question are not Wesnoth's point of view either.)
20140302 12:50:19< mordante> Yasin, any luck with compiling?
20140302 12:56:25< Yasin> No. Getting cannot open file 'VCOMPD.lib' fatal error. I looked it up and found it's supposed to be part of msvc but not the express edition, which I have.
20140302 12:57:18< mordante> Yasin, look under your linker settings, there you can probably find the name of that library and remove it
20140302 12:57:27< Yasin> And codeblocks tells me it cannot find libraries, even though they exist, and I did add the search path to both linker and compiler.
20140302 12:57:46< shadowm> Yasin: Are you on 64-bit Windows?
20140302 12:58:04< Yasin> No 32.
20140302 12:58:17< Yasin> It's not in the linker included libs either.
20140302 12:58:28< Yasin> I added it to the ignore libs but no difference.
20140302 12:58:30< shadowm> Hm, okay, then what are the actual errors?
20140302 12:59:14< Yasin> In msvc?
20140302 12:59:23< shadowm> No, codeblocks.
20140302 12:59:40< Yasin> F:\C++\Wesnoth\wesnoth\src\lua\lmathlib.cpp|9|fatal error: boost/math/constants/constants.hpp: No such file or directory|
20140302 12:59:40< Yasin> ||=== Build finished: 1 errors, 0 warnings (0 minutes, 1 seconds) ===|
20140302 12:59:40< Yasin> ||=== Build finished: 1 errors, 0 warnings (0 minutes, 1 seconds) ===|
20140302 12:59:58< Yasin> Sorry, meant to copy only the error.
20140302 13:00:58< shadowm> Unless the zip file download was corrupted and you only got the partial contents... that sounds like the paths you configured aren't quite right.
20140302 13:01:45< Yasin> Yeah. I just noticed, I switched the compiler and linker paths.
20140302 13:03:12-!- irker924 [~irker@ai0867.net] has quit [Quit: transmission timeout]
20140302 13:04:08< Yasin> Compiling now.
20140302 13:04:33< mordante> Yasin, but the compilation errors in smart_list are gone now? could you post your patch, then I'll commit it
20140302 13:05:26< Yasin> Yes, they are gone.
20140302 13:06:17-!- kex [~kex@89.205.75.19] has joined #wesnoth-dev
20140302 13:10:54-!- kex [~kex@89.205.75.19] has quit [Ping timeout: 252 seconds]
20140302 13:16:35-!- happygrue [~happygrue@wesnoth/developer/wintermute] has joined #wesnoth-dev
20140302 13:18:10< shadowm> happygrue: Feel free to draft a News post about http://forums.wesnoth.org/viewtopic.php?p=567480#p567480 ! I'm too lazy and not particularly enthusiastic about this kind of thing!
20140302 13:18:45-!- Matei [~razvan@82.76.24.247] has quit [Remote host closed the connection]
20140302 13:22:43-!- Velensk [~Velensk@cpe-75-187-86-211.woh.res.rr.com] has joined #wesnoth-dev
20140302 13:23:01< happygrue> Okay, I'll see if inspiration strikes me. ;)
20140302 13:23:04-!- spoffy [~spoffy@152.78.175.8] has joined #wesnoth-dev
20140302 13:23:15< shadowm> (Drafts go in Moderators'.)
20140302 13:24:15< happygrue> ah, right.  THere isn't  "draft", just post it there and mention it's a draft, right?
20140302 13:24:23< shadowm> Yes.
20140302 13:24:32< happygrue> busy for a bit, but I should find a few minutes later on.
20140302 13:25:59< shadowm> I noticed that they started doing this only like three months ago and I haven't been able to measure their userbase yet, seems tiny.
20140302 13:26:24< shadowm> OTOH that may well be due to lack of proper promotion.
20140302 13:27:11-!- wesbot changed the topic of #wesnoth-dev to: release of 1.11.11 maybe on Sunday | string+feature freeze active on master | 213 bugs, 350 feature requests, 28 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
20140302 13:27:25< shadowm> So I'm not sure, perhaps a tweet by @Wesnoth would be less work and work better.
20140302 13:27:47< shadowm> @Wesnoth has 830 followers, but who knows how many of those are bots.
20140302 13:27:52< happygrue> Yes it seems quite small, though also worth mentioning - good news is good news
20140302 13:28:14< happygrue> shadowm: sure, since it's in the forums already, a tweet seems like a fine offical response
20140302 13:28:18< shadowm> Or how many of those were obtained from @frogatto.
20140302 13:28:22< happygrue> hehe
20140302 13:29:23< happygrue> that reminds me I've been meaning to download the latest version of Frogatto.
20140302 13:29:23-!- albince [~albince@117.221.145.23] has joined #wesnoth-dev
20140302 13:29:28< Yasin> mordante where should I submit the patch?
20140302 13:33:06-!- Gallaecio [~quassel@84.120.219.139.dyn.user.ono.com] has quit [Read error: Connection reset by peer]
20140302 13:33:09< mordante> Yasin, http://paste.debian.net/ will do
20140302 13:33:27< shadowm> happygrue: Done. Less work for everyone \o/
20140302 13:34:25-!- Gallaecio [~quassel@84.120.219.139.dyn.user.ono.com] has joined #wesnoth-dev
20140302 13:35:53-!- Gallaecio [~quassel@84.120.219.139.dyn.user.ono.com] has quit [Read error: Connection reset by peer]
20140302 13:36:39-!- Gallaecio [~quassel@84.120.219.139.dyn.user.ono.com] has joined #wesnoth-dev
20140302 13:37:21< Yasin> Link: http://paste.debian.net/84884/
20140302 13:38:49< Yasin> Wait, mordante, there is an error in codeblocks from the patch.
20140302 13:40:03< mordante> Yasin, I only will commit the smart list changes, but they won't compile for me
20140302 13:40:20< mordante> Yasin, can you try to change
20140302 13:40:22< mordante> typename inline smart_list<Data>::iterator_base<Value, Reversed> 
20140302 13:40:24< mordante> to
20140302 13:40:29-!- albince [~albince@117.221.145.23] has quit [Quit: Leaving]
20140302 13:40:33< mordante> typename inline smart_list<Data>::template iterator_base<Value, Reversed> 
20140302 13:40:47< mordante> that compiles for me (at least it did before ;-))
20140302 13:41:02< Yasin> Will check.
20140302 13:41:33< happygrue> shadowm: perfect!
20140302 13:42:25< happygrue> sachith500|2: I think your propsal page is shaping up nicely.  I assume the brainstorm section will eventually turn into the full-on plan.
20140302 13:42:39< sachith500|2> yeah that's the plan :D
20140302 13:43:05< sachith500|2> I havent put down some of the ideas
20140302 13:43:11< sachith500|2> what I was thinking is
20140302 13:43:12< Yasin> mordante, so far so good.
20140302 13:43:14< happygrue> one thing though, as I actually don't have GSoC experience before, since it does get submitted to Google as well as us, I'd suggest going over the questions they provide as you finish up and just make sure you have hit everything that they asked for
20140302 13:43:19< sachith500|2> provide the basic functionality
20140302 13:43:40< sachith500|2> oh the questionnaire?
20140302 13:43:45< sachith500|2> yeah I will fill that
20140302 13:43:48< Yasin> Do I patch again?
20140302 13:43:54< sachith500|2> I was being ambitious
20140302 13:43:58< mordante> Yasin, thanks then I'll commit the fix, no I just change it locally
20140302 13:44:02< sachith500|2> :P playing with media wiki etc
20140302 13:44:04< sachith500|2> happygrue
20140302 13:44:11< sachith500|2> I was thinking
20140302 13:44:11< happygrue> yeah.  Just I guess read it over and make sure it's not asking for some things that arn't there that seem imporant
20140302 13:44:20< sachith500|2> of getting working components
20140302 13:44:24< sachith500|2> for each of the major parts
20140302 13:44:29< sachith500|2> well basically parts 1 and 2
20140302 13:44:42< sachith500|2> and then improving them iteratively
20140302 13:44:54< sachith500|2> instead of just going head on, at part 1?
20140302 13:44:57< sachith500|2> what do you think?
20140302 13:45:13< sachith500|2> basically this would be something like
20140302 13:45:26< sachith500|2> since we know for definite games that end with leader death
20140302 13:45:29< sachith500|2> of all enemies
20140302 13:45:37< sachith500|2> we know the result
20140302 13:45:45< sachith500|2> so record that
20140302 13:45:55< sachith500|2> get a bit of data analysis with that
20140302 13:46:03< sachith500|2> as in actually implement it
20140302 13:46:13< sachith500|2> and then go back and improve the victory detection part
20140302 13:46:19< sachith500|2> what do you think? :)
20140302 13:46:44< happygrue> That seems a fair way to do it
20140302 13:47:07< sachith500|2> ok :D
20140302 13:48:10< happygrue> one advantage of that way is that you can be testing what is working as you go along, rather than thinking you have part one all done nicely only to find you have to change stuff because it's harder to pass along for some reason
20140302 13:48:37< sachith500|2> right :D
20140302 13:51:34< sachith500|2> halfway through the project, if progress is satisfactory, I was thinking of doing some basic presentation of it.
20140302 13:51:42< sachith500|2> havent worked out the details yet
20140302 13:52:07< sachith500|2> presentation of the analysed data that is.
20140302 13:53:08-!- irker106 [~irker@ai0867.net] has joined #wesnoth-dev
20140302 13:53:08< irker106> wesnoth: Mark de Wever wesnoth:master bf03a7c77f3e / src/utils/smart_list.hpp: Fix compilation with MSVC. http://git.io/vu8jWA
20140302 13:53:16< mordante> Yasin, committed ^
20140302 13:54:17< mordante> anonymissimus, jamit, MSVC should compile again^^
20140302 13:54:31< shadowm> iceiceice: No news from iceiceice3?
20140302 13:55:11-!- Gallaecio [~quassel@84.120.219.139.dyn.user.ono.com] has quit [Read error: Connection reset by peer]
20140302 13:55:23< happygrue> That seems reasonable sachith500|2.  There may be some open source tools already that exist and would help with display, I don't know.
20140302 13:56:09-!- Gallaecio [~quassel@84.120.219.139.dyn.user.ono.com] has joined #wesnoth-dev
20140302 13:56:24< Yasin> Not sure it would. Someone who has msvc not express should check it out.
20140302 13:58:26< mordante> I'm quite confident anonymissimus could compile before, I still expect the leakdetector to be a configuration issue
20140302 13:58:53< mordante> you could ask anonymissimus whether he can help you with that issue
20140302 13:59:09< Yasin> I'm not combiling with the leak detector anymore, but there is still the VCOMP lib error.
20140302 13:59:27< Yasin> Which is usually part of msvc, except not express.
20140302 13:59:39< Yasin> Ok, I will.
20140302 14:01:14< mordante> do you have a switch for OMP?
20140302 14:03:24< Yasin> What is omp?
20140302 14:04:46< mordante> http://msdn.microsoft.com/en-us/library/tt15eb9t.aspx
20140302 14:09:30< Yasin> Yes, it is set to "Yes (/openmp)"
20140302 14:11:23< mordante> can you try to disable it, openmp doesn't work really well for Wesnoth, so disabling it has little disadvantages
20140302 14:11:26-!- sachith500|2 [~kvirc@112.135.85.77] has quit [Ping timeout: 253 seconds]
20140302 14:11:38-!- sachith500 [~kvirc@61.245.168.44] has joined #wesnoth-dev
20140302 14:13:31-!- sachith500 [~kvirc@61.245.168.44] has quit [Client Quit]
20140302 14:14:52< Yasin> Building.
20140302 14:15:44< Yasin> Something happened in tutorial, but I can't reproduce it. When choosing to attack the quintain, the first dummy, the game exits abruptly.
20140302 14:17:25-!- kex [~kex@89.205.75.19] has joined #wesnoth-dev
20140302 14:17:52< shadowm> mordante: And yet I think the OS X builds are done with OpenMP enabled.
20140302 14:18:29< shadowm> The codeblocks project also passes -fopenmp by default.
20140302 14:21:31-!- sk94 [~sk@77-254-169-239.adsl.inetia.pl] has joined #wesnoth-dev
20140302 14:22:32< vultraz> IIRC they are
20140302 14:23:08-!- sk94 [~sk@77-254-169-239.adsl.inetia.pl] has quit [Client Quit]
20140302 14:23:23< mordante> shadowm, oke, the last thing I heard it wasn't working great
20140302 14:31:18< Yasin> mordante, no change.
20140302 14:32:33-!- werlley [~werlley@200-199-29-16.user.veloxzone.com.br] has joined #wesnoth-dev
20140302 14:35:40< mordante> Yasin, :-(
20140302 14:36:26-!- sachith500 [~kvirc@112.135.85.77] has joined #wesnoth-dev
20140302 14:37:05-!- Gambit [~derek@wesnoth/developer/grickit] has joined #wesnoth-dev
20140302 14:38:15< Yasin> It's ok. It's running on codeblocks atm.
20140302 14:39:01< Yasin> I like msvc better naturally, but I can tinker with it later, I'd like to play the game for a bit.
20140302 14:39:12-!- stikonas_ [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20140302 14:39:49-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Ping timeout: 264 seconds]
20140302 14:40:43< mordante> ok
20140302 14:48:07< Ivanovic> AI0867, boucman, Coffee_irc, elias, fabi, Gallaecio, iceiceice, loonycyborg, mordante, shadowm, Soliton, thunderstruck, trademark_, zookeeper, everyone else who cares: please get ready for 1.11.11
20140302 14:48:20< mordante> Ivanovic, I'm ready
20140302 14:48:22< Ivanovic> i am currently applying utils/wesnoth-optipng on the checkout
20140302 14:48:29< Ivanovic> once that is done i want to work on releasing
20140302 14:48:41-!- Gallaecio [~quassel@84.120.219.139.dyn.user.ono.com] has quit [Remote host closed the connection]
20140302 14:49:25-!- Gallaecio [~quassel@84.120.219.139.dyn.user.ono.com] has joined #wesnoth-dev
20140302 14:50:03< shadowm> I don't have anything pending to do at this time.
20140302 14:50:22< fabi> Ivanovic: okay
20140302 14:51:48< Gallaecio> I won’t have the translations ready for .11, will hold for .12.
20140302 14:53:11< fabi> zookeeper: I did some work on your map (the one you derive from) regarding the south. 
20140302 14:54:10< shadowm> > Updated translations: Slovak
20140302 14:54:25< shadowm> Sounds like not much happened in the translation department since 1.11.10.
20140302 14:54:51< zookeeper> fabi, ok... are you going to show it too? :p
20140302 14:55:21< shadowm> "Updated maps for scenario 8, 11, 12, 14, 16 and 17a." Uh, yes, everyone knows what these numbers mean...?
20140302 14:55:25< fabi> zookeeper: Yes, I am converting it from xcf to jpeg right now.
20140302 14:57:11< fabi> zookeeper: http://imagebin.org/296658
20140302 14:57:30< irker106> wesnoth: Ignacio R. Morelle wesnoth:master da47ad32b139 / changelog: Update a changelog entry http://git.io/LzwqKg
20140302 14:58:40< irker106> wesnoth: Ignacio R. Morelle wesnoth:master 751dab2bd8b2 / changelog: Update a changelog entry http://git.io/8VChxA
20140302 14:58:48< zookeeper> fabi, i think you mean "strait", not "street" :P
20140302 14:59:49< shadowm> Update: after multiple attempts to get this working, I've concluded [variation] [female] is just an unsupported border case and I'm on my own -- boo.
20140302 15:00:08< fabi> zookeeper: Oh, in German we say "Straße von Hormus". I took the name from that.
20140302 15:00:45< fabi> shadowm: No, it should work. Fill a proper bug report.
20140302 15:01:26< fabi> zookeeper: Straße == street
20140302 15:01:31< shadowm> fabi: Yes, of course it should, but it doesn't.
20140302 15:01:44< mordante> fabi, Straße == strait, it can mean both ;-)
20140302 15:02:14< fabi> strait sound better in my ears :-)
20140302 15:02:29< shadowm> As for filing a "proper" bug report: I'd have to come up with a generic test case so as to not bother people with nonsensical use case details, and that's a tedious chore I'm not going to do now.
20140302 15:03:11< mordante> strait is the proper word as zookeeper already mentioned
20140302 15:03:11< shadowm> And for now I have to decide what to do with the excess baseframes I have.
20140302 15:04:22< fabi> shadowm: What is wrong with your use case?
20140302 15:04:25-!- Yasin [~Yasin@41.129.34.167] has quit [Ping timeout: 264 seconds]
20140302 15:04:49< shadowm> It's a campaign-specific thing, thus doesn't belong in the bug tracker.
20140302 15:06:12< fabi> shadowm: So we won't get a bug report until there is a female variant in mainline?
20140302 15:07:24< shadowm> If you put it that way, I suddenly sound like a villain, yes.
20140302 15:07:54< shadowm> "Mwahaha, I'm not giving you this precious bug report!" et cetera.
20140302 15:08:09< fabi> :-)
20140302 15:09:15< zookeeper> see, this is why i hate git: i made a changelog edit. committed. tried to push - oh noes, someone's already pushed something else, so git just errors out. ok, fine, i pull. oh noes, git can't pull because there's a merge conflict. and tortoisesvn won't let me resolve that conflict right there and then, but instead i have to do things in an obnoxious way.
20140302 15:09:50< matthiaskrgr> zookeeper: git fetch, git rebase origin/master    ? 
20140302 15:09:58< fabi> zookeeper: But all this can happen with svn as well?
20140302 15:10:06< zookeeper> fabi, no, it can't
20140302 15:10:10< shadowm> SVN and Git both throw conflict markers at you, I don't see what the problem is.
20140302 15:10:35< matthiaskrgr> ( git fetch ; git rebase origin/master ; git push ) 
20140302 15:10:48< zookeeper> no, with SVN i can just edit the conflict and then re-commit, problem solved
20140302 15:11:20< fabi> same procedure with git
20140302 15:12:28< zookeeper> no it's not
20140302 15:12:40-!- mattsc [~mattsc@154.20.32.245] has joined #wesnoth-dev
20140302 15:12:43< fabi> at least my commandline git implementation and the eclipse integration work pretty much the same in merge conflict situation.
20140302 15:13:59< zookeeper> i dismissed the error dialog, checked the local changes, and changelog just had some nonsensical unrelated change in it not by me. ok, so i rever that to try to get back to a clean slate; it won't let me pull because there's still a merge conflict somewhere, except that i can't see it
20140302 15:19:11< fabi> zookeeper: Well, I can't say how your git frontend wants to be used. The commandline procedure is pretty much straight forward. git pull --rebase; edit conflicts with editor; git add filename; git pull --rebase-contine (or similar); git push 
20140302 15:19:38< shadowm> It's git rebase --continue/--skip/--abort.
20140302 15:20:01< zookeeper> can't push, can't pull
20140302 15:20:10< zookeeper> oh well
20140302 15:21:13-!- Octalot [~noct@31.185.149.167] has quit [Ping timeout: 240 seconds]
20140302 15:21:27< fabi> Do you have a commandline client? The output of "git status" would be of help then.
20140302 15:24:17< AI0867> he's using tortoisegit, which is a problem in itself
20140302 15:25:10< zookeeper> i can't even find a way to just wipe out all local commits and start over
20140302 15:25:47< zookeeper> ...which would always be the easiest and quickest way to solve these problems
20140302 15:27:49< shadowm> Oh, there is a way to do that in plain, vanilla Git.
20140302 15:30:38< zookeeper> well i don't want to fight with this crap, so i'll just forget about it for now and hope i can fix my clone someday
20140302 15:34:49 * vultraz uses tgit
20140302 15:35:49-!- dragonofair0 [~dragonofa@2601:7:8800:359:3c8c:fbe2:126:3962] has joined #wesnoth-dev
20140302 15:37:39-!- goblinThing [44bd8c2c@gateway/web/freenode/ip.68.189.140.44] has joined #wesnoth-dev
20140302 15:37:53< happygrue> zookeeper: if it's not a lot of files you can usegit checkout -- <file> to just overwrite changes to them I think?
20140302 15:37:54< happygrue> http://git-scm.com/book/en/Git-Basics-Undoing-Things
20140302 15:39:28< irker106> wesnoth: Nils Kneuper wesnoth:master e308bf55c570 / / (11 files in 10 dirs): updated Italian translation http://git.io/85sa0w
20140302 15:39:30< irker106> wesnoth: Nils Kneuper wesnoth:master dff2303bb200 / changelog players_changelog po/wesnoth-ai/de.po po/wesnoth/de.po: updated German translation http://git.io/ZmxZPg
20140302 15:39:49-!- shadowm_desktop [ignacio@wesnoth/developer/shadowmaster] has quit [Ping timeout: 264 seconds]
20140302 15:42:19< goblinThing> SoC question: If I have multiple project ideas, would it be prefferable to have them all on one page, or on separate pages?
20140302 15:42:31< zookeeper> in case someone wants the changelog to be up-to-date before the release, then this is what it should say:
20140302 15:42:32< zookeeper>    * Eastern Invasion:
20140302 15:42:32< zookeeper>      * Updated maps for scenarios 12, 14, 16 and 17a.
20140302 15:42:32< zookeeper>      * Fixed a bug in 'Captured' which can cause the beginning to make no sense.
20140302 15:43:02< mordante> goblinThing, separate and have one of them point to the other page for the questionaire
20140302 15:43:32< goblinThing> Thanks. 
20140302 15:44:12< goblinThing> With current WML capabilities, would it be possible (or even useful) to put more campaigns in a multiplayer format?
20140302 15:45:07< goblinThing> I liked the idea of playing... that one with kalenz, I always seem to forget the name... in multiplayer
20140302 15:46:18< zookeeper> assuming there's nothing inherently broken WRT MP campaigns (and there often has been, dunno about right now), then surely it'd be possible, but personally i don't see many mainline campaigns which could be converted to MP without extensive rewrites
20140302 15:47:47< mordante> zookeeper, I'll update the changelog
20140302 15:48:10< zookeeper> and by that i mean that i don't see them being playable by 2 or more players if the plot and gameplay is still supposed to make sense.
20140302 15:48:14< zookeeper> mordante, cool, thanks
20140302 15:48:44< goblinThing> I'm not thinking about changing the story at all, just about giving the omnipresent "adviser" character to a second player. And, yes, I'm thinking along the lines of extensive rewrites. I'm trying to come up with potential projects for SoC that I can pull off without C++.
20140302 15:49:43< mordante> you're welcome
20140302 15:50:01< goblinThing> Of course, if the devs in general don't like the idea, I won't try to implement it...
20140302 15:50:21< irker106> wesnoth: Mark de Wever wesnoth:master b88aa5a23f23 / changelog: Update the changelog at Zookeeper's request. http://git.io/-Zun_w
20140302 15:50:22< mordante> zookeeper, ^
20140302 15:50:56-!- DCW [~Thunderbi@cpc66863-finc15-2-0-cust393.4-2.cable.virginmedia.com] has joined #wesnoth-dev
20140302 15:51:09< mordante> goblinThing, you want to work on rewriting/changing a campaign as GSoC idea?
20140302 15:51:55< goblinThing> It seems like a possibility. Right now I'm just throwing stuff out here to see what the general feel on it is.
20140302 15:52:34< mordante> well we expect students to work on coding not only on modifying WML
20140302 15:52:41< goblinThing> I have not "decided" on any one thing. I've written one proposal, but want to do whatever is most helpful for the devs within my capabilities.
20140302 15:52:59< mordante> I'm quite sure we rejected this kind of ideas in the past
20140302 15:53:15< goblinThing> Good to know.
20140302 15:53:56< mordante> just want to prevent you from spending time on an idea that will not be accepted ;-)
20140302 15:54:31< mordante> in general if you have ideas not listed, first discuss them here to see whether they have a chance to be accepted
20140302 15:54:38< goblinThing> Yeah, that's why I threw it out here first.
20140302 15:54:47< zookeeper> goblinThing, sure, but the thing is that in most campaigns there is no omnipresent advisor; they're usually gone for a least some scenarios. anyway, i'm sure there's some which would work, i'm just saying that most would need to be more heavily adapted.
20140302 15:56:13< goblinThing> I'm juggling ideas a bit because I have not learned any C++... I'd be happy to learn, and will learn it at some point, but I'm not sure it will be in time to be useful.
20140302 15:56:45< zookeeper> liberty could work, for instance. but then there's the question of whether splitting the player side into two sides which simply start next to each other makes for very fun scenarios. maybe it's ok, maybe it's just boring. i dunno.
20140302 15:58:15< goblinThing> @zookeeper: Since It is not likely to be accepted for GSoc, I'll have to try that some other time.
20140302 16:00:08< mordante> goblinThing, what languages do you know?
20140302 16:02:49< goblinThing> Java, Flash Actionscript. I don't really know either of them super in-depth, but coding comes naturally to me. I'm currently getting deeper into Java since Flash has fallen out of favor in the tech world.
20140302 16:04:31< goblinThing> note: I've finished the page for the wiki... the project is subject to change if I find something more useful to do.
20140302 16:05:14< Ivanovic> argh, okay, i got to rerun the wesnoth-optipng step
20140302 16:05:21< Ivanovic> will take about an hour until it is done...
20140302 16:10:02< mordante> ok, well if you're comfortable in Java, I think learning the basics C++ shouldn't be too hard
20140302 16:10:18< mordante> (just unlearn to new everything first ;-) )
20140302 16:10:49-!- werlley [~werlley@200-199-29-16.user.veloxzone.com.br] has quit [Ping timeout: 252 seconds]
20140302 16:11:01< goblinThing> Someone else, I think Ashika, said that as well.
20140302 16:11:26-!- ahi [~ahi@41.34.240.50] has quit [Read error: Operation timed out]
20140302 16:11:27< goblinThing> The first language is always the hardest. After that, they all make sense. : )
20140302 16:13:24< sachith500> oh ;)
20140302 16:13:27< sachith500> you should try befunge
20140302 16:13:39< mordante> that's evil :-P
20140302 16:13:44< sachith500> hahah
20140302 16:14:06< goblinThing> ?
20140302 16:14:09< mordante> but depends, after you know C/C++/C#/Jave etc LISP makes little sense ;-)
20140302 16:14:25< mordante> http://de.wikipedia.org/wiki/Befunge
20140302 16:14:37< sachith500> haha there was a round of codeforces
20140302 16:14:40< sachith500> that used befunge
20140302 16:14:45< mordante> or better http://en.wikipedia.org/wiki/Befunge
20140302 16:14:46< sachith500> epic
20140302 16:15:52< mordante> I believe that, I've never tried to use these languages
20140302 16:16:07< goblinThing> 0_0      It appears I was too hasty in my assesment.
20140302 16:16:25< mordante> but my most favourite is the printout of a whitespace program
20140302 16:19:32< goblinThing> It appears my internet has died...
20140302 16:21:56 * Ivanovic is going crazy now!
20140302 16:21:59< Ivanovic> for i in wesnoth wesnoth-ai wesnoth-anl wesnoth-aoi wesnoth-did wesnoth-dm wesnoth-dw wesnoth-editor wesnoth-ei wesnoth-help wesnoth-httt wesnoth-l wesnoth-lib wesnoth-low wesnoth-manpages wesnoth-manual wesnoth-multiplayer wesnoth-sof wesnoth-nr wesnoth-sotbe wesnoth-tb wesnoth-test wesnoth-thot wesnoth-trow wesnoth-tsg wesnoth-tutorial wesnoth-units wesnoth-utbs; do for j in wesnoth wesnoth-ai wesnoth-anl wesnoth-aoi wesnoth-did 
20140302 16:22:00< Ivanovic> wesnoth-dm wesnoth-dw wesnoth-editor wesnoth-ei wesnoth-help wesnoth-httt wesnoth-l wesnoth-lib wesnoth-low wesnoth-manpages wesnoth-manual wesnoth-multiplayer wesnoth-sof wesnoth-nr wesnoth-sotbe wesnoth-tb wesnoth-test wesnoth-thot wesnoth-trow wesnoth-tsg wesnoth-tutorial wesnoth-units wesnoth-utbs; do ./utils/po2po $i $j de; done ; done
20140302 16:23:58-!- goblinThing [44bd8c2c@gateway/web/freenode/ip.68.189.140.44] has quit [Ping timeout: 245 seconds]
20140302 16:24:25< mordante> Ivanovic, you shouldn't start programming in befunge :-)
20140302 16:24:29< mordante> :-P
20140302 16:25:49-!- goblinThing [44bd8c2c@gateway/web/freenode/ip.68.189.140.44] has joined #wesnoth-dev
20140302 16:30:27-!- werlley [~werlley@187-41-128-108.user.veloxzone.com.br] has joined #wesnoth-dev
20140302 16:31:31< irker106> wesnoth: Nils Kneuper wesnoth:master abbad834164f / po/ (27 files in 27 dirs): updated German translation using evil foo http://git.io/I6yx8A
20140302 16:42:33-!- Matei [~razvan@p16.eregie.pub.ro] has joined #wesnoth-dev
20140302 16:44:55< goblinThing> sorry, my internet is inconsistent. @mordante: do you think my proposed project has a chance of being accepted? If not, would it stand a higher chance if I made integration with the map editor a focus? http://wiki.wesnoth.org/GoblinThing_Wizard  (you don't have to read the whole thing, just the description and implementation. ^_^)
20140302 16:47:12-!- vorobeez [~quassel@85.142.148.12] has quit [Ping timeout: 240 seconds]
20140302 16:49:34< mordante> goblinThing, since you know Java, maybe talk to fabi or timotei there is an eclipse project which seems to be related to this
20140302 16:51:01< goblinThing> ...an eclipse project? Do you mean eclipse, the java editor/compiler, or some other eclipse?
20140302 16:51:38< mordante> http://wiki.wesnoth.org/SummerofCode2011_Timotei21
20140302 16:52:21< mordante> but I know next to nothing about it, but both fabi and timotei have worked on this project
20140302 16:55:43< goblinThing> Good god. That looks very, very comprehensive... but my guess is that this is not what I'm trying to do. I'll have to contact them and see if I can look at what they've done. Thanks for the info!
20140302 16:58:31< goblinThing> Hey, do you know of a way to contact timotei? Preferrably something that lets me leave a message? A forum name, perhaps?
20140302 16:59:31< mordante> maybe not exactly, but this we already have, so having somebody improving it might be a good idea
20140302 16:59:37< mordante> the code is in utils/umc_dev
20140302 17:00:01< goblinThing> Aaaaah, thank you. I'll look into that.
20140302 17:00:21< mordante> he might be reading the logs, he lurks here and he uses the same handle in the forums
20140302 17:08:35-!- kex [~kex@89.205.75.19] has quit [Remote host closed the connection]
20140302 17:09:12-!- kex [~kex@89.205.75.19] has joined #wesnoth-dev
20140302 17:13:32-!- kex [~kex@89.205.75.19] has quit [Ping timeout: 252 seconds]
20140302 17:16:01-!- jamit [~jamit@wesnoth/developer/jamit] has joined #wesnoth-dev
20140302 17:21:17< jamit> mordante: Thanks.
20140302 17:21:18-!- sachith500 [~kvirc@112.135.85.77] has quit [Quit: KVIrc 4.2.0 Equilibrium http://www.kvirc.net/]
20140302 17:22:08< jamit> Still not sure why clang accepted those lines, but MSVC did not.
20140302 17:23:48< irker106> wesnoth: Nils Kneuper wesnoth:master 8009ed38b3cf / po/ (28 files in 28 dirs): updated German translation http://git.io/IDIZJw
20140302 17:29:05< mordante> jamit, me neither also not why msvc was happy with the typename in front of it, clang wants the template keyword when I use typename
20140302 17:30:38< mordante> then again I've heard often enough the namelookup is rather difficult to implement and it's not uncommon for compilers to get it wrong
20140302 17:32:56< jamit> gcc also complained with either typename or template alone, but is fine with them both present.
20140302 17:33:55< jamit> Meh, I can see using both keywords. It's just annoying when someone else's compiler is being more picky than mine. (Since I won't know about the problem until they complain.)
20140302 17:34:05< mordante> well at least there is a combination that works on all three compilers ;-)
20140302 17:35:02< jamit> That is good.
20140302 17:35:03< mordante> yes it is, but well such is life
20140302 17:37:49< mordante> Ivanovic, nice to see the German translation picking up speed again :-)
20140302 17:38:13< Ivanovic> mordante: i just wanted to have something to do while running optipng on all of mainline
20140302 17:38:21< mordante> :-)
20140302 17:38:30< Ivanovic> mordante: and i just phoned chrber to have him update the other campaigns which are "close to complete"
20140302 17:38:49< Ivanovic> looks like it should be possible to get many textdomain to completion with about one or two hours of work
20140302 17:38:57< mordante> nice
20140302 17:40:03< Ivanovic> about 20 textdomains might be completed once someone takes those 2h
20140302 17:40:30< Ivanovic> right now 9 textdomains are complete
20140302 17:43:31< mordante> is chrber healthy again?
20140302 17:43:51< irker106> wesnoth: Nils Kneuper wesnoth:master 3ca6988963d2 / / (172 files in 28 dirs): applied utils/wesnoth-optipng http://git.io/5Afcuw
20140302 17:44:04< mordante> so we're moving to 100% again :-)
20140302 17:44:12-!- lipkab [~lipkab@host-91-147-212-189.biatv.hu] has joined #wesnoth-dev
20140302 17:44:47-!- DCW [~Thunderbi@cpc66863-finc15-2-0-cust393.4-2.cable.virginmedia.com] has quit [Read error: Connection reset by peer]
20140302 17:45:21< Ivanovic> he is able to walk again
20140302 17:47:53< mordante> but not fully healed yet?
20140302 17:48:08< Ivanovic> his doc says that he should not be doing sports yet
20140302 17:48:12< Ivanovic> but he can go back to work
20140302 17:50:35< mordante> ok
20140302 17:51:12< mordante> btw is there a reason several fuzzy have remained, where the #| text is gone?
20140302 17:52:08< mordante> just because the translation is not updated?
20140302 17:52:53< Ivanovic> this happens if there is another update
20140302 17:52:59< Ivanovic> this just makes the old comments vanish
20140302 17:53:33< mordante> ah ok
20140302 17:53:37< Ivanovic> okay i'll release 1.11.11 now
20140302 17:56:37-!- lipkab [~lipkab@host-91-147-212-189.biatv.hu] has quit [Ping timeout: 264 seconds]
20140302 17:58:19< jamit> 1.11.11 -- the "one" release? :)
20140302 17:58:52< irker106> wesnoth: Nils Kneuper wesnoth:master 28c8ab9815b0 / changelog players_changelog src/wesconfig.h: version bump to 1.11.11 http://git.io/tOu7sw
20140302 17:58:54< irker106> wesnoth: Nils Kneuper wesnoth:master 2e2b6a7295f7 / Doxyfile: updated Version in Doxyfile http://git.io/C0ar9Q
20140302 17:58:54< mordante> the one starting the branching :-)
20140302 17:58:56< irker106> wesnoth: Nils Kneuper wesnoth:master c2d278500ee0 / / (111 files in 29 dirs): pot-update and regenerated doc files http://git.io/AhLmwA
20140302 17:59:08-!- lipkab [~lipkab@host-91-147-212-189.biatv.hu] has joined #wesnoth-dev
20140302 17:59:55< vultraz> The One Ring
20140302 18:05:38-!- Ivanovic changed the topic of #wesnoth-dev to: releasing 1.11.11 | string+feature freeze active on master | 213 bugs, 350 feature requests, 28 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
20140302 18:08:28-!- anonymissimus [~chatzilla@HSI-KBW-149-172-228-192.hsi13.kabel-badenwuerttemberg.de] has joined #wesnoth-dev
20140302 18:10:13< anonymissimus> mordante: unfortunately still not fixed
20140302 18:10:46< anonymissimus> (no idea what that guy did, he said MSVC2008...
20140302 18:10:48< mordante> anonymissimus, not :-(
20140302 18:10:48-!- kex [~kex@89.205.75.19] has joined #wesnoth-dev
20140302 18:11:03< mordante> anonymissimus, what error do you get now?
20140302 18:11:32< anonymissimus> mordante: it changed something though: http://pastebin.com/xrGfBNVj
20140302 18:12:26< anonymissimus> your patch is the same as what jamit wrote I see
20140302 18:12:59< mordante> except that I have a small change in the parameter type, also made it a template
20140302 18:13:04< anonymissimus> (and for me none of those compile errors makes or made any sense nor do I understand that templatish stuff)
20140302 18:13:43< anonymissimus> okay, my smart list is unmodified however
20140302 18:13:55< anonymissimus> so i just have your commit
20140302 18:15:49-!- kex [~kex@89.205.75.19] has quit [Ping timeout: 264 seconds]
20140302 18:16:50< anonymissimus> mordante: i copy-paste and search, it's exactly what jamit wrote
20140302 18:17:01< anonymissimus> perhaps you forgot something ?
20140302 18:18:18< jamit> anonymissimus: Probably some other issue that MSVC just didn't get to before.
20140302 18:19:35< anonymissimus> jamit: the problem is somewhere else you think ?
20140302 18:21:12< mordante> anonymissimus, I also had this part http://paste.debian.net/84924/
20140302 18:21:21< jamit> Not sure. Those line numbers are the ends of the functions that return an iterator_base. So related somehow, but it might be a different technicality.
20140302 18:21:45-!- werlley [~werlley@187-41-128-108.user.veloxzone.com.br] has quit [Ping timeout: 252 seconds]
20140302 18:22:46< jamit> Hmmm... what if you remove that part? That's in the assignment operator, which is what the new messages are complaining about.
20140302 18:22:46-!- ujdf [~V@72-69-ftth.on.nl] has joined #wesnoth-dev
20140302 18:23:35-!- {V} [~V@72-69-ftth.on.nl] has quit [Disconnected by services]
20140302 18:23:41-!- ujdf is now known as {V}
20140302 18:23:43< jamit> Only one of them. Maybe not.
20140302 18:25:25-!- lipkab [~lipkab@host-91-147-212-189.biatv.hu] has quit [Ping timeout: 264 seconds]
20140302 18:27:40-!- lipkab [~lipkab@host-91-147-212-189.biatv.hu] has joined #wesnoth-dev
20140302 18:28:38< zookeeper> yay, fixed my clone
20140302 18:29:14< anonymissimus> jamit: which part ?
20140302 18:30:27< jamit> anonymissimus: The part mordante mentioned:  http://paste.debian.net/84924/
20140302 18:32:17 * grzywacz imagines zookeeper transplanting a new liver ;)
20140302 18:33:05< anonymissimus> jamit: i tried with and without already
20140302 18:33:23< irker106> wesnoth: ivanovic wesnoth: c2d278500ee0 tagged as 1.11.11
20140302 18:33:27< anonymissimus> and errors are for all 3, operator++, -- and =
20140302 18:35:25< Ivanovic> loonycyborg, Rhonda: 1.11.11 is tagged
20140302 18:35:42< Ivanovic> currently uploading to sf.net
20140302 18:37:04< anonymissimus> hm, interesting, commenting them out shows no compile errors
20140302 18:37:17< irker106> wesnoth: Nils Kneuper wesnoth:master 271961939439 / Doxyfile changelog players_changelog src/wesconfig.h: post release version bump to 1.11.11+dev http://git.io/gWQtxg
20140302 18:37:22< anonymissimus> where are the operators used ? I would assume lots of places
20140302 18:38:26< anonymissimus> I suspect you may be right about the problem being somewhere else; putting things into .hpps and using inline tends to lead to such errors
20140302 18:40:44< jamit> Those operators might not be used anywhere yet. I had defined them so that if someone were to try to use them in the future they would work. (And I think I had done a compile test where they were all instantiated. But I could have missed something.)
20140302 18:44:53< zookeeper> fabi, so what's the intent behind those southlands map edits? to flesh out the khalifate lore, or something else?
20140302 18:46:22< anonymissimus> jamit: ok, so a bad workaround could be to remove them ?
20140302 18:46:48< anonymissimus> and I should try compiling that way to check whether any other problems have been introduced
20140302 18:47:12< anonymissimus> the longer I don't the more accumulates usually
20140302 18:47:41< mordante> anonymissimus, maybe conditionally disable them, based on the MSVC version
20140302 18:48:01-!- lipkab [~lipkab@host-91-147-212-189.biatv.hu] has quit [Ping timeout: 240 seconds]
20140302 18:48:03-!- lipkab_ [~lipkab@host-91-147-212-189.biatv.hu] has joined #wesnoth-dev
20140302 18:49:13< anonymissimus> mordante: you suspect 2010 or newer compiles ?
20140302 18:49:50< anonymissimus> fabi: would be nice if you could fix http://gna.org/bugs/index.php?21467 I'm stashing/unstashing my workaround all the time
20140302 18:52:17< mordante> anonymissimus, more like I don't like to delete the code it is useful
20140302 18:53:11< anonymissimus> mordante: then its better to disable it in another way
20140302 18:53:30< anonymissimus> at some point some Linux dev uses it...dang
20140302 18:53:39< mordante> maybe try to move the cofe in the class
20140302 18:53:43< mordante> code*
20140302 18:53:57-!- Ivanovic changed the topic of #wesnoth-dev to: released 1.11.11, announcing "soon" | string+feature freeze active on master | 213 bugs, 350 feature requests, 28 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
20140302 18:55:29< anonymissimus> yes, such things can work (for no apparent reason)
20140302 18:55:40< jamit> Yes, moving the code into the class should fix the error. It just makes the code uglier.
20140302 18:56:33< mordante> well you can add a comment it's as work-around for an MSVC error
20140302 18:57:05< mordante> anyway I'm off bye
20140302 18:57:33-!- mordante [~mordante@wesnoth/developer/mordante] has quit [Quit: Leaving]
20140302 19:02:25-!- spoffy [~spoffy@152.78.175.8] has quit [Ping timeout: 240 seconds]
20140302 19:03:43-!- spoffy [~spoffy@152.78.175.8] has joined #wesnoth-dev
20140302 19:05:14< irker106> wesnoth: Nils Kneuper wesnoth:master 5c07956e597f / changelog players_changelog po/wesnoth/de.po: updated German translation http://git.io/UN9Nyw
20140302 19:06:30< jamit> anonymissimus: Could you try changing line 115 in smart_list.hpp to iterator_base<Value, Reversed> & operator=(const iterator_base & that);
20140302 19:06:51< jamit> (That adds "<Value, Reversed>" to what is already there.)
20140302 19:07:22< jamit> That's the only thing I've come up with that could be confusing MSVC.
20140302 19:08:19< jamit> Answers on Stack Overflow suggest this might be a bug in MSVC, but that doesn't mean there might not be a workaround.
20140302 19:09:16-!- travis-ci [~travis-ci@ec2-23-22-100-248.compute-1.amazonaws.com] has joined #wesnoth-dev
20140302 19:09:16< travis-ci> [travis-ci] wesnoth/wesnoth#1914 (1.11.11 - c2d2785 : Nils Kneuper): The build passed.
20140302 19:09:16< travis-ci> [travis-ci] Build details : http://travis-ci.org/wesnoth/wesnoth/builds/19927009
20140302 19:09:16-!- travis-ci [~travis-ci@ec2-23-22-100-248.compute-1.amazonaws.com] has left #wesnoth-dev []
20140302 19:14:28< Ivanovic> looks as if sf.net is slow propagating the downloads today
20140302 19:14:39< Ivanovic> it only lists "will be available for download shortly"
20140302 19:15:24< Ivanovic> i will probably create the 1.12 branch some time tomorrow in the evening
20140302 19:15:31< Ivanovic> so don't worry, be happy!
20140302 19:15:32< Ivanovic> ;)
20140302 19:21:24-!- Delfinisko [~Delfinisk@ip-85.163.0.232.o2inet.sk] has joined #wesnoth-dev
20140302 19:26:01-!- spoffy [~spoffy@152.78.175.8] has quit [Ping timeout: 264 seconds]
20140302 19:27:11-!- wesbot changed the topic of #wesnoth-dev to: released 1.11.11, announcing "soon" | string+feature freeze active on master | 213 bugs, 349 feature requests, 28 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
20140302 19:28:40< Delfinisko> anyone skilled in translation of usermade campaign?
20140302 19:28:47< Delfinisko> is this PBL okay?
20140302 19:29:11< Delfinisko> http://pastebin.com/PWPB4dyg
20140302 19:29:31< anonymissimus> jamit: doesnt work
20140302 19:29:42< jamit> :(
20140302 19:30:09< jamit> <insert rant about the creators of MSVC, who shall remain nameless>
20140302 19:30:20< anonymissimus> there might actually be some bug in MSVC, I tried something with templates in another project some time ago, it only worked when I did put it into the class
20140302 19:30:56< anonymissimus> i dont recall whether it was the same compile errors but maybe
20140302 19:31:36< jamit> Yeah, I found some discussion on Stack Overflow that suggest it's a bug in MSVC. Oh well, I guess separating the definition and declaration will have to go.
20140302 19:34:05< anonymissimus> jamit: okay, if you pastebin a patch based on current trunk i can test it
20140302 19:34:26< anonymissimus> Delfinisko: looks okay, but that means nothing based on experience
20140302 19:35:33< anonymissimus> I wonder why that yasin or how the nick was said it compiles for him though
20140302 19:35:57< anonymissimus> one of use makes a mistake or the build systems differ in some way
20140302 19:38:03< Delfinisko> thanks
20140302 19:38:23< Delfinisko> working on it for a week and it is still not on wescamp...
20140302 19:40:35< anonymissimus> ah that you dotn need to wait for
20140302 19:41:00< anonymissimus> some script needs to be run, it may only appear after several months
20140302 19:41:32< anonymissimus> that is, it's a person who needs to do it (AI ? mordante ?) and they only do that....whenever
20140302 19:42:35< anonymissimus> if you kindly ask though...
20140302 19:43:12-!- Delfinisko [~Delfinisk@ip-85.163.0.232.o2inet.sk] has quit [Ping timeout: 240 seconds]
20140302 19:44:47-!- ALourenco [c189db30@gateway/web/freenode/ip.193.137.219.48] has joined #wesnoth-dev
20140302 19:45:03< ALourenco> Hi, I'm back. My application is already fixed. I think it's ok now
20140302 19:45:54< ALourenco> About the SpriteSheet project, It's meant to just load and use Sprite Sheets or it also has to get all assets in a SpriteSheet automaticaly
20140302 19:46:01< ALourenco> like a compresser or something?
20140302 19:46:05-!- iceiceice [~chris@cpe-66-108-20-80.nyc.res.rr.com] has quit [Read error: Operation timed out]
20140302 19:46:55-!- stikonas__ [~gentoo@cpc18-sgyl27-2-0-cust35.18-2.cable.virginm.net] has joined #wesnoth-dev
20140302 19:47:01-!- stikonas_ [~gentoo@wesnoth/translator/stikonas] has quit [Ping timeout: 264 seconds]
20140302 19:52:00-!- lipkab [~lipkab@host-91-147-212-189.biatv.hu] has joined #wesnoth-dev
20140302 19:52:32-!- lipkab_ [~lipkab@host-91-147-212-189.biatv.hu] has quit [Ping timeout: 252 seconds]
20140302 19:55:10-!- ancestral [~ancestral@75-161-229-57.mpls.qwest.net] has joined #wesnoth-dev
20140302 19:59:03-!- kex [~kex@89.205.75.19] has joined #wesnoth-dev
20140302 20:00:21< jamit> anonymissimus: http://pastebin.com/T8cY1tGB
20140302 20:00:40< jamit> Although I see no reason why this wouldn't work.
20140302 20:03:29-!- kex [~kex@89.205.75.19] has quit [Ping timeout: 240 seconds]
20140302 20:08:56-!- cib0 [~cib@p20030067CE0E5A01267703FFFEE75B84.dip0.t-ipconnect.de] has joined #wesnoth-dev
20140302 20:22:56-!- Gallaecio [~quassel@84.120.219.139.dyn.user.ono.com] has quit [Read error: Connection reset by peer]
20140302 20:23:01-!- timotei [~timotei@wesnoth/developer/timotei] has quit [Ping timeout: 264 seconds]
20140302 20:23:53-!- lipkab [~lipkab@host-91-147-212-189.biatv.hu] has quit [Ping timeout: 252 seconds]
20140302 20:24:28-!- Gallaecio [~quassel@84.120.219.139.dyn.user.ono.com] has joined #wesnoth-dev
20140302 20:24:52-!- timotei [~timotei@wesnoth/developer/timotei] has joined #wesnoth-dev
20140302 20:25:19< mattsc> Ivanovic: OS X package is currently uploading to SF
20140302 20:26:24-!- Jetrel_ [~Jetrel@c-75-73-180-126.hsd1.mn.comcast.net] has joined #wesnoth-dev
20140302 20:28:24-!- jamit [~jamit@wesnoth/developer/jamit] has quit [Ping timeout: 240 seconds]
20140302 20:29:05-!- ALourenco [c189db30@gateway/web/freenode/ip.193.137.219.48] has quit [Quit: Page closed]
20140302 20:29:37-!- Jetrel [~Jetrel@c-75-73-180-126.hsd1.mn.comcast.net] has quit [Ping timeout: 264 seconds]
20140302 20:30:00-!- lipkab [~lipkab@host-91-147-212-189.biatv.hu] has joined #wesnoth-dev
20140302 20:30:50< anonymissimus> jamit: works
20140302 20:31:24-!- anonymissimus [~chatzilla@HSI-KBW-149-172-228-192.hsi13.kabel-badenwuerttemberg.de] has quit [Quit: ChatZilla 0.9.90.1 [Firefox 12.0/20120423122928]]
20140302 20:32:48-!- Gambit [~derek@wesnoth/developer/grickit] has quit [Ping timeout: 240 seconds]
20140302 20:34:27< irker106> wesnoth: mattsc wesnoth:master 8f748f9bd2ca / projectfiles/Xcode/ (3 files in 2 dirs): Xcode project update for 1.11.11 http://git.io/cQP1PQ
20140302 20:38:24-!- lipkab [~lipkab@host-91-147-212-189.biatv.hu] has quit [Ping timeout: 240 seconds]
20140302 20:42:15-!- lipkab [~lipkab@host-91-147-212-189.biatv.hu] has joined #wesnoth-dev
20140302 20:42:33-!- Matei [~razvan@p16.eregie.pub.ro] has quit [Remote host closed the connection]
20140302 20:43:46-!- lipkab [~lipkab@host-91-147-212-189.biatv.hu] has quit [Read error: Connection reset by peer]
20140302 20:44:57-!- iceiceice [~chris@cpe-66-108-20-80.nyc.res.rr.com] has joined #wesnoth-dev
20140302 20:53:56-!- jamit [~jamit@wesnoth/developer/jamit] has joined #wesnoth-dev
20140302 20:56:08< mattsc> iceiceice: hi
20140302 20:56:12< iceiceice> hi
20140302 20:56:18-!- Aishiko_laptop [~unknown@cpe-065-191-176-226.nc.res.rr.com] has joined #wesnoth-dev
20140302 20:56:22< mattsc> any luck yet on the animation crash bug?
20140302 20:56:31< iceiceice> i keep forgetting about that
20140302 20:56:40< iceiceice> are you thinking about it as "assigned to me" right now?
20140302 20:56:42< Aishiko_laptop> there is an animation crash bug?
20140302 20:57:01-!- Gallaecio [~quassel@84.120.219.139.dyn.user.ono.com] has quit [Read error: Connection reset by peer]
20140302 20:57:17< iceiceice> i can look at it again now, i have some time
20140302 20:57:18< mattsc> iceiceice: well, not officially.  You said that you were going to look into it; and it is definitely over my head.
20140302 20:57:56< mattsc> So if you want to look at it, great.  If not, I'll put in a work-around for the multi-kill crash, because that one's really bugging me.  :)
20140302 20:57:57-!- Gallaecio [~quassel@84.120.219.139.dyn.user.ono.com] has joined #wesnoth-dev
20140302 20:58:06< mattsc> *look into it
20140302 20:58:41< Aishiko_laptop> iceiceice, I'd be interested in following how you investigate so I can learn a bit more about the animation subsystem and hunting down bugs
20140302 20:58:45< mattsc> Aishiko_laptop: yes - the easiest way to reproduce it is to press shift-k on a unit twice quickly in debug mode.
20140302 20:59:08< iceiceice> hmm do you remember what day it was that we told jamit about it ?
20140302 20:59:14< mattsc> If you press it the second time while the animation is still playing for the first, you get an assert.
20140302 20:59:15< iceiceice> i remember there were some notes pasted in irc
20140302 20:59:17< Aishiko_laptop> mattsc, OK I need to recompile in debug mode then
20140302 20:59:38< iceiceice> aishiko -- actually you can just execute with --debug flag
20140302 20:59:38< iceiceice> no need to recompile
20140302 20:59:43< jamit> iceiceice: It was one of the days before today. :P
20140302 20:59:48< Aishiko_laptop> iceiceice, ohh thats good
20140302 20:59:53< iceiceice> hehe
20140302 21:00:39< jamit> Two weeks ago? It was probably a weekend.
20140302 21:00:47< Aishiko_laptop> well I've never learned how to apply a flag like that so I have to recompile my previous projects with a change of debug to on
20140302 21:00:53< mattsc> I think it was less than that.
20140302 21:01:10< mattsc> But if you give me a moment, I can point you to the relevant parts in the code again.
20140302 21:02:29-!- aquileia [52d4193f@gateway/web/freenode/ip.82.212.25.63] has joined #wesnoth-dev
20140302 21:02:51-!- Gambit [~derek@wesnoth/developer/grickit] has joined #wesnoth-dev
20140302 21:02:54< iceiceice> its from 2/23
20140302 21:03:53< jamit> OK, one week ago then. This much still holds: It was probably a weekend. :)
20140302 21:04:02< mattsc> So, the double-kill crash gets trigger here: https://github.com/wesnoth/wesnoth/blob/master/src/menu_events.cpp#L1140
20140302 21:04:41< mattsc> If you put a test in for whether an animation is already playing, it does not happen.  But this is just one manifestation of a bigger problem.
20140302 21:05:05-!- Samual [~dioteckte@xonotic/core-team/Samual] has quit [Remote host closed the connection]
20140302 21:05:30-!- Samual [~dioteckte@c-71-195-88-69.hsd1.pa.comcast.net] has joined #wesnoth-dev
20140302 21:05:30-!- Samual [~dioteckte@c-71-195-88-69.hsd1.pa.comcast.net] has quit [Changing host]
20140302 21:05:30-!- Samual [~dioteckte@xonotic/core-team/Samual] has joined #wesnoth-dev
20140302 21:06:31< mattsc> The actual crash happens here: https://github.com/wesnoth/wesnoth/blob/master/src/unit_animation.cpp#L1428
20140302 21:06:56< iceiceice> yeah thats from the notes
20140302 21:07:06< Aishiko_laptop> hmmmm mine failed same file but at L1416
20140302 21:07:42< mattsc> Aishiko_laptop: well, that's because AI0867 put in an assert in that line.  If you take that out, you get a segfault in l.1428.
20140302 21:08:38< mattsc> What's going on (as far as I understand these things; which is not very far), you start one animation, interrupt it to start the second, play through the second, at which point both get dereferenced and you get the crash.
20140302 21:08:40< irker106> wesnoth: JaMiT wesnoth:master 237a1a0dc4e4 / src/utils/smart_list.hpp: Fix compilation on MSVC. http://git.io/8bYVKQ
20140302 21:09:25< iceiceice> hmm so i remember the line 1428
20140302 21:09:32< mattsc> So anyways, I do know how to prevent the double-kill (or double-creation) crash/assert, but there are still other ways of producing this, as e.g. pointed out by shadowm 
20140302 21:09:32< iceiceice> and the issue is this second pointer dereference
20140302 21:09:37< iceiceice> the unit might not exist if has already died
20140302 21:09:51< iceiceice> so you get segfault by bad pointer dereference
20140302 21:09:54< aquileia> Aishiko_laptop: Would you prefer if I post a suggestion for the animation WML side of the spritesheet project here, or at your wiki page?
20140302 21:10:03-!- Samual [~dioteckte@xonotic/core-team/Samual] has quit [Remote host closed the connection]
20140302 21:10:07< iceiceice> so then shadowm and i were like hmm maybe a smart pointer... and then the thought trails off
20140302 21:10:17< iceiceice> but maybe thats not the right idea,
20140302 21:10:20< iceiceice> maybe the right idea is 
20140302 21:10:27< iceiceice> instead of storing a pointer to a unit,
20140302 21:10:27< iceiceice> store a unit id
20140302 21:10:32< iceiceice> and ask the unit_map to find the unit
20140302 21:10:34< iceiceice> and abort if it is gone
20140302 21:10:38-!- Samual [~dioteckte@xonotic/core-team/Samual] has joined #wesnoth-dev
20140302 21:11:47< iceiceice> it would require to refactor this animation code i guess ...
20140302 21:11:59< mattsc> iceiceice: I really can't comment on what's the best here, so I'll just watch you and jamit and Aishiko_laptop and ... discuss this.
20140302 21:12:01< iceiceice> and i'm not familiar with those objects so i would have to read all the headers etc.
20140302 21:12:05< iceiceice> does that sound like it might be a reasonable fix?
20140302 21:12:18-!- justinzane [~justinzan@tiny.justinzane.com] has quit []
20140302 21:12:36-!- ALourenco [c189db30@gateway/web/freenode/ip.193.137.219.48] has joined #wesnoth-dev
20140302 21:13:58< Aishiko_laptop> aquileia, either would be fine with me
20140302 21:15:17< aquileia> I think it would be possible to use mostly the old syntax and have the code decide whether the unit already was converted
20140302 21:15:44< Aishiko_laptop> aquileia, actually that was the whole idea!
20140302 21:16:30< Aishiko_laptop> instead of the [n~m] being image files they'd become frames
20140302 21:16:35< aquileia> If I understand correctly you can easily access ->w and ->h of a SDL surface to get it's size?
20140302 21:17:33< aquileia> 1. if (w=h) old code
20140302 21:17:47-!- trewe [~trewe@65.160.54.77.rev.vodafone.pt] has joined #wesnoth-dev
20140302 21:18:28< Aishiko_laptop> aquileia, maybe but since all the units are 72x72, we'd really not need to get them, but I think that (w=h) is for when its looking at the whole image
20140302 21:18:46< aquileia> 2. else y_offset = index*h
20140302 21:18:58< aquileia> not all are 72x72
20140302 21:19:08< aquileia> there are e.g. 100x100 images
20140302 21:19:41< Aishiko_laptop> lovely
20140302 21:19:43< aquileia> that's the point - if the image is square, it's obviously no spritesheet
20140302 21:20:05< Aishiko_laptop> ohhh I see where your going
20140302 21:20:36< aquileia> else you can index the image in multiplies of the height, because that's also the width of a single image
20140302 21:20:49< Aishiko_laptop> but it could be example a 5x5 image sprite sheet with a square image would still end up being square
20140302 21:21:21< aquileia> For my suggestion to work, it would need linear spritesheets
20140302 21:21:41< Aishiko_laptop> See your going with a 1x tall with each image concatenated on to the end
20140302 21:21:57< Aishiko_laptop> it would work
20140302 21:22:41< Aishiko_laptop> the artists might not like it too much, all that scrolling to examine and work on each unit however, its a step forward
20140302 21:22:53< aquileia> and the syntax could stay exactly the same, because the index in [n~m] corresponds to a x position
20140302 21:23:06-!- kex [~kex@89.205.75.19] has joined #wesnoth-dev
20140302 21:23:50< aquileia> The artists don't have to work on a 1x20 spritesheet... it can be split back for them
20140302 21:24:18< Aishiko_laptop> and having it go wide instead of tall makes it easier to work with the current trend toward wider and wider monitors
20140302 21:24:25< aquileia> If you had in mind to write a conversion script, splitting should be a lot sompler than merging
20140302 21:25:12< aquileia> oh, and of course I meant x_offset some lines above...
20140302 21:25:22< aquileia> s/sompler/simpler
20140302 21:25:45< zookeeper> do you have your spritesheet implementation/API plans written down somewhere?`
20140302 21:26:01< Aishiko_laptop> aquileia, I was actually going to make that a goal, a splitting and merging script (it'd likely depend on a third-party tool) but since it would only affect the artists that make or want to make sprites I think it would be easier to use something they likely already have
20140302 21:26:03< zookeeper> i can't figure those out from the discussion
20140302 21:26:05< jamit> iceiceice: I haven't looked through the code to see what's a good way to fix it, but constantly finding a unit by id sounds like a resource-intensive procedure in a time-starved process.
20140302 21:26:18-!- PeterKeyser [827f4a23@gateway/web/freenode/ip.130.127.74.35] has joined #wesnoth-dev
20140302 21:26:30< iceiceice> idk whats the point of having this highly optimized unit_map object if we just store these crappy pointers everywhere?
20140302 21:26:35< Aishiko_laptop> zookeeper, I have but at the moment they are random and out of order
20140302 21:27:09< Aishiko_laptop> What I;d like to do is follow a unit from creation until it is displayed
20140302 21:28:19< Aishiko_laptop> that way I can check that nothing is going to break on its way to display, as I know and can follow that part but, before it gets there its like looking at the cloud, you don't need to know what happens here, but if there is even a small change in the syntax than something might break before that point
20140302 21:28:30< aquileia> Aishiko_laptop: I thought an additional 'size' attribute would be useful, but in fact it seems you get the information directly from you SDL surface
20140302 21:28:32< iceiceice> i mean ideally i guess we would never store any unit * pointers and just use the unit_map i would think?
20140302 21:28:43< iceiceice> i dont really have a sense of how fast / slow that actually is though
20140302 21:29:36-!- Gallaecio_ [~quassel@84.120.219.139.dyn.user.ono.com] has joined #wesnoth-dev
20140302 21:29:50< aquileia> Aishiko_laptop: the ->w and ->h calls are cheaper than reading it from WML I guess
20140302 21:30:15< aquileia> So without any syntax changes, it'd work
20140302 21:30:55< Aishiko_laptop> aquileia, I would think so too
20140302 21:31:25-!- Gallaecio [~quassel@84.120.219.139.dyn.user.ono.com] has quit [Ping timeout: 264 seconds]
20140302 21:34:15< zookeeper> well, looks like you're planning on some kind of smart indexing system to refer to individual images in the sheet, which is good
20140302 21:34:17< Aishiko_laptop> aquileia, I was thinking of having a function that's whole purpose was to determine spritesheet or not and if it was get the correct image out of the sheet and then pass it along to the next step(s), if it wasn't just pass out the orginal image and again just let it do what needs to be done to it
20140302 21:34:41-!- PeterKeyser [827f4a23@gateway/web/freenode/ip.130.127.74.35] has quit [Quit: Page closed]
20140302 21:34:51< Aishiko_laptop> zookeeper, I was thinking of having them X wide and then have multiple rows
20140302 21:35:31< iceiceice> jamit: if i wanted to profile the code, what do you think would be the best way to go about it?
20140302 21:35:46< Aishiko_laptop> so if 5 wide have it get the width/5 and then use that to calculate where an individual images is (mergine aquileia's idea and my orginalish plan)
20140302 21:36:03< aquileia> Aishiko_laptop: But then both the checking and indexing are less efficient - of course it's your choice
20140302 21:36:32< Aishiko_laptop> aquileia, and I'm for efficiency for sure
20140302 21:37:18< Aishiko_laptop> I was honestly having issues deciding the best way to find a particular image
20140302 21:37:44< aquileia> I mean it's a trifle (some modulo operations and divisions), so you should choose what you prefer
20140302 21:37:46< Aishiko_laptop> I might try it both ways and have some sort of test
20140302 21:38:24-!- justinzane [~justinzan@tiny.justinzane.com] has joined #wesnoth-dev
20140302 21:39:03< Aishiko_laptop> aquileia, don't get me started on the modu and divisions I was banging my head against the wall with a 3 hour tutoring session to write a little console program that changes random user number of seconds into hours, minutes, seconds displayed
20140302 21:39:09< aquileia> linear sheets are better internally I think (memory, speed) and 5xN sheets are nicer to look at
20140302 21:39:53< mattsc> jamit: cc bug #21258: https://github.com/wesnoth/wesnoth/commit/657f3c4c8056e4e0a4feba61e0a90a43463a7b13
20140302 21:39:59< Aishiko_laptop> {} not in the right spots, and trying to get him to grasp simple concepts (told several times) that / returns ints and % returns the remainer
20140302 21:40:26< jamit> iceiceice: I don't know what would be good to profile this. In the past, I've set a loop to do a small task 100,000 times and timed how long that took.
20140302 21:40:58< iceiceice> ok
20140302 21:40:58 * Aishiko_laptop feels her school should require any programming class to require either a previous programming class or the school's own intro to logic and programming
20140302 21:41:33< aquileia> Aishiko_laptop: There are border cases
20140302 21:42:20< Aishiko_laptop> aquileia, which is why I wanted to go with the multi row ones, its not that hard to implement, and would please the artists and anyone else that works with the sheets directly outside the game
20140302 21:42:31-!- Octalot [~noct@31.185.149.167] has joined #wesnoth-dev
20140302 21:42:38< aquileia> I have a fellow student (physics bachelor) who really wouldn't need a programming primer
20140302 21:42:58< jamit> mattsc: I hope it didn't take long for you to find that. :) 
20140302 21:43:21< zookeeper> Aishiko_laptop, why do you need to be able to tell whether an image is a ss or not? can't you just treat it as a ss if it's used as a ss? in any case, i think making constraints like how wide or tall they can be isn't a good idea.
20140302 21:43:29< aquileia> There was a programming contest (including master students in CS) - and he won it
20140302 21:43:44< zookeeper> and note that i really don't know your plans so that question might be kinda irrelevant
20140302 21:43:44< aquileia> And he programs just for fun...
20140302 21:43:52< Aishiko_laptop> aquileia, sorry but after having sat through a basic electronics course (which I some how missed for my Computer engineering degree) and having sat next to students that couldn't do fractions or percentages
20140302 21:46:16< mattsc> jamit: went onto github and checked out the history of the paladin unit file.  A minute or so.  :)
20140302 21:47:19< zookeeper> if i got to decide how ss's work, i might go for something like this: 1) individual frames and their dimensions indicated by surrounding them with magic-color boxes 2) sheets have to be arranged in a row/column format 3) WML-side, you can simply use x,y to index the image at the given row,column
20140302 21:47:25< Aishiko_laptop> I don't mind beginner, in many ways I am still one, but I could wish that explaining that { function declaration() statements} is not the right way only once
20140302 21:47:50< zookeeper> then when you first access an image in an ss, you just need to cache the metadata about the individual frames somewhere
20140302 21:48:54< zookeeper> anyway, i gotta go to bed but i'll read the logs
20140302 21:48:55< Aishiko_laptop> zookeeper, it would be done at the same time in the same function that gets the image out of the sheet and sends it on
20140302 21:48:57< aquileia> the metadata for a linear spritesheet being only a single index
20140302 21:49:23< zookeeper> ok
20140302 21:49:25-!- trademark_ [~trademark@nsg93-8-88-175-59-164.fbx.proxad.net] has quit [Ping timeout: 264 seconds]
20140302 21:49:46< Aishiko_laptop> once the image is gotten and passed on its the same as a single image being passed on
20140302 21:49:52< aquileia> Jetrel_: ^ I guess you might want to give a statement
20140302 21:51:10< zookeeper> i don't think it's a big deal at all if the frames of one row needs to have identical dimensions, but if you have no magic-color markers then you need to pass the info about number of rows/columns/etc from WML and that's kinda icky
20140302 21:52:36< aquileia> zookeeper: All sprites are square IIRC, so you can reconstruct it from the height of the spritesheet for linear ones
20140302 21:52:55< aquileia> no lines, no worries
20140302 21:53:07-!- justinzane [~justinzan@tiny.justinzane.com] has quit []
20140302 21:53:19< Aishiko_laptop> gabba
20140302 21:54:02< Aishiko_laptop> gabba is the one that said that and I think it makes sense, as it puts a copy into memory, modifies it (zoom, colour effects, etc) and then displays it
20140302 21:55:02< Aishiko_laptop> aquileia, and that gives the added benefit of not having to change the size of anything, like the arrows or magical attacks or large units that are not 72x72 as long as they are square
20140302 21:55:26< Aishiko_laptop> to begin with then it won't effect display once grabed from a sprite sheet
20140302 21:56:01-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [Ping timeout: 264 seconds]
20140302 21:56:38< aquileia> Aishiko_laptop: I'll leave it to you, I have to go now, bye
20140302 21:57:23< Aishiko_laptop> ok thank you for the suggestions aquileia I apperciate them, bye have a nice day
20140302 21:57:41< aquileia> Perhaps Jetrel will drop in and offer an artists view
20140302 21:57:50< Aishiko_laptop> I would love that
20140302 22:01:20-!- shadowm_desktop [ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20140302 22:02:26< Aishiko_laptop> and now I have to go myself back later =)
20140302 22:03:17< shadowm> iceiceice: Asking the unit_map to find the unit might might result in performance drops with lots of units.
20140302 22:03:42< iceiceice> hmm
20140302 22:04:58-!- dragonofair0 [~dragonofa@2601:7:8800:359:3c8c:fbe2:126:3962] has quit [Remote host closed the connection]
20140302 22:07:25-!- Aishiko_laptop [~unknown@cpe-065-191-176-226.nc.res.rr.com] has quit [Ping timeout: 264 seconds]
20140302 22:07:40-!- bumbadadabum_ [~bumbadada@d155109.upc-d.chello.nl] has joined #wesnoth-dev
20140302 22:09:49-!- bumbadadabum [~bumbadada@d155109.upc-d.chello.nl] has quit [Ping timeout: 264 seconds]
20140302 22:11:58-!- aquileia [52d4193f@gateway/web/freenode/ip.82.212.25.63] has quit [Quit: Page closed]
20140302 22:14:25< shadowm> aquileia: There are rectangular sprites.
20140302 22:14:50-!- grzywacz [~grzywacz@wesnoth/developer/grzywacz] has quit [Quit: :wq]
20140302 22:15:41-!- ancestral [~ancestral@75-161-229-57.mpls.qwest.net] has quit [Quit: End Transmission.]
20140302 22:16:02-!- aquileia [52d4193f@gateway/web/freenode/ip.82.212.25.63] has joined #wesnoth-dev
20140302 22:17:06< ALourenco> If you have some parameters like spriteWidth and spriteHeight, you can easily handle them
20140302 22:17:27< ALourenco> It might get confusing if you have both squared and rectangular sprites in the same SpriteSheet
20140302 22:17:43< ALourenco>  but that doesn't make a lot of sense. I think...
20140302 22:17:56< irker106> wesnoth: Ignacio R. Morelle wesnoth:master 780cced9c2bd / data/tools/terrain2wiki.py: terrain2wiki: wesnoth-old.git -> wesnoth.git http://git.io/fh-6Gw
20140302 22:18:25-!- timotei [~timotei@wesnoth/developer/timotei] has quit [Ping timeout: 240 seconds]
20140302 22:19:23-!- timotei [~timotei@wesnoth/developer/timotei] has joined #wesnoth-dev
20140302 22:20:08< irker106> wesnoth: Ignacio R. Morelle website:master f1e2dfc0e64f / gettext.wesnoth.org/ (index.extra.php index.lang.php index.php): gettext.w.o: wesnoth-old.git -> wesnoth.git http://git.io/h9PgOA
20140302 22:23:34-!- ancestral [~ancestral@75-161-229-57.mpls.qwest.net] has joined #wesnoth-dev
20140302 22:27:21< Jetrel_> Well, you need to stay online if you want to reach me
20140302 22:27:53< shadowm> Jetrel_: People read the logs.
20140302 22:28:12< Jetrel_> shadowm:  those don't replace discourse
20140302 22:28:33< Jetrel_> they're fine for monologues, they're not suitable for dialogues.
20140302 22:28:57< shadowm> Unfortunately they're often the only way to get things done here since we all live in different timezones.
20140302 22:29:21< iceiceice> shadowm: what do you suggest about the pointers?
20140302 22:29:41< iceiceice> there must be something like a unit id / pointer into an optimized data structure we can use, that is better than a smart pointer...
20140302 22:29:57< iceiceice> or like a quick "is this unit dead" check in the unit_map object or something
20140302 22:30:01< shadowm> iceiceice: I declare myself incompetent on this matter.
20140302 22:30:05< iceiceice> T_T
20140302 22:30:29< shadowm> I can't say "better than a smart pointer" sounds like a thing to me.
20140302 22:35:08-!- thunderstruck [~zaibotren@cpc13-sgyl31-2-0-cust696.18-2.cable.virginm.net] has quit [Quit: leaving]
20140302 22:37:23< iceiceice> what is we store a unit_map:iterator instead of a pointer?
20140302 22:37:29< iceiceice> that should be sort of the same thing, right?
20140302 22:37:39< iceiceice> will the iterator do some kind of safety check in case the unit is dead?
20140302 22:39:16-!- ancestral [~ancestral@75-161-229-57.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20140302 22:40:28< iceiceice> it looks like it will...
20140302 22:40:38< iceiceice> will have to read more to be sure
20140302 22:45:23< shadowm> Iterators are a generalization of pointers as a concept, as they may or may not (generally do not) implement smart pointer functionality.
20140302 22:45:28< shadowm> *and they may or may not
20140302 22:50:21-!- ancestral [~ancestral@75-161-229-57.mpls.qwest.net] has joined #wesnoth-dev
20140302 23:03:27< aquileia> Jetrel_: Well, I didn't know she would be off as well
20140302 23:03:36< Jetrel_> egad!
20140302 23:03:45< Jetrel_> I've been pinged. :o
20140302 23:04:00< aquileia> Yeah, sorry for that ;)
20140302 23:04:07< Jetrel_> So what's cooking?
20140302 23:04:19-!- justinzane [~justinzan@67.21.190.177] has joined #wesnoth-dev
20140302 23:05:23< aquileia> For the spritesheet project the question arose whether the sheets should contain a single line of images (1xN) or several
20140302 23:05:50< iceiceice> shadowm: the thing is that "smart pointer" functionality isnt actually what we want here
20140302 23:05:53< Jetrel_> I propose doing spritesheets like we've done in frogatto.
20140302 23:06:01< iceiceice> the issue is not units failing to be deallocated
20140302 23:06:13< Jetrel_> aquileia:  which means "several", to be helpful. :)
20140302 23:06:17< iceiceice> its that we dont have a way to check if a unit pointer is valid
20140302 23:06:22< aquileia> ok
20140302 23:06:22< shadowm> iceiceice: I thought the issue was that we had dangling pointers to units that were already deallocated.
20140302 23:08:52< iceiceice> sure
20140302 23:09:12< iceiceice> does a smart pointer actually solve that problem?
20140302 23:09:15< shadowm> iceiceice: Except in the rare cases where a type's destructor is manually called, when an object is destroyed, its memory is deallocated. Any pointers that continue to point to that memory become dangling pointers, since they may 1) point to memory that no longer belongs to the program, or 2) point to memory that now represents something completely different.
20140302 23:09:26< iceiceice> im not sure i guess that there are many kinds of smart pointers
20140302 23:09:53< shadowm> I remember suggesting a shared ownership pointer, i.e. shared_ptr.
20140302 23:10:16< iceiceice> heres the thing: will that require overhaul of unit_map?
20140302 23:10:23< shadowm> Sure, the problem is that the objects managed that way won't truly be destroyed until the last pointer is destroyed.
20140302 23:10:26< iceiceice> i'm reading unit_map.h, and it says that "the container owns the unit"
20140302 23:11:06< shadowm> Yes, it'd certainly require an overhaul of some sort.
20140302 23:11:10< iceiceice> https://github.com/wesnoth/wesnoth/blob/master/src/unit_map.hpp#L31
20140302 23:11:24< iceiceice> the container is designed so that looking up units by underlying id should be O(1)
20140302 23:11:30< iceiceice> i.e. like a hasthable look up
20140302 23:11:37< iceiceice> so not significantly worse than a pointer dereference
20140302 23:11:55< iceiceice> if we can't do that, wouldnt storing an iterator be more kosher with how unit_map is currently set up?
20140302 23:12:34< shadowm> Okay, in my memory unit_map was implemented using a std::map, I guess that changed at some point.
20140302 23:12:36< aquileia> shadowm: I stand corrected - but halos like light-beam-7 aren't the norm, so assuming square ones if nothing else is given seems acceptable
20140302 23:12:54< iceiceice> shadowm: it sounds like unit map is already doing reference counting
20140302 23:13:09< iceiceice> so if we use a shared ownership pointer i think we will be like reinventing the wheel
20140302 23:13:26< iceiceice> or duplicating existing functionality rather
20140302 23:13:47< shadowm> Does unit_map or its ancillary types have a way to check whether an existing iterator has been invalidated?
20140302 23:16:16< iceiceice> it looks like the iterator does some sanity checking when it is dereferenced: https://github.com/wesnoth/wesnoth/blob/master/src/unit_map.hpp#L179
20140302 23:17:16< iceiceice> not sure exaclty what checks those are yet...
20140302 23:17:40-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Remote host closed the connection]
20140302 23:19:19< shadowm> 	///Checks invariants.  For debugging purposes only.  Doesn't do anything in non-debug mode.
20140302 23:19:22< shadowm> 	bool self_check() const
20140302 23:19:31< aquileia> shadowm, vultraz: Is the help section on orbs still outdated? I checked it on 1.11.9 and the newsystem isn't there
20140302 23:20:02< vultraz> IIRC Ivanovic was updating it
20140302 23:20:08< aquileia> Ah, ok
20140302 23:20:08< shadowm> aquileia: It's updated.
20140302 23:20:08-!- justinzane [~justinzan@67.21.190.177] has quit [Ping timeout: 244 seconds]
20140302 23:20:35< aquileia> Time to update to 1.11.11 for me
20140302 23:20:57< shadowm> iceiceice: And the other check is only for underlying pointers to NULL or end iterators.
20140302 23:21:19< iceiceice> yeah... but it says in the documentation at the top that "iterators to dead units point to NULL"
20140302 23:21:23< jamit> iceiceice: Would it be better to validate the unit, or to cancel an existing animation when a new one is added?
20140302 23:21:54-!- spoffy [~spoffy@152.78.175.8] has joined #wesnoth-dev
20140302 23:22:51< iceiceice> my feeling is that, in this, or any other place in the code where we have a unit *, it should actually be a unit iterator
20140302 23:23:01< iceiceice> and we can check for valid before using it
20140302 23:23:14< iceiceice> but as to whether the animation logic should say i dont have an informed opinion
20140302 23:24:41< iceiceice> *whether the animation logic should change
20140302 23:25:00< shadowm> Who sets iterators to dead units to point to NULL?
20140302 23:25:20< iceiceice> from this comment, https://github.com/wesnoth/wesnoth/blob/master/src/unit_map.hpp#L104, i assume the unit_map does?
20140302 23:25:37-!- fabi_ [~fabi@88-134-21-163-dynip.superkabel.de] has joined #wesnoth-dev
20140302 23:25:37-!- fabi_ [~fabi@88-134-21-163-dynip.superkabel.de] has quit [Changing host]
20140302 23:25:37-!- fabi_ [~fabi@wesnoth/developer/fendrin] has joined #wesnoth-dev
20140302 23:26:12< shadowm> But only for its own iterators, not external iterators, or I'm failing to see some greater change propagation framework here?
20140302 23:28:19< iceiceice> i think i should read unit_map.cpp to make sure i have it right...
20140302 23:28:47< iceiceice> but what i'm taking away is that unit_map is intended to be this optimized data structure for us, and that the unit_map iterators are like a smart pointer essentially, implemented back in the day before boost etc.
20140302 23:28:58-!- fabi [~fabi@wesnoth/developer/fendrin] has quit [Ping timeout: 240 seconds]
20140302 23:29:20< shadowm> The checks I mentioned above do not really qualify as smart pointer functionality.
20140302 23:30:39< shadowm> unit_map::extract() (which is called by unit_map::erase()) appears to be in charge of pointing iterators to NULL.
20140302 23:30:58< iceiceice> ok, but i think that separately from those checks, it is doing some kind of reference counting / freeing memory of units no longer pointed to
20140302 23:31:17< iceiceice> or at least the documentation claims this
20140302 23:31:47< shadowm> It looks like this may work as long as iterators aren't copied.
20140302 23:32:06< shadowm> But iterators are usually copied.
20140302 23:32:54< shadowm> Hm.
20140302 23:33:30< shadowm> Okay, maybe the best way to know what's really going on is to actually destroy a unit from the map and see what happens to an iterator previously held by the caller.
20140302 23:34:19< iceiceice> yeah thats a good test
20140302 23:34:36< iceiceice> i think separately of this i'm also going to try to read unit_map.cpp and make sense of it
20140302 23:34:41< iceiceice> but we should do both
20140302 23:34:50-!- justinzane [~justinzan@67.21.190.177] has joined #wesnoth-dev
20140302 23:35:02< iceiceice> unit_map.cpp is only like 400 lines
20140302 23:35:08< iceiceice> and alot of it is error handling
20140302 23:35:21< shadowm> Oh, I was suggesting things, because I'm not really in the mood for C++ coding today. :p
20140302 23:35:54< shadowm> (Unless you happen to know why [variation] [female] is broken.)
20140302 23:37:27-!- mjs-de [~mjs-de@f049178011.adsl.alicedsl.de] has joined #wesnoth-dev
20140302 23:37:46-!- ALourenco [c189db30@gateway/web/freenode/ip.193.137.219.48] has quit [Quit: Page closed]
20140302 23:39:16-!- justinzane [~justinzan@67.21.190.177] has quit [Client Quit]
20140302 23:45:03< aquileia> shadowm: You probably already checked if the number of [frame]s is equal for one animation for all paths?
20140302 23:45:33< shadowm> aquileia: You mean for the variation thing?
20140302 23:45:38< aquileia> Yes
20140302 23:46:47< shadowm> I went and copied the entirety of the base type's animation WML  into the female variation with replaced paths and it's still broken.
20140302 23:47:17< aquileia> ok
20140302 23:48:47< shadowm> *variation's female variation
20140302 23:50:52< aquileia> The desert archer in UtBS might offer an ugly walkaround...
20140302 23:51:32< aquileia> There, the unit definition is packed in a macro and the different part (the unit cost) passed as an argument
20140302 23:51:53< shadowm> Yes, I know I could use derived unit types instead.
20140302 23:52:34< shadowm> But that doesn't work nice for the help system, nor for unit filtering.
20140302 23:53:05< aquileia> Which reminds me... is this acceptable? I mean it could be replaced with base_unit
20140302 23:53:33< shadowm> Wait, what are you talking about? I'm looking at the Desert Archer and it uses base_unit.
20140302 23:53:46< aquileia>     hide_help=true     do_not_list=yes
20140302 23:54:22< shadowm> Because we don't want the Desert Archer<N> units in the help.
20140302 23:54:26< aquileia> Oh right, I was wrong
20140302 23:54:39< aquileia> Yes, I know
20140302 23:55:12< shadowm> Yeah, I already checked for mainline instances of units with two genders and variations and there are none. :(
20140302 23:56:22< aquileia> Sorry I couldn't help
20140302 23:57:05< jamit> shadowm: have you described what's wrong with [variation] [female] anywhere? (Meaning someplace I can read later since I wouldn't be looking at it now.)
20140302 23:57:26< shadowm> The problem I see is that [male] and [female] might be always constructed from the base unit type without acknowledging that they may be constructed for a variation.
20140302 23:58:01< shadowm> Looking at unit_type::build() and the methods it calls.
20140302 23:59:51< shadowm> But the behavior between image/animations and description labels seems incongruous, which means that might not be a valid hypothesis.
--- Log closed Mon Mar 03 00:00:06 2014