--- Log opened Fri Oct 30 00:00:15 2015
--- Day changed Fri Oct 30 2015
20151030 00:00:15-!- louis94 [~~louis94@109.129.245.154] has quit [Quit: Konversation terminated!]
20151030 00:01:00-!- mjs-de [~mjs-de@x4db59880.dyn.telefonica.de] has quit [Remote host closed the connection]
20151030 00:07:23-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Ping timeout: 244 seconds]
20151030 00:11:21-!- Appleman1234 [~Appleman1@KD111239023035.au-net.ne.jp] has quit [Ping timeout: 250 seconds]
20151030 00:14:08-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20151030 00:16:40-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 246 seconds]
20151030 00:16:40-!- wedge010 is now known as wedge009
20151030 00:29:26-!- TheJJ [~rofl@ipbcc36ea9.dynamic.kabel-deutschland.de] has quit [Ping timeout: 250 seconds]
20151030 00:30:35-!- TheJJ [~rofl@ipbcc36ea9.dynamic.kabel-deutschland.de] has joined #wesnoth-dev
20151030 00:34:58-!- legoktm is now known as legoktm[NE]
20151030 00:44:14-!- Portaljacker [~Portaljac@modemcable081.139-178-173.mc.videotron.ca] has joined #wesnoth-dev
20151030 01:01:03-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20151030 01:08:45-!- Appleman1234 [~Appleman1@KD111239011115.au-net.ne.jp] has joined #wesnoth-dev
20151030 01:16:46-!- Portaljacker [~Portaljac@modemcable081.139-178-173.mc.videotron.ca] has quit [Quit: Leaving]
20151030 01:26:37-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20151030 01:37:52-!- iceiceice [~chris@wesnoth/developer/iceiceice] has quit [Quit: Leaving]
20151030 01:59:55-!- prkc [~prkc@51B78F27.dsl.pool.telekom.hu] has quit [Ping timeout: 252 seconds]
20151030 02:01:19-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20151030 02:11:33< mattsc> celticminstrel: does adding that flag fix the linking error for you also?
20151030 02:14:19< celticminstrel> Ah... somehow I got distracted and forgot...
20151030 02:15:26< celticminstrel> Oh huh. It was already there, but only for release builds. Weird.
20151030 02:15:52< mattsc> Hmm…  I didn’t even check that.
20151030 02:16:06< celticminstrel> It was under Preprocessor Macros rather than Other Flags.
20151030 02:16:16< mattsc> I see.
20151030 02:19:53-!- gfgtdf_ [~chatzilla@f054131148.adsl.alicedsl.de] has joined #wesnoth-dev
20151030 02:21:43-!- gfgtdf [~chatzilla@f054155221.adsl.alicedsl.de] has quit [Ping timeout: 244 seconds]
20151030 02:21:47-!- gfgtdf_ is now known as gfgtdf
20151030 02:38:16-!- gfgtdf [~chatzilla@f054131148.adsl.alicedsl.de] has quit [Quit: ChatZilla 0.9.92 [Firefox 41.0.2/20151014143721]]
20151030 02:55:10-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Quit: wedge009]
20151030 02:55:30-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20151030 02:57:57-!- ancestral [~ancestral@71-220-62-196.mpls.qwest.net] has joined #wesnoth-dev
20151030 03:00:48-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Quit: wedge009]
20151030 03:01:05-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20151030 03:02:35-!- aquileia [~chatzilla@HSI-KBW-149-172-208-211.hsi13.kabel-badenwuerttemberg.de] has joined #wesnoth-dev
20151030 03:03:30< aquileia> shadowm, loonycyborg: 1.12.4a is internally still handled as 1.12.4 (both by NSIS and Wesnoth), right?
20151030 03:04:21-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Client Quit]
20151030 03:05:02< shadowm> aquileia: I don't know about either for sure, I only know that I recommended not changing the in-game version because there weren't any code or content differences (1.12.4a is identical to 1.12.4 except for a few bundled DLLs.)
20151030 03:06:12-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20151030 03:06:31< shadowm> The installer's identification is a completely different matter. Also, the network lag is killing me right now.
20151030 03:06:37< aquileia> shadowm: Changing the version number (A.B, A.B.Ca, A.B.C.D) would mean that my NSIS patch script won't be able to find the registry entry for the previous version...
20151030 03:06:54-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Client Quit]
20151030 03:07:11-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20151030 03:07:41< shadowm> As I said, for 1.12.x you can safely assume all version numbers will be either A.B.C or A.B.Cx. I don't intend to change things in 1.13.x or 1.15.x without consulting and warning people well in advance.
20151030 03:08:17< shadowm> (You'd be surprised how much other stuff is likely to break with a version format change beyond just installers.)
20151030 03:09:47-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Client Quit]
20151030 03:09:58< shadowm> Also don't forget that the previous announced version does not always equal the previous installed version and viceversa.
20151030 03:10:31< aquileia> shadowm: If you indend to change the format down the line, I'd rather harden it now - there's an optional VersionCompare header that supports 4 numbers
20151030 03:10:59< shadowm> For example, for all intents and purposes versions 1.6.0 and 1.12.3 do not actually properly exist, being superseded by 1.6.0a and 1.12.4 (critical AI and security bugs, respectively).
20151030 03:12:44-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20151030 03:12:57< aquileia> shadomw: I iterate over all installed versions, pick those of the same major release, and compare the minor version numbers
20151030 03:13:44< shadowm> The reason I don't have a definite answer to the format question s (and I'm not speaking in an official fashion here) over the years it's become increasingly pointless to have that leading 1.
20151030 03:13:47< aquileia> Thus they don't have to be continuous, it's sufficient if they're integers
20151030 03:14:28< shadowm> So sometimes I'm tempted to drop it, but I know that many people will be displeased by such a move.
20151030 03:15:24-!- Appleman1234 [~Appleman1@KD111239011115.au-net.ne.jp] has quit [Ping timeout: 255 seconds]
20151030 03:17:13< shadowm> In fact, it wasn't more than three years ago that I proposed the idea.
20151030 03:17:14< aquileia> As long as we don't inflate the numbering like Firefox, I don't see a problem (even though I'd have to adjust the installer & patch scripts)
20151030 03:17:35-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Client Quit]
20151030 03:17:43< shadowm> Firefox doesn't inflate the numbering, they just have a different development process and pacing.
20151030 03:17:54-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20151030 03:18:30< shadowm> A new stable every month, features migrate over time across three staging tiers (nightly, aurora/devedition, beta) towards stable.
20151030 03:19:12< shadowm> They can afford this because they don't have the incompatible stable branches situation we have.
20151030 03:19:21-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Client Quit]
20151030 03:21:04< aquileia> Well, compared to their previous numbering scheme it's quite a leap forward, which I meant by inflated (I didn't write 'purposely')
20151030 03:22:08< shadowm> Only following in the footsteps of Chrome in that regard.
20151030 03:22:59< shadowm> I feel that it's much easier for end users this way for such a fast-paced development target (i.e. the Web).
20151030 03:24:07< shadowm> (HTML and CSS change all the time post HTML 4.01 and CSS 2.1, so there isn't much point in establishing hard and fast "major upgrade" thresholds this way.)
20151030 03:24:57 * aquileia looks at the ever-changing WML...
20151030 03:25:57-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20151030 03:26:46-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Client Quit]
20151030 03:27:31< aquileia> generalizing looping in WML isn't less revolutionary than generalizing media content in HTML 5
20151030 03:31:37-!- Necrosporus [~Necrospor@unaffiliated/necrosporus] has joined #wesnoth-dev
20151030 03:32:53< shadowm> Right, but those features in our case are part of a much longer queue of stuff that needs to be tested and fixed before it can be pushed to the next stable.
20151030 03:33:03< aquileia> Well, if the most likely change will be removing the leading 1, rsplit('.',1) should be somewhat future-proof
20151030 03:34:08< shadowm> I meant backend changes.
20151030 03:34:38< shadowm> Plus there's the stable API guarantee needed primarily for MP content.
20151030 03:35:38< shadowm> It's quite annoying, really, but until someone comes up with a viable alternative there isn't much I can do other than keep slipping more minor things past the radar.
20151030 03:35:50-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20151030 03:36:49< shadowm> vultraz: There were spelling fixes in master, have those been ported and pofixed to 1.12 yet?
20151030 03:37:03< vultraz> no
20151030 03:37:06< shadowm> -3:37 hours to do so.
20151030 03:37:12< shadowm> Okay, string freeze begins now then.
20151030 03:38:04-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Client Quit]
20151030 03:38:12< aquileia> shadowm: As to the announcement to the i18n list...
20151030 03:38:48< shadowm> Announcement? There hasn't been an announcement in recent weeks.
20151030 03:39:55< shadowm> The last post was on August 18.
20151030 03:40:16< aquileia> The announcement you might write now that the string freeze is in place. Could you add that the Windows installer has a separate packaging/windows/translations.nsh which will need translating as soon as the patcher is ready?
20151030 03:41:10< shadowm> And when would that be and how? I'd rather let you post that yourself so people know who to ask questions to.
20151030 03:41:31-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20151030 03:41:52< aquileia> It's theoretically possible to generate it from po-files, but that'd require to run another python script every time we build the installer
20151030 03:42:17< shadowm> It can't be made part of the usual pot-update targets?
20151030 03:42:25< aquileia> I could finalize the strings this weekend if needed
20151030 03:43:07< shadowm> (Read strings from .nsh, generate pots, merge pots into pos, write po strings back to .nsh.)
20151030 03:43:10-!- legoktm[NE] is now known as legoktm
20151030 03:43:21< aquileia> We can of course add it to these targets, but it's extra code nontheless
20151030 03:44:38< shadowm> So these strings are still likely to change before release is what you are saying?
20151030 03:45:34< shadowm> Also, how would people submit translations and to whom?
20151030 03:45:51< aquileia> It's basic error messages as "Download failed" "[version x] or higher required" and so on
20151030 03:46:27< shadowm> Okay, so strings that aren't even in place right now.
20151030 03:47:15< aquileia> If you take a look at the file (it currently contains all strings necessary for the normal installer) you'll see the Wesnoth-relevant strings are already in place
20151030 03:47:17-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: So long and thanks for all the fish.]
20151030 03:47:41< shadowm> Yes, I saw them, but I don't see any basic error messages (or error messages of any sort really) right now.
20151030 03:48:12< aquileia> That's because we don't have these yet, I'll do it this WE if I can
20151030 03:49:26< shadowm> Anyway, the important points that you need to decide on first: 1) to whom will translations be submitted and how? 2) when can translations begin to be submitted?
20151030 03:50:21< shadowm> (Also, to clarify a thing: I'm not in charge of translations at all, I just do the initial and final pot-updates and enforce the freeze rules.)
20151030 03:50:44< shadowm> (bbl)
20151030 03:51:50< aquileia> 1) either a) we add all of them to a textdomain and use the regular gettext approach or b) we consider it a one-time effort and just ask the translators to provide them in any form they want
20151030 03:56:14< aquileia> For reference: http://nsis.sourceforge.net/Creating_language_files_and_integrating_with_MUI#Another_solution:_Native_PO_File_support
20151030 04:05:28-!- TC01 [~quassel@london.acm.jhu.edu] has quit [Remote host closed the connection]
20151030 04:06:03< aquileia> shadowm: I just checked, 1.12.4a is handled as 1.12.4 by the installer, so integer comparision of minor version numbers in the patcher will work without a hitch
20151030 04:09:36-!- TC01 [~quassel@london.acm.jhu.edu] has joined #wesnoth-dev
20151030 04:12:28-!- Appleman1234 [~Appleman1@KD106161091208.au-net.ne.jp] has joined #wesnoth-dev
20151030 04:16:59< aquileia> loonycyborg, shadowm: Should the icon of the installer be the Wesnoth logo or the NSIS logo? I could use ours if desired
20151030 04:19:22< vultraz> former
20151030 04:19:25< vultraz> definitely 
20151030 04:21:19< aquileia> One potential reason against is that currently installer and wesnoth.exe are easily identified by their logo, but I guess most people won't save the installer to their wesnoth dir, so it's probably irrelevant
20151030 04:21:50< aquileia> But I'll add the define to use the Wesnoth logo
20151030 04:22:16< aquileia> Won't push upstram before I hear from loonycyborg though
20151030 04:23:48-!- Shackra [~Jorge@186.177.2.148] has quit [Ping timeout: 272 seconds]
20151030 04:34:33< shadowm> Wesnoth logo, yes.
20151030 04:54:41-!- ancestral [~ancestral@71-220-62-196.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20151030 04:58:15-!- irker891 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20151030 04:58:15< irker891> wesnoth: Ignacio R. Morelle wesnoth:1.12 604d9c53be23 / / (338 files in 28 dirs): pot and documentation update http://git.io/vlfL5
20151030 04:58:19-!- shadowm changed the topic of #wesnoth-dev to: String freeze ACTIVE on 1.12 branch | 1.12.5 release ETA: November 6 00:00 UTC | Wesnoth Developers Channel | >>> Want to help? Go here: http://r.wesnoth.org/t42911 (and thanks!) <<< | Logs: http://irclogs.wesnoth.org | Bug tracker: http://bugs.wesnoth.org
20151030 05:02:16-!- ancestral [~ancestral@71-220-62-196.mpls.qwest.net] has joined #wesnoth-dev
20151030 05:05:40< shadowm> Mail sent.
20151030 05:26:30-!- rayblade53 [~chatzilla@124.109.10.167] has joined #wesnoth-dev
20151030 05:26:49-!- Necrosporus [~Necrospor@unaffiliated/necrosporus] has left #wesnoth-dev ["Leaving"]
20151030 05:29:13-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Disconnected by services]
20151030 05:29:23-!- rayblade53 is now known as vultraz
20151030 05:29:26-!- vultraz [~chatzilla@124.109.10.167] has quit [Changing host]
20151030 05:29:26-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20151030 05:31:12< aquileia> loonycyborg: A bit more scons-y, but still a mess: http://pastebin.com/PRXwcu31
20151030 05:31:52< aquileia> But well, I've got to go now
20151030 05:32:01-!- aquileia [~chatzilla@HSI-KBW-149-172-208-211.hsi13.kabel-badenwuerttemberg.de] has quit [Quit: ChatZilla 0.9.92 [Firefox 41.0.2/20151014143721]]
20151030 05:43:20< irker891> wesnoth: Ignacio R. Morelle wesnoth:master 851eddfd5e40 / data/tools/ (wmlindent wmllint wmlscope): Restore executable bit on WML tools http://git.io/vlfWR
20151030 05:43:39< celticminstrel> Huh, how'd that disappear?
20151030 05:44:09< irker891> wesnoth: Ignacio R. Morelle wesnoth:master ba8da9213511 / data/tools/GUI.pyw: Restore executable bit on the WML tools GUI frontend http://git.io/vlfWF
20151030 05:44:20< shadowm> Elvish_Hunter, probably something about his set-up.
20151030 06:05:47-!- ancestral [~ancestral@71-220-62-196.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20151030 06:16:31< shadowm> "/home/shadowm/src/projects/Invasion_from_the_Unknown/about.cfg", line 45: possible misspelling "English"
20151030 06:16:47< shadowm> ♥
20151030 06:18:11< celticminstrel> ?
20151030 06:20:29< shadowm> wmllint is telling me "English" is not a word.
20151030 06:21:38< celticminstrel> Ahh, heh.
20151030 06:57:12< Aginor> evening
20151030 06:58:33< vultraz> evening 
20151030 06:59:41< Aginor> 'sup?
20151030 07:02:44< vultraz> biology coursework 
20151030 07:04:31-!- travis-ci [~travis-ci@ec2-50-17-32-147.compute-1.amazonaws.com] has joined #wesnoth-dev
20151030 07:04:31< travis-ci> wesnoth/wesnoth#7743 (master - ba8da92 : Ignacio R. Morelle): The build has errored.
20151030 07:04:32< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/88278830
20151030 07:04:32-!- travis-ci [~travis-ci@ec2-50-17-32-147.compute-1.amazonaws.com] has left #wesnoth-dev []
20151030 07:04:43< Aginor> vultraz: :p
20151030 07:08:58< vultraz> Aginor: you?
20151030 07:09:28< Aginor> just had dinner and now I'm back to chasing my elusive crash for a while
20151030 07:15:53-!- celticminstrel is now known as celmin|sleep
20151030 07:29:26-!- aeonskelton [enchilado@defocus/yummy/enchilado] has quit [Ping timeout: 240 seconds]
20151030 07:31:40-!- aeonskelton [enchilado@gateway/shell/blinkenshell.org/x-ixehncnzycbuabuq] has joined #wesnoth-dev
20151030 07:32:07-!- aeonskelton [enchilado@gateway/shell/blinkenshell.org/x-ixehncnzycbuabuq] has quit [Changing host]
20151030 07:32:07-!- aeonskelton [enchilado@defocus/yummy/enchilado] has joined #wesnoth-dev
20151030 07:37:49-!- horrowind1 [~Icedove@2a02:810a:8b00:1c54:21b:fcff:fee3:c3ff] has joined #wesnoth-dev
20151030 07:42:30-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20151030 07:55:48< Aginor> cause of crash found; integer underflow
20151030 07:56:29< shadowm> If it affects 1.12 and master I'm interested to know all the details about it.
20151030 07:56:51< Aginor> sdl_rects use signed ints whereas gui2 uses  unsigned ints, turning all arithmetics into unsigned
20151030 07:56:58< shadowm> (I mean "affect" as in "the issue exists regardless of whether it has visible harmful effects right now".)
20151030 07:57:11< Aginor> sdl1 seems to cope with getting silly values, sdl2 crashes and burns
20151030 07:57:25< Aginor> the issue is there in master, so presumably 1.12
20151030 07:57:44-!- boucman_work [~jrosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20151030 07:57:56< shadowm> So in other words SDL 2 doesn't do bounds checking when processing blits?
20151030 07:58:29< Aginor> explicitly casting the unsigned int to a signed int fixes the crash, but there will be many places in the code having this potential underflow
20151030 08:00:15< Aginor> shadowm: SDL2 does do bounds checking, I'm surprised this is not being picked up
20151030 08:00:25< Aginor> it may be an issue in SDL2_ttf
20151030 08:00:32< Aginor> I have not read that source
20151030 08:01:47< Aginor> hmm
20151030 08:01:58< Aginor> this is gui1, not gui2,, my bad
20151030 08:02:28< Aginor> src/widgets/windget.cpp relies on unsigned, ironically there's no need
20151030 08:03:47< Aginor> I need to mull over this a bit
20151030 08:03:56< Aginor> SDL2 fix is simple, SDL1 fix slightly less so
20151030 08:12:04< Aginor> ironically, fixing the types properly could break UI all over the place where position calculations implicitly depend on unsigned arithmetics without the person writing the code realising
20151030 08:12:26< Aginor> but not fixing the types will continue to court crashes in sdl2
20151030 08:16:10< wedge009> Aginor: First of all, well done. Secondly, I think it's worth fixing even if it means a lot of testing.
20151030 08:16:38< Aginor> I think so too, I am currently doing a fix "plain" fix without ifdefs
20151030 08:16:58< wedge009> And I'll be happy to help with that testing... though perhaps not right away.
20151030 08:17:58< Aginor> the interesting bit will be defining the appropriate behaviour where in the past there could have been overflows on the uint16_t by assigning a uint32_t/uint64_t to it
20151030 08:18:32< Aginor> the appropriate solution might simply be to add a cast inside the method for SDL1 to restore the old behaviour there
20151030 08:19:01< vultraz> really, don't focus too much on sdl1 for master
20151030 08:19:20< Aginor> vultraz: this potentially affects 1.12 too
20151030 08:19:29< Aginor> although "if it aint broken..."
20151030 08:20:04< Aginor> vultraz: I still need to make sure sdl1 works too, we need to have the option of reverting if things go wrong
20151030 08:25:46< Aginor> code's running
20151030 08:25:53< Aginor> and UI is looking sensible
20151030 08:25:59 * Aginor can feel a commit coming on
20151030 08:33:01< vultraz> is this the last issue prior to merge?
20151030 08:37:27< Aginor> vultraz: afaik it's the last crash issue
20151030 08:38:06< Aginor> there's resize issues that aren't resolved (exists on master too but less pronounced) and something is looking odd with the text in gui1 textfields in sdl2
20151030 08:39:18< vultraz> blagh, gui1
20151030 08:39:24< vultraz> it sucks that you have to deal with that POS
20151030 08:39:41< Aginor> vultraz: I don't really have a problem with it
20151030 08:39:48< Aginor> it's pretty straightforward
20151030 08:40:51< Aginor> hmm
20151030 08:41:33< Aginor> I will not make any special fixes for sdl1 here. This code is already flaky in terms of integer sizes and positions so I see no value in trying to fix it up partially
20151030 08:44:26-!- irker891 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20151030 08:46:40-!- irker108 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20151030 08:46:40< irker108> wesnoth: Andreas Löf wesnoth:sdl2 afde826f226a / src/ (7 files in 2 dirs): Fix bug #23908 - Crash in SDL_BlitSurface caused by integer underflow http://git.io/vlJVx
20151030 08:47:18< Aginor> wedge009: I would appreciate it if you could confirm that it is fixed for you too
20151030 08:47:24< Aginor> when time allows :)
20151030 08:47:53< Aginor> shadowm: are you still seeing crashes at game start with sdl2?
20151030 08:49:53< shadowm> I'll check in 18 hours approx.
20151030 08:50:21< Aginor> shadowm: cool, just ping me and provide me with a stack trace if it breaks
20151030 08:50:36< Aginor> I probably won't be around at that time, but I'll look at it when I am
20151030 08:50:43< Aginor> and thanks :)
20151030 08:50:49< shadowm> (I also want to check on my GM45 laptop to see if it makes any difference.)
20151030 08:52:14< Aginor> shadowm: afde826f226a fixes a real bug, but I do not think it's needed in 1.12.x unless someone's experienced problems
20151030 09:02:27 * Aginor is now feeling smug
20151030 09:22:36-!- shadowm_desktop [ignacio@wesnoth/developer/shadowm] has quit [Ping timeout: 255 seconds]
20151030 09:28:38-!- zombah [~zombah@2a02:28:3:1:214:4fff:fe47:5920] has joined #wesnoth-dev
20151030 09:35:21-!- horrowind1 [~Icedove@2a02:810a:8b00:1c54:21b:fcff:fee3:c3ff] has quit [Quit: horrowind1]
20151030 09:38:29-!- yann_ [~dwitch@nan92-1-81-57-214-146.fbx.proxad.net] has joined #wesnoth-dev
20151030 09:40:04-!- yann_ is now known as yann
20151030 09:54:59< irker108> wesnoth: Andreas Löf wesnoth:sdl2 55f7b28720f2 / src/widgets/textbox.cpp: Fix alpha blending of text in gui1 textbox http://git.io/vlUEC
20151030 09:59:06-!- aquileia [~chatzilla@eduroam172-145.wlan.uni-ulm.de] has joined #wesnoth-dev
20151030 10:04:57< aquileia> wedge009: I looked at your vcproj backport - none of the changes are problematic, I'd even suggest to expand the PR to basically backport the entire first half of the project files (everything but the File nodes)
20151030 10:07:06< aquileia> A quick glance shows me that packaging/windows/wesnoth.rc needs to be backported as well
20151030 10:07:58-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Quit: wedge009]
20151030 10:08:19-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20151030 10:09:27< aquileia> shadowm: ^ The 1.12 resource file is basically blank...
20151030 10:11:33-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Client Quit]
20151030 10:13:25-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20151030 10:15:46-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20151030 10:15:56< aquileia> wedge009, gfgtdf, anonymissimus: I'd like to do a drastic change in the vcproj structure - move basically all code generation to wesnothlib. This would allow to execute the boost unit tests and link the binary in separate projects, which means we no longer need those Test_* configurations. The disadvantage of configurations is that they don't share .obj files, so we currently have to compile...
20151030 10:15:58< aquileia> ...twice to get the boost tests
20151030 10:16:39< aquileia> Feedback & other suggestions are welcome
20151030 10:17:26-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 240 seconds]
20151030 10:17:26-!- wedge010 is now known as wedge009
20151030 10:17:50< aquileia> It would also allow to execute the boost unit tests automatically with every build, just as with the WML tests
20151030 10:24:35< aquileia> Other news: I'm removing Release Debug and instead generate a .pdb for Release (the binary remains basically untouched as the pdb contains all the debug information)
20151030 10:28:45< aquileia> The Debug build is noticeably faster, though still sluggish (see https://randomascii.wordpress.com/2011/07/22/visual-c-debug-buildsfast-checks-cause-5x-slowdowns/ )
20151030 10:38:09-!- zombah [~zombah@2a02:28:3:1:214:4fff:fe47:5920] has quit [Quit: Leaving]
20151030 10:38:46-!- zombah [~zombah@2a02:28:3:1:214:4fff:fe47:5920] has joined #wesnoth-dev
20151030 10:43:46-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection]
20151030 10:43:52-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev
20151030 10:57:42< aquileia> anyone: When did someone last use the schema-generator tool?
20151030 11:20:09< loonycyborg> aquileia: uhh, just write patch generation in fully procedural way(without using any scons call) and confirm it works, then based on it I could show you how to decompose it into scons rules
20151030 11:20:30-!- aquileia [~chatzilla@eduroam172-145.wlan.uni-ulm.de] has quit [Ping timeout: 255 seconds]
20151030 11:41:42-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20151030 11:45:22< wedge009> aquileia: I'm using VC14/2015, but used the VC9 files as a base when I started working on the 1.12 branch recently. I didn't want to disrupt anything so I kept changes to a minimum, only changing things I felt were necessary to get it working for me. eg I don't know what difference it will make (for SDL?) to change the wesnothd subsystem from Windows to Console.
20151030 11:46:23< wedge009> aquileia: I don't mind what you do with the project files... though porting any changes to my VC14 working copy could be a pain. I don't use schema-generator because it's not enabled in the build settings and I don't know what it does anyway.
20151030 11:46:52< loonycyborg> oh crud. logging functionality depends on SDL
20151030 11:48:31< wedge009> aquileia: I'll see about back-porting the rest of the project file stuff for 1.12, that I'm reasonably confident of not breaking anything. Feel free check the remainder yourself.
20151030 11:49:55< loonycyborg> it uses SDL_timer.h
20151030 11:50:14< loonycyborg> guess I'll need to change it to use some other time library..
20151030 12:00:52< wedge009> aquileia: Done.
20151030 12:09:25< wedge009> Aginor: You deserve to feel smug in this case, mission seemingly accomplished. Tested the choose-faction-MP-join-screen in Windows and Linux and no trouble on either. I think aside from the resizing issues (which were always a problem even in SDL1.2 builds) I think we're just about done!
20151030 12:09:54< wedge009> Aginor: Are you planning on any more work for this underflow bug? What other things might need to be tested/changed?
20151030 12:11:31-!- aquileia [~chatzilla@eduroam172-145.wlan.uni-ulm.de] has joined #wesnoth-dev
20151030 12:21:41-!- aquileia [~chatzilla@eduroam172-145.wlan.uni-ulm.de] has quit [Ping timeout: 265 seconds]
20151030 12:23:59-!- aquileia [~chatzilla@eduroam172-145.wlan.uni-ulm.de] has joined #wesnoth-dev
20151030 12:29:38-!- Appleman1234 [~Appleman1@KD106161091208.au-net.ne.jp] has quit [Ping timeout: 250 seconds]
20151030 12:50:59-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection]
20151030 12:51:05-!- janebot_ [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev
20151030 12:52:56-!- louis94 [~~louis94@109.129.245.154] has joined #wesnoth-dev
20151030 12:55:16-!- irker108 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20151030 13:09:52< celmin|sleep> There's no such thing as "integer underflow".
20151030 13:10:05< celmin|sleep> Underflow is something that applies only to floating-point numbers.
20151030 13:12:49< celmin|sleep> It's when the number is too small to represent and becomes zero.
20151030 13:26:00-!- Appleman1234 [~Appleman1@KD111239008011.au-net.ne.jp] has joined #wesnoth-dev
20151030 13:26:29-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20151030 13:43:28-!- abhi [b4d68483@gateway/web/freenode/ip.180.214.132.131] has joined #wesnoth-dev
20151030 13:54:17-!- prkc [~prkc@51B78F27.dsl.pool.telekom.hu] has joined #wesnoth-dev
20151030 14:04:38-!- abhi [b4d68483@gateway/web/freenode/ip.180.214.132.131] has quit [Quit: Page closed]
20151030 14:09:06-!- prkc [~prkc@51B78F27.dsl.pool.telekom.hu] has quit [Read error: Connection reset by peer]
20151030 14:16:43-!- prkc [~prkc@51B78F27.dsl.pool.telekom.hu] has joined #wesnoth-dev
20151030 14:44:35-!- louis94 [~~louis94@109.129.245.154] has quit [Ping timeout: 264 seconds]
20151030 15:03:23-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: mattsc]
20151030 15:12:14-!- ancestral [~ancestral@71-220-62-196.mpls.qwest.net] has joined #wesnoth-dev
20151030 15:13:46< aquileia> What is loadscreen_empty.cpp used for, if at all? Other OS's?
20151030 15:15:02-!- gfgtdf [~chatzilla@f054131148.adsl.alicedsl.de] has joined #wesnoth-dev
20151030 15:15:30-!- nurupo [~nurupo.ga@unaffiliated/nurupo] has quit [Ping timeout: 272 seconds]
20151030 15:15:55< gfgtdf> aquileia: isnt wesnothlib also user by wesnothd ?
20151030 15:17:10< aquileia> Yes, but it shouldn't hurt build performance too much to go through a larger lib. I doubt anyone wants to build wesnothd without wesnoth
20151030 15:17:43< gfgtdf> aquileia: you want to put wesnothd and wesnoth.exe ina  single executable ?
20151030 15:18:08< aquileia> No
20151030 15:18:19< gfgtdf> aquileia: so what exactly do you want to do ?
20151030 15:19:07< aquileia> I want to move 99% of the source files of wesnoth.vcproj to wesnothlib.vcproj
20151030 15:19:33< aquileia> wesnoth.vcproj only compiles wesnoth.cpp & wesnoth.rc and gets everything else from the lib
20151030 15:19:38< gfgtdf> aquileia: i think that would collide with the plan to move the sdl dependency out of wesnothd
20151030 15:21:04< aquileia> As the official builds don't use MSVC, the vcproj doesn't really matter
20151030 15:21:38< aquileia> But we can of course split off the wesnothd-relevant subset if preferred
20151030 15:22:54< aquileia> That can come in a later fixup commit though
20151030 15:22:56< gfgtdf> aquileia: that subset woudl then be what wesnothlib currently is ?
20151030 15:23:48< gfgtdf> aquileia: wouldnt it be easier to just split the wesnoth project ?
20151030 15:24:19< celmin|sleep> I think aquileia's plan is a good one.
20151030 15:24:25-!- celmin|sleep is now known as celticminstrel
20151030 15:24:32< celticminstrel> More or less.
20151030 15:24:45< aquileia> according to anonymissimus (back when I took over maintenance) wesnothlib isn't actually the lowest common denominator but just a bunch of files  for basic API that were supposed to change less frequently
20151030 15:25:23< celticminstrel> With source files that are used by both wesnoth and wesnothd, it's better to include them in a static lib than in two separate projects, to avoid compiling them twice.
20151030 15:25:37-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20151030 15:25:41< celticminstrel> That said, wesnothd only has around 40 files while the game has around 600.
20151030 15:26:23< celticminstrel> I think it would be silly to put everything in wesnothlib, because that could bloat the wesnothd.
20151030 15:26:37< gfgtdf> wasnt that aquileias plan ?
20151030 15:26:38< celticminstrel> ...though, maybe unused symbols would be eliminated at link time?
20151030 15:26:51< celticminstrel> I'm not quite sure about that.
20151030 15:26:59< aquileia> yes, there's a linker option for that /OPT:REF
20151030 15:29:19< gfgtdf> loonycyborg: when replacing the SDL_GetTicks() in log.cpp you should also check that it performs good, for example the print_precise_timestamp() function (in log.cpp) is quiet slow compared to the SDL_GetTicks() based version
20151030 15:29:57< gfgtdf> loonycyborg: hmm wait
20151030 15:29:59< aquileia> The link time I just experienced for my first try of the new configuration is insanely long, but then I'm testing it on VC9. I might find some options to get it down
20151030 15:30:00< celticminstrel> I guess a bloated wesnothd wouldn't be a problem with debug builds.
20151030 15:30:35< celticminstrel> I do still think it's silly to put everything except for main in a wesnothlib project.
20151030 15:31:05< gfgtdf> loonycyborg: hmm it seems like log.cpp uses time(NULL) AND SDL_GetTicks() hmm
20151030 15:31:15< celticminstrel> But I guess anything that seems lib-like is fine?
20151030 15:32:11< celticminstrel> For example, GUI2, Lua (as in src/lua, not src/scripting), the stuff that interfaces directly with SDL, etc
20151030 15:32:14< loonycyborg> gfgtdf: it seems to use Get_Ticks only to get a time interval, guess it can be replaced with difftime
20151030 15:32:31< aquileia> celticminstrel: The motivation is to simplify the boost test execution - if the code isn't bundled in a .lib, we have to link wesnoth_test against every file it uses
20151030 15:32:44< celticminstrel> Ooh.
20151030 15:32:49< celticminstrel> I see your point.
20151030 15:33:05< celticminstrel> The tests use a ton of files not used by the server.
20151030 15:33:07< gfgtdf> loonycyborg: i tzhink difftime is onyl seconds while Get_Ticks is milliseconds ?
20151030 15:33:49< celticminstrel> Actually, do the tests use every file?
20151030 15:34:00< loonycyborg> seems so indeed
20151030 15:34:14< celticminstrel> In XCode, wesnoth has 578 source files while the unit tests have 592.
20151030 15:34:50< aquileia> No, they don't test everything, but anonymissimus mentioned 400 linker errors when trying the naive implementation
20151030 15:35:00< aquileia> So it's not far off I guess
20151030 15:35:09< gfgtdf> aquileia: cant we just maybe  have a commandline flag to run the tests witgh the normal executable? That way user coudl alos rin the boost test if they only have the normal executable
20151030 15:35:23< celticminstrel> If I consider the files in common, there's 577 of them, so it does look like they need everything.
20151030 15:35:38< celticminstrel> Though it's possible some of the linked files are not actually used.
20151030 15:35:57< aquileia> gfgtdf: Boost compiles it's own .exe from the source files, it can't test existing binaries
20151030 15:36:25< celticminstrel> gfgtdf: The Boost testing framework provides its own main().
20151030 15:36:32< aquileia> celticminstrel: src/lua is in a separate lualib project, forgot to mention that
20151030 15:36:39< celticminstrel> Really?
20151030 15:36:42< gfgtdf> celticminstrel: maybe it provides some way zo overwrite it ?
20151030 15:36:52< gfgtdf> to*
20151030 15:36:56< celticminstrel> I should probably consider refactoring the XCode project to use some libs too.
20151030 15:37:02< celticminstrel> gfgtdf: Possibly, not sure.
20151030 15:37:11< aquileia> s/lualib/liblua
20151030 15:37:20< celticminstrel> gfgtdf: However, I think it's silly to put the unit tests in the main executable.
20151030 15:37:27< gfgtdf> celticminstrel: why?
20151030 15:37:38< celticminstrel> I don't think there are any users of the game who will want to run the unit tests.
20151030 15:37:47< celticminstrel> The unit tests are solely for developers.
20151030 15:37:49< aquileia> CMake makes use of extensive source grouping (basically libs) IIUC
20151030 15:38:11< gfgtdf> celticminstrel: well if they have some bug it might be useful
20151030 15:38:28< celticminstrel> How?
20151030 15:38:46< aquileia> gfgtdf: The WML tests, sure, but not the C++ tests
20151030 15:39:08< celticminstrel> Well, those are run with the main executable though.
20151030 15:39:10< gfgtdf> why shoudl zhe wml test make more sense then the cüü tests? :o
20151030 15:40:23< aquileia> Any bug the c++ tests find should surface on the dev's system as well
20151030 15:41:45< gfgtdf> aquileia: uhm why? the devepoers system migth be different
20151030 15:42:08< celticminstrel> I think the point is more...
20151030 15:42:28< celticminstrel> If someone has a bug, it's probably something the C++ tests haven't accounted for.
20151030 15:44:02< gfgtdf> celticminstrel: i'd to look at every test to check that....  . But on the other side i dont see advantages of having it on 2 exectuables.
20151030 15:44:07< aquileia> If you're only testing function return values (correct me if I'm wrong), the system they run on doesn't interfere as much, so as celticminstrel said, these bugs will be found before releasing and the users will find other bugs
20151030 15:44:32< celticminstrel> There's no advantage to combining them.
20151030 15:47:48-!- louis94 [~~louis94@109.129.245.154] has joined #wesnoth-dev
20151030 15:48:40< aquileia> loonycyborg: Would a powershell script be ok for the prototype?
20151030 15:49:57< gfgtdf> aquileia: well these function can call whaever specialyl the qui2 and filesystem tests to test sureley interfere with the system
20151030 15:50:37< aquileia> I can't find a 'split' equivalent for .cmd, and .sh isn't an option on Windows
20151030 15:51:05< celticminstrel> Unless you're targeting MinGW.
20151030 15:51:23< aquileia> Oh, right, he has the MinGW shell
20151030 15:51:28< celticminstrel> It's probably better if you can do it as .bat
20151030 15:51:53< celticminstrel> Is PowerShell installed by default on Windows systems though? If so, maybe that would be okay.
20151030 15:51:56< aquileia> Well, it's only a prototype which should be merged into scons later on
20151030 15:52:04< celticminstrel> Oh.
20151030 15:52:26< celticminstrel> Sorry, I don't actually know what you're talking about, so take what I said with a grain of salt. >_>
20151030 15:52:54< aquileia> I'm grateful for a second opinion
20151030 15:56:41< aquileia> I don't really know what to use - GitPython, PowerShell, cmd, sh, ...
20151030 15:57:16< loonycyborg> aquileia: python would be the best
20151030 16:16:14-!- gfgtdf [~chatzilla@f054131148.adsl.alicedsl.de] has quit [Quit: ChatZilla 0.9.92 [Firefox 41.0.2/20151014143721]]
20151030 16:51:43< aquileia> loonycyborg: a bit of progress: http://pastebin.com/nir7UkHv
20151030 17:02:25-!- Shackra [~Jorge@186.177.2.148] has joined #wesnoth-dev
20151030 17:03:40-!- boucman_work [~jrosen@wesnoth/developer/boucman] has quit [Ping timeout: 244 seconds]
20151030 17:04:11-!- ancestral [~ancestral@71-220-62-196.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20151030 17:06:26-!- Appleman1234 [~Appleman1@KD111239008011.au-net.ne.jp] has quit [Ping timeout: 240 seconds]
20151030 17:11:16-!- ancestral [~ancestral@71-220-62-196.mpls.qwest.net] has joined #wesnoth-dev
20151030 17:17:28-!- mjs-de [~mjs-de@f049064004.adsl.alicedsl.de] has joined #wesnoth-dev
20151030 17:23:00< aquileia> I want to read subprocess.Popen().stdout line by line, but it doesn't seem to wait until the command finished. Any ideas that retain the output?
20151030 17:23:36< celticminstrel> What doesn't wait?
20151030 17:25:25< aquileia> celticminstrel: http://pastebin.com/M4FjAcvb - I don't get any output 
20151030 17:26:26< aquileia> I need to parse the full stdout stream
20151030 17:30:54< celticminstrel> Not sure.
20151030 17:33:37-!- ancestral [~ancestral@71-220-62-196.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20151030 17:36:07< loonycyborg> aquileia: you need to explicitly wait for it to inish
20151030 17:36:10< loonycyborg> finish
20151030 17:36:38< loonycyborg> iirc with communicate() method
20151030 17:37:06< aquileia> I tried proc.communicate and it didn't seem to work though... maybe I misused it
20151030 17:37:47< aquileia> ValueError: I/O operation on closed file
20151030 17:38:51< loonycyborg> https://docs.python.org/2/library/subprocess.html?highlight=popen#subprocess.Popen.communicate
20151030 17:39:12-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20151030 17:39:38< aquileia> Ah, ok
20151030 17:39:45< o11c> aquileia: you might need to change the buffering mode
20151030 17:39:49-!- Kwandulin [~Miranda@p200300760F18C0D30D2DC345679F3A54.dip0.t-ipconnect.de] has joined #wesnoth-dev
20151030 17:40:14< o11c> by default, reading from the inferior will only return what is currently available, which might be incomplete
20151030 17:40:46< loonycyborg> aquileia: there are some examples of using communicate in that page
20151030 17:42:00< o11c> communicate forces slurping though
20151030 17:42:06< o11c> which is not nice with large data
20151030 17:44:30< aquileia> Well, proc.communicate()[0] is empty... I'm still doing sth. wrong
20151030 17:45:32-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Ping timeout: 250 seconds]
20151030 17:47:48< o11c> aquileia: make sure you're not getting errors ...
20151030 17:48:17< celticminstrel> Check the return code.
20151030 17:48:25< aquileia> ... you're right, my command isn't parsed correctly
20151030 17:48:32< celticminstrel> Or stderr even.
20151030 17:48:38< celticminstrel> Oh...
20151030 17:50:00-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20151030 18:03:58-!- Appleman1234 [~Appleman1@KD111239024039.au-net.ne.jp] has joined #wesnoth-dev
20151030 18:04:23< aquileia> Found the bug - 'data fonts images doc/manual' is passed as a single CL option, and git searches for a file with exactly that name
20151030 18:04:46< aquileia> Splitting the option fixes it
20151030 18:08:03-!- iceiceice [~chris@50.245.222.235] has joined #wesnoth-dev
20151030 18:08:03-!- iceiceice [~chris@50.245.222.235] has quit [Changing host]
20151030 18:08:03-!- iceiceice [~chris@wesnoth/developer/iceiceice] has joined #wesnoth-dev
20151030 18:09:07< celticminstrel> Oh, yeah.
20151030 18:09:40< celticminstrel> When calling exec in that way, shell parsing is not applied.
20151030 18:10:57< celticminstrel> ...well, exec doesn't apply shell parsing at all, but Python's wrapper offers the option to do so.
20151030 18:15:24-!- louis94 [~~louis94@109.129.245.154] has quit [Quit: Konversation terminated!]
20151030 18:32:48< aquileia> celticminstrel, o11c, loonycyborg: Thanks, it works now! http://pastebin.com/HZhzhym0
20151030 18:33:19< loonycyborg> cool
20151030 18:33:25< celticminstrel> Is there a reason why the list of files is commented out?
20151030 18:34:32< aquileia> The full patch is pretty big, so I commented out part of it
20151030 18:38:46< aquileia> Fun fact - the 7z version shipped with git is ancient, but comes first in PATH, so I used 7z.exe instead of 7za.exe
20151030 18:39:05< aquileia> Thanks again and bye
20151030 18:39:44-!- aquileia [~chatzilla@eduroam172-145.wlan.uni-ulm.de] has quit [Quit: ChatZilla 0.9.92 [Firefox 41.0.2/20151014143721]]
20151030 18:59:52-!- Kwandulin [~Miranda@p200300760F18C0D30D2DC345679F3A54.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20151030 19:15:28-!- nurupo [~nurupo.ga@unaffiliated/nurupo] has joined #wesnoth-dev
20151030 19:34:20-!- Ivanovic_ [~ivanovic@p4FC53FEB.dip0.t-ipconnect.de] has joined #wesnoth-dev
20151030 19:34:20-!- Ivanovic_ [~ivanovic@p4FC53FEB.dip0.t-ipconnect.de] has quit [Changing host]
20151030 19:34:20-!- Ivanovic_ [~ivanovic@wesnoth/developer/ivanovic] has joined #wesnoth-dev
20151030 19:34:47-!- Ivanovic [~ivanovic@wesnoth/developer/ivanovic] has quit [Ping timeout: 256 seconds]
20151030 19:36:47-!- Ivanovic [~ivanovic@p4FC53FED.dip0.t-ipconnect.de] has joined #wesnoth-dev
20151030 19:37:19-!- Ivanovic [~ivanovic@p4FC53FED.dip0.t-ipconnect.de] has quit [Changing host]
20151030 19:37:19-!- Ivanovic [~ivanovic@wesnoth/developer/ivanovic] has joined #wesnoth-dev
20151030 19:37:22-!- EliDupree [~quassel@idupree.com] has quit [Remote host closed the connection]
20151030 19:37:37-!- EliDupree [~quassel@idupree.com] has joined #wesnoth-dev
20151030 19:39:10-!- Ivanovic_ [~ivanovic@wesnoth/developer/ivanovic] has quit [Ping timeout: 240 seconds]
20151030 19:43:22-!- EliDupree [~quassel@idupree.com] has quit [Remote host closed the connection]
20151030 19:43:29-!- EliDupree [~quassel@idupree.com] has joined #wesnoth-dev
20151030 19:46:15-!- EliDupree [~quassel@idupree.com] has quit [Remote host closed the connection]
20151030 19:46:22-!- EliDupree [~quassel@idupree.com] has joined #wesnoth-dev
20151030 19:51:14-!- EliDupree [~quassel@idupree.com] has quit [Remote host closed the connection]
20151030 19:51:24-!- EliDupree [~quassel@idupree.com] has joined #wesnoth-dev
20151030 20:00:57-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20151030 20:13:04-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: So long and thanks for all the fish.]
20151030 20:28:12-!- EliDupree [~quassel@idupree.com] has quit [Remote host closed the connection]
20151030 20:28:19-!- EliDupree [~quassel@idupree.com] has joined #wesnoth-dev
20151030 20:28:45-!- EliDupree [~quassel@idupree.com] has quit [Remote host closed the connection]
20151030 20:28:52-!- EliDupree [~quassel@idupree.com] has joined #wesnoth-dev
20151030 20:28:58-!- EliDupree [~quassel@idupree.com] has quit [Remote host closed the connection]
20151030 20:29:08-!- EliDupree [~quassel@idupree.com] has joined #wesnoth-dev
20151030 20:29:52-!- EliDupree [~quassel@idupree.com] has quit [Remote host closed the connection]
20151030 20:29:59-!- EliDupree [~quassel@idupree.com] has joined #wesnoth-dev
20151030 20:36:33-!- EliDupree [~quassel@idupree.com] has quit [Remote host closed the connection]
20151030 20:36:49-!- EliDupree [~quassel@idupree.com] has joined #wesnoth-dev
20151030 20:39:46-!- EliDupree [~quassel@idupree.com] has quit [Remote host closed the connection]
20151030 20:39:53-!- EliDupree [~quassel@idupree.com] has joined #wesnoth-dev
20151030 20:39:55-!- EliDupree [~quassel@idupree.com] has quit [Remote host closed the connection]
20151030 20:40:02-!- EliDupree [~quassel@idupree.com] has joined #wesnoth-dev
20151030 20:42:41-!- gfgtdf [~chatzilla@x55b19e79.dyn.telefonica.de] has joined #wesnoth-dev
20151030 20:50:13-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20151030 21:09:43-!- horrowind [~Icedove@2a02:810a:8b00:1c54:21b:fcff:fee3:c3ff] has joined #wesnoth-dev
20151030 21:15:27< Aginor> celticminstrel: you're right, the correct terminology would have been overflow
20151030 21:50:07-!- ancestral [~ancestral@63.92.240.233] has joined #wesnoth-dev
20151030 21:52:01-!- ypnos [~ypnos@lme51.informatik.uni-erlangen.de] has quit [Ping timeout: 246 seconds]
20151030 21:57:23< celticminstrel> Ugh.
20151030 21:57:57< celticminstrel> The dynamic link libraries use @executable_path.
20151030 21:59:48-!- ancestral [~ancestral@63.92.240.233] has quit [Quit: i go nstuf kthxbai]
20151030 22:00:36< celticminstrel> Either I can make them use @rpath (and set the search paths for that) or copy the libraries to ../Frameworks.
20151030 22:01:08< celticminstrel> I don't like the latter, but I'm not sure how to do the former (other than with post-build use of install_name_tool)
20151030 22:02:26< celticminstrel> Another option is building the unit tests straight to the application package, but that'll probably confuse people since XCode can't comprehend the idea that the products go anywhere other than the official build products directory (at least in my version, maybe that's improved in later versions).
20151030 22:08:42-!- daMark [5439aa5b@gateway/web/freenode/ip.84.57.170.91] has joined #wesnoth-dev
20151030 22:15:49-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20151030 22:17:51-!- ypnos [~ypnos@lme51.informatik.uni-erlangen.de] has joined #wesnoth-dev
20151030 22:18:27-!- horrowind [~Icedove@2a02:810a:8b00:1c54:21b:fcff:fee3:c3ff] has quit [Quit: horrowind]
20151030 22:18:58-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 250 seconds]
20151030 22:18:58-!- wedge010 is now known as wedge009
20151030 22:21:40-!- louis94 [~~louis94@109.129.245.154] has joined #wesnoth-dev
20151030 22:24:19< celticminstrel> I guess I'll have to go with #2.
20151030 22:29:26-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20151030 22:35:58-!- ancestral [~ancestral@83.sub-70-197-204.myvzw.com] has joined #wesnoth-dev
20151030 22:36:15< celticminstrel> Ugh, that doesn't even work for some reason...
20151030 22:37:58< celticminstrel> Maybe the only way is to get them to use @rpath... which I think means editing all the shared libraries to use @rpath... :|
20151030 22:38:23< celticminstrel> And I'm still not sure how to make the main executable reference libraries with @rpath. Unless that's also something in the library itself.
20151030 22:42:11< celticminstrel> ...what? Cairo and expat are referencing Macports-installed libs by the absolute path. :|
20151030 22:42:41< celticminstrel> ...though, that could just be me...
20151030 22:42:58< celticminstrel> Several others are doing that too... and I think I replaced some with MacPorts versions...
20151030 22:47:47< celticminstrel> mattsc: Any objections to adding the four Pango modules to the libs download, with instructions to put them in Resources/pango/1.8.0/modules?
20151030 22:50:18< celticminstrel> Also, what do you think about changing the install names in all the libraries to use @rpath instead of @executable_path/../Frameworks? I could imagine it making things harder for someone who for some reason can't use the downloadable libs archive, but I think it would make things quite a bit easier otherwise.
20151030 22:50:23< celticminstrel> Regarding the unit tests.
20151030 22:51:01< celticminstrel> I think it's currently impossible to run the unit tests from within XCode. The only way to run them would be to move the executable into the MacOS folder and call it from the command-line.
20151030 22:51:34< celticminstrel> (The same is also true of the server, by the way.)
20151030 22:53:28-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20151030 22:56:30-!- ancestral [~ancestral@83.sub-70-197-204.myvzw.com] has quit [Read error: Connection reset by peer]
20151030 23:12:18-!- DDR is now known as DDR_machine
20151030 23:15:00-!- daMark [5439aa5b@gateway/web/freenode/ip.84.57.170.91] has left #wesnoth-dev []
20151030 23:16:46-!- Appleman1234 [~Appleman1@KD111239024039.au-net.ne.jp] has quit [Ping timeout: 260 seconds]
20151030 23:55:25< mattsc> celticminstrel: I have no objection to any of that, nor do I really have an opinion on it
20151030 23:55:58< mattsc> In fact, I wouldn’t object to you taking over responsibility for all this
20151030 23:56:23< mattsc> since you clearly know a lot more about it than I do (as I’ve said before, I think)
--- Log closed Sat Oct 31 00:00:10 2015