--- Log opened Thu May 29 00:00:40 2014
20140529 00:03:22-!- c74d [~c74d3a4eb@2002:4404:712c:0:76de:2bff:fed4:2766] has quit [Remote host closed the connection]
20140529 00:04:56-!- gfgtdf [~chatzilla@e176189241.adsl.alicedsl.de] has quit [Quit: ChatZilla 0.9.90.1 [Firefox 29.0.1/20140506152807]]
20140529 00:06:31-!- c74d [~c74d3a4eb@2002:4404:712c:0:76de:2bff:fed4:2766] has joined #wesnoth-dev
20140529 00:12:56-!- Ivanovic_ [~ivanovic@frnk-5f74dbd9.pool.mediaWays.net] has joined #wesnoth-dev
20140529 00:12:56-!- Ivanovic_ [~ivanovic@frnk-5f74dbd9.pool.mediaWays.net] has quit [Changing host]
20140529 00:12:56-!- Ivanovic_ [~ivanovic@wesnoth/developer/ivanovic] has joined #wesnoth-dev
20140529 00:16:55-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20140529 00:17:00-!- ShikadiLord [~ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20140529 00:17:30-!- shadowm [~ignacio@wesnoth/developer/shadowmaster] has quit [Killed (verne.freenode.net (Nickname regained by services))]
20140529 00:17:30-!- ShikadiLord is now known as shadowm
20140529 00:17:51-!- iceiceice_ [~chris@cpe-66-108-20-80.nyc.res.rr.com] has joined #wesnoth-dev
20140529 00:18:19-!- shadowm_desktop [ignacio@wesnoth/developer/shadowmaster] has quit []
20140529 00:18:35-!- shadowm_desktop [ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20140529 00:19:00-!- Ivanovic [~ivanovic@wesnoth/developer/ivanovic] has quit [Write error: Broken pipe]
20140529 00:19:18-!- [Relic] [~relic@99-58-54-211.lightspeed.milwwi.sbcglobal.net] has quit [Quit: Leaving]
20140529 00:20:55-!- Ivanovic_ is now known as Ivanovic
20140529 00:21:22-!- kex [~kex@89.205.75.19] has quit [Remote host closed the connection]
20140529 00:33:44-!- ancestral [~ancestral@17.114.45.98] has quit [Quit: ancestral]
20140529 00:55:38-!- ancestral [~ancestral@12.23.74.29] has joined #wesnoth-dev
20140529 00:56:01-!- tomreyn [~tomreyn@megaglest/team/tomreyn] has quit [Quit: tomreyn]
20140529 01:00:41< aquileia> iceiceice_: I decided for the PR as there are 5 commits now
20140529 01:00:45-!- kex [~kex@89.205.75.19] has joined #wesnoth-dev
20140529 01:02:02< aquileia> Feel free to merge #173, it's feature-complete now
20140529 01:09:32< aquileia> anonymissimus: BTW, are those Test_Debug & Test_Release configurations useful at all? As wesnothlib doesn't have them, they can't be selected anyhow
20140529 01:11:10< aquileia> I applied my changes to them as well for now, but I'd prefer to clean them up if you don't object
20140529 01:18:40-!- groggy [~chatzilla@75-131-166-139.dhcp.spbg.sc.charter.com] has joined #wesnoth-dev
20140529 01:21:51-!- Necrosporus_ [~Necrospor@unaffiliated/necrosporus] has joined #wesnoth-dev
20140529 01:24:41-!- Necrosporus [~Necrospor@unaffiliated/necrosporus] has quit [Ping timeout: 255 seconds]
20140529 01:33:26-!- groggy [~chatzilla@75-131-166-139.dhcp.spbg.sc.charter.com] has quit [Remote host closed the connection]
20140529 01:39:37-!- un214 [~un214@2602:306:ce78:1ac9:224:8cff:fed2:ef57] has joined #wesnoth-dev
20140529 01:39:44< RiftWalker> Neat note about std::map, its operator[] default constructs the value. So for preproc_map, calling, e.g. defines_map_[define] is the same as calling defines_map_[define] = preproc_define();
20140529 01:44:46-!- [Relic] [~relic@99-58-54-211.lightspeed.milwwi.sbcglobal.net] has joined #wesnoth-dev
20140529 01:50:17-!- un214 [~un214@2602:306:ce78:1ac9:224:8cff:fed2:ef57] has quit [Read error: Connection reset by peer]
20140529 01:52:48-!- aquileia [2edf524c@gateway/web/freenode/ip.46.223.82.76] has quit [Ping timeout: 240 seconds]
20140529 02:12:27-!- prophile [~alynn@oftn/member/prophile] has joined #wesnoth-dev
20140529 02:16:48-!- gfgtdf [~chatzilla@f054158248.adsl.alicedsl.de] has joined #wesnoth-dev
20140529 02:17:09< gfgtdf> is there a reason why we only allow stop network processing for observers ?
20140529 02:18:02< gfgtdf> i mean it could also be usful for players to watch their enemies movements, and press "stop network processing" while they are afk
20140529 02:18:50< gfgtdf> iceiceice_, happygrue : ^
20140529 02:23:34< irker318> wesnoth: gfgtdf wesnoth:master 73de2e684a5d / src/ (8 files): move is_host_ to playturn.hpp http://git.io/KTjhvQ
20140529 02:23:36< irker318> wesnoth: gfgtdf wesnoth:master dce39b93e149 / src/ (5 files): add maybe_linger to playsingle/mp_controller http://git.io/7ITSUw
20140529 02:28:54< happygrue> gfgtdf: I don't know about the logistics, but that sounds like it would be useful to some users
20140529 02:29:12< happygrue> though I wouldn't expect it to be used a lot either.
20140529 02:33:15< gfgtdf> hm ok i wait until i am in the situation when i need it and think about it again then
20140529 02:37:07-!- happygrue [~happygrue@wesnoth/developer/wintermute] has quit [Ping timeout: 240 seconds]
20140529 02:45:11< irker318> wesnoth: gfgtdf wesnoth:master 0ea2ef482121 / src/playturn.cpp: remove outdated code. http://git.io/3kiORQ
20140529 02:46:06-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20140529 02:52:22-!- shadowm_desktop [ignacio@wesnoth/developer/shadowmaster] has quit [Ping timeout: 245 seconds]
20140529 02:55:26-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20140529 03:01:20-!- shadowm_desktop [ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20140529 03:02:48-!- riksteri [~riksteri@dsl-tkubrasgw3-54f96b-216.dhcp.inet.fi] has joined #wesnoth-dev
20140529 03:10:30-!- gfgtdf [~chatzilla@f054158248.adsl.alicedsl.de] has quit [Quit: ChatZilla 0.9.90.1 [Firefox 29.0.1/20140506152807]]
20140529 03:26:26-!- prophile [~alynn@oftn/member/prophile] has quit [Quit: The Game]
20140529 03:26:50-!- c74d [~c74d3a4eb@2002:4404:712c:0:76de:2bff:fed4:2766] has quit [Excess Flood]
20140529 03:31:50-!- c74d [~c74d3a4eb@2002:4404:712c:0:76de:2bff:fed4:2766] has joined #wesnoth-dev
20140529 03:49:01-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20140529 03:57:22-!- shadowm_desktop [ignacio@wesnoth/developer/shadowmaster] has quit [Ping timeout: 245 seconds]
20140529 04:01:31-!- kex [~kex@89.205.75.19] has quit [Remote host closed the connection]
20140529 04:02:07-!- kex [~kex@89.205.75.19] has joined #wesnoth-dev
20140529 04:06:18-!- kex [~kex@89.205.75.19] has quit [Ping timeout: 240 seconds]
20140529 04:18:28-!- shadowm_desktop [ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20140529 04:25:18-!- Gambit [~derek@wesnoth/developer/grickit] has quit [Remote host closed the connection]
20140529 04:57:22-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20140529 05:31:39< irker318> wesnoth: aquileia wesnoth:master 483272456571 / projectfiles/VC9/ (WindowsTimeout.cpp WindowsTimeout.vcproj wesnoth.sln): Windows WML tests: Add WindowsTimeout project https://github.com/wesnoth/wesnoth/commit/4832724565715e7454d0f972f58bd242c235e609
20140529 05:31:41< irker318> wesnoth: aquileia wesnoth:master 8d91cef48879 / projectfiles/VC9/WML_tests.cmd: Windows WML tests: Port .sh test script to .cmd http://git.io/DApRkg
20140529 05:31:43< irker318> wesnoth: aquileia wesnoth:master f90569ccee06 / projectfiles/VC9/WML_tests.cmd: Windows WML tests: Improve stdout output http://git.io/PTeoFQ
20140529 05:31:45< irker318> wesnoth: aquileia wesnoth:master 412aca56d770 / projectfiles/VC9/WML_tests.cmd: Windows WML tests: Add optional argument for executable path http://git.io/H-rdmw
20140529 05:31:47< irker318> wesnoth: aquileia wesnoth:master 64d72a63a91e / projectfiles/VC9/wesnoth.vcproj: Windows WML tests: Enable automatic tests in VC http://git.io/X8r6FQ
20140529 05:31:49< irker318> wesnoth: Chris Beck wesnoth:master 8306aa9353ad / projectfiles/VC9/ (5 files): Merge pull request #173 from aquileia/WML_test http://git.io/q92ypw
20140529 05:32:07< iceiceice_> aquileia: thanks for this :)
20140529 05:33:47-!- Sulfur [~Miranda@p5B008432.dip0.t-ipconnect.de] has joined #wesnoth-dev
20140529 05:52:38-!- shadowm_desktop [ignacio@wesnoth/developer/shadowmaster] has quit [Ping timeout: 240 seconds]
20140529 05:59:27-!- [Relic] [~relic@99-58-54-211.lightspeed.milwwi.sbcglobal.net] has quit [Quit: Leaving]
20140529 06:00:55-!- cib [~cib@p20030067CE126501267703FFFEE75B84.dip0.t-ipconnect.de] has joined #wesnoth-dev
20140529 06:01:19-!- cib is now known as Guest63621
20140529 06:09:00-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20140529 06:22:49-!- ancestral [~ancestral@12.23.74.29] has quit [Quit: i go nstuf kthxbai]
20140529 07:20:16-!- ancestral [~ancestral@12.23.74.29] has joined #wesnoth-dev
20140529 07:23:01-!- Necrosporus_ is now known as Necrosporus
20140529 07:24:58-!- exciton [chuck-the-@89.208.170.132] has joined #wesnoth-dev
20140529 07:25:12< Necrosporus> shadowm, It seems unlikely, though, since (as far as I know) wesnoth.require() and wesnoth.do_file() will fail unless the target file is syntactically-valid Lua. // Just make user download a lua file, it's not that hard, then maybe that file could use ./ to retrieve other files in the directory?
20140529 07:26:19-!- yann [~dwitch@nan92-1-81-57-214-146.fbx.proxad.net] has quit [Remote host closed the connection]
20140529 07:29:23-!- exciton [chuck-the-@89.208.170.132] has quit [Ping timeout: 255 seconds]
20140529 07:32:23-!- RiftWalker [~nathan@ip24-252-126-205.no.no.cox.net] has quit [Quit: leaving]
20140529 07:39:55-!- c74d [~c74d3a4eb@2002:4404:712c:0:76de:2bff:fed4:2766] has quit [Ping timeout: 240 seconds]
20140529 07:39:55-!- _8680_ [~8680@2002:4404:712c:0:986d:a83b:dd8e:b1f4] has quit [Ping timeout: 240 seconds]
20140529 07:43:21-!- mordante [~mordante@wesnoth/developer/mordante] has joined #wesnoth-dev
20140529 07:43:37< mordante> servus
20140529 07:47:36-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20140529 07:51:15-!- ancestral [~ancestral@12.23.74.29] has quit [Quit: i go nstuf kthxbai]
20140529 07:52:57-!- ancestral [~ancestral@12.23.74.29] has joined #wesnoth-dev
20140529 07:59:21< mordante> iceiceice_, thanks for the report, I tried it but can't reproduce the problem :-
20140529 07:59:27< mordante> :-( *
20140529 08:05:55-!- shadowm_desktop [ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20140529 08:13:23-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20140529 08:15:58-!- shadowm_desktop [ignacio@wesnoth/developer/shadowmaster] has quit [Ping timeout: 240 seconds]
20140529 08:18:17-!- shadowm_desktop [ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20140529 08:21:27-!- _8680_ [~8680@2002:4404:712c:0:804e:87cb:6903:deb6] has joined #wesnoth-dev
20140529 08:24:58-!- c74d [~c74d3a4eb@2002:4404:712c:0:76de:2bff:fed4:2766] has joined #wesnoth-dev
20140529 08:31:54-!- irker318 [~irker@fehu.ai0867.net] has quit [Quit: transmission timeout]
20140529 08:33:41-!- _8680_ [~8680@2002:4404:712c:0:804e:87cb:6903:deb6] has quit [Ping timeout: 252 seconds]
20140529 08:33:57-!- c74d [~c74d3a4eb@2002:4404:712c:0:76de:2bff:fed4:2766] has quit [Ping timeout: 265 seconds]
20140529 08:34:52-!- _8680_ [~8680@2002:4404:712c:0:804e:87cb:6903:deb6] has joined #wesnoth-dev
20140529 08:38:19-!- c74d [~c74d3a4eb@2002:4404:712c:0:76de:2bff:fed4:2766] has joined #wesnoth-dev
20140529 08:43:20-!- c74d [~c74d3a4eb@2002:4404:712c:0:76de:2bff:fed4:2766] has quit [Ping timeout: 252 seconds]
20140529 08:43:37-!- _8680_ [~8680@2002:4404:712c:0:804e:87cb:6903:deb6] has quit [Ping timeout: 265 seconds]
20140529 08:58:24-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20140529 08:59:51-!- _8680_ [~8680@2002:4404:712c:0:edb0:cc62:2b62:3058] has joined #wesnoth-dev
20140529 09:03:45-!- shadowm_desktop [ignacio@wesnoth/developer/shadowmaster] has quit [Ping timeout: 252 seconds]
20140529 09:07:03-!- _8680_ [~8680@2002:4404:712c:0:edb0:cc62:2b62:3058] has quit [Ping timeout: 252 seconds]
20140529 09:12:16-!- _8680_ [~8680@2002:4404:712c:0:81b6:5048:a06:7a14] has joined #wesnoth-dev
20140529 09:12:40-!- _8680_ is now known as Guest9003
20140529 09:15:57-!- mjs-de [~mjs-de@f048083245.adsl.alicedsl.de] has joined #wesnoth-dev
20140529 09:32:47-!- Guest9003 [~8680@2002:4404:712c:0:81b6:5048:a06:7a14] has quit [Ping timeout: 245 seconds]
20140529 09:36:44-!- _8680_ [~8680@2002:4404:712c:0:19ef:dcf9:d9a5:84a] has joined #wesnoth-dev
20140529 09:39:36-!- sachith500 [~kvirc@112.135.7.69] has joined #wesnoth-dev
20140529 09:48:03-!- Sulfur [~Miranda@p5B008432.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20140529 09:51:54-!- c74d [~c74d3a4eb@2002:4404:712c:0:76de:2bff:fed4:2766] has joined #wesnoth-dev
20140529 10:09:08-!- Sulfur [~Miranda@p5B008432.dip0.t-ipconnect.de] has joined #wesnoth-dev
20140529 10:09:52-!- tomreyn [~tomreyn@p5795480A.dip0.t-ipconnect.de] has joined #wesnoth-dev
20140529 10:10:00-!- tomreyn [~tomreyn@p5795480A.dip0.t-ipconnect.de] has quit [Changing host]
20140529 10:10:00-!- tomreyn [~tomreyn@megaglest/team/tomreyn] has joined #wesnoth-dev
20140529 10:11:22-!- aquileia [2edf524c@gateway/web/freenode/ip.46.223.82.76] has joined #wesnoth-dev
20140529 10:11:43< aquileia> iceiceice_: Thanks for merging.
20140529 10:12:19< aquileia> Just wondering... could we di-implement --timeout, now that its use is gone on both platforms?
20140529 10:12:31< aquileia> s/di/de
20140529 10:13:20-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20140529 10:26:48-!- sachith500|2 [~kvirc@112.135.33.128] has joined #wesnoth-dev
20140529 10:29:39-!- sachith500 [~kvirc@112.135.7.69] has quit [Ping timeout: 276 seconds]
20140529 11:01:23< AI0867> aquileia: sure?
20140529 11:09:50< aquileia> AI0867: Why not?
20140529 11:10:17< aquileia> unix based platforms have it built in, and I added it for Windows
20140529 11:10:52< aquileia> I don't see a particular reason to preserve the cmd line option
20140529 11:11:12< aquileia> Especially as it's a little buggy
20140529 11:13:56< aquileia> With --timeout, some tests timed out after 30s which now run in less than 20s
20140529 11:15:00< aquileia> Reason: wesnoth.exe was stalled directly before exiting
20140529 11:27:03-!- EdB [~edb@85.69.242.6] has joined #wesnoth-dev
20140529 11:34:15-!- spoffy [~spoffy@152.78.175.8] has joined #wesnoth-dev
20140529 11:39:02< aquileia> zookeeper: Would you mind two questions?
20140529 11:40:17< aquileia> 1) Which parts of Wesnoth would you show a tourist? That is, are there any breathtaking views of cities/natural sites?
20140529 11:40:45< aquileia> 2) Would Eldred encounter Delfadors troops directly outside of Weldyn, or do we storm the city wall first? In one case most of the terrain is grassland with a few suburbs with the wall at the edge of the map, in the other we fight on the streets of the capital - refreshingly new, but we don't really have the terrain graphics for that...
20140529 11:40:50-!- c74d [~c74d3a4eb@2002:4404:712c:0:76de:2bff:fed4:2766] has quit [Read error: Connection reset by peer]
20140529 11:42:46< aquileia> The only place for 1) I know is Halstead...
20140529 11:43:27-!- _8680_ [~8680@2002:4404:712c:0:19ef:dcf9:d9a5:84a] has quit [Remote host closed the connection]
20140529 11:43:46-!- _8680_ [~8680@2002:4404:712c:0:a8:a7ab:401d:237d] has joined #wesnoth-dev
20140529 11:44:41-!- c74d [~c74d3a4eb@2002:4404:712c:0:76de:2bff:fed4:2766] has joined #wesnoth-dev
20140529 11:44:50-!- Bodhi-Baum [~Bodhi@dslb-084-063-011-251.pools.arcor-ip.net] has joined #wesnoth-dev
20140529 11:47:27< zookeeper> aquileia, hah, before i read your last line i thought of halstead too
20140529 11:48:04< zookeeper> aquileia, i've always assumed the battle was outside the city, but i don't know if that's really supported by anything
20140529 11:49:04< zookeeper> looks like it is: in HttT delfador says that after the battle he "made haste to Weldyn"
20140529 11:52:06< zookeeper> i don't know why delfador and the loyalists didn't just stay in weldyn and defend the city, but instead "marched north to meet Eldred" (from the intro)
20140529 11:58:48< iceiceice_> aquileia: so the thing is, even if wesnoth --timeout is bad its maybe ok just as a fallback
20140529 11:58:51< zookeeper> anyway, WRT 1), i can't really think of anything. weldyn and elensefar i guess would be worth seeing?
20140529 11:59:51< iceiceice_> im not really sure if it makes sense, there's certainly a good argument to remove
20140529 12:00:30< iceiceice_> but for instance i think macs maynot have unix timeout unless you jump through hoops to configure them so
20140529 12:00:41< iceiceice_> not realy sure
20140529 12:02:36< aquileia> zookeeper: Thanks, I didn't know that there's info in HttT about that battle
20140529 12:02:52< aquileia> I concentrated on DM
20140529 12:03:05< iceiceice_> it also might be more buggy on other platforms, i really have no idea
20140529 12:04:15< aquileia> iceiceice_: it seems OS X has several options to do a timeout: http://stackoverflow.com/questions/601543/command-line-command-to-auto-kill-a-command-after-a-certain-amount-of-time
20140529 12:05:10< iceiceice_> y i mean wesnoth --timeout is not recommended
20140529 12:06:08< iceiceice_> but using it is still easier than saying "learn how to write scripts for your platform"
20140529 12:06:30< iceiceice_> if the quality is sometimes atrocious then it should go though
20140529 12:06:53< iceiceice_> in my system it seems to work like 95% of the time
20140529 12:07:18< iceiceice_> and the other 5% i have to manually kill the window / kill the process
20140529 12:07:36< iceiceice_> but i mean, its using a deprecated SDL function
20140529 12:07:56< iceiceice_> and in SDL 2 i have to do something also sort of weird
20140529 12:08:45< aquileia> iceiceice_: On windows it always kills, but it can delay a normal exit up to the point where it's killed
20140529 12:09:38< aquileia> With --showgui I saw that the test had finished, but wesnoth.exe didn't close; it was stalled till the timeout applied
20140529 12:10:06< iceiceice_> hmmmm
20140529 12:10:25< iceiceice_> thats wierd
20140529 12:10:33< iceiceice_> and the test failed then?
20140529 12:10:57< iceiceice_> that sounds like an error in SDL join thread or something
20140529 12:11:35< aquileia> the test was registered as a timeout although it finished - it just never returned
20140529 12:11:56< aquileia> or with never I rather mean "after the timeout"
20140529 12:12:03< iceiceice_> so you were using the windows timeout utility, but also with wesnoth --timeout?
20140529 12:12:55< aquileia> No, I used --timeout alone and had these problems. That's the reason I bothered to dive into WinAPI for my own WindowsTimeout
20140529 12:13:16< aquileia> After the switch it works perfectly
20140529 12:14:19< iceiceice_> ok i think i understand
20140529 12:14:28< iceiceice_> yeahi think we are talking about the same issue
20140529 12:14:51< iceiceice_> y its probably a good idea to get rid of the wesnoth --timeout option
20140529 12:26:43< iceiceice_> how hard is it for someone to build your windows timeout separately of VC / separately of wesnoth
20140529 12:27:29< aquileia> iceiceice_: Just compile the cpp and you're done
20140529 12:27:38< iceiceice_> i see
20140529 12:27:44< aquileia> it's completely independant of VC and Wesnoth
20140529 12:27:56< aquileia> It's integrated, but separate
20140529 12:37:39-!- ancestral [~ancestral@12.23.74.29] has quit [Quit: i go nstuf kthxbai]
20140529 12:39:21-!- Bodhi-Baum [~Bodhi@dslb-084-063-011-251.pools.arcor-ip.net] has quit [Ping timeout: 252 seconds]
20140529 12:41:02-!- markus_ [~mjs-de@g228033017.adsl.alicedsl.de] has joined #wesnoth-dev
20140529 12:44:29-!- mjs-de [~mjs-de@f048083245.adsl.alicedsl.de] has quit [Ping timeout: 252 seconds]
20140529 12:58:22-!- irker343 [~irker@fehu.ai0867.net] has joined #wesnoth-dev
20140529 12:58:22< irker343> wesnoth: Elvish_Hunter wesnoth:master 5480cca466f9 / data/tools/GUI.pyw: wmllint gui: fixed a bug in regexp handling http://git.io/lp25Ug
20140529 12:58:22< irker343> wesnoth: Elvish_Hunter wesnoth:master e3fc8f13b83f / data/tools/GUI.pyw: wmllint gui: made some options in the wmlscope tab evenly spaced http://git.io/w7V7NA
20140529 13:08:40-!- Bodhi-Baum [~Bodhi@dslb-084-063-011-251.pools.arcor-ip.net] has joined #wesnoth-dev
20140529 13:13:38-!- markus_ is now known as mjs-de
20140529 13:28:05-!- fabi_ [~fabi@wesnoth/developer/fendrin] has quit [Quit: Konversation terminated!]
20140529 13:35:46-!- fabi [~fabi@wesnoth/developer/fendrin] has joined #wesnoth-dev
20140529 13:42:32-!- EdB [~edb@85.69.242.6] has quit [Quit: Konversation terminated!]
20140529 14:03:41-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20140529 14:23:57-!- sachith500|2 [~kvirc@112.135.33.128] has quit [Read error: Connection reset by peer]
20140529 14:24:16-!- sachith500|2 [~kvirc@112.135.33.128] has joined #wesnoth-dev
20140529 14:36:04-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20140529 14:37:26< AI0867> 13:09 < aquileia> AI0867: Why not? ← er, I agreed?
20140529 14:38:11-!- Netsplit *.net <-> *.split quits: crimson_penguin
20140529 14:38:25-!- Netsplit over, joins: crimson_penguin
20140529 14:51:58< aquileia> AI0867: Ah, then I misunderstood, sorry
20140529 14:58:30-!- sachith500 [~kvirc@61.245.163.41] has joined #wesnoth-dev
20140529 15:01:21-!- sachith500|2 [~kvirc@112.135.33.128] has quit [Ping timeout: 252 seconds]
20140529 15:07:07-!- Bodhi-Baum [~Bodhi@dslb-084-063-011-251.pools.arcor-ip.net] has quit [Ping timeout: 252 seconds]
20140529 15:14:18-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20140529 15:16:34-!- apoi [~andi@85-126-180-242.volume.xdsl-line.inode.at] has quit [Quit: leaving]
20140529 15:22:51< aquileia> mattsc, the AI in DM 21 seemed to pretty much ignore one of the two goals... just out of interest, how much work would it be to convert it to a micro AI?
20140529 15:23:16-!- Samual [diotecktec@xonotic/core-team/Samual] has joined #wesnoth-dev
20140529 15:30:12< iceiceice_> gfgtdf: i am looking at the replay controller code,
20140529 15:30:19< iceiceice_> it looks like when you reset the replay the variables may not be reset
20140529 15:30:33-!- ancestral [~ancestral@12.23.74.29] has joined #wesnoth-dev
20140529 15:30:34< iceiceice_> not sure though, just thought i would point it out
20140529 15:31:59< iceiceice_> hmm no i guess the gamestate object is reset so its probably fine
20140529 15:33:03< mattsc> aquileia: I looked at the code that  does that: https://github.com/wesnoth/wesnoth/blob/master/data/campaigns/Delfadors_Memoirs/scenarios/21_Clash_at_the_Manor.cfg#L200
20140529 15:33:56< mattsc> I think that method works just fine (within limits) as it is, but it only sets the goals for (on average) half of the riders, and then with a random chance between the two signposts.
20140529 15:34:24< mattsc> So the fact that you often see the AI apparently ignoring one of the posts makes sense.
20140529 15:34:59< mattsc> It might make sense to change l.210 to ‘rand=1..2’ though, to set goals for all rides.
20140529 15:35:09< mattsc> I don’t know, it’s been a while since I played that scenario.
20140529 15:35:09< aquileia> I just wondered because none moved there during several tries - but then, I don't remember to have seen any riders at that point...
20140529 15:35:50< mattsc> That might also be a difference - the new recruiting method might not recruit enough riders.  That could be changed as well.
20140529 15:35:57< aquileia> So it's probably that the AI didn't recruit them yet
20140529 15:37:15< aquileia> ok
20140529 15:37:26< mattsc> Yeah, I think changing recruiting might be more important than the way the goals are set.
20140529 15:40:02-!- Bodhi-Baum [~Bodhi@dslb-084-063-011-251.pools.arcor-ip.net] has joined #wesnoth-dev
20140529 15:49:54< iceiceice_> does anyone know why we have this line?
20140529 15:49:56< iceiceice_> https://github.com/wesnoth/wesnoth/blob/master/src/play_controller.cpp#L752
20140529 15:50:07< iceiceice_> it causes end of scenario saves to be stripped of all units
20140529 15:50:15-!- apoi [~andi@85-126-180-242.volume.xdsl-line.inode.at] has joined #wesnoth-dev
20140529 15:50:23< iceiceice_> my theory is that it was some kind of hack for some point in time before carryover was implemented properly
20140529 15:50:34< iceiceice_> i propose to remove it in 1.12 and master
20140529 15:51:09-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20140529 15:53:12-!- ancestral [~ancestral@12.23.74.29] has quit [Quit: i go nstuf kthxbai]
20140529 15:54:04-!- sachith500|2 [~kvirc@112.135.33.128] has joined #wesnoth-dev
20140529 15:56:22-!- sachith500 [~kvirc@61.245.163.41] has quit [Read error: Connection reset by peer]
20140529 15:58:00-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20140529 15:58:21-!- irker343 [~irker@fehu.ai0867.net] has quit [Quit: transmission timeout]
20140529 16:16:46-!- irker675 [~irker@fehu.ai0867.net] has joined #wesnoth-dev
20140529 16:16:46< irker675> wesnoth: Chris Beck wesnoth:master c08409855686 / src/ (unit_map.cpp unit_map.hpp): add const qualifier to unit_map::has_unit, which was missing it http://git.io/1_pxTQ
20140529 16:17:17< irker675> wesnoth: Chris Beck wesnoth:1.12 5efa0696fbb8 / src/ (unit_map.cpp unit_map.hpp): add const qualifier to unit_map::has_unit, which was missing it http://git.io/1gihNA
20140529 16:20:27< Necrosporus> Does Emacs plugin contain something like a version of wesnoth' wiki with descriptions of WML tags and attributes?
20140529 16:20:54< iceiceice_> no
20140529 16:20:57< iceiceice_> afaik there is a umc idea
20140529 16:20:58< iceiceice_> *ide
20140529 16:21:16< iceiceice_> also various posts on the forums about finding ways to support this, i think with a web-based text editor?
20140529 16:21:32< iceiceice_> fwiw i think its better to just to... use a regular text editor, and have a web browser open to the documentation
20140529 16:22:37-!- ancestral [~ancestral@17.114.45.98] has joined #wesnoth-dev
20140529 16:24:17-!- sachith500|2 [~kvirc@112.135.33.128] has quit [Read error: Connection reset by peer]
20140529 16:31:25< irker675> wesnoth: Chris Beck wesnoth:master c663cdcaba98 / src/play_controller.cpp: remove code made dead by commit d808b7c7874465e314c433029d709236bea4fab9 http://git.io/A_clTQ
20140529 16:32:47-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20140529 16:37:07-!- Sulfur [~Miranda@p5B008432.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20140529 16:57:11< irker675> wesnoth: ln-zookeeper wesnoth:master cf3699dabda0 / / (4 files in 3 dirs): Prevent allied leaders' suicidal behavior http://git.io/v_qHrg
20140529 17:01:29< irker675> wesnoth: ln-zookeeper wesnoth:1.12 b8be2faa8c51 / data/campaigns/Son_Of_The_Black_Eye/ (3 files in 2 dirs): Prevent allied leaders' suicidal behavior http://git.io/R50YoA
20140529 17:14:12-!- gfgtdf [~chatzilla@f054158248.adsl.alicedsl.de] has joined #wesnoth-dev
20140529 17:14:27< gfgtdf> mordante: 20140524 16:50:54< gfgtdf> mordante: do you think https://gna.org/bugs/?22046 can be fixed by replacing (height -20) with if(height > 20, height -20, 0) in _initial.cfg ?
20140529 17:14:42< AI0867> iceiceice_: https://github.com/wesnoth/wesnoth/commit/aa2cf7d635c49246e50011fe6f2d3e71fa1ce017
20140529 17:15:33< iceiceice_> hmm
20140529 17:15:39< iceiceice_> AI0867: idk if that fix makes much sense though
20140529 17:16:04< AI0867> gfgtdf: don't we have a function like: max(heigh-20, 0)
20140529 17:16:06< AI0867> ?
20140529 17:16:40< gfgtdf> iceiceice_: i think in the thead on gna he said this might not be the est solution
20140529 17:16:52< AI0867> we did have quite a bit of carryover refactoring since 1.8
20140529 17:17:14< mattsc> iceiceice_: this reminds me of something I have been meaning to ask about
20140529 17:17:28< gfgtdf> AI0867: ye that might work too. I think i copied that if from smewhere else
20140529 17:18:30< mattsc> If I let an AI vs. AI test scenario runs its course until one side is defeated, then save the replay, then save a certain turn during the replay (for testing), loading that turn save comes up with the “you’ve been defeated” dialog.
20140529 17:18:52< iceiceice_> hmm i guess i should just test end fo scenario saves and see what is actually going wrong
20140529 17:18:54< gfgtdf> AI0867: i think i copied from here: https://github.com/wesnoth/wesnoth/blob/master/data/gui/default/macros/_initial.cfg#L26
20140529 17:18:55< mattsc> I have to manually change the status from “defeated” to “running” (or something along those lines) in the save to be able to play from there.
20140529 17:19:32< iceiceice_> mattsc: that's wierd
20140529 17:20:15< iceiceice_> hmm wait what exactly do you have to change?
20140529 17:20:22< mattsc> iceiceice_: it’s easy to reproduce: run ‘./wesnoth -t fred’ from the CL and just wait it out.  Then do the steps I just described.
20140529 17:20:23< iceiceice_> do you change defeat_condition in the save file?
20140529 17:20:40< iceiceice_> oh and does the replay have to end or can i just save at any point
20140529 17:22:00< mattsc> iceiceice_: no, there’s a line “completion=…” at the beginning of the save files.
20140529 17:22:15< mattsc> I have to change that from “defeated” (I think) to “running”
20140529 17:22:59< mattsc> iceiceice_: you can save at any point
20140529 17:25:20< iceiceice_> hmm this sounds like a minor bug in the repay viewer i guess
20140529 17:25:51< iceiceice_> most likely replay_controller should reset the value of that guy?
20140529 17:26:00< iceiceice_> idk what the "completion" field does anyways
20140529 17:26:12< mattsc> iceiceice_: yeah, it’s nothing major, not a big deal (that’s why I kept forgetting to ask about it)
20140529 17:26:31< iceiceice_> hmm
20140529 17:26:40< iceiceice_> when i type wesnoth -t fred i get "scenario not found
20140529 17:26:52< iceiceice_> in 1.13
20140529 17:27:00< gfgtdf> iceiceice_:  i currnelty cannot find teh cde that might reset teh ariables in a replay
20140529 17:27:06< gfgtdf> the code
20140529 17:27:23< iceiceice_> gfgtdf: i was looking at it recently
20140529 17:27:27< mattsc> iceiceice_: oh, crap, right …
20140529 17:27:37< gfgtdf> iceiceice_: ye i said becaus you asked me
20140529 17:27:48< mattsc> That’s a scenario in AI-demos, not mainline  :P  Sorry for that.
20140529 17:28:02< iceiceice_> so i guess in replay_controller::reset_replay()
20140529 17:28:12< iceiceice_> theres a line gamestate_ = gamestate_start
20140529 17:28:18< iceiceice_> i think completion is in game_state?
20140529 17:28:28< mattsc> iceiceice_: let me test quickly if this can be reproduced with one of the MAI test scenarios as well.
20140529 17:28:33< iceiceice_> so maybe the replay viewer should also set game_state_start.completion to running
20140529 17:28:34< iceiceice_> in its constructor
20140529 17:28:39< iceiceice_> *the replay_controller
20140529 17:28:44< iceiceice_> should do this in its constructor
20140529 17:28:46< gfgtdf> iceiceice_: hm the completion maybe is but i think the variables are in gamedata
20140529 17:28:58< iceiceice_> hmmm...
20140529 17:29:01< iceiceice_> right about my earlier quesitons
20140529 17:29:11< gfgtdf> iceiceice_: 20140529 15:30:12< iceiceice_> gfgtdf: i am looking at the replay controller code, ...
20140529 17:29:44< iceiceice_> i think "replay_controller::reset_replay" is what correspodns to the stop button
20140529 17:30:04< iceiceice_> hmm it might be a good idea if the constructor calls that also though
20140529 17:30:08< iceiceice_> so we only have to write setup code once
20140529 17:30:12< irker675> wesnoth: ln-zookeeper wesnoth:master bab0915c9469 / / (5 files in 3 dirs): Prevent allied leaders' suicidal behavior http://git.io/Y8eF0g
20140529 17:30:37< iceiceice_> they should be pretty much the same, no?
20140529 17:31:22< gfgtdf> iceiceice_: i think teh contructor does call it
20140529 17:31:43< iceiceice_> aha right you are
20140529 17:32:14-!- riksteri [~riksteri@dsl-tkubrasgw3-54f96b-216.dhcp.inet.fi] has quit [Quit: riksteri]
20140529 17:32:40< irker675> wesnoth: ln-zookeeper wesnoth:1.12 e78ed21ea312 / data/campaigns/Son_Of_The_Black_Eye/ (4 files in 2 dirs): Prevent allied leaders' suicidal behavior http://git.io/khJSvg
20140529 17:33:40< mattsc> iceiceice_: interestingly, the same does not happen when you do an AI vs. AI MP game.
20140529 17:33:47-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20140529 17:35:16< iceiceice_> hmm
20140529 17:35:21< iceiceice_> it looks like game_data is constructed from level
20140529 17:35:25< iceiceice_> in the play_controller constructor
20140529 17:35:35< iceiceice_> https://github.com/wesnoth/wesnoth/blob/master/src/play_controller.cpp#L115
20140529 17:35:57< iceiceice_> maybe there needs to be a "game_data_start" variable in replay_controller?
20140529 17:37:53< mattsc> iceiceice_: okay, way to reproduce: wesnoth -t protect_unit
20140529 17:38:12< mattsc> Take control of side 2 (Koorzhar) and move the leader toward the enemy to get him killed.
20140529 17:38:27< mattsc> Then do the steps described above.
20140529 17:38:34< gfgtdf> iceiceice_: hm i think we have a  variable level_ so we can jst recreate from that ?
20140529 17:38:44< mattsc> Btw, the defeat dialog does not show up, the game just goes into linger mode directly.
20140529 17:39:16< gfgtdf> iceiceice_: hm o that wont wokr i guess
20140529 17:39:19< gfgtdf> no*
20140529 17:39:34< gfgtdf> iceiceice_: ye a game_data_start sounds better
20140529 17:40:05< Necrosporus> btw could you compare sum of several variables with a number? like if (x+y > 5) {} ?
20140529 17:40:10< Necrosporus> in WML
20140529 17:40:33< Necrosporus> or only way is to make a temporary variable?
20140529 17:41:34< iceiceice_> necrosporus:
20140529 17:41:36< gfgtdf> no its possible [set_variable] name="is_gereater" value = "$($x + $y < 5)" [set_varaible]
20140529 17:42:08< iceiceice_> y i think thats the best you will get
20140529 17:42:11< Necrosporus> And to check it you still need [variable] ?
20140529 17:42:17< iceiceice_> Necrosporus:
20140529 17:42:24< iceiceice_> you shoul be using hte macros {VARIABLE}
20140529 17:42:27< iceiceice_> and {VARIABLE_CONDITIONAL}
20140529 17:42:31< iceiceice_> otherwise your life will be hell
20140529 17:42:38< iceiceice_> also {IF_VAR}
20140529 17:43:27< gfgtdf> iceiceice_: i think we choudl mave ad a condition that just takes an expression
20140529 17:43:32< gfgtdf> which rturns a bool
20140529 17:44:13< iceiceice_> coudl make a "boolean_conditional" macro i guess
20140529 17:44:15< gfgtdf> iceiceice_: like [conditonal] value = "$($x + $y < 5)" [/conditonal]
20140529 17:45:14< iceiceice_> hmm y i guess maybe
20140529 17:45:21< iceiceice_> [is_true] value = ... [/is_true]
20140529 17:45:31< iceiceice_> which could be used at places where [variable] normally appears
20140529 17:45:49< gfgtdf> iceiceice_: y like that
20140529 17:45:49< iceiceice_> its pretty small issue thouh
20140529 17:46:12< gfgtdf> iceiceice_: i have another idea:  merge contitional tags with action tags, so that "conditional tags" like [have_unit] are just special action tagss that return abool value
20140529 17:46:47< iceiceice_> hmm i think tags dont usually return a value though
20140529 17:46:55< iceiceice_> values have to go in temporary variables
20140529 17:47:05< gfgtdf> iceiceice_: ye i mean internaly from the c++/lua core
20140529 17:47:20< iceiceice_> y so i think [have_unit] does already do what you say
20140529 17:47:27< iceiceice_> i mean its all in the implementatin of [if] in c++
20140529 17:47:32< iceiceice_> sry
20140529 17:47:33< iceiceice_> in lua i guess
20140529 17:48:15< iceiceice_> have_unit is a valid conditional expression in wml right now iiuc, im pretty sure i used it that way in a unit test :)
20140529 17:48:34< iceiceice_> maybe my test is borked? :O
20140529 17:48:55< gfgtdf> iceiceice_: but my thought was to have conditional tags like [have_unit] just beeing a specialy cases of action tags like [move_unit] that return a value whether the action shoudl continue
20140529 17:49:16< iceiceice_> hmm i see
20140529 17:49:24< iceiceice_> so you want move_unit and other tags to be valid conditional expressions?
20140529 17:49:31-!- lipkab [~the_new_l@host-91-147-212-189.biatv.hu] has joined #wesnoth-dev
20140529 17:49:44< lipkab> mordante: Hi.
20140529 17:50:04< gfgtdf> hmm ye, now that you say it it sounds strane :)
20140529 17:50:16< iceiceice_> gfgtdf: i think while its sensible from a normal programming language perspective, it would be awkward in implementation
20140529 17:50:27< iceiceice_> because everything in wml is actually configs, and not strongly ordered sort of
20140529 17:50:37< iceiceice_> [else] and [then] are always interchangeable 
20140529 17:50:49< iceiceice_> because the implementation is based on like ".child("else")"
20140529 17:51:16< iceiceice_> if you want to allow arbitray things to be condition in [if], then i think the [if] evlauator code would have to consider all possible children
20140529 17:51:29< gfgtdf> iceiceice_: yes it should
20140529 17:51:47< iceiceice_> i mean maybe not, 
20140529 17:51:57< iceiceice_> maybe it could like parse out the [then] and [else] separetly
20140529 17:51:59< iceiceice_> and [elseif] 
20140529 17:52:02< iceiceice_> and then see whats left
20140529 17:52:19< iceiceice_> but i think thats not how it works right now
20140529 17:53:17< gfgtdf> iceiceice_: it snds that i is not a code that is spready over many unrelaed files so it sould be relertively eeasy to chane
20140529 17:53:25< iceiceice_> y
20140529 17:54:05< iceiceice_> i think for the most part wml scenarios can't handle errors,
20140529 17:54:14< iceiceice_> its just not designed that way that they should recover when stuff doesnt work
20140529 17:55:06< iceiceice_> hmm ok
20140529 17:55:13< iceiceice_> im going to try making game_data_start i think
20140529 17:55:16< iceiceice_> and test that
20140529 17:55:20< iceiceice_> and also the thing we did earlier
20140529 17:55:25< gfgtdf> iceiceice_: i think about commention out these 2 lines+commit to master: https://github.com/wesnoth/wesnoth/blob/master/src/playcampaign.cpp#L498
20140529 17:55:27< iceiceice_> about init sides being saved
20140529 17:55:48< gfgtdf> commentiong out
20140529 17:56:18< iceiceice_> really?
20140529 17:56:28< iceiceice_> will that disable the singlepalyer contorller?
20140529 17:56:44< gfgtdf> iceiceice_: yes, i currently doesnt see what could go wrong
20140529 17:56:55< iceiceice_> hmm
20140529 17:57:10< iceiceice_> from what i understand there are small differences between sp and mp in terms of like how carryover is calculated
20140529 17:57:22< iceiceice_> idk if that changed when thunderstruck did gsoc
20140529 17:57:30< iceiceice_> but i would ask him
20140529 17:58:30< gfgtdf> iceiceice_: well but teh mwthods playsingle_scenario and playmp_scenario do the same epect that one uses a playmp-controller and teh ther uses a playsingle_controller
20140529 17:58:34< gfgtdf> methods
20140529 17:58:38< gfgtdf> ecept
20140529 17:58:41< gfgtdf> except
20140529 17:59:59< iceiceice_> hmm maybe could make a template function?
20140529 18:00:53< iceiceice_> i dont think we should eliminate playsingle_controller though right now, it might confuse RiftWalker if he is trying to test and see what the current differences are
20140529 18:01:51-!- gfgtdf [~chatzilla@f054158248.adsl.alicedsl.de] has quit [Quit: ChatZilla 0.9.90.1 [Firefox 29.0.1/20140506152807]]
20140529 18:21:59< irker675> wesnoth: Chris Beck wesnoth:master c8699e2f057b / .travis.yml run_wml_tests src/commandline_options.cpp src/game.cpp: remove built-in wesnoth timeout http://git.io/ocjv4w
20140529 18:22:16-!- spoffy [~spoffy@152.78.175.8] has quit [Ping timeout: 258 seconds]
20140529 18:23:18< iceiceice_> aquileia: ^
20140529 18:25:35-!- janebot_ is now known as janebot
20140529 18:30:19-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20140529 18:32:29< mordante> hi lipkab 
20140529 18:39:06< lipkab> mordante: Now I'm here again.
20140529 18:39:37< vultraz> Could someone backport the two commits by ElvishHunter earlier to 1.12?
20140529 18:40:12< lipkab> vultraz: How about you? :P
20140529 18:40:25< vultraz> No clone  
20140529 18:40:37< lipkab> Oh.
20140529 18:40:42< vultraz> I haven't bothered since it takes days and I have to reinstall my OS soon 
20140529 18:40:53< lipkab> Aha.
20140529 18:42:35< vultraz> Which is why I don't want to bother with a dev environment yet
20140529 18:42:52< vultraz> (plus I don't know how to use git cherry-pick ._. )
20140529 18:43:23< lipkab> vultraz: Well, I'm having a problem with conflicting commits over here, so later.
20140529 18:43:24-!- shadowm_desktop [ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20140529 18:43:40< lipkab> ...but here comes shadowm to help you!
20140529 18:44:07< shadowm> No.
20140529 18:47:15-!- prophile [~alynn@oftn/member/prophile] has joined #wesnoth-dev
20140529 18:50:50< iceiceice_> mordante: what resolutions did you try to reproduce the assertion at?
20140529 18:51:00< iceiceice_> i have now tried all my options,
20140529 18:51:09< iceiceice_> i dont get failure at 1024 x 768 or 800 x 600
20140529 18:51:22< iceiceice_> only 1360 x 768, and 1366 x 768
20140529 18:53:58< mordante> iceiceice_, I tried with 1364x718x24 since that was in your bug report
20140529 18:54:33< iceiceice_> hmm
20140529 18:54:51< iceiceice_> oh i think i know what thats about
20140529 18:55:04< iceiceice_> in my prefernces it says this: xresolution=1360
20140529 18:55:04< iceiceice_> xwindowsize=1364
20140529 18:55:04< iceiceice_> yresolution=768
20140529 18:55:04< iceiceice_> ywindowsize=718
20140529 18:55:26< iceiceice_> im not sure which number appeasr in the console output
20140529 18:56:08< mordante> Checking video mode: 1364x718x32... \n setting mode to 1364x718x32 
20140529 18:56:22< mordante> will test 1360 x 768 in a sec
20140529 18:56:36< iceiceice_> ok thx
20140529 19:05:49-!- shadowm_desktop [ignacio@wesnoth/developer/shadowmaster] has quit [Ping timeout: 252 seconds]
20140529 19:06:12-!- shadowm_desktop [ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20140529 19:09:37< mordante> iceiceice_, thanks for the additional info I can now reproduce it (after setting my language to English)
20140529 19:09:56< mordante> I'm working on another patch at the moment, so will look at it closer later
20140529 19:10:02< mordante> Necrosporus, ^
20140529 19:11:42< Necrosporus> good for you... maybe you could reproduce my problem too ?
20140529 19:12:14< mordante> Necrosporus, I'm quite sure it is the same problem
20140529 19:12:43< iceiceice_> ok, thanks for looking at it again mordante :)
20140529 19:12:57< mordante> no problem
20140529 19:16:20< irker675> wesnoth: Elvish_Hunter wesnoth:1.12 b89a04056a50 / data/tools/GUI.pyw: wmllint gui: fixed a bug in regexp handling http://git.io/o2sHQw
20140529 19:17:48-!- spoffy [~spoffy@152.78.175.146] has joined #wesnoth-dev
20140529 19:18:10< mattsc> vultraz: ^  (and argh …  now I am getting a conflict with the second of Elvish_Hunter’s commits)
20140529 19:20:02< irker675> wesnoth: Elvish_Hunter wesnoth:1.12 a8199e9406ac / data/tools/GUI.pyw: wmllint gui: made some options in the wmlscope tab evenly spaced http://git.io/9x5LCw
20140529 19:20:09< mattsc> ah, no, that’s just because I’m an idiot, as usual
20140529 19:20:17< mattsc> vultraz: should be done ^
20140529 19:20:29< vultraz> thanks
20140529 19:21:55-!- aboutGod [~aboutGod@static-72-66-66-50.washdc.fios.verizon.net] has joined #wesnoth-dev
20140529 19:25:21-!- spoffy [~spoffy@152.78.175.146] has quit [Ping timeout: 252 seconds]
20140529 19:27:01-!- aboutGod [~aboutGod@static-72-66-66-50.washdc.fios.verizon.net] has left #wesnoth-dev []
20140529 19:28:30< iceiceice_> gfgtdf: http://wiki.wesnoth.org/ScenarioWML#Scenario_End_Conditions
20140529 19:28:38< iceiceice_> actually, anyone feel free to comment / edit
20140529 19:28:40< iceiceice_> ^
20140529 19:29:12< mordante> Necrosporus, I had a look at your patch. It fixes the crash, but makes the output horrible so the && seems not to be a mistake
20140529 19:29:34< mordante> will have to look at it later this week
20140529 19:29:35< Necrosporus> output?
20140529 19:29:45< mordante> that is shown on the screen
20140529 19:29:49< mordante> what*
20140529 19:30:03< Necrosporus> screen?
20140529 19:30:37< Necrosporus> I mean give me a snapshot if you can
20140529 19:30:44-!- Tobarja [~Tobarja@kazi.atholdings.us] has joined #wesnoth-dev
20140529 19:32:05-!- Bodhi-Baum [~Bodhi@dslb-084-063-011-251.pools.arcor-ip.net] has quit [Ping timeout: 264 seconds]
20140529 19:32:37-!- Bodhi-Baum [~Bodhi@dslb-084-063-011-251.pools.arcor-ip.net] has joined #wesnoth-dev
20140529 19:33:19-!- Guest63621 [~cib@p20030067CE126501267703FFFEE75B84.dip0.t-ipconnect.de] has quit [Ping timeout: 252 seconds]
20140529 19:41:42< mordante> Necrosporus, http://tinypic.com/view.php?pic=28b4rpd&s=8
20140529 19:42:24< Necrosporus> it's not that bad for me
20140529 19:45:54< iceiceice_> y i see
20140529 19:46:17< iceiceice_> hehe i guess the || was a pretty silly idea, if that was what the author intended they wuldnt have had an "if" block at all probably...
20140529 19:46:25< iceiceice_> idk hard to say
20140529 19:46:57< iceiceice_> or maybe just on the assert or something
20140529 19:49:29< mordante> well I have look further, just not now
20140529 19:50:53< Necrosporus> What if you just remove assert, would it look better?
20140529 19:51:13-!- gfgtdf [~chatzilla@f054158248.adsl.alicedsl.de] has joined #wesnoth-dev
20140529 19:51:15< Necrosporus> so action in if won't be performed, as it stll && but no crash
20140529 19:51:19< irker675> wesnoth: Boldizsár Lipka wesnoth:master 7684fe06fe1e / src/sdl/ (texture.cpp texture.hpp): Add a lot of functions to sdl::ttexture. http://git.io/K5pMlQ
20140529 19:53:44< gfgtdf> iceiceice_: do you see why we dont use state_of_game.carryover_sides.child_or_add in https://github.com/wesnoth/wesnoth/blob/master/src/playcampaign.cpp#L296 ?
20140529 19:54:21< iceiceice_> i think the reason is because i am a noob :)
20140529 19:54:32< iceiceice_> it used to be just ".child"
20140529 19:54:41< iceiceice_> and that caused crashes in a certain bug reported by EP...
20140529 19:54:45< iceiceice_> so i changed it to what ti is now
20140529 19:54:51< iceiceice_> but what you say is better 
20140529 19:55:31< mordante> Necrosporus, no idea, but as said I'll look later
20140529 19:56:01< iceiceice_> https://gna.org/bugs/?21797
20140529 19:56:35< iceiceice_> gfgtdf: it might be better to figure out why there is no end_level_data
20140529 19:56:40< iceiceice_> this was sort of just a bandaid
20140529 19:56:50< iceiceice_> i should have added a code comment to that effect
20140529 19:57:17< gfgtdf> hm i wonder why there shoudol be any end_level_data
20140529 19:59:17< iceiceice_> also, i am testing now all of the saving and reloading issues that we pushed fixes to master for...
20140529 19:59:32< iceiceice_> if it looks good it hink i am going to backport to 1.12, even if no one else tested in the past few days
20140529 19:59:47< gfgtdf> iceiceice_: i have a question: does timeout work in a local mp game ?
20140529 20:00:01< iceiceice_> y it has in the past
20140529 20:00:17< iceiceice_> i have noticed that when i have run testing local mp games,
20140529 20:00:24< iceiceice_> in 1.12 or master,
20140529 20:00:32< iceiceice_> if i leave the game running after i see the bug and go look at code,
20140529 20:00:37< iceiceice_> i will get a bell 5 minutes later
20140529 20:00:41< iceiceice_> so i think that is working
20140529 20:00:54< gfgtdf> iceiceice_: then io_type is IO_SERVER ofr local mp games ?
20140529 20:01:01< gfgtdf> in playcampaign ?
20140529 20:01:23< iceiceice_> i think the local mp game path goes throguh multiplayer.cpp?
20140529 20:01:30< iceiceice_> i could be wrong
20140529 20:01:37< iceiceice_> i'm not sure if it uses playcampaign at all
20140529 20:01:44< gfgtdf> iceiceice_: all plaing of games goes though playcampaign
20140529 20:01:47< gfgtdf> aplying
20140529 20:01:51< gfgtdf> all playing
20140529 20:01:51< iceiceice_> you would hve to look at game.cpp, that is what controls what happens when you click a button on title screen
20140529 20:02:18< gfgtdf> playcampaign is teh only files that instanciases playymp_controller/playsing_controller
20140529 20:03:57< iceiceice_> looksl ike it uses: mp::start_client
20140529 20:06:01< gfgtdf> iceiceice_: also in mp i think it woul be better if the hosts sends data to teh other client on whether waiting in th mp lobby instead of trying to calculate that on every client
20140529 20:06:09< gfgtdf> iceiceice_: idk if tgat currently happens
20140529 20:06:11< gfgtdf> if that
20140529 20:06:36< iceiceice_> hmm y thats not a bad idea
20140529 20:06:41< iceiceice_> gfgtdf:
20140529 20:06:47< iceiceice_> i think all the buttons on the title screen
20140529 20:06:51< iceiceice_> use a path other than playcampaign.cpp
20140529 20:07:03< iceiceice_> they may have predated that code path
20140529 20:07:18< gfgtdf> no every game runs though playcampaign.cpp
20140529 20:07:33< iceiceice_> hmm well i would like to see that line by line
20140529 20:07:40< gfgtdf> becasue every game uses the _controller clases and that file is the onyl one that contructs them
20140529 20:08:48< iceiceice_> hmm ok, i think you are right, it eventually gets to there
20140529 20:08:55< mordante> I'm off bye
20140529 20:09:03< gfgtdf> iceiceice_: i think the server should send whether to wait in mp lobby inside the [notify_next_scenario]
20140529 20:09:13-!- mordante [~mordante@wesnoth/developer/mordante] has quit [Quit: Leaving]
20140529 20:11:54-!- ancestral [~ancestral@17.114.45.98] has quit [Quit: ancestral]
20140529 20:11:57< iceiceice_> gfgtdf: do you want to add that to 1.12 or just master?
20140529 20:12:12< iceiceice_> gfgtdf: it looks like local games and commandline games are started with io_type::SERVER
20140529 20:12:15< iceiceice_> not io_type::NONE
20140529 20:13:02< gfgtdf> iceiceice_: i didn't plan to put it in 1.12. But we can think about it again when i tis on master
20140529 20:13:43< iceiceice_> hmm im not sure
20140529 20:13:55< iceiceice_> so i guess io_type::NONE is the default value of play_game
20140529 20:14:30< iceiceice_> i think unit tests run in io_none
20140529 20:15:35< iceiceice_> hmm
20140529 20:15:35< iceiceice_> ok
20140529 20:15:42< iceiceice_> the only time that anything runs in io_type::NONE is 
20140529 20:15:46< iceiceice_> --test
20140529 20:15:48< irker675> wesnoth: gfgtdf wesnoth:master 28ffcc544839 / src/playcampaign.cpp: use config::child_or_add http://git.io/6eg3lQ
20140529 20:15:48-!- spoffy [~spoffy@152.78.175.8] has joined #wesnoth-dev
20140529 20:15:48< iceiceice_> --unit 
20140529 20:15:52< iceiceice_> and...
20140529 20:16:13< iceiceice_> from game.cpp launch_game
20140529 20:16:19< gfgtdf> iceiceice_: i think normal campaign should also use IO_NONE
20140529 20:16:24< gfgtdf> iceiceice_: ah ok
20140529 20:16:26< iceiceice_> because launch_game is the only place that that function uses default value
20140529 20:16:36< gfgtdf> iceiceice_: ^
20140529 20:16:46< gfgtdf> (the commit)
20140529 20:18:19< iceiceice_> hmm why did you move end_level so much further down?
20140529 20:18:39< iceiceice_> it is referred to in the case else if (res == victory)
20140529 20:19:02< iceiceice_> i think its not a refactor
20140529 20:19:20-!- lipkab [~the_new_l@host-91-147-212-189.biatv.hu] has quit [Quit: Vannak idők, mikor menni kell]
20140529 20:20:39< gfgtdf> iceiceice_: ough i didnt ee that one
20140529 20:21:04< gfgtdf> iceiceice_: i'll correct this
20140529 20:22:38< gfgtdf> iceiceice_: the intention was for linger mode to be able to to change end_level, because i wanted to add a flag "wait_in_lobby" to end_level_data.transient
20140529 20:23:06< gfgtdf> iceiceice_: because we receive the "notify_next_scenario" in this time
20140529 20:23:58< iceiceice_> hmm
20140529 20:24:57< iceiceice_> y idk maybe it should send a different message?
20140529 20:25:12< gfgtdf> iceiceice_: also i dont see why we only store carryover when we have avictory
20140529 20:25:21< gfgtdf> iceiceice_: (in teh situation of a mp campaign)
20140529 20:25:23< gfgtdf> tha*
20140529 20:25:49< iceiceice_> i think that it is all designed with the idea that you will go back to lobby if you are defeated
20140529 20:26:00< iceiceice_> this like competitive mp campaign is just not at all considered
20140529 20:26:21< iceiceice_> it probaly doesnt hurt to store carryover anyways though
20140529 20:26:24< iceiceice_> even if you go back to lobby
20140529 20:26:29< iceiceice_> it will just be discarded anyways?
20140529 20:26:55< gfgtdf> i think yes
20140529 20:27:51< irker675> wesnoth: gfgtdf wesnoth:master b6efcec496d4 / src/playcampaign.cpp: fixup 28ffcc54483942c598d24cecd4f65534f0185c51 http://git.io/EoERnQ
20140529 20:29:47< iceiceice_> hmm
20140529 20:29:59< iceiceice_> i am still reproducing the bug where sides are not initialized on reload
20140529 20:30:05< iceiceice_> at revision: c8699e2f057b14b19ba60dbd4e71f4647b45ca78
20140529 20:30:08< iceiceice_> that we thoguht that we fixed...
20140529 20:30:39< iceiceice_> i think i will go back to the commit that supposedly fixed it and try
20140529 20:30:50< iceiceice_> maybe my tests are invalid for some reason but it seems wierd
20140529 20:34:01< gfgtdf> iceiceice_: i think we should make a field "continie_to_next_scenario" in end_level_data.transient and depend whether to store cayyrovr to that
20140529 20:34:37< iceiceice_> why not to just always save carryover?
20140529 20:36:06< gfgtdf> iceiceice_: hmm maybe, but thre is still this: https://github.com/wesnoth/wesnoth/blob/master/src/playcampaign.cpp#L538
20140529 20:37:40< iceiceice_> hmm
20140529 20:37:49< iceiceice_> next to the game_state_.snapshot = config() lines
20140529 20:37:53< iceiceice_> culd also set carryover to blank
20140529 20:38:23< iceiceice_> hmm i think we didnt fix the init side bug
20140529 20:38:36< iceiceice_> or something went wrong in what i acutally committed, not sure
20140529 20:38:43-!- fabi [~fabi@wesnoth/developer/fendrin] has quit [Remote host closed the connection]
20140529 20:39:19< iceiceice_> the replays have init_side_done=no...
20140529 20:39:53-!- shadowm_desktop [ignacio@wesnoth/developer/shadowmaster] has quit [Ping timeout: 258 seconds]
20140529 20:42:01-!- Bodhi-Baum [~Bodhi@dslb-084-063-011-251.pools.arcor-ip.net] has quit [Quit: Verlassend]
20140529 20:42:13< iceiceice_> hmm i should have tested after i merged
20140529 20:42:16< gfgtdf> iceiceice_: you reloaded tem in sp or mp ?
20140529 20:42:18< iceiceice_> maybe i will test at the commit maybe just before the merge
20140529 20:42:20< iceiceice_> reload in mp
20140529 20:42:33< iceiceice_> it might be that we changed maybe_do_init_side or soething?
20140529 20:42:47< iceiceice_> after my branch departed?
20140529 20:42:50< iceiceice_> so it fooled my tests?
20140529 20:45:03< gfgtdf> iceiceice_:  hmm #ifver WESNOTH_VERSION >= 1.12.999 gives me an error now :S
20140529 20:45:31< iceiceice_> hmm idk what that means
20140529 20:45:41< iceiceice_> uhm...
20140529 20:45:46< iceiceice_> i merged vc project file changes from aquileia
20140529 20:45:53< iceiceice_> like early this morning
20140529 20:46:47< gfgtdf> iceiceice_: hmm idk form when my ersion is i better make freh rebuild
20140529 20:46:50< gfgtdf> version
20140529 20:47:12< aquileia> gfgtdf, iceiceice_ : I completely rebuild from the project files in the repo to assure everything works...
20140529 20:47:23-!- prophile [~alynn@oftn/member/prophile] has quit [Quit: The Game]
20140529 20:47:42< iceiceice_> hmm ok
20140529 20:47:50< iceiceice_> wesnoth is successfully getting init_side_done = no when it reloads the save
20140529 20:47:53< iceiceice_> but it doesnt init the side then
20140529 20:48:16< iceiceice_> at this commit: 28d0e2df0e8d190c3660b2a09d9b4b5a374738fd
20140529 20:48:27< iceiceice_> (https://github.com/wesnoth/wesnoth/commit/28d0e2df0e8d190c3660b2a09d9b4b5a374738fd)
20140529 20:48:58< iceiceice_> going to try at this commit now
20140529 20:49:04< iceiceice_> https://github.com/wesnoth/wesnoth/commit/77ead906927bec69d9136d349d637458f7a553f0
20140529 20:49:10-!- ancestral [~ancestral@17.114.45.98] has joined #wesnoth-dev
20140529 20:51:28< iceiceice_> hmm it doesnt work there either :/
20140529 20:51:31< iceiceice_> so my testing was flawed somehow
20140529 20:53:41< gfgtdf> iceiceice_: hm i didn nothing an now it works ...
20140529 20:54:03< gfgtdf> (the ivver)
20140529 20:54:06< gfgtdf> ifver
20140529 20:54:18< iceiceice_> ide voodoo :O
20140529 20:55:03< gfgtdf> iceiceice_: maybe it was some error relaetd to 2 wesnoth instances trying to read teh file at the same time or something
20140529 20:59:52< iceiceice_> hmmm
20140529 21:00:08< iceiceice_> it gets confused about the value of init side done by the time it enters maybe do init side
20140529 21:01:24< iceiceice_> ooh
20140529 21:01:27< iceiceice_> look in do_init_side
20140529 21:02:32< iceiceice_> there is special code to block it from happening on a reload
20140529 21:02:33< iceiceice_> i think
20140529 21:02:39-!- travis-ci [~travis-ci@ec2-54-82-224-50.compute-1.amazonaws.com] has joined #wesnoth-dev
20140529 21:02:39< travis-ci> [travis-ci] wesnoth/wesnoth#2860 (master - 28ffcc5 : gfgtdf): The build was broken.
20140529 21:02:39< travis-ci> [travis-ci] Build details : http://travis-ci.org/wesnoth/wesnoth/builds/26332068
20140529 21:02:39-!- travis-ci [~travis-ci@ec2-54-82-224-50.compute-1.amazonaws.com] has left #wesnoth-dev []
20140529 21:03:46< gfgtdf> hmm the victory_never_fail_two  timeouted
20140529 21:04:14-!- mattsc_ [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20140529 21:05:27-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Ping timeout: 252 seconds]
20140529 21:05:28-!- mattsc_ is now known as mattsc
20140529 21:06:43-!- gfgtdf_ [~chatzilla@e177118255.adsl.alicedsl.de] has joined #wesnoth-dev
20140529 21:08:16< iceiceice_> i think all that code in do_init_side needs to change, the assumptions are bad
20140529 21:08:35< iceiceice_> it should defer to "init_side_done" 
20140529 21:08:36-!- happygrue [~happygrue@wesnoth/developer/wintermute] has joined #wesnoth-dev
20140529 21:08:41-!- gfgtdf [~chatzilla@f054158248.adsl.alicedsl.de] has quit [Ping timeout: 264 seconds]
20140529 21:08:45< iceiceice_> and not make stupid guesses based on is_loading or is_replay most liekly
20140529 21:09:07< iceiceice_> i guess it is probably very old...
20140529 21:09:18-!- gfgtdf_ is now known as gfgtdf
20140529 21:09:18< iceiceice_> this is probably why this case has been broken
20140529 21:09:26< iceiceice_> gfgtdf: i reran your travis build,
20140529 21:09:30< iceiceice_> if it happens again... idk
20140529 21:09:44< iceiceice_> if it doesnt happen again i will increase travis timelimit for unit etsts
20140529 21:12:05-!- Gambit [~derek@wesnoth/developer/grickit] has joined #wesnoth-dev
20140529 21:15:31-!- fabi [~fabi@91-67-44-12-dynip.superkabel.de] has joined #wesnoth-dev
20140529 21:15:31-!- fabi [~fabi@91-67-44-12-dynip.superkabel.de] has quit [Changing host]
20140529 21:15:31-!- fabi [~fabi@wesnoth/developer/fendrin] has joined #wesnoth-dev
20140529 21:16:13< iceiceice_> hmm ok
20140529 21:16:23< iceiceice_> i'm going to delete the "it_is_a_new_turn" variable i think... ?
20140529 21:16:36< iceiceice_> idk will have to think about this...
20140529 21:16:54< iceiceice_> i think both loading_game and it_is_a_new_turn should go away
20140529 21:20:47< iceiceice_> hmm well this is a bit odd...
20140529 21:21:01< gfgtdf> iceiceice_: why do you want to delete it_is_a_new_turn ?
20140529 21:21:03< iceiceice_> it looks like loading_game and init_side_done_ are essentially the same somehow
20140529 21:21:11< iceiceice_> they both are read from config,
20140529 21:21:22< iceiceice_> and boht set only, to false, and true, respectively, in do_init_side
20140529 21:21:36< iceiceice_> init_side_done is read from config, directly on construction,
20140529 21:21:49< iceiceice_> loading_game is derived based on whether or not there is a "playing_side" variable appearing in the level...
20140529 21:21:59< gfgtdf> i think it_is_a_new_turn is mainly use to calculate wheh new turn events shoudl be fored
20140529 21:22:28< iceiceice_> y maybe it_is_a_new_turn should live but i think those events should fire if init side was not done!
20140529 21:22:33< iceiceice_> or those will be skipped on reload
20140529 21:23:03< iceiceice_> i think we should perhaps delete "loading_game"
20140529 21:23:10< iceiceice_> and replace with "!init_side_done" everywhere
20140529 21:23:21< iceiceice_> or else delete "init_side_done" and replace with "!loading_game" everywhere
20140529 21:23:23< gfgtdf> idk what loading_game does
20140529 21:23:40< iceiceice_> idk there mgiht be some corner case where the logic of these shold be different but i dont see it right now
20140529 21:24:14< gfgtdf> init_side_done not related  to reloasing i think
20140529 21:24:16< gfgtdf> reloading
20140529 21:24:28< gfgtdf> gfgtdf: it least it was not before
20140529 21:24:45< iceiceice_> all the logic in do_init_side_done is assuming that loading_game => not init_side_done
20140529 21:24:54< iceiceice_> *do_init_side
20140529 21:25:05-!- sea [~sea@fsf/member/sea4ever] has joined #wesnoth-dev
20140529 21:25:34< sea> Hi all, I'm looking at wesnoth AI making. It looks so far like the AI programs have to be compiled into the wesnoth binary. Is there any way so far to develop them separately? 
20140529 21:25:59< sea> It's much more useful to have them separate, so that they can be transferred around for AI tournaments and so on.
20140529 21:26:01< iceiceice_> sea: yes, you can use lua
20140529 21:26:23< iceiceice_> since like wesnoth 1.8 or something
20140529 21:26:25< sea> Ah, there's a lua interpreter in there. 
20140529 21:26:31< iceiceice_> look on the wiki
20140529 21:27:03< iceiceice_> unfortunately alot of AI documentation is badly out of date, but most of the stuff related to lua ai is accurate because it is very actively developed still
20140529 21:27:23< gfgtdf> hm ye i just looked at teh code an i think loading_game is strange
20140529 21:27:33< iceiceice_> it doesnt make sense,
20140529 21:27:41< iceiceice_> they are making decisions like "if we are loading the game then dont heal the sides units"
20140529 21:27:46< iceiceice_> we have to do that if init side didnt happen!
20140529 21:28:24< iceiceice_> maybe shold try just in this fucntion for now but i wonder if more broadly those two fields can be merged
20140529 21:28:24< gfgtdf> ye spo delete loading_game ?
20140529 21:28:42< iceiceice_> because play_controller.cpp is pretty long and i dont know what all the consequences are atm...
20140529 21:28:52< iceiceice_> but if we can make a branch that fixes this bug then we are on the right track
20140529 21:30:12< iceiceice_> im going to run git blame to see hwo old these lines
20140529 21:30:13< iceiceice_> are
20140529 21:31:28< iceiceice_> the whole block "if !loading game { decide to fire events or not } " is from 2009 - 2011
20140529 21:37:08-!- sea [~sea@fsf/member/sea4ever] has left #wesnoth-dev ["ERC Version 5.3 (IRC client for Emacs)"]
20140529 21:38:25< iceiceice_> hmm so i think 
20140529 21:38:33< iceiceice_> the only reason why the work around for the side not inited bug exists
20140529 21:38:39< iceiceice_> is because in replay_controller constructor, 
20140529 21:38:43< iceiceice_> loading_game is set to false
20140529 21:38:56< iceiceice_> which is also a bit bizarre :/
20140529 21:39:04< gfgtdf> not its not
20140529 21:39:15< gfgtdf> we warnt reloaing a game in replac controller
20140529 21:39:17< gfgtdf> arent
20140529 21:39:23< gfgtdf> are a playing it form the beginning
20140529 21:39:35< iceiceice_> hmm
20140529 21:39:57< iceiceice_> i mean clearly it changes which events are fired in the replay vs the original one
20140529 21:40:15< gfgtdf> no why should it
20140529 21:40:23< iceiceice_> i mean thats the work around
20140529 21:40:35< iceiceice_> if you get a game where player dc'd without getting initialized and you save it,
20140529 21:40:42< iceiceice_> you can repair it by watching it in the replay viewer and saving that
20140529 21:40:51< iceiceice_> because when you load normally, you wont be inited
20140529 21:40:57< gfgtdf> ye
20140529 21:40:59< iceiceice_> but when you watch in replay viewer you will be
20140529 21:41:13< iceiceice_> idk it depends how loading_game_ variable is used,
20140529 21:41:17< iceiceice_> i shold have read the code before i said it
20140529 21:41:28< gfgtdf> its wring in teh replay and it was fron in teh original game
20140529 21:41:29< iceiceice_> play_controller.cpp is long 
20140529 21:41:31< gfgtdf> wrong
20140529 21:41:43< gfgtdf> s/fron/wrng
20140529 21:41:44< iceiceice_> so i think it might be that,
20140529 21:41:49< iceiceice_> no!
20140529 21:41:52< iceiceice_> i dont think thats true
20140529 21:41:58< iceiceice_> i think the reason it inits the side is differnet
20140529 21:42:17< iceiceice_> i dont thnk there is init_side in the replay when it does that,
20140529 21:42:25< iceiceice_> it does it because it knows that that it suppsoed to happen next
20140529 21:42:29< iceiceice_> and there is no latency when you watch the replay
20140529 21:43:03< iceiceice_> i mean clearly it wasnt in the original game, when you saved that side had not been inited yet
20140529 21:43:09< gfgtdf> replays call do_init_side directly which is unrelated to init_side_done
20140529 21:43:25< gfgtdf> at least i think that
20140529 21:43:31< iceiceice_> y but the impl of do_init_side only acutally happens if loading_game = false apparently
20140529 21:43:45< iceiceice_> and replays have serendepitiously set this to be false always
20140529 21:43:49< iceiceice_> even if it was true in the original game
20140529 21:44:05< gfgtdf> it was never true in teh original game
20140529 21:44:13< iceiceice_> no it is true in turn 1 i think
20140529 21:44:17< iceiceice_> hmm
20140529 21:44:21< iceiceice_> y you are right
20140529 21:44:22< iceiceice_> only if you reload
20140529 21:44:26< gfgtdf> no its one true when you relaod a game
20140529 21:44:32< iceiceice_> but maybe that has something to do with why replays of reloaded are screwy
20140529 21:44:50< iceiceice_> the replay viewer assumes basically that it is not watching a reloaded game
20140529 21:45:10< iceiceice_> regardless of what the level file tells it
20140529 21:45:16< gfgtdf> if reloading changes the gamestate than this not a bug withreplays
20140529 21:45:46< gfgtdf> reloadinag should not change the game
20140529 21:45:46< iceiceice_> it depends on what the meaning of variables in play controlelr is supposed to be
20140529 21:45:50< iceiceice_> are those part of the game state or not?
20140529 21:45:58< iceiceice_> i thikn some are and some are not
20140529 21:46:00< gfgtdf> reloading shouldnt change teh variables neiugher
20140529 21:46:07< iceiceice_> it has to for some of them,
20140529 21:46:23< iceiceice_> clearly the things related to mouse event handlers and such
20140529 21:46:37< iceiceice_>  im ean there are some member variables that it wouldtn even make sense to save
20140529 21:46:43< iceiceice_> and i dont think all of the play_controller variables are serialized.
20140529 21:46:46< iceiceice_> in to_config
20140529 21:47:02< iceiceice_> loading_game_ is not serialized
20140529 21:47:09< iceiceice_> therefore it is not considered part of the game_state i guess
20140529 21:47:14< iceiceice_> but it affects the game state directly
20140529 21:47:14< gfgtdf> ye
20140529 21:47:21< iceiceice_> by deciding whether or not some events are fired
20140529 21:47:47< gfgtdf> not realy it it just a wirkarount to not fir these event alwas as soon as relaoing a game
20140529 21:47:53< gfgtdf> workaround
20140529 21:48:02< iceiceice_> ok 
20140529 21:48:11< gfgtdf> becsue we always call init_side in play_side begin
20140529 21:48:13< iceiceice_> do you not agree with me that, whent eh replay_controller writes in its constructor "loading_game = false"
20140529 21:48:22< iceiceice_> that means that the replay_viewer is assuming tha tit is not watching a reloaded_Game?
20140529 21:48:38< iceiceice_> and will thereafter fire events differnetly based on that assumption?
20140529 21:48:46-!- shadowm_desktop [ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20140529 21:49:13< gfgtdf> no the replay can laways safeley asume that this is not a relaoded game becasue relaosing doesnt cahnge the gamestate so it doesnt matter
20140529 21:49:17< gfgtdf> always
20140529 21:50:13< iceiceice_> what do you mean by "reloading doesn't change the gamestate"
20140529 21:50:23< iceiceice_> precisely
20140529 21:50:44< iceiceice_> you dont consider the "is_reloading" variable to be part of the gamestate?
20140529 21:50:47< iceiceice_> sorry
20140529 21:50:49< iceiceice_>  "is loading"?
20140529 21:50:52< gfgtdf> no i dont
20140529 21:51:00< iceiceice_> hmm then why does it determine whether we fire events or not?
20140529 21:51:23< gfgtdf> it prvent event form fireing that are already foired
20140529 21:51:41< iceiceice_> but is_reloading doesn't indicate that
20140529 21:51:46< iceiceice_> init_side_done does
20140529 21:52:04< iceiceice_> because init_side_done_ is properly part of the game state and is actually serialized
20140529 21:53:13-!- [Relic] [~relic@99-58-54-211.lightspeed.milwwi.sbcglobal.net] has joined #wesnoth-dev
20140529 21:53:37< gfgtdf> hm ye bt then we would have to set init_side_done to false on turn end and not shortl before init_side, also there are gui related things ini do_init_side that arenot part of teh gamestate
20140529 21:53:55< iceiceice_> y so i think the gui related things should be done by loading_game
20140529 21:54:03< iceiceice_> but questions liek whether to fire an event or not
20140529 21:54:10< iceiceice_> shold be based only on variables in the gamestate i think
20140529 21:54:20< gfgtdf> iceiceice_: ye that would be preferable
20140529 21:54:36< iceiceice_> there should be like, a less confusing way to arrange the code so this kind of stuff is clearer...
20140529 21:54:42< iceiceice_> like, move all the variables that get serialized to game_state
20140529 21:54:51< iceiceice_> and never serialize anything else in play_controller
20140529 21:54:54< iceiceice_> idk how feasible that is though
20140529 21:55:02< iceiceice_> wuold have to review
20140529 21:55:36< iceiceice_> hmm maybe its not bad if the replay_viewer writes loading_game = false though
20140529 21:55:41< iceiceice_> i guess you are right about that
20140529 21:56:04< iceiceice_> i think the loading_game should not affect whether events are fired or not though
20140529 21:57:02< gfgtdf> iceiceice_: loading_game_ &= init_side_done_; at teh begining of play_controller::init_side would maybe be a eways fix for the problem
20140529 21:57:57< gfgtdf> an eays fix*
20140529 21:58:03< gfgtdf> eays*
20140529 21:59:13-!- ancestral [~ancestral@17.114.45.98] has quit [Quit: ancestral]
20140529 21:59:58< gfgtdf> iceiceice_:  hm but that still not perfect i agree
20140529 22:01:12< iceiceice_> gfgtdf: i made a branch yesterday to make a game_board object to sit between play_controller and the unit_map and game_map
20140529 22:01:27< gfgtdf> iceiceice_: i think its maybe just wrng to call init_side alyways on relaoding
20140529 22:01:33< gfgtdf> wrong
20140529 22:01:44< iceiceice_> hmm
20140529 22:02:22< iceiceice_> because there are [init_side] records in the replay?
20140529 22:02:57< gfgtdf> iceiceice_: i also think you'll get most likeley get an error related to synced_checkup on realong with tzh curent implementation
20140529 22:03:04< gfgtdf> in teh log i mean
20140529 22:04:05< iceiceice_> i wonder what was happening that i thought that we fixed this
20140529 22:04:33< iceiceice_> i remember compiling and having it work at some point.... leader getting green light and having moves...
20140529 22:05:05< iceiceice_> should figure out how to make a unit test to test reloaded games
20140529 22:05:12< gfgtdf> iceiceice_: why is init_side virtual?
20140529 22:05:45< iceiceice_> idk
20140529 22:06:32< iceiceice_> does that have any effect?
20140529 22:06:48< iceiceice_> resources::controler is always a play_controller anyways, right?
20140529 22:07:02< iceiceice_> does naything override?
20140529 22:08:57< iceiceice_> idk i am going to get lunch and see if i can think of what to do
20140529 22:09:04< gfgtdf> no i think init_side is never overridden
20140529 22:09:11< iceiceice_> it might be that changing this is too risky close to release
20140529 22:09:31< gfgtdf> the virutual ?
20140529 22:09:37< iceiceice_> no the init_side_done vs loading
20140529 22:09:52< gfgtdf> hm but the more i look at it th more i think it's bugy
20140529 22:10:35< iceiceice_> i think that the bugs are limited to these mp disconnect scenarios
20140529 22:10:39< iceiceice_> and they have workarounds
20140529 22:10:46< iceiceice_> so i would like to not make the situation worse
20140529 22:10:56< iceiceice_> its very hard to test all the cases :/
20140529 22:12:18< iceiceice_> idk if we can convince ourselves that the distinction between is_loading and init_side_done definitely matters only in edge cases then i would be happy to change things around
20140529 22:12:26< gfgtdf> hmm no even normal reloaings are bugged (a little)
20140529 22:12:46< gfgtdf> save any game after attacking a unit and then reload it
20140529 22:12:47< iceiceice_> all i know is, the last time we changed a core function in play_controller that was this old was with check_victory
20140529 22:12:57< gfgtdf> and then look in teh stderr file
20140529 22:12:58< iceiceice_> and that became the zombie bug that wouldn't die
20140529 22:13:37< gfgtdf> you'll most likley get a report like We called random 0 times, but the original game called random -99 times
20140529 22:13:42< iceiceice_> i guess we changed the constructor of play controller just before this...
20140529 22:14:22< iceiceice_> hmm ok
20140529 22:14:24< iceiceice_> i will come back
20140529 22:14:26-!- iceiceice_ [~chris@cpe-66-108-20-80.nyc.res.rr.com] has quit [Quit: Leaving]
20140529 22:26:30-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20140529 22:55:21-!- wesbot changed the topic of #wesnoth-dev to: string+feature freeze active on 1.12 | 245 bugs, 345 feature requests, 29 patches | Logs: http://irclogs.wesnoth.org | Alternate logs: http://wesnoth.debian.net | Don't paste on IRC! Use a pastebin: http://pastebin.com | http://imagebin.org
20140529 23:01:35< irker675> wesnoth: gfgtdf wesnoth:master e50a7785fdf1 / src/ (play_controller.cpp play_controller.hpp): fix reloading http://git.io/zwRBvA
20140529 23:06:38-!- vultraz [~chatzilla@124.109.10.167] has quit [Read error: Connection reset by peer]
20140529 23:06:56-!- ancestral [~ancestral@17.114.45.98] has joined #wesnoth-dev
20140529 23:07:15-!- vultraz [~chatzilla@124.109.10.167] has joined #wesnoth-dev
20140529 23:09:54-!- mjs-de [~mjs-de@g228033017.adsl.alicedsl.de] has quit [Remote host closed the connection]
20140529 23:13:35-!- vultraz [~chatzilla@124.109.10.167] has quit [Changing host]
20140529 23:13:35-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20140529 23:17:17-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20140529 23:21:26-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: Traveling until May 31]
20140529 23:25:23< gfgtdf> iceiceice: i think we shoudl backport the fix above to 1.12
20140529 23:28:55-!- iceiceice_ [~chris@cpe-66-108-20-80.nyc.res.rr.com] has joined #wesnoth-dev
20140529 23:29:06< iceiceice_> gfgtdf: i see your commit
20140529 23:29:12< iceiceice_> does it fix the thing about stderr?
20140529 23:29:25< gfgtdf> iceiceice_: yes it does
20140529 23:29:43< iceiceice_> ok i think we should consider refactoring from here then to try to simplify things
20140529 23:30:17< iceiceice_> because it was already complicated and now theres another variable added... and i think theres some assertions we can make
20140529 23:30:43< gfgtdf> iceiceice_: hm ok
20140529 23:31:49-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20140529 23:35:50-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [Ping timeout: 258 seconds]
20140529 23:36:14< gfgtdf> iceiceice_: is shoudl also fix reload games with init_side_done =no.
20140529 23:36:45< gfgtdf> iceiceice_: you have a reason against backporting it into 1.12 in it's current form ?
20140529 23:37:24< iceiceice_> i would just like to see if it can be simpler without changing it
20140529 23:37:37< iceiceice_> no particular reason not to backport esp. if it fixes bugs
20140529 23:40:07< iceiceice_> i only actually read this function like an hour ago, and i dont know the rest of it as well as you, so right now i think i dont fully understand it, that's all
20140529 23:42:58< gfgtdf> iceiceice_: wel it's just that in case that we reload a scenario we visualy need to initilize the turn gui. And to do that we reuse some code of do_init_side with only_visual = true.
20140529 23:50:00-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20140529 23:55:07< iceiceice_> gfgtdf: is game_data.phase serialized?
--- Log closed Fri May 30 00:00:34 2014