--- Log opened Wed Mar 09 00:00:03 2016
20160309 00:11:49-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20160309 00:34:15-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has quit [Quit: Connection closed for inactivity]
20160309 00:36:31-!- louis94 [~~louis94@91.178.240.71] has quit [Quit: Konversation terminated!]
20160309 00:51:18< shadowm> I know what the equivalent for phpBB 3.0.x is/was. I'm not interested in installing any phpBB mods until I can get around to upgrading to 3.1.x (which uses extensions instead of file edit mods).
20160309 00:51:56< shadowm> (And the reason I never installed the forum mod is precisely because it was far too involved in terms of edit count and extent.)
20160309 00:52:35< shadowm> MW uses proper extensions too so maybe I'll take a look later.
20160309 01:01:31-!- mjs-de [~mjs-de@f048105065.adsl.alicedsl.de] has quit [Remote host closed the connection]
20160309 01:03:03-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Ping timeout: 240 seconds]
20160309 01:17:12< celticminstrel> By "equivalent" I kinda meant "uses the same underlying syntax highlighting framework".
20160309 01:22:07-!- prkc [~prkc@gateway/vpn/privateinternetaccess/prkc] has quit [Remote host closed the connection]
20160309 01:24:05< shadowm> vultraz: "It also provides us a platform for working on several exciting changes in the months ahead." What changes? People will ask.
20160309 01:24:28< shadowm> Just pointing that out. Maybe you want to keep it a surprise (though I'm pretty sure I know what you are thinking of).
20160309 01:24:55-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160309 01:28:07< shadowm> vultraz I ... ugh.
20160309 01:28:19< shadowm> You just copy-pasted the contributors section from the 1.13.2 release.
20160309 01:30:50< shadowm> loonycyborg: Do you still have the Xdelta SHA256? (No, I won't calculate it from one of the uploaded files)
20160309 01:31:16< shadowm> Oh, also, there's no Xdelta directly from 1.13.2 to 1.13.4.
20160309 01:31:47< shadowm> Okay, bnever mind, we don't do that.
20160309 01:32:16< shadowm> Look, nobody will miss the Xdelta links from the topic.
20160309 01:32:35< shadowm> In that case we don't need the SHA256s for them either.
20160309 01:36:52< shadowm> vultraz: I added the two screenshots I took, fixed the formatting of the collapsible blocks section, and fixed the links for the source and Windows packages, and the changelog links (they were pointing to 1.13.2).
20160309 01:37:08< shadowm> I believe that's all I'm supposed to do for now.
20160309 01:37:43< shadowm> Oh and removed the contributors section since I doubt you are planning on keeping track of that, and it's not my responsibility anymore.
20160309 01:38:31< shadowm> I wonder if you'll even read this.
20160309 02:13:16-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has joined #wesnoth-dev
20160309 02:18:57-!- gfgtdf_ [~chatzilla@f054060056.adsl.alicedsl.de] has joined #wesnoth-dev
20160309 02:20:49-!- gfgtdf [~chatzilla@f054139219.adsl.alicedsl.de] has quit [Ping timeout: 240 seconds]
20160309 02:20:52-!- gfgtdf_ is now known as gfgtdf
20160309 02:42:36-!- gfgtdf [~chatzilla@f054060056.adsl.alicedsl.de] has quit [Quit: ChatZilla 0.9.92 [Firefox 44.0.2/20160210153822]]
20160309 03:46:18-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20160309 04:25:27-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160309 04:28:42-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has joined #wesnoth-dev
20160309 04:30:07-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 260 seconds]
20160309 04:31:21< ancestral> vultraz: Finishing the build tonight. Sorry it’s taking longer than I thought.
20160309 04:31:33< ancestral> I literally bought a house this morning.
20160309 04:31:47 * ancestral signed his life away. More or less.
20160309 04:31:51< celticminstrel> Vultraz seems to be missing recently.
20160309 04:31:55< celticminstrel> Weird.
20160309 04:32:14 * celticminstrel has an almost working Pygments lexer for WML.
20160309 04:32:35< ancestral> I am also getting an error with Boost 1.60
20160309 04:32:39< celticminstrel> Aw.
20160309 04:32:41< ancestral> But, worst case, I revert back down
20160309 04:32:52< celticminstrel> Is it about boost::none_t?
20160309 04:33:28< celticminstrel> Or perhaps I should just ask... what's the error?
20160309 04:33:57< shadowm> IIRC only the unit tests can't be built with Boost 1.60, the game and server should be good.
20160309 04:34:15-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has quit [Quit: Connection closed for inactivity]
20160309 04:34:42< ancestral> celticminstrel: Wait, didn’t you want me to build Boost so you could run unit tests
20160309 04:35:32< ancestral> one sec
20160309 04:36:06< ancestral> celticminstrel: https://paste.ee/p/UXCj5
20160309 04:36:14< celticminstrel> ancestral: Yes, but as long as the deps for that are actually there, I can probably deal with getting them to actually build.
20160309 04:36:23< celticminstrel> Eventually.
20160309 04:36:44< celticminstrel> That looks like Boost.Regex.
20160309 04:37:20< celticminstrel> Are you linking with libboost_regex?
20160309 04:37:37< wedge009> Ivanovic: Okay, thanks. Just wanted confirmation that it was okay to do it this way as I understand po changes are normally supposed to go through you.
20160309 04:37:44< celticminstrel> I suppose it'd be weird if you weren't, given you're using XCode...
20160309 04:38:02< shadowm> https://gna.org/bugs/index.php?23897 -- "Neither 1.12.x nor 1.13.x unit tests build with boost 1.59.0"
20160309 04:38:46< ancestral> Okay so
20160309 04:38:57< ancestral> I think I need to build new boost frameworks then
20160309 04:39:36< celticminstrel> The unit tests can't be built at all on XCode because the XCode project has not been properly updated for them, and the downloadable package is missing the dependencies.
20160309 04:39:42< shadowm> What you pasted looks like something else entirely, not really sure what though.
20160309 04:39:45< celticminstrel> So, I wanted to resolve that.
20160309 04:39:51< celticminstrel> Eventually.
20160309 04:40:04< ancestral> Alright, so basically, keep 1.57 or whatever I had before
20160309 04:40:15< celticminstrel> Probably fine.
20160309 04:40:22< celticminstrel> I think the current version included is 1.54.
20160309 04:40:23< ancestral> Build this, and then figure out how to make 1.60 wrk later
20160309 04:41:59< ancestral> celticminstrel: What is MPI support?
20160309 04:42:08< celticminstrel> Eh?
20160309 04:42:40< ancestral> http://stackoverflow.com/questions/13143409/
20160309 04:42:56< ancestral> Probably better *described* elsewhere
20160309 04:43:19< ancestral> I’m guessing Wesnoth doesn’t use it
20160309 04:43:20< celticminstrel> I'm not sure what that is, but it's probably some sort of processor feature or something.
20160309 04:43:20< ancestral> http://www.boost.org/doc/libs/1_60_0/doc/html/mpi.html
20160309 04:43:30< ancestral> Multi processor stuff
20160309 04:43:37< ancestral> No
20160309 04:43:41< celticminstrel> Ah okay.
20160309 04:43:47< ancestral> http://www.mcs.anl.gov/research/projects/mpi/
20160309 04:43:55< ancestral> Alright, ignoring
20160309 04:50:01-!- nurupo [~nurupo.ga@unaffiliated/nurupo] has quit [Ping timeout: 240 seconds]
20160309 04:52:04-!- nurupo [~nurupo.ga@unaffiliated/nurupo] has joined #wesnoth-dev
20160309 05:10:09< celticminstrel> Okay, I give up.
20160309 05:11:00< celticminstrel> It fails on at least one edge case, but if anyone's interested, it's at http://celmin.pwcsite.com/wesnoth/WmlLexer.py
20160309 05:12:08< celticminstrel> The edge case it was failing at was something like {MACRO_NAME (type=Giant Spider) 80 ()}
20160309 05:12:26< celticminstrel> The parenthesized key=attribute, to be precise.
20160309 05:13:11< celticminstrel> I don't know if it handles $x?y| syntax correctly; my test files were for 1.12, so there wasn't any of that.
20160309 05:13:26< celticminstrel> It does have allowance for it, I'm just not certain it actually works.
20160309 05:13:48 * celticminstrel might continue hacking on it later. Still want to add WFL support.
20160309 05:13:58< celticminstrel> ...oh yeah, I forgot $() syntax. Oh well. Later.
20160309 05:17:34< ancestral> Built Boost 1.60 through Homebrew (took a while)
20160309 05:17:41< celticminstrel> Fun.
20160309 05:17:52< celticminstrel> I wonder if I should post this on the forum...
20160309 05:19:40< ancestral> Shoot, I know I asked this before
20160309 05:19:48< ancestral> celticminstrel: Do you know what -mt is for?
20160309 05:20:10< ancestral> Like, libboost_regex-mt.dylib vs. libboost_regex.dylib
20160309 05:20:22< ancestral> Multi-threaded?
20160309 05:20:51< celticminstrel> I think so. I don't really see the point of the suffix.
20160309 05:21:32< celticminstrel> I mean, why are there even two different versions.
20160309 05:21:48-!- SigurdFD [~SigurdFD@dynamic-acs-72-23-176-151.zoominternet.net] has quit []
20160309 05:24:02< ancestral> I think is answering my questions: http://stackoverflow.com/questions/11526193/proper-linking-of-boost-multithreaded-libraries-on-linux
20160309 05:28:56< ancestral> Basically, use all -mt, or none
20160309 05:29:33< ancestral> Hmm, yeah, getting more errors
20160309 05:29:56< ancestral> gettext_boot.o is not happy
20160309 05:30:02< ancestral> *gettext_boost.o
20160309 05:36:21-!- janebot_ [~Gambot@grickit.us] has quit [Remote host closed the connection]
20160309 05:36:27-!- janebot_ [~Gambot@grickit.us] has joined #wesnoth-dev
20160309 05:37:09< celticminstrel> And I assume libboost_thread-mt.dylib doesn't exist, so really, use only -mt.
20160309 05:40:19< ancestral> I have libboost_thread-mt.dylib
20160309 05:40:24< ancestral> Yeah
20160309 05:40:35< celticminstrel> I meant libboost_thread.dylib, whoops
20160309 05:40:39< ancestral> Right
20160309 05:40:43< ancestral> You are correct
20160309 05:41:15< ancestral>       (anonymous namespace)::translation_manager::update_locale_internal() in gettext_boost.o
20160309 05:41:15< ancestral>       (anonymous namespace)::default_utf8_locale_name::default_utf8_locale_name() in gettext_boost.o
20160309 05:44:00< wedge009> Some pretty crazy stuff happening after PR #622: https://gna.org/bugs/?24226 (read comment #10)
20160309 05:51:09-!- Kwandulin [~Miranda@p200300760F7D7338A029F89888C96B5B.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160309 05:59:04< ancestral> Okay, so there are problems with source/gettext_boost.cpp
20160309 05:59:26< ancestral> src/gettext_boost.cpp
20160309 05:59:39< ancestral> Anyway, revertong back to previous boost
20160309 06:00:49-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20160309 06:04:33< ancestral> bug report filed
20160309 06:04:43< ancestral> vultraz: Hi, I’m building right now
20160309 06:04:49< vultraz> hm?
20160309 06:05:08< celticminstrel> vultraz: Check logs, shadowm said some stuff about the announcement.
20160309 06:05:16< vultraz> gah
20160309 06:06:15< vultraz> celticminstrel: can you paste them, I can't get to the logs site
20160309 06:06:29< vultraz> (DNS errors)
20160309 06:10:01-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Quit: wedge009]
20160309 06:11:19< ancestral> vultraz: https://paste.ee/p/xEUok
20160309 06:11:35< celticminstrel> Whoops, should've posted that I did it already.
20160309 06:11:59< vultraz> thanks, both of you
20160309 06:12:34< celticminstrel> That first bit was addressed to me though, not you..
20160309 06:12:53< ancestral> 20160309 04:31:21< ancestral> vultraz: Finishing the build tonight. Sorry it’s taking longer than I thought.
20160309 06:12:53< ancestral> 20160309 04:31:33< ancestral> I literally bought a house this morning.
20160309 06:13:05< vultraz> :O
20160309 06:13:11< vultraz> like, a house house?
20160309 06:13:57< ancestral> Well, a detached townhouse, but for all intents and purposes, taxes, responsibilities… yes. A house house ;-)
20160309 06:15:02< ancestral> Yeah, we made got a cashier’s check out for about $65k
20160309 06:15:40< ancestral> (that’s just the downpayment of course)
20160309 06:16:11< ancestral> Okay, built successfully. (I segwayed into trying Boost 1.60 but it didn’t go well.)
20160309 06:17:41< vultraz> that must be a hella good job you have
20160309 06:20:32< ancestral> Nah, my fiancée and I just saved a lot of money
20160309 06:20:48< ancestral> We’ll be making payments for up to 30 years
20160309 06:21:31-!- celticminstrel is now known as celmin|sleep
20160309 06:22:22< celmin|sleep> vultraz: See what I did today. http://celmin.pwcsite.com/wesnoth/WmlLexer.py
20160309 06:27:44-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20160309 06:33:55-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20160309 06:52:17-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160309 06:57:07-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 260 seconds]
20160309 06:57:28-!- mjs-de [~mjs-de@x4db6bd0c.dyn.telefonica.de] has joined #wesnoth-dev
20160309 06:58:03-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Ping timeout: 240 seconds]
20160309 07:19:18 * Aginor is done haunting the issuetracker for now
20160309 07:22:53< Aginor> it looks like there's no major fallout so far from SDL2
20160309 07:23:06< Aginor> or am I missing things?
20160309 07:26:17< zookeeper> Aginor, the issues with labels?
20160309 07:27:01< zookeeper> i guess there's no report for those
20160309 07:27:22< Aginor> zookeeper: pretend I'm unaware of it 
20160309 07:27:25< Aginor> please tell me :)
20160309 07:28:20< zookeeper> labels disappear when a menu is opened (right-click menu for instance), and also when you have a GUI2 dialog like preferences open, the labels get redrawn over the dialog when a tooltip in the dialog activates (test scenario is great for reproducing).
20160309 07:28:32< Aginor> hmm
20160309 07:28:44< Aginor> the former is old behaviour (pre SDL2)
20160309 07:28:48< Aginor> the latter is a regression
20160309 07:28:51< zookeeper> can't be
20160309 07:29:00< Aginor> I am pretty sure it is
20160309 07:29:12< Aginor> I'll investugate
20160309 07:29:17< Aginor> but spelled better
20160309 07:29:20< zookeeper> not in 1.12
20160309 07:29:34< Aginor> 1.13 though
20160309 07:29:38< Aginor> I'll compare
20160309 07:29:52< Aginor> zookeeper: thank you :)
20160309 07:30:55< zookeeper> i'm pretty sure it's relatively new in 1.13 too, since... i think i would have noticed, seeing how i've used the test scenario a lot
20160309 07:33:03< Aginor> I was messing about with it, not noticing I broke GUI2 dialogs
20160309 07:33:24< Aginor> I'm pretty sure labels are not supposed to be drawn while the GUI is active though
20160309 07:33:38< Aginor> but I'll compare to 1.12
20160309 07:33:46< Aginor> it shouldn't be hard to fix
20160309 07:36:46< zookeeper> oh? well, that's good
20160309 07:36:54< Aginor> someone who knows more about the origin of the wesnoth world might be interested in this: https://www.reddit.com/r/wesnoth/comments/49jiig/how_much_are_specific_wesnoth_characteristics/
20160309 07:37:11< Aginor> zookeeper: the draw-layering *should* make those things much easier :)
20160309 07:37:49-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20160309 07:40:20< Aginor> what's up with the forums? they appear to be down
20160309 07:40:27< Aginor> vultraz: do you know?
20160309 07:40:30< vultraz> all wesnoth services are down
20160309 07:40:38< Aginor> what's happened?
20160309 07:40:51< vultraz> I don't know
20160309 07:40:54< Aginor> www.wesnoth.org is still up
20160309 07:41:06< Aginor> who's looking after them?
20160309 07:41:18< vultraz> it's down for me
20160309 07:41:26< vultraz> shadowm might know
20160309 07:41:55< zookeeper> forums work for me, wiki doesn't
20160309 07:42:42< Aginor> interesting
20160309 07:42:50< Aginor> might be an isp/routing issue
20160309 07:43:11< Aginor> zealot:~ andreas$ ping www.wesnoth.org
20160309 07:43:11< Aginor> PING www.wesnoth.org (144.76.5.6): 56 data bytes
20160309 07:43:11< Aginor> 64 bytes from 144.76.5.6: icmp_seq=0 ttl=49 time=451.215 ms
20160309 07:43:12< Aginor> 64 bytes from 144.76.5.6: icmp_seq=1 ttl=49 time=473.041 ms
20160309 07:44:17< Aginor> wiki appears down
20160309 07:44:18< vultraz> I'm getting DNS errors here
20160309 07:45:23< Aginor> good call, so do I now when I'm digging more
20160309 07:53:25-!- boucman_work [~jrosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20160309 08:21:16-!- midzer [~quassel@p5B312EF2.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160309 08:37:45-!- Kwandulin [~Miranda@p200300760F7D7338A029F89888C96B5B.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160309 08:39:14< Aginor> https://gna.org/bugs/index.php?23488 closed :)
20160309 08:39:47< vultraz> yay, closed bugs :D
20160309 08:40:34< Aginor> speaking of which, have we gone through and closed all of the fixed bugs that were in the release?
20160309 08:41:14< vultraz> yes
20160309 08:41:20< vultraz> didn't you get the email spam not too long ago
20160309 08:41:39< Aginor> I may have forgotten
20160309 08:42:02< Aginor> it's been a stressfull weekend for me, I may not have paid that much attention
20160309 08:42:16< Aginor> I did
20160309 08:42:26< Aginor> I must just have marked them all as read and carried on
20160309 08:42:38< Aginor> good job :)
20160309 08:42:49< vultraz> we closed around 90 bugs
20160309 08:42:52< vultraz> probably more
20160309 08:45:55< vultraz> some of them from the dev cycle internally, some of them ancient, some of them wedge009 found to be no longer issues
20160309 08:51:38-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20160309 08:53:22-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has joined #wesnoth-dev
20160309 09:01:44< Aginor> wedge009 did a great job with that
20160309 09:04:07-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160309 09:08:13-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 248 seconds]
20160309 09:10:52< Soliton> looks like the dns issue is already resolved.
20160309 09:28:15< zookeeper> leader carryover in campaigns is broken
20160309 09:28:38< zookeeper> i hope that's happened since the release -.-
20160309 09:29:21< zookeeper> such as due to gfgtdf's changes
20160309 09:30:48< zookeeper> for example: start AOI, proceed to second scenario. erlornas gets created anew from the second scenario's [side] contents (and thus in this case his dialogue won't appear), whereas the one from the first scenario ends up on the recall list.
20160309 09:33:00< zookeeper> (i believe his dialogue won't appear because he won't actually be created with id=Erlornas as the [side] contents dictate, because there is already a unit with that id on the recall list, so he ends up with a generated id instead)
20160309 09:35:57-!- zombah [~zombah@2a02:28:3:1:214:4fff:fe47:5920] has quit [Quit: Leaving]
20160309 09:37:56< zookeeper> vultraz, ^ better not announce before we know if the release is affected
20160309 09:38:05< vultraz> :|
20160309 09:38:14< vultraz> ._______________________.
20160309 09:42:33-!- Appleman1234 [~Appleman1@KD106161154152.au-net.ne.jp] has quit [Ping timeout: 240 seconds]
20160309 09:42:51< zookeeper> your eyes are like stars
20160309 09:43:47-!- irker760 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20160309 09:43:47< irker760> wesnoth: Charles Dang wesnoth:master 29e19cd854e3 / / (11 files in 7 dirs): GUI2: implemented a Unit Preview Pane widget https://github.com/wesnoth/wesnoth/commit/29e19cd854e3089050caebc8b0f0683211678f64
20160309 09:43:50< irker760> wesnoth: Charles Dang wesnoth:master 99ea0aa71680 / / (3 files in 2 dirs): tunit_create: made use of unit preview pane widget https://github.com/wesnoth/wesnoth/commit/99ea0aa716802eb1e547a508e6621597d1c1dfd4
20160309 09:43:53< irker760> wesnoth: Charles Dang wesnoth:master 2b46bb7e46b8 / data/gui/default/widget/unit_preview_pane.cfg: tunit_preview_pane: added minimal definition (no image/title) https://github.com/wesnoth/wesnoth/commit/2b46bb7e46b83af9dc659c50147b9828f1f29bcc
20160309 09:43:56< irker760> wesnoth: Charles Dang wesnoth:master 42816df6f640 / / (7 files in 4 dirs): Converted Recruit dialog to GUI2 https://github.com/wesnoth/wesnoth/commit/42816df6f64022ffb48693eeec85f34e7be097a8
20160309 09:43:59< irker760> wesnoth: Charles Dang wesnoth:master 1265ee1fc9ac / / (18 files in 9 dirs): Merge pull request #623 from Vultraz/unit_preview_pane https://github.com/wesnoth/wesnoth/commit/1265ee1fc9ac3dd69ca57c78b84368965146fc03
20160309 09:46:19< loonycyborg> I'm getting a compile error from master
20160309 09:46:59< loonycyborg> src/actions/../team.hpp:306:72: error: operands to ?: have different types 'const string {aka const std::__cxx11::basic_string<char>}' and 'const t_string'
20160309 09:47:13< vultraz> noticed in travis as well
20160309 09:50:09< vultraz> let me test this fix
20160309 09:50:33< loonycyborg> guess just change to info_.side_name.str()
20160309 09:50:49< vultraz> that's what I'm testing
20160309 09:53:14-!- travis-ci [~travis-ci@ec2-50-17-73-77.compute-1.amazonaws.com] has joined #wesnoth-dev
20160309 09:53:15< travis-ci> wesnoth/wesnoth#8805 (master - 1265ee1 : Charles Dang): The build failed.
20160309 09:53:15< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/114736634
20160309 09:53:15-!- travis-ci [~travis-ci@ec2-50-17-73-77.compute-1.amazonaws.com] has left #wesnoth-dev []
20160309 09:56:28-!- zombah [~zombah@2a02:28:3:1:214:4fff:fe47:5920] has joined #wesnoth-dev
20160309 09:57:56< irker760> wesnoth: Charles Dang wesnoth:master cab4d9fe4803 / src/team.hpp: Attempt to fix a compile error in master https://github.com/wesnoth/wesnoth/commit/cab4d9fe48036f0e4079060a43d090d38479ff41
20160309 09:59:00-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20160309 10:00:19< Soliton> let's hope the compiler tells you that that is a bad idea.
20160309 10:03:49< zookeeper> vultraz, so what are you planning with these race+alignment icons? are you insisting on keeping adding them everywhere unless players protest them, or what?
20160309 10:05:08< vultraz> well, currently they're only used in 2 places
20160309 10:05:15< vultraz> I wasn't able to make them work in Attack
20160309 10:05:18< zookeeper> i know
20160309 10:05:20< vultraz> (work = look good)
20160309 10:05:33< vultraz> but yes, I also intend to use them in Recall
20160309 10:05:41< vultraz> and Unit List
20160309 10:05:49< zookeeper> the recruit dialog is completely off-sync with the right panel unit info, so presumably you intend to add them there too, because that kind of off-sync is unacceptable
20160309 10:08:20< vultraz> I don't want to delve into ThemeWML, no
20160309 10:08:22< vultraz> sorry
20160309 10:08:43< zookeeper> well then they can't stay, that's kind of obvious
20160309 10:08:56< zookeeper> so why are you putting time and work into them in the first place?
20160309 10:09:12< vultraz> The widget can be easily converted to use labels, you know
20160309 10:09:48< zookeeper> good! i hope we see that soon
20160309 10:13:53< vultraz> No
20160309 10:13:56< zookeeper> Aginor, there's also another probably related issue: if you start a campaign and try to recall, you get the message saying you have no recallable units. however, that dialog disappears almost immediately with a flicker, presumably due to a redraw or something.
20160309 10:14:41< zookeeper> i bet that applies to all similar message boxes, but that's the easiest to reproduce.
20160309 10:16:07< zookeeper> vultraz, well, at some point then?
20160309 10:16:15< vultraz> we'll see
20160309 10:16:26< vultraz> I'd like to get some community feedback on the use of icons, too
20160309 10:16:44< zookeeper> because you make it sound like you're just changing our interface in a few places but with no intention to actually do it everywhere, as such a change would require
20160309 10:17:53< vultraz> I didn't say the change can't happen in the main UI
20160309 10:17:58< vultraz> just that I don't want tobe the one to do it
20160309 10:18:25< vultraz> #PoliticianTalk
20160309 10:20:20-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has joined #wesnoth-dev
20160309 10:21:39< zookeeper> yes, delegating something like that to a hypothetical other person who would really enjoy doing it has maybe a 1% chance of working out
20160309 10:22:31< vultraz> ancestral is the only person on the team who knows ThemeWML
20160309 10:23:38< vultraz> I guinely have absolutely no idea what the person who designed themewml was thinking when they did so
20160309 10:23:41< vultraz> genuinely 
20160309 10:24:08< zookeeper> in any case, it would be reassuming if you'd promise to not invoke "but it's been there for a while and i put a lot of work into it and not everyone hates it" if/when the issue of interface discrepancy comes up
20160309 10:24:13< zookeeper> reassuring, even
20160309 10:25:37< vultraz> it would take less than 5 minutes to change those icons to labels
20160309 10:25:37< vultraz> in fact, I could add a special definition right now that uses only labels
20160309 10:25:56< zookeeper> sure, but it's not just the icons, it's this whole new unit info display thing you have going on
20160309 10:26:35< vultraz> waaiiittt wait wait wait
20160309 10:26:42< vultraz> I thought you were just on about the icons
20160309 10:27:06< vultraz> what's wrong with the display!
20160309 10:28:03< zookeeper> the discrepancy. i don't particularly like the new display, but if it's at least consistent with the right panel then i can live with it.
20160309 10:28:21< vultraz> what do you not like about the new display
20160309 10:28:27< vultraz> it's better than the old display :|
20160309 10:28:30< zookeeper> the discrepancy
20160309 10:28:34< vultraz> BESIDES that
20160309 10:28:42< vultraz> the layout?
20160309 10:28:44< vultraz> the spacing?
20160309 10:30:01< zookeeper> eh, i guess i don't specifically _dislike_ anything about it
20160309 10:30:35-!- travis-ci [~travis-ci@ec2-54-166-189-246.compute-1.amazonaws.com] has joined #wesnoth-dev
20160309 10:30:35< travis-ci> wesnoth/wesnoth#8806 (master - cab4d9f : Charles Dang): The build is still failing.
20160309 10:30:35< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/114739313
20160309 10:30:35-!- travis-ci [~travis-ci@ec2-54-166-189-246.compute-1.amazonaws.com] has left #wesnoth-dev []
20160309 10:39:11-!- Appleman1234 [~Appleman1@KD106161141205.au-net.ne.jp] has joined #wesnoth-dev
20160309 11:04:29-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has joined #wesnoth-dev
20160309 11:39:02-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20160309 11:46:05-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160309 11:46:40-!- ChipmunkV [~vova@d0017-2-88-172-31-68.fbx.proxad.net] has joined #wesnoth-dev
20160309 11:50:35-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 244 seconds]
20160309 12:10:45-!- ChipmunkV [~vova@d0017-2-88-172-31-68.fbx.proxad.net] has quit [Quit: ChipmunkV]
20160309 12:13:23-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160309 12:27:01-!- Samual [~Samual@xonotic/core-team/Samual] has quit [Ping timeout: 240 seconds]
20160309 12:27:02-!- clavi [~clavi@163.172.10.77] has quit [Ping timeout: 240 seconds]
20160309 12:27:02-!- _laco [~laco@static.95.25.4.46.clients.your-server.de] has quit [Ping timeout: 240 seconds]
20160309 12:27:02-!- fabi [~quassel@wesnoth/developer/fendrin] has quit [Ping timeout: 240 seconds]
20160309 12:27:07-!- fendrin [~quassel@176.7.53.120] has joined #wesnoth-dev
20160309 12:27:07-!- fendrin [~quassel@176.7.53.120] has quit [Changing host]
20160309 12:27:07-!- fendrin [~quassel@wesnoth/developer/fendrin] has joined #wesnoth-dev
20160309 12:27:12< wedge009> zookeeper, Aginor: Any chance your label issue is related to this one? https://gna.org/bugs/?22176
20160309 12:27:50< zookeeper> yes, sounds like the same thing
20160309 12:28:23< wedge009> Cool, I'll let you two sort out the details then.
20160309 12:29:08< wedge009> vultraz, Aginor: I think there are many more bugs that could be closed but I was being conservative, sticking to what I was quite sure of. And also for vultraz's sake, so he didn't go into meltdown closing too many at once.
20160309 12:32:38< wedge009> zookeeper, Aginor: I also see the flickering after the no-units-to-recall message, but only on menu buttons, mini-map toggle icons, End Turn button, that sort of thing.
20160309 12:32:53< zookeeper> yep
20160309 12:33:10< wedge009> vultraz: It's okay, I haven't seen anything to complain about the UI changes. Yet. ;)
20160309 12:33:55< wedge009> No, seriously, it all looks fine to me so far. But so many things have changed lately, I don't have a list of what's changed. o.O
20160309 12:34:09-!- Samual [~Samual@xonotic/core-team/Samual] has joined #wesnoth-dev
20160309 12:34:23 * wedge009 prepares to sleep
20160309 12:37:10< vultraz> wedge009, zookeeper: if the flickering is in the UI that's to be expected
20160309 12:37:21< vultraz> if the flickering is in the dialog it is not
20160309 12:39:58-!- _laco [~laco@static.95.25.4.46.clients.your-server.de] has joined #wesnoth-dev
20160309 12:40:01< vultraz> zookeeper: do you want me to commit a contrast bump for the Rockbound hills?
20160309 12:40:08< vultraz> it makes them a bit more distinguishable 
20160309 12:40:59< zookeeper> well, contrast bump or anything that fixes the issue
20160309 12:41:07< zookeeper> i don't know before seeing it
20160309 12:43:00< irker760> wesnoth: Charles Dang wesnoth:master e7c2533c5e54 / data/core/images/terrain/cave/ (9 files): Slightly bumped contrast on Rockbound Cave https://github.com/wesnoth/wesnoth/commit/e7c2533c5e5475c6354561612431c27165d10c8a
20160309 12:43:03< vultraz> zookeeper: see if that improves it
20160309 12:44:30< zookeeper> seems to, yes
20160309 12:47:09-!- clavi [~clavi@163.172.10.77] has joined #wesnoth-dev
20160309 13:14:15-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has quit [Quit: Connection closed for inactivity]
20160309 13:16:05-!- travis-ci [~travis-ci@ec2-50-17-73-77.compute-1.amazonaws.com] has joined #wesnoth-dev
20160309 13:16:06< travis-ci> wesnoth/wesnoth#8807 (master - e7c2533 : Charles Dang): The build has errored.
20160309 13:16:06< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/114776167
20160309 13:16:06-!- travis-ci [~travis-ci@ec2-50-17-73-77.compute-1.amazonaws.com] has left #wesnoth-dev []
20160309 13:22:02-!- prkc [~prkc@gateway/vpn/privateinternetaccess/prkc] has joined #wesnoth-dev
20160309 13:46:40< celmin|sleep> Great, I get to clean up after gfgtdf.
20160309 13:47:08< celmin|sleep> Why does side_name() return a std::string?
20160309 13:47:21< zookeeper> or you can wait for him to clean up whatever that is
20160309 13:47:39< celmin|sleep> It looks like he did at least fix the compiler error, probably.
20160309 13:57:14< celmin|sleep> I have a feeling WFL is exclusively left associative.
20160309 13:57:52< celmin|sleep> Which is better than being exclusively right associative, considering it has an exponentiation operator, but...
20160309 13:58:31< celmin|sleep> Or wait. That's wrong.
20160309 13:58:49< celmin|sleep> It's the exponentiation operator that's supposed to be right associative. Sigh...
20160309 14:00:56< celmin|sleep> While subtraction and division (and modulus I suppose, not sure on that one) are supposed to be left associative.
20160309 14:01:42< celmin|sleep> Dot should also be left associative. Dice, well... uh... who cares/
20160309 14:01:44< celmin|sleep> ^.
20160309 14:07:37< irker760> wesnoth: Celtic Minstrel wesnoth:master 89ef1d318a7f / src/ (team.cpp team.hpp): Fix team_info::side_name() returning wrong type https://github.com/wesnoth/wesnoth/commit/89ef1d318a7f219586ee894da46514dbb312cfe7
20160309 14:08:22< celmin|sleep> Oh yeah, vultraz broke my build too.
20160309 14:10:28< celmin|sleep> So uh, I need a widget_definition, a window_builder, and widget, right? And maybe a dialog?
20160309 14:13:37 * celmin|sleep found all those, hopes it works.
20160309 14:15:18< irker760> wesnoth: Celtic Minstrel wesnoth:master 26be1be486d0 / projectfiles/Xcode/Wesnoth.xcodeproj/project.pbxproj: Update XCode project https://github.com/wesnoth/wesnoth/commit/26be1be486d0a6ef3817bedc4cb4a3835d1ee0b3
20160309 14:16:22< celmin|sleep> Ugh, merge conflict on XCode project. :/
20160309 14:16:55< celmin|sleep> At least it's an easy one (keep both sides).
20160309 14:21:01< vultraz> celmin|sleep: I fixed the compiler error, and yes, those are the files you need to add
20160309 14:21:18< celmin|sleep> Oh, that was you?
20160309 14:21:20< celmin|sleep> I see.
20160309 14:21:48< vultraz>  https://github.com/wesnoth/wesnoth/commit/cab4d9fe48036f0e4079060a43d090d38479ff41 if that's what you're referring to
20160309 14:22:07< celmin|sleep> Yes, uh... that might just make it worse.
20160309 14:22:14< vultraz> well it builds
20160309 14:22:34< celmin|sleep> Didn't Soliton or someone say that binding a "const std::string&" to a t_string was dangerous?
20160309 14:23:01< celmin|sleep> And why on earth is it returning a const std::string& in the first place when side_name is translatable?
20160309 14:23:08< vultraz> ask gfgtdf
20160309 14:23:10< celmin|sleep> But whatever.
20160309 14:23:12< vultraz> I didn't write that  code
20160309 14:23:25< vultraz> but you're right, now that you mention it
20160309 14:23:26< celmin|sleep> I changed it to return a t_string, and do the fallback at load instead of on the fly.
20160309 14:23:40< vultraz> what?
20160309 14:23:45< vultraz> oh
20160309 14:23:54< celmin|sleep> See my commit if you're confused.
20160309 14:24:16< vultraz> ah
20160309 14:25:32< vultraz> ....hm
20160309 14:29:34< celmin|sleep> I'm probably just gonna watch anime this morning, because dentist.
20160309 14:29:43< vultraz> ...aw crap. I think I broken recruitment in a really weird way :(
20160309 14:29:47< celmin|sleep> Yay!
20160309 14:29:57< vultraz> no sense, does this make
20160309 14:30:03< vultraz> I thought it maybe was custom units, but it's not that
20160309 14:30:19< celmin|sleep> The Force is weak with you.
20160309 14:30:35< celmin|sleep> ...sorry.
20160309 14:30:37< vultraz> heh
20160309 14:32:38< vultraz> ah, got it
20160309 14:32:43< vultraz> fast debugging!
20160309 14:33:59< irker760> wesnoth: Charles Dang wesnoth:master f4590e6bda99 / src/menu_events.cpp: Fixed recruitment accidentally using type name instead of id https://github.com/wesnoth/wesnoth/commit/f4590e6bda995bf5430acb76fde1a52ac6491aa7
20160309 14:34:34< celmin|sleep> Yeah okay, that would be a problem.
20160309 14:41:35-!- travis-ci [~travis-ci@ec2-54-234-9-116.compute-1.amazonaws.com] has joined #wesnoth-dev
20160309 14:41:36< travis-ci> wesnoth/wesnoth#8808 (master - 89ef1d3 : Celtic Minstrel): The build failed.
20160309 14:41:36< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/114794384
20160309 14:41:36-!- travis-ci [~travis-ci@ec2-54-234-9-116.compute-1.amazonaws.com] has left #wesnoth-dev []
20160309 14:41:44< celmin|sleep> :|
20160309 14:42:19-!- janebot_ is now known as janebot
20160309 14:42:48-!- janebot is now known as Guest15979
20160309 14:42:50< vultraz> oh, yeah, I forgot to update the unit tests
20160309 14:43:06< celmin|sleep> Ah.
20160309 14:43:27< celmin|sleep> Lately I've been adding all new GUI2 files to the unit test build, even though that build does not currently work.
20160309 14:55:04-!- gfgtdf [~chatzilla@f054060056.adsl.alicedsl.de] has joined #wesnoth-dev
20160309 14:55:34< gfgtdf> celmin|sleep: uuhm why did you change the side_name() ?
20160309 14:55:46< gfgtdf> celmin|sleep: i mean you simply remove teh empoty() check
20160309 14:55:51< gfgtdf> celmin|sleep: remoed
20160309 14:55:57< gfgtdf> celmin|sleep: shich sureley boreak that function
20160309 14:56:00< gfgtdf> creaks
20160309 14:56:02< celmin|sleep> gfgtdf: I moved it to the read() function.
20160309 14:56:02< gfgtdf> breaks*
20160309 14:56:46< gfgtdf> celmin|sleep: hmm but this will not wokr in case that current_player changes during th game
20160309 14:56:51< gfgtdf> celmin|sleep: work*
20160309 14:56:57< celmin|sleep> True.
20160309 14:57:23< celmin|sleep> I didn't think of that.
20160309 14:57:31< celmin|sleep> Hmm.
20160309 14:57:36< gfgtdf> celmin|sleep: was there a bug before your commit ?
20160309 14:58:07< celmin|sleep> Well, it was actually triggered by the compiler error that vultraz fixed.
20160309 14:59:04< gfgtdf> celmin|sleep: hmm it did work for me, and the travis for soem reason failes for travis silliness
20160309 14:59:20< gfgtdf> celmin|sleep: butmy question was 'before your commit' not 'after my commit'
20160309 15:00:37< celmin|sleep> Basically the question is whether vultraz's commit introduced a bug, right?
20160309 15:01:03< celmin|sleep> I'm not entirely sure, but I suspect so.
20160309 15:01:26< gfgtdf> celmin|sleep: no my question was menat as 'is there any reason not to revert your commit'
20160309 15:02:15< gfgtdf> vultraz: about the preview pane i think it migth mase senst to pass the team color to the constructor instead of querrying it from resources::controller, this migth be useful when using the unit preview pane in the editor
20160309 15:02:37< gfgtdf> make sense*
20160309 15:03:14< celmin|sleep> By the way, why did side_name() return std::string?
20160309 15:03:28< celmin|sleep> Instead of t_string.
20160309 15:03:57< gfgtdf> celmin|sleep: beasue t_string& is convertible to a std::string& but not te opposute way
20160309 15:04:09< gfgtdf> celmin|sleep: note the '&'
20160309 15:04:49< celmin|sleep> Both are std::string and t_string are convertible to each other.
20160309 15:04:57< celmin|sleep> ^-are
20160309 15:05:03< gfgtdf> celmin|sleep: no not in this way
20160309 15:05:17< gfgtdf> celmin|sleep: you cannot create a const t_string& from a const std::string
20160309 15:05:41< gfgtdf> celmin|sleep: you can only create a temporary t_string, not a reference
20160309 15:06:17< celmin|sleep> Which is in fact possible, but okay, you don't want to do that when returning it.
20160309 15:06:32< celmin|sleep> What about the reverse then?
20160309 15:06:45< gfgtdf> celmin|sleep: what do you mean by reverse ?
20160309 15:06:51< celmin|sleep> The one you say is possible.
20160309 15:07:20< gfgtdf> celmin|sleep: the t_string -> std::strign operator retruns a std::string& pointing to a memeber of the t_string
20160309 15:07:53< celmin|sleep> So t_string::str() returns some sort of cached string?
20160309 15:07:59< gfgtdf> celmin|sleep: yes
20160309 15:08:26< celmin|sleep> Hmm, okay. In that case, and considering what you said about what if current_player changes mid-game, I can't think of any reason not to revert my commit.
20160309 15:08:46< gfgtdf> celmin|sleep: this is also why codes like const std::string& a = t_string("str"); will cause segfaults
20160309 15:09:09< gfgtdf> celmin|sleep: we had that case recanelty in one of vukltraz codes
20160309 15:09:14< gfgtdf> recently*
20160309 15:09:15< celmin|sleep> Yeah, I knew there was something like that, and it confused me.
20160309 15:09:26< celmin|sleep> But I guess in this case the t_string is not a temporary, so it's fine.
20160309 15:14:42< irker760> wesnoth: gfgtdf wesnoth:master 4810e2c0ff53 / src/ (saved_game.cpp saved_game.hpp): add warning if current_player attribute is specified by wml https://github.com/wesnoth/wesnoth/commit/4810e2c0ff531248e56aa1b710ae4ba918be3b9b
20160309 15:14:44< irker760> wesnoth: gfgtdf wesnoth:master 49ee9daad7aa / src/game_initialization/connect_engine.cpp: dont have reserved sides in singleplayer connect engine. https://github.com/wesnoth/wesnoth/commit/49ee9daad7aa3160bcef178c6aa2394d67e3d968
20160309 15:14:46< irker760> wesnoth: gfgtdf wesnoth:master d424911eeb1d / src/ (team.cpp team.hpp): Revert "Fix team_info::side_name() returning wrong type" https://github.com/wesnoth/wesnoth/commit/d424911eeb1d89aed83911e1fe24c31c55748775
20160309 15:32:37-!- travis-ci [~travis-ci@ec2-54-166-189-246.compute-1.amazonaws.com] has joined #wesnoth-dev
20160309 15:32:38< travis-ci> wesnoth/wesnoth#8809 (master - 26be1be : Celtic Minstrel): The build has errored.
20160309 15:32:38< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/114796291
20160309 15:32:38-!- travis-ci [~travis-ci@ec2-54-166-189-246.compute-1.amazonaws.com] has left #wesnoth-dev []
20160309 15:36:57< celmin|sleep> vultraz: Fix Travis please.
20160309 15:39:33-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has joined #wesnoth-dev
20160309 15:56:34-!- travis-ci [~travis-ci@ec2-50-17-73-77.compute-1.amazonaws.com] has joined #wesnoth-dev
20160309 15:56:35< travis-ci> wesnoth/wesnoth#8810 (master - f4590e6 : Charles Dang): The build has errored.
20160309 15:56:35< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/114801049
20160309 15:56:35-!- travis-ci [~travis-ci@ec2-50-17-73-77.compute-1.amazonaws.com] has left #wesnoth-dev []
20160309 15:56:49< zookeeper> ooookay... something about messages is mightily broken in master
20160309 15:57:17< celmin|sleep> Messages as in... [message]?
20160309 15:57:26< celmin|sleep> Is this my fault or Aginor's or?
20160309 15:57:28< zookeeper> just start liberty; most of the starting dialogue doesn't show up
20160309 15:57:48< zookeeper> namely, Fal Khag's and Baldras's lines
20160309 15:57:55< celmin|sleep> I see.
20160309 15:59:18< zookeeper> no idea who's fault that might be. something to do with sides and id's i imagine
20160309 16:02:47< zookeeper> gfgtdf, did you notice what i said in the log about leader carryover?
20160309 16:03:04< gfgtdf> zookeeper: hmm no i didnt let me see
20160309 16:04:57< zookeeper> maybe it's related to the abovementioned issue too, if it botches up leader id's (since baldras and fal khag are side 1 and 2 leaders)
20160309 16:17:35-!- Kwandulin [~Miranda@p200300760F7D73B4A19A81CAD4EBE545.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160309 16:18:51-!- travis-ci [~travis-ci@ec2-50-17-73-77.compute-1.amazonaws.com] has joined #wesnoth-dev
20160309 16:18:52< travis-ci> wesnoth/wesnoth#8811 (master - d424911 : gfgtdf): The build has errored.
20160309 16:18:52< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/114812198
20160309 16:18:52-!- travis-ci [~travis-ci@ec2-50-17-73-77.compute-1.amazonaws.com] has left #wesnoth-dev []
20160309 16:28:34-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160309 16:28:41-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160309 16:28:48-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160309 16:30:48< irker760> wesnoth: Charles Dang wesnoth:master 09bb648bfa98 / src/tests/gui/test_gui2.cpp: Add Unit Recruit to gui2 unit tests https://github.com/wesnoth/wesnoth/commit/09bb648bfa98e032b214431de0e3fd7c9c7387b7
20160309 16:37:17< irker760> wesnoth: gfgtdf wesnoth:master fce35e3ad6c9 / src/team.cpp: attempt to fix bug related to leaders id attribute https://github.com/wesnoth/wesnoth/commit/fce35e3ad6c9ba8e6e46b6cc1b316e9b1caff135
20160309 16:37:21< gfgtdf> zookeeper: try that ^
20160309 16:40:20-!- Guest15979 is now known as janebot
20160309 16:40:50-!- janebot is now known as Guest74474
20160309 16:42:48-!- Guest74474 [~Gambot@grickit.us] has quit [Changing host]
20160309 16:42:48-!- Guest74474 [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev
20160309 16:43:29-!- Guest74474 is now known as janebot
20160309 16:49:22-!- ChipmunkV [~vova@d0017-2-88-172-31-68.fbx.proxad.net] has joined #wesnoth-dev
20160309 17:02:16-!- fendrin [~quassel@wesnoth/developer/fendrin] has quit [Read error: Connection reset by peer]
20160309 17:02:30-!- fendrin [~quassel@wesnoth/developer/fendrin] has joined #wesnoth-dev
20160309 17:06:23-!- travis-ci [~travis-ci@ec2-54-163-65-175.compute-1.amazonaws.com] has joined #wesnoth-dev
20160309 17:06:24< travis-ci> wesnoth/wesnoth#8812 (master - 09bb648 : Charles Dang): The build failed.
20160309 17:06:24< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/114833638
20160309 17:06:24-!- travis-ci [~travis-ci@ec2-54-163-65-175.compute-1.amazonaws.com] has left #wesnoth-dev []
20160309 17:10:09-!- Nobun [~nobun@host62-49-dynamic.12-79-r.retail.telecomitalia.it] has joined #wesnoth-dev
20160309 17:18:02< gfgtdf> vultraz: is it intentionaly that the image is now missing in the recruit unit  dialog ?
20160309 17:18:26< gfgtdf> vultraz: the name is missing too..
20160309 17:23:09-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20160309 17:23:09-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Remote host closed the connection]
20160309 17:24:40-!- Elvish_Hunter [~elvish_hu@wesnoth/developer/elvish-hunter] has joined #wesnoth-dev
20160309 17:26:27< vultraz> gfgtdf: yes
20160309 17:28:52-!- louis94 [~~louis94@91.178.242.211] has joined #wesnoth-dev
20160309 17:30:00-!- travis-ci [~travis-ci@ec2-50-17-73-77.compute-1.amazonaws.com] has joined #wesnoth-dev
20160309 17:30:01< travis-ci> wesnoth/wesnoth#8813 (master - fce35e3 : gfgtdf): The build failed.
20160309 17:30:01< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/114835379
20160309 17:30:01-!- travis-ci [~travis-ci@ec2-50-17-73-77.compute-1.amazonaws.com] has left #wesnoth-dev []
20160309 17:31:45< celmin|sleep> That t_string situation most often crops up with configs, right?
20160309 17:32:06< celmin|sleep> ie const std::string xyz = config["some_key"].str();
20160309 17:32:11< celmin|sleep> ^&
20160309 17:32:22< celmin|sleep> ...
20160309 17:32:27< gfgtdf> celmin|sleep: hm no afaik it doesnt
20160309 17:32:30< celmin|sleep> ie const std::string& xyz = cfg["some_key"].str();
20160309 17:32:45< gfgtdf> .str() returns teh strign by value
20160309 17:32:50< gfgtdf> the string*
20160309 17:32:57< celmin|sleep> Hmm, is that so?
20160309 17:32:58< gfgtdf> so no problem there
20160309 17:33:06< celmin|sleep> I thought I'd heard about something there.
20160309 17:33:27< celmin|sleep> Maybe it was the more obviously stupid const std::string& xyz = cfg["some_key"].t_str()...
20160309 17:33:42< gfgtdf> celmin|sleep: i in some anivent wesnoth version all allrtibure_values were stores as t_string
20160309 17:33:45< gfgtdf> anicent*
20160309 17:33:54< celmin|sleep> Oh, really? Huh.
20160309 17:34:00< celmin|sleep> Why would you do that.
20160309 17:34:23< gfgtdf> celmin|sleep: so that problem was there at some point, but that was before i was in the team
20160309 17:34:42< celmin|sleep> I wonder if there's a nice config file that exhibits almost everything you can do in WML, to show off my syntax highlighter...
20160309 17:37:23-!- SigurdFD [~SigurdFD@dynamic-acs-72-23-176-151.zoominternet.net] has joined #wesnoth-dev
20160309 17:40:33-!- iwaim___ [~iwaim@rasteenie.alib.jp] has quit [Ping timeout: 246 seconds]
20160309 17:41:21< irker760> wesnoth: Charles Dang wesnoth:master 7b23a3940b7a / src/tests/gui/test_gui2.cpp: Fixup 09bb648bfa98 https://github.com/wesnoth/wesnoth/commit/7b23a3940b7a492795952c85124e2c1460a64f99
20160309 17:41:43< vultraz> celmin|sleep: the test scenario is probably the biggest config fileblock
20160309 17:42:05< celmin|sleep> I'll see how that looks.
20160309 17:42:20< celmin|sleep> I was actually considering just applying it to the whole blob though. >_
20160309 17:42:22< celmin|sleep> ^ >_>
20160309 17:43:33< gfgtdf> vultraz: did you update the schema.cfg ?=
20160309 17:43:38< vultraz> gfgtdf: no
20160309 17:43:44< vultraz> I couldn't figure out the syntax
20160309 17:43:54< vultraz> it kept warning me about stuff
20160309 17:43:54< celmin|sleep> Oh, I should add a case for negative numbers.
20160309 17:44:59< celmin|sleep> Uh-oh, booleans are detected in preference to strings.
20160309 17:45:21-!- midzer [~quassel@p5B312EF2.dip0.t-ipconnect.de] has quit [Remote host closed the connection]
20160309 17:46:37< celmin|sleep> (I had the "no" of "not_living" highlighted as a boolean.)
20160309 17:46:47-!- iwaim___ [~iwaim@rasteenie.alib.jp] has joined #wesnoth-dev
20160309 17:47:57< celmin|sleep> Oh hey, the test scenario also hits the macro edge case that I couldn't solve.
20160309 17:48:07< celmin|sleep> I wonder if I can get pygments to number the lines.
20160309 17:48:59-!- iwaim___ [~iwaim@rasteenie.alib.jp] has quit [Remote host closed the connection]
20160309 17:49:00< celmin|sleep> Heh, the test scenario checks for WLP.
20160309 17:49:19< celmin|sleep> Okay, there's a problem with the Lua detection.
20160309 17:50:19-!- iwaim [~iwaim@rasteenie.alib.jp] has joined #wesnoth-dev
20160309 17:56:54< celmin|sleep> My fix works as long as the Lua code does not use the > operator. Hmm...
20160309 17:56:58< zookeeper> gfgtdf, yes, that seems to fix the leader bugs
20160309 17:57:00-!- louis94 [~~louis94@91.178.242.211] has quit [Read error: Connection reset by peer]
20160309 17:57:08< zookeeper> gfgtdf, do you know if 1.13.3 was affected by that?
20160309 17:57:20< gfgtdf> zookeeper: hmm i dont think it is
20160309 17:57:23-!- louis94 [~~louis94@91.178.242.211] has joined #wesnoth-dev
20160309 18:02:55< zookeeper> gfgtdf, well, that's good, because it'd be a blocker :p
20160309 18:09:43< celmin|sleep> Could make it highlight Pango markup and HTML entities in strings...
20160309 18:09:53< celmin|sleep> Dunno if that's really worth it.
20160309 18:10:06< gfgtdf> celmin|sleep: for which editor is your syntax highjlighting ?
20160309 18:10:20< celmin|sleep> It's Pygments. My hope is that it can be used on the wiki.
20160309 18:10:28< celmin|sleep> (I also made one for TextWrangler.)
20160309 18:10:44< celmin|sleep> (Pygments is a framework, not an editor.)
20160309 18:11:27< celmin|sleep> Oh, I suppose I could also highlight IPF and animation progressive syntax if I really wanted to. I'll put those all down as "maybe".
20160309 18:11:54-!- travis-ci [~travis-ci@ec2-50-17-73-77.compute-1.amazonaws.com] has joined #wesnoth-dev
20160309 18:11:55< travis-ci> wesnoth/wesnoth#8814 (master - 7b23a39 : Charles Dang): The build is still failing.
20160309 18:11:55< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/114852515
20160309 18:11:55-!- travis-ci [~travis-ci@ec2-50-17-73-77.compute-1.amazonaws.com] has left #wesnoth-dev []
20160309 18:14:53-!- midzer [~quassel@p5B312EF2.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160309 18:16:04-!- midzer [~quassel@p5B312EF2.dip0.t-ipconnect.de] has quit [Remote host closed the connection]
20160309 18:16:13< celmin|sleep> Once I solve the encoding issues I'll upload a sample.
20160309 18:22:46-!- zhifen [863c69b8@gateway/web/freenode/ip.134.60.105.184] has joined #wesnoth-dev
20160309 18:22:53-!- Elvish_Hunter [~elvish_hu@wesnoth/developer/elvish-hunter] has left #wesnoth-dev ["Ciao!"]
20160309 18:28:49< celmin|sleep> It works if I manually add a meta tag, but I can't figure out how to make pygments add that tag.
20160309 18:29:00< celmin|sleep> Oh well. It'll do.
20160309 18:29:08< vultraz> what are you even doing
20160309 18:29:42< gfgtdf> vultraz: did you see ma message bafore about image and name missing in unt recruit dialouge ?
20160309 18:29:57< vultraz> gfgtdf: yes, it's intentional 
20160309 18:30:28-!- zhifen [863c69b8@gateway/web/freenode/ip.134.60.105.184] has quit [Quit: Page closed]
20160309 18:30:33< gfgtdf> vultraz: hmm if you want to treat the oreview pan like a 'details page' then you shodul consider putting it to the right side of the dialog
20160309 18:30:46< gfgtdf> vultraz: the rrent way looks suite trange to me
20160309 18:30:56< celmin|sleep> vultraz: See for yourself: http://celmin.pwcsite.com/wesnoth/test_syntax.html
20160309 18:31:13< celmin|sleep> And the code is add http://celmin.pwcsite.com/wesnoth/WmlLexer.py
20160309 18:31:16< celmin|sleep> ^at
20160309 18:31:30< vultraz> gfgtdf: but all our dialogs have it on the left
20160309 18:31:43< vultraz> Load Game, Attack, Create Unit, Recall..
20160309 18:31:57-!- louis94 [~~louis94@91.178.242.211] has quit [Ping timeout: 248 seconds]
20160309 18:32:39< celmin|sleep> Attack doesn't have it on the left.
20160309 18:32:42< celmin|sleep> It has it on the top.
20160309 18:33:11< gfgtdf> vultraz: hmm also whats is the reason for ommitting the image on the first place ?
20160309 18:33:29< vultraz> gfgtdf: the image is in the list, why do we need it again
20160309 18:33:47< vultraz> plus this format takes more space, and since recruit lists can be small it's a good idea to conserve space
20160309 18:33:51< celmin|sleep> vultraz: I'm hoping this can eventually be used for highlighting WML samples on the wiki.
20160309 18:33:52< gfgtdf> vultraz: becasue we'd have t search it one the list whiel on the panel its aleays on the same location
20160309 18:33:57< gfgtdf> to*
20160309 18:34:00< celmin|sleep> (Exact colour scheme is easily changed.)
20160309 18:34:15< vultraz> gfgtdf: ...what?
20160309 18:34:37< gfgtdf> vultraz: for specailyl when the list is a t a scrolling position where i cannot see it, but even if visible its still an isue
20160309 18:34:47< celmin|sleep> I'm not sure what gfgtdf is saying, but I don't think putting the image in the preview is bad.
20160309 18:35:02< celmin|sleep> It's duplication, sure, but duplication is not always bad.
20160309 18:35:11< vultraz> :|
20160309 18:35:14< vultraz> guys...
20160309 18:35:20< celmin|sleep> What?
20160309 18:37:01< celmin|sleep> Oh wow, the test scenario lets you inject arbitrary Lua code into the scenario.
20160309 18:37:03< vultraz> I dunno, just seems whenever I want to remove something there's backlash]
20160309 18:37:08< gfgtdf> vultraz: my point is, that in the current layout i don't know which unit is currently selected unless i seach in the whole recruit listbox for the currently selected (togelled) item.
20160309 18:37:22< celmin|sleep> vultraz: Well, you should expect removing things to create a lot of backlash.
20160309 18:37:41< vultraz> gfgtdf does have a point..
20160309 18:37:59< celmin|sleep> People get used to how things are. Also, things that you find unnecessary, others might consider essential.
20160309 18:38:17< celmin|sleep> Any comments on my syntax highlighter? I have to go in a few minutes.
20160309 18:39:28< vultraz> Not really
20160309 18:39:30< vultraz> it looks like it works
20160309 18:39:37< vultraz> dont really like the colors
20160309 18:40:14< celmin|sleep> Yeah, I just threw together colours that allowed me to easily distinguish things, without really putting much thought into them.
20160309 18:40:25< celmin|sleep> The provided schemes made it hard to distinguish stuff.
20160309 18:41:14-!- atarocch [~atarocch@31.159.149.240] has joined #wesnoth-dev
20160309 18:43:51-!- celmin|sleep is now known as celmin|away
20160309 19:01:25-!- Kwandulin [~Miranda@p200300760F7D73B4A19A81CAD4EBE545.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160309 19:37:38-!- Kwandulin [~Miranda@p200300760F7D73B47D1779C7A7D4C004.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160309 19:45:38-!- louis94 [~~louis94@91.178.242.211] has joined #wesnoth-dev
20160309 19:46:30-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20160309 19:49:22-!- ancestral [~ancestral@63.92.240.233] has joined #wesnoth-dev
20160309 19:51:26< ancestral> Okay, I think the app is good, so I’m making a package
20160309 20:03:49< ancestral> vultraz: Do you have a screenshot of what your unit preview window looks like?
20160309 20:14:30-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160309 20:19:12-!- louis94 [~~louis94@91.178.242.211] has quit [Ping timeout: 260 seconds]
20160309 20:28:42< ancestral> Updated the changelog in the app
20160309 20:28:54< ancestral> Gotta find a better way to format this.
20160309 20:31:05-!- ancestral [~ancestral@63.92.240.233] has quit [Quit: i go nstuf kthxbai]
20160309 20:38:57-!- louis94 [~~louis94@91.178.242.211] has joined #wesnoth-dev
20160309 20:41:36-!- irker760 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20160309 20:41:52< celmin|away> Ugh. Okay, back.
20160309 20:42:02-!- celmin|away is now known as celticminstrel
20160309 20:46:21-!- louis94 [~~louis94@91.178.242.211] has quit [Ping timeout: 248 seconds]
20160309 20:48:54-!- midzer [~quassel@p5B312EF2.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160309 21:00:38< Nobun> celticminstrel: I fear that the \2 is not supported on python regexpes
20160309 21:00:53< celticminstrel> Uh.
20160309 21:01:07< celticminstrel> I'm pretty sure it's supported, but the syntax may be different, let me check.
20160309 21:01:38< Nobun> I will send you the regexp I setted for Lua single-line bracketed strings... you can verify if I mispelled something:
20160309 21:02:14< celticminstrel> Hmm, no, the syntax is indeed \2
20160309 21:02:15< Nobun> celticminstrel:  r'''(?:[^["']*?)(_?)\s*\[(=*)\[(.*?)]\2]'''
20160309 21:02:36< Nobun> this is the regexp I am testing for lua single line bracketed strings
20160309 21:03:17< Nobun> note: I am implicitly using re.match, so the starting ^ is not required
20160309 21:03:35< celticminstrel> I just tested that regex in the Python REPL, and it works.
20160309 21:03:50< celticminstrel> >>> pat = re.compile(r'''(?:[^["']*?)(_?)\s*\[(=*)\[(.*?)]\2]''')
20160309 21:03:51< celticminstrel> >>> m = pat.match('[===[abcdef]===]')>>> m.group()
20160309 21:03:52< celticminstrel> '[===[abcdef]===]'
20160309 21:04:03< celticminstrel> Gah, I missed a line break in the middle there.
20160309 21:04:51< celticminstrel> m.group(2) yields '===' and m.group(3) yields 'abcdef'.
20160309 21:05:09< Nobun> well... it is very strange
20160309 21:05:23< Nobun> here is my (very fake) lua file I am using for testing
20160309 21:05:28< celticminstrel> Are you perhaps using m.group(2) where you should use m.group(3)?
20160309 21:05:45-!- SigurdFD [~SigurdFD@dynamic-acs-72-23-176-151.zoominternet.net] has quit []
20160309 21:05:52< celticminstrel> Since with your previous expression, m.group(2) would've been correct.
20160309 21:06:22< Nobun> I should have changed it... let me check again
20160309 21:07:02< gfgtdf> i think we need a way to tell wesnoth.get_units to limit the numberof results, so that it will stop searching if it already found n units
20160309 21:07:02< Nobun> nope... it is still match.group(3)
20160309 21:07:25< gfgtdf> i wonder whether it woudl bebetter as a secodn paraeter or as a memeber max_macthes= in the filter
20160309 21:07:46< celticminstrel> gfgtdf: Not sure how easy that is. It sounds simple, but I believe the unit_filter class doesn't have an iterable interface.
20160309 21:08:05-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has quit []
20160309 21:08:10< celticminstrel> The latter option would allow non-Lua filters to also use it.
20160309 21:08:12< gfgtdf> celticminstrel: ihm yes seems lei your are right
20160309 21:08:18< gfgtdf> like
20160309 21:08:33< gfgtdf> celticminstrel: don't know why i assumed it had a iterator interface
20160309 21:08:46< celticminstrel> (Personally I'd call the key "limit" rather than "max_matches", but it's not really important.)
20160309 21:09:15< gfgtdf> celticminstrel: hmm 'limit' sounds like it coudl already be used by somethign.
20160309 21:09:32< celticminstrel> (Also, if you're adding it to unit_filter you might as well also add it to terrain_filter. You could even do side_filter too if you wanted to.)
20160309 21:09:42< gfgtdf> celticminstrel: i guess i'll try to make a key in te filter and implement it in the unut_filter class
20160309 21:09:56-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has joined #wesnoth-dev
20160309 21:10:03< gfgtdf> celticminstrel: y for terrain filter it woudl also make sense i guess
20160309 21:10:07< celticminstrel> I'm guessing it would be checked in a completely different place from all the other keys.
20160309 21:11:10< Nobun> celticminstrel: a test done directly in the interpreter seems to confirm what you stated... I have to figure why the string is not actually added -.-
20160309 21:12:24-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit []
20160309 21:12:50< Nobun> what a stupid I am... I forgot the _ before bracketed string in sample lua file... so my interpeter parse those strings as non-translatable :P
20160309 21:12:58< celticminstrel> Ah, heh.
20160309 21:13:02< Nobun> lol
20160309 21:15:33< Nobun> now I have to think a solution about another problem, but this time is related to import system
20160309 21:15:59< Nobun> I have to change the structure of LuaStr30 in order to allow it to check external variable without problems
20160309 21:16:38< Nobun> probably I will change it to an always-run state with internal regexp check
20160309 21:17:29< celticminstrel> The way I showed you means it does consume the text regardless of the check - it just marks it differently.
20160309 21:18:08< celticminstrel> Does it really need to be an always-run state?
20160309 21:18:59-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160309 21:19:38< Nobun> yep since when the state is added to the state machine, the regexp value (inited on __init__ function) records the existing value on machine setup phase, and will ignore any future changes
20160309 21:20:06< Nobun> (infact it cannot checks for a value non-existing at the moment of the initialization)
20160309 21:20:30< Nobun> so the best thing is to move the regexp definition at very run-time
20160309 21:21:53< Nobun> I did some tests and defining directly the regexp at state.regexp works in a weird way
20160309 21:22:23< celticminstrel> What do you mean by "records the existing value"?
20160309 21:25:46< Nobun> I tryed to initialize the regexp value to a string cuncatenation: fixed regexp part + variable value 
20160309 21:25:59< Nobun> at the start the variable value is 0
20160309 21:26:13< celticminstrel> Huh?
20160309 21:26:28< celticminstrel> Why are you appending 0 to the regex?
20160309 21:26:35< Nobun> if you try to test the script, the regexp will work only for value=0 even if the external value changed in the meantime
20160309 21:26:54< celticminstrel> Well obviously, but I don't understand why you're doing that in the first place.
20160309 21:27:00-!- atarocch [~atarocch@31.159.149.240] has quit [Ping timeout: 246 seconds]
20160309 21:27:03< celticminstrel> The regex I gave you doesn't depend on any variables.
20160309 21:27:07< Nobun> since the default value is 0 equals sign
20160309 21:27:27< celticminstrel> It merely matches anything that might be a string ending, and then once it's matched, you check the variable to see if it really is a string ending.
20160309 21:27:35< Nobun> I am speaking now about LuaStr30 (end of multi-line bracketed string)
20160309 21:27:52< celticminstrel> Yes, I suggest you look at my pastebin again, if possible.
20160309 21:27:56< Nobun> yeah... but that code could be error-prone
20160309 21:28:03< celticminstrel> Error-prone how?
20160309 21:28:19< Nobun> since it could end the string in a bad way... I am not a good coder :P
20160309 21:28:30< celticminstrel> Did you look at my pastebin?
20160309 21:28:33< Nobun> its not your fault... I am not sure on myself :P
20160309 21:28:46< Nobun> yep... but I take a look again
20160309 21:29:30< celticminstrel> (If you have specific questions about it and you still have the link, please post it here. If you don't, I can look it up somewhere.)
20160309 21:30:16< Nobun> I am reading the log of our discussion to find the link now :)
20160309 21:31:14-!- atarocch [~atarocch@31.159.149.240] has joined #wesnoth-dev
20160309 21:37:07< Nobun> yep I confirm my state
20160309 21:37:15< Nobun> I re-read your idea
20160309 21:37:24< Nobun> the problem is this one:
20160309 21:37:48< Nobun> the regexp will end at the first ](=*)] it founds
20160309 21:38:09< Nobun> the if() on the run function can check if it is the correct and or not
20160309 21:38:24< Nobun> but, if it is not correct a infinite loop could be caused
20160309 21:38:41< Nobun> since the regexp, when matched again, will stops again at the same point
20160309 21:39:03< Nobun> not sure if I explained well my doubt
20160309 21:39:06< celticminstrel> Maybe I was mistaken about something.
20160309 21:39:21< celticminstrel> First of all... you're parsing line by line, right?
20160309 21:39:28< Nobun> yep
20160309 21:39:44< Nobun> I have to mark that your idea is right
20160309 21:39:59< Nobun> only the pseudo-code cannot be applied as-is
20160309 21:40:03< celticminstrel> Okay, then you probably have a state for when a multi-line string is currently pending, and the current line does not end the stirng.
20160309 21:40:31< Nobun> yep
20160309 21:40:51< celticminstrel> And, sometimes, your regex will match part of a line, and the states will be rerun for the remainder of the line, right?
20160309 21:41:15< celticminstrel> Could you also post the pastebin link quickly, so I can see it?
20160309 21:41:30< Nobun> no properly
20160309 21:41:38< celticminstrel> What?
20160309 21:41:51< Nobun> the regex and the state does not see part of line
20160309 21:42:15< Nobun> the non-consumed part of the line is passed by statemachine as it was returned previously by a state
20160309 21:42:43< Nobun> the run function see all the non-consumed line 
20160309 21:42:51< Nobun> passed by statemachine
20160309 21:43:04< celticminstrel> So for example, if you have a line like this: [this is the end of a multiline double-quoted string" # and this is a comment afterwards]
20160309 21:43:23< celticminstrel> That will be matched by the regex for ending a multiline double-quoted string, correct?
20160309 21:43:43< Nobun> yep
20160309 21:43:47< celticminstrel> (Sorry, I used the wrong comment syntax, but you get the idea.)
20160309 21:43:56< Nobun> the match will end at the quote sign
20160309 21:44:01< Nobun> the xline is the entire line
20160309 21:44:16< celticminstrel> Okay, and then, if I understand your design correctly, the remainder of that line will be checked against the regex in the next state, right?
20160309 21:44:22< Nobun> than run return to the statemachine the non-consumed part of the line (after the #)
20160309 21:44:35< Nobun> yep
20160309 21:45:04< celticminstrel> Okay, give me a few minutes and I'll write up an annotated example.
20160309 21:45:38< Nobun> the next state will try its regexp with the non-consumed part of the line (in your example the #.... code)
20160309 21:46:04< Nobun> imho I see two possible solution without doing the things too tricky
20160309 21:46:39< Nobun> 1) delegate the regex to an external function defined in a statemachine... this should solve the problems around the initialization
20160309 21:47:47< celticminstrel> What are your state names for start, middle, and end of multiline bracketed string?
20160309 21:47:47< Nobun> 2) change the LuaStr30 as an always-run state and move the regex in the internal run state, so the regexp is builded surely on runtime and will know exactly what it should exactly end the string, without having to write other ifs
20160309 21:48:15< Nobun> LuaStr03o is start of multiline bracketed string
20160309 21:48:24< Nobun> LuaStr30 is for multiline end
20160309 21:48:54< Nobun> if the regex not match (multiline not ended), LuaStr31 (always-run state) store all the code line as a line of the string
20160309 21:49:33< celticminstrel> And each state returns its next state, right?
20160309 21:49:59< celticminstrel> Hmm, wait.
20160309 21:50:11< celticminstrel> (Do answer the question though.)
20160309 21:51:23< celticminstrel> It's possible that I've misunderstood your design somehow, and I'm trying to work out if that's the case.
20160309 21:51:41< Nobun> nope... It seems you understood correctly
20160309 21:51:51< celticminstrel> So, what state does LuaStr03o return?
20160309 21:52:21< Nobun> LuaStr03o, if matches, returns LuaStr31 and xline=None
20160309 21:52:45< celticminstrel> LuaStr31 is middle-of-string... okay, what state does it return?
20160309 21:52:49< Nobun> so the line is consumed, statemachine will read next line, and the next line will be parsed by LuaStr30
20160309 21:52:59< Nobun> sorry LuaStr30 not LuaStr31
20160309 21:53:14< celticminstrel> So LuaStr03o returns LuaStr30?
20160309 21:53:25< Nobun> yep
20160309 21:53:39< Nobun> if matches
20160309 21:53:44< celticminstrel> Then if LuaStr30 fails to match, LuaStr31 is run?
20160309 21:54:13< Nobun> yep... but LuaStr30 will read a different line than LuaStr03o
20160309 21:54:14< celticminstrel> Is that your iffail thing?
20160309 21:54:26< Nobun> exactly
20160309 21:54:36< Nobun> LuaStr03o iffail return nextstate on design
20160309 21:54:43< Nobun> but we don't care
20160309 21:54:47< celticminstrel> Okay.
20160309 21:55:01< celticminstrel> So then, does LuaStr31 also return LuaStr30?
20160309 21:55:08< Nobun> yep
20160309 21:55:25< Nobun> LuaStr31 ALWAYS return LuaStr30
20160309 21:55:39< Nobun> while LuaStr30, if matches, return LuaIdle
20160309 21:56:25-!- ChipmunkV [~vova@d0017-2-88-172-31-68.fbx.proxad.net] has quit [Quit: ChipmunkV]
20160309 21:56:28< Nobun> LuaIdleState is the starting state of the desig idea described on chapter 4 of my documentation
20160309 21:56:36< Nobun> (for lua code)
20160309 21:57:52< Nobun> celticminstrel: this should be a possible solution
20160309 21:57:59< Nobun> on LuaState30
20160309 21:58:08< Nobun> self.regex = re.compile(pywmlx.state.machine.lua3_endrx())
20160309 21:58:16< celticminstrel> Okay, so my idea is basically that LuaStr30 returns LuaStr30 if the number of equals is wrong.
20160309 21:58:21< celticminstrel> With xline set to None.
20160309 21:58:28< celticminstrel> Or.
20160309 21:58:36< celticminstrel> Well, with xline set to whatever's after the closing brackets.
20160309 21:58:53< celticminstrel> Does that make sense?
20160309 21:59:06< Nobun> the problem on setting None is that the statemachine will read the next line
20160309 21:59:07< Nobun> but
20160309 21:59:21< Nobun> if you have a string that ends like ]==] ]] on the same line
20160309 21:59:32< Nobun> where  ]] is the right end of bracket string
20160309 21:59:39< celticminstrel> Yeah, I realized None is incorrect.
20160309 21:59:41< Nobun> it misses the end of line
20160309 21:59:49< celticminstrel> Sorry.
20160309 22:00:12< Nobun> don't worry... I appreciate your hard work and hard think :)
20160309 22:00:47< Nobun> now I try the solution I already said before (external function)
20160309 22:00:56< Nobun> and see if it works
20160309 22:01:17< celticminstrel> http://pastebin.com/wswRzK0D
20160309 22:01:48< celticminstrel> ...oh heh, I accidentally ended the Lua comment. Doesn't really matter though, the syntax highlighting isn't really needed anyway.
20160309 22:02:18< celticminstrel> Does that make any sense to you?
20160309 22:02:29-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160309 22:02:33< celticminstrel> The main point is that LuaStr30 does not have to always return LuaIdle.
20160309 22:02:51< celticminstrel> Oh, I see an error near the end there too.
20160309 22:03:19-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160309 22:05:53< Nobun> work suspended, celticminstrel: I have an emergency
20160309 22:06:06-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160309 22:06:12-!- Greg-Bog_ [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160309 22:16:13-!- louis94 [~~louis94@91.178.242.211] has joined #wesnoth-dev
20160309 22:31:15-!- Kwandulin [~Miranda@p200300760F7D73B47D1779C7A7D4C004.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160309 22:41:17< Nobun> comed back
20160309 22:42:03< Nobun> sorry... in those days I have to take an eye to my girlfriend, hoping she will feel better as soon as possible
20160309 22:42:56< Nobun> celticminstrel: your idea is nice, but I fear is not good if the ]===] and the closing ]==] are in the same line
20160309 22:43:18< celticminstrel> Why?
20160309 22:43:24< celticminstrel> Maybe I'm missing something.
20160309 22:43:44< celticminstrel> Hmm, I can think of at least one way it could fail, though.
20160309 22:43:52< celticminstrel> [===[Some text]==]===]
20160309 22:44:07< Nobun> yep
20160309 22:44:19< celticminstrel> You could get around that by not consuming the final bracket if the number of equals is wrong.
20160309 22:44:28< Nobun> yep
20160309 22:44:35< Nobun> this is my doubt
20160309 22:45:00< Nobun> this why perhaps the best idea is the alwyas run state with regexp in its run function
20160309 22:45:29< Nobun> with run that returns LuaIdle if the end reached, while returns LuaStr31 if end not reached
20160309 22:46:00< Nobun> I'm coding this solution now and I will see if it works
20160309 22:59:11< Nobun> celticminstrel: the solution of always-run state works
20160309 22:59:27-!- elias [~allefant@allegro/developer/allefant] has quit [Ping timeout: 250 seconds]
20160309 22:59:38< celticminstrel> I really think that shouldn't be necessary, but if it works, well, yay.
20160309 23:00:00-!- elias [~allefant@allegro/developer/allefant] has joined #wesnoth-dev
20160309 23:00:56< Nobun> celticminstrel: http://pastebin.com/X9w5X0Qd
20160309 23:01:04< Nobun> this is the code of LuaStr30
20160309 23:01:57< Nobun> the run function will evaluate the "numequals" value of the pending string (new parameter added for this purpose)
20160309 23:02:23< Nobun> the numequals was initialized on LuaStr03o
20160309 23:03:20< Nobun> moving the re.match to the run function, allows to avoid the problems and the limitations of having regex at init when the regexp (like in this case) must be calculated on runtime
20160309 23:03:48< Nobun> hoping that this process will not slow-down the parsing too much
20160309 23:04:46< celticminstrel> This sounds like it could eliminate LuaStr31, too.
20160309 23:05:01< Nobun> now... I will fix the last thing remained around lua string... the final escape \ on multiline quoted strings
20160309 23:05:12< Nobun> yep
20160309 23:05:21< Nobun> we could do it
20160309 23:05:28< celticminstrel> [Mar 09@6:03:20pm] Nobun: moving the re.match to the run function, allows to avoid the problems and the limitations of having regex at init when the regexp (like in this case) must be calculated on runtime
20160309 23:05:29< celticminstrel> I am still confused why you think you need to calculate the regex at runtime. You don't need to do this.
20160309 23:05:35< Nobun> I will try to see if I can erase LuaStr31
20160309 23:06:07< Nobun> becouse we have a fixed regexp
20160309 23:06:19< celticminstrel> I see, you're storing the string of equals.
20160309 23:06:29< Nobun> instead of having a general regexp than must be verified later with ifs
20160309 23:06:36< celticminstrel> You're storing match.group(2) rather than len(match.group(2)).
20160309 23:06:42< Nobun> nope
20160309 23:06:42-!- Greg-Bog_ [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160309 23:06:48< Nobun> I stored len
20160309 23:06:48< celticminstrel> Ah wait, you're not.
20160309 23:06:56< Nobun> but I used len on {}
20160309 23:07:21< Nobun> I could also store equals string
20160309 23:07:25< Nobun> it is more or less the same
20160309 23:11:04-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [Ping timeout: 260 seconds]
20160309 23:11:04< celticminstrel> So why can't you do something like this? http://pastebin.com/3JfP8vAJ
20160309 23:11:09< Nobun> I could remove LuaStr31 as you suggested
20160309 23:11:42< Nobun> I did more or less like you suggested
20160309 23:12:51< Nobun> http://pastebin.com/NwY23LkY
20160309 23:13:10< Nobun> as you see it is more or less the same thing you suggested
20160309 23:14:23< celticminstrel> It's very similar, yes; however, mine is not dynamically generating the regex and could be implemented with a static regex. Try clicking the "diff" button in the corner of mine.
20160309 23:14:57< celticminstrel> (Mine may not be quite correct, as well, but that should be easily fixed if so.)
20160309 23:15:09< celticminstrel> (Unless I've misunderstood something.)
20160309 23:17:23< Nobun> your idea is fine too. but it can lead to errors if the programmer is (like me) not careful
20160309 23:17:38< celticminstrel> Maybe. What sort of errors?
20160309 23:17:58< celticminstrel> But anyway, if it works, whatever.
20160309 23:18:09< Nobun> for example:
20160309 23:18:58< Nobun> I can consume the part of the line in order to check the right endblock, but I introduce some bad code and I forgot what LuaStr30 matched without introducing in pendingstring
20160309 23:19:32< Nobun> without adding *
20160309 23:20:20< Nobun> in my code, however, I could remove LuaStr31 requirement since it is now an always-run state that execute internally a regexp
20160309 23:21:02< Nobun> (wich is a thing also done on LuaFinalState for function regexp (that requires re.search instead of re.match... but it is a very buggy feature at the moment)
20160309 23:35:44< Nobun> tomorrow I will add the commit... So I will have time to fix the escape \ problem on multiline quoted (" ') strings
20160309 23:36:29-!- louis94 [~~louis94@91.178.242.211] has quit [Quit: Konversation terminated!]
20160309 23:36:35-!- mjs-de [~mjs-de@x4db6bd0c.dyn.telefonica.de] has quit [Remote host closed the connection]
20160309 23:38:44< celticminstrel> I thought you fixed that already.
20160309 23:39:03< Nobun> nope... I started directly with bracket strings xD
20160309 23:39:20< Nobun> the escaping \ is quicker xD
20160309 23:43:10< Nobun> however you are almost right.. while I am writing the fix I remembered I already written this code yesterday... perhaps I forgot to save it
20160309 23:44:46-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160309 23:45:18< Nobun> the fix should fix the final \\n" in \n", right?
20160309 23:47:22< celticminstrel> Ideally.
20160309 23:47:54< celticminstrel> If the string contains an actual newline, it needs to be replaced with r'\n'
20160309 23:48:21< celticminstrel> If the string contains r'\n' then it would probably need to be replaced with r'\\n', though I suspect this case is unlikely to occur.
20160309 23:48:36< celticminstrel> \n is not the sort of thing you use in dialog.
20160309 23:48:51< celticminstrel> (As a literal string, I mean.)
20160309 23:49:13< celticminstrel> So my WFL parser works, but nesting it in the WML parser is having problems. Hmm.
20160309 23:50:17-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 244 seconds]
20160309 23:51:34< Nobun> ok now I will save the changes in the local tmp branch. commit there, merge to master and then I will update the commit in the online branch master
--- Log closed Thu Mar 10 00:00:32 2016