--- Log opened Tue Aug 16 00:00:10 2016
--- Day changed Tue Aug 16 2016
20160816 00:00:10< vultraz> i really don't know what to do
20160816 00:00:32< vultraz> it's an inherent flaw in the design of WML, really.
20160816 00:00:49< vultraz> the fact that we need a preprocessor
20160816 00:00:51< gfgtdf> celmin: thor mp settigns don't apply to LoW (since ut used for_lock_settings=yes)
20160816 00:01:05< gfgtdf> celmin: and this is likeley to be true for other mp campaigns aswell
20160816 00:01:05< celmin> Okay?
20160816 00:01:16< celmin> But does that prevent the loading screen from appearing?
20160816 00:01:21< gfgtdf> force_lock_settings
20160816 00:01:22< vultraz> gfgtdf: yeah, but in the gui1 version there's still a loading screen because it needs the list of scenarios for the Entry Points list
20160816 00:01:24< vultraz> :|
20160816 00:01:54-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160816 00:02:02< gfgtdf> vultraz: i rdont relal liek antry point, i mean, since you can just load a start-of-scenario save, whats teh point of having them ?
20160816 00:02:18< gfgtdf> afaik fabi added them so me might know.
20160816 00:02:24< celmin> That sounds like a terrible question.
20160816 00:02:39< vultraz> i guess so you can start the campaign anywhere without a SoS save
20160816 00:03:28-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [Ping timeout: 252 seconds]
20160816 00:04:00< gfgtdf> vultraz: so in case you accidnetly deelted your startofscenario save? I mean people usuall wouldn't start a campaign with chapter 2 of a campaign so i think it reasonable to assume they have one.
20160816 00:04:49< celmin> There might be reasons to start in chapter 2 though.
20160816 00:05:20< gfgtdf> vultraz: another idea migth be to instad making the starting saves with allow_new_game=yes, to ahev  alist of starting scenarios in teh [campaign] tag
20160816 00:05:47< celmin> What?
20160816 00:05:50< gfgtdf> vultraz: so like [campaign]first_scenario= but a list
20160816 00:05:58< celmin> Sigh...
20160816 00:06:03< vultraz> i dunno...
20160816 00:06:04-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 250 seconds]
20160816 00:06:08< vultraz> that's possible
20160816 00:06:19< vultraz> entry points isn't really great but it seems fine to keep
20160816 00:06:28< celmin> Changing WML is still not a good solution.
20160816 00:06:29< gfgtdf> vultraz: i actuall think its rather stnag anyways tha in mp we makes teh scenarios as 'start' while in sp we specify the starting point in the [campaign]
20160816 00:06:57< vultraz> since i assume one can specify a campaign scenario with allow_new_game = no
20160816 00:07:08< vultraz> and it won't show up in the entry points
20160816 00:07:46< vultraz> celmin: it's not out of the question, though. It's a mistake to restrict ourselves only to the way WML currently does things.
20160816 00:08:31< celmin> It's not out of the question to change WML, but it is out of the question to drop support for the old way of doing things.
20160816 00:08:48< celmin> Which makes changing WML not a (short term) solution to the actual problem here.
20160816 00:09:08< vultraz> I disagree, but whatever
20160816 00:09:26< Aginor> vultraz: any wml changes should deprecate the old API for a <while> while providing the new
20160816 00:09:31< celmin> ^
20160816 00:09:41< gfgtdf> well if its rareley used and easy to update then i dont really see aprobem with chaning wml.
20160816 00:09:59< Aginor> where I'd say that <while> depends on whether it was already present in a stable release or not
20160816 00:10:02< vultraz> new stuff from the dev series is alright to change immediately
20160816 00:10:07< vultraz> anyway
20160816 00:10:07< celmin> Being easy to update is irrelevant.
20160816 00:10:12< celmin> Being rarely used is irrelevant too.
20160816 00:10:21< celmin> If it's never been used, then maybe.
20160816 00:10:27< vultraz> I think keeping the entry points thing is alright
20160816 00:10:30< celmin> But LoW is already a counterargument to that here.
20160816 00:10:35< gfgtdf> celmin: no its not, it sureley a differnt wheter you break all addons or 1 addon,
20160816 00:10:37< Aginor> vultraz: no, if it's been included in a dev release we still have to manage the API
20160816 00:10:45< gfgtdf> celmin: we can change Low as we want
20160816 00:10:48< gfgtdf> LoW
20160816 00:11:02< celmin> But you can't change any other MP campaigns that might be on the add-on server.
20160816 00:11:02< vultraz> Aginor: no, users are supposed to expect any and all things to break in dev releases
20160816 00:11:41< celmin> And there is no difference between breaking all add-ons and breaking one add-on – both are unacceptable.
20160816 00:11:52< vultraz> false
20160816 00:11:57< vultraz> breaking one is better than breaking all
20160816 00:12:08< celmin> It should be possible to grab any add-on from the 1.12 server, install it on 1.13, and run it without game-breaking issues.
20160816 00:12:18< vultraz> you're kidding, right :|
20160816 00:12:25< celmin> No, I'm serious.
20160816 00:12:31 * vultraz groans
20160816 00:12:35< Aginor> I fully agree with that
20160816 00:12:43< vultraz> addons always needs updating when they move to a new stable series
20160816 00:12:49< Aginor> they should get deprication warnings, but it should fully work
20160816 00:12:53< celmin> ^
20160816 00:13:04< Aginor> vultraz: that means we've goofed up
20160816 00:13:12< celmin> Add-ons don't need updating unless they've skipped a stable series.
20160816 00:13:15 * vultraz grits teeth
20160816 00:13:30< vultraz> yes, perhaps it's great if that can happen
20160816 00:13:35< vultraz> but it's not a necessity 
20160816 00:13:37< celmin> eg, add-ons from the 1.10 server can't be expected to run on 1.13 without issues.
20160816 00:13:50< celmin> It is a necessity.
20160816 00:13:52< vultraz> and in any case, we're not discussing breaking campaigns
20160816 00:14:01< vultraz> we're trying to figure out what to do with the bloody reload step!
20160816 00:14:41< vultraz> the peprocessor design is a shitty, yes, everyone agrees, but it's not going anywhere.
20160816 00:16:18< vultraz> shitty one*
20160816 00:18:04< vultraz> gfgtdf: which part of ANL that's needed for settings is behind preprocessor defines?
20160816 00:18:08< vultraz> all the [side]s are outside
20160816 00:18:11< vultraz> and settings
20160816 00:18:21< gfgtdf> vultraz: hmm then maybe no changes to ANl must be made
20160816 00:18:55< vultraz> gfgtdf: it looks like any scenario with a define= triggers a loading screen
20160816 00:19:04< gfgtdf> vultraz: yes
20160816 00:19:05< vultraz> gfgtdf: even if the define doesn't guard any necessary settings
20160816 00:19:28< vultraz> IMO, this is bad design and should be fixed.
20160816 00:19:36< gfgtdf> vultraz: yes becasue teh game currentl loads the 'complete' scenario to protn the setting
20160816 00:19:53< vultraz> that shouldn't be necessary
20160816 00:19:58< gfgtdf> vultraz: also currently the game enigne cannot now ehtehr there which values are inisde the #if and which not
20160816 00:19:58< vultraz> we only want the settings
20160816 00:20:34< celmin> vultraz: It's impossible to fix.
20160816 00:20:35< gfgtdf> vultraz: well yes that mor or less what i suggeste above, don'd load teh full scenario to display the settings, and assum not settings relevant information is inside ifdefs
20160816 00:20:40< vultraz> in the old prototype version of the dialog it read directly from the [multiplayer] tag
20160816 00:20:49< vultraz> tags
20160816 00:20:51< celmin> vultraz: The game has no idea what'll change when parsing with different preprocessor defines.
20160816 00:21:33< gfgtdf> vultraz: yes mabye change it to load them directl from the multiplaer tags, but this them means you need to fill the staet_ object correctly when you press start
20160816 00:21:44< vultraz> right, that can be done
20160816 00:21:51< vultraz> gfgtdf: but what about campaigns?
20160816 00:22:08< vultraz> i guess i can read from there too?
20160816 00:22:40< gfgtdf> vultraz: hmm no for campaigsn its ussualyl all [scenario]s insie the #if
20160816 00:23:02< vultraz> well, yes
20160816 00:23:08< vultraz> true..
20160816 00:23:34< vultraz> we cannot get a list of scenarios without parsing the campaign :|
20160816 00:23:34< gfgtdf> vultraz: but we could 1) just don'T show these settings for campaogns (at lest low doesnt use thme anyways)
20160816 00:23:47< gfgtdf> vultraz: 2_) change teh campaigsn wml to previde this information for exampel in [campaign]
20160816 00:24:04< vultraz> gfgtdf: well the 'use map settings' section is already disabled in gui1 for campaigns
20160816 00:24:06< vultraz> it seems
20160816 00:24:12< vultraz> so we'd just need the 'time limit' stuff
20160816 00:24:16< vultraz> AFAICT
20160816 00:24:23< vultraz> that isn't defined in WML at all
20160816 00:24:28< celmin> I don't see why everyone insists on ignoring the obvious solution - keep the dialogs separate like they were before.
20160816 00:24:43< gfgtdf> vultraz: well in 1.12 it cna be enables by using force_lock_ettings=no in th first scneario
20160816 00:24:49< vultraz> hmm
20160816 00:25:01-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has quit [Quit: Connection closed for inactivity]
20160816 00:25:43< vultraz> celmin: because that's a workaround, not a solution
20160816 00:26:15< vultraz> it might make things easier, i suppose
20160816 00:26:23< gfgtdf> i dont have a strong opinion on whther two or one screens is better, but i do think the new gui2 dialog looks better then the old one, speciall when you see the gui1 dialog right after the gui2 one disappears.
20160816 00:26:48< celmin> It's no more of a workaround than anything you're discussing.
20160816 00:29:26-!- celmin [~celticmin@unaffiliated/celticminstrel] has quit [Quit: And lo! The minstrel departs, to spread the music to the masses!]
20160816 00:31:08< celticminstrel> Whoooa, it's insane. The description isn't wrapping at all.
20160816 00:31:13< celticminstrel> It's all on one line.
20160816 00:31:45< celticminstrel> vultraz: What was the thing to enable the "missed events" messages?
20160816 00:32:41< gfgtdf> celticminstrel: this onyl happens on very low resulutions
20160816 00:34:25< vultraz> celticminstrel: some kinds of bug
20160816 00:34:31< vultraz> celticminstrel: uh can't remeber
20160816 00:35:20< vultraz> GUI2_SHOW_UNHANDLED_EVENT_WARNINGS
20160816 00:36:21< vultraz> gfgtdf: the Random Factions setting isn't scenario-dependent, is it?
20160816 00:36:35< vultraz> if, I were to split the dialog, it could stay in Create?
20160816 00:36:50< gfgtdf> vultraz: hmm no its not
20160816 00:36:59< gfgtdf> vultraz: timing isnt eigher
20160816 00:37:20< gfgtdf> vultraz: we coudl but i wonder why? If we already plit it mkae sesne to me to keep those settigns together
20160816 00:37:47< vultraz> i feel like it does better grouped with eras and mods
20160816 00:38:17< celticminstrel> Time limit feels better where it is, I think. Random factions is debatable.
20160816 00:38:45< vultraz> i just dunno what i'd fill all the empty space with if i split off all the settings
20160816 00:39:53< vultraz> everything on the right would move, except Eras and Mods
20160816 00:41:19< vultraz> and that's a problem
20160816 00:41:25< vultraz> it'd be empty
20160816 00:41:45< celticminstrel> So, should I or should I not bother fiddling with the WML to try and get a working low-res version?
20160816 00:41:50< vultraz> do not
20160816 00:42:03< celticminstrel> Wait until you've decided whether to split?
20160816 00:42:09< vultraz> i guess
20160816 00:42:29< vultraz> there's no easy solution here
20160816 00:42:58< celticminstrel> So consider all the alternatives and come up with something. :)
20160816 00:43:01< vultraz> sure, we can read the necessary settings data from the wml
20160816 00:43:09< vultraz> like it was doing before
20160816 00:43:46< vultraz> but that doesn't really interface with the create engine..
20160816 00:43:54< vultraz> and there's still the matter of entry points
20160816 00:46:02< celticminstrel> What the heck is this tlock?
20160816 00:46:11< celticminstrel> Oh right, that's a reference, never mind.
20160816 00:46:27< celticminstrel> Still, it's disgustingly assert-based.
20160816 00:47:36< vultraz> what?
20160816 00:50:10-!- travis-ci [~travis-ci@ec2-54-226-126-59.compute-1.amazonaws.com] has joined #wesnoth-dev
20160816 00:50:11< travis-ci> wesnoth/wesnoth#10418 (master - 60caf25 : gfgtdf): The build has errored.
20160816 00:50:11< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/152537313
20160816 00:50:11-!- travis-ci [~travis-ci@ec2-54-226-126-59.compute-1.amazonaws.com] has left #wesnoth-dev []
20160816 00:51:08< celticminstrel> I wonder if anyone can fix the Travis.
20160816 00:51:16< celticminstrel> Seems to be something wrong with the MP tests.
20160816 00:51:50< celticminstrel> I can't help much since the server still crashes with invalid gzip data.
20160816 00:52:20< celticminstrel> The local server does. I can connect to the remote one no problem.
20160816 00:53:15< gfgtdf> celticminstrel: hmm if it crashes then that seem likeley to be related to mp travis tests failing.
20160816 00:53:25< celticminstrel> It might be related.
20160816 00:53:33< celticminstrel> But I haven't the slightest idea how to fix it.
20160816 00:53:38< gfgtdf> celticminstrel: which boost version do you use ?
20160816 00:53:38< celticminstrel> It might not be related, too.
20160816 00:53:44< celticminstrel> I think 1.58.
20160816 01:02:59-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20160816 01:04:41-!- gfgtdf [~chatzilla@x4e368e1d.dyn.telefonica.de] has quit [Quit: ChatZilla 0.9.92 [Firefox 48.0/20160726073904]]
20160816 01:06:44-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 265 seconds]
20160816 01:06:45-!- wedge010 is now known as wedge009
20160816 01:07:31-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20160816 01:16:29< vultraz> celticminstrel: i suppose bringing a chat are back to create would be alright if it were split
20160816 01:16:43< vultraz> could you possibly move the chat area from the lobby into a widget?
20160816 01:16:56< celticminstrel> Eh, that sounds like a huge amount of work..
20160816 01:17:48< vultraz> Depends
20160816 01:18:07< vultraz> But it will be very useful in any case
20160816 01:20:21< vultraz> celticminstrel: also, would you recommend keeping everything that used to be in configure together should the dialogs be split?
20160816 01:20:37< vultraz> including custom options
20160816 01:23:09< celticminstrel> I dunno, probably?
20160816 01:23:57 * vultraz ponders
20160816 01:25:55< vultraz> it's really a choice between no random loading screens
20160816 01:25:57< vultraz> or...
20160816 01:26:26-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has joined #wesnoth-dev
20160816 01:26:34< vultraz> the usability of keeping setup options together
20160816 01:28:30< tad_> voltraz: I finally got a chance to test your change for the recall/recruit panels. My issue #27 has the results.
20160816 01:30:33< vultraz> your what?
20160816 01:30:48< tad_> Issues on my fork.
20160816 01:30:57< tad_> https://github.com/GregoryLundberg/wesnoth/issues/27
20160816 01:31:48< vultraz> what scenario was that from?
20160816 01:32:26< tad_> AToTG
20160816 01:32:56< vultraz> looks fine to me
20160816 01:33:01< tad_> Looks like the start of S04
20160816 01:33:50< vultraz> still looks fine
20160816 01:34:21< vultraz> do you not see this? https://drive.google.com/file/d/0B-mR9s8FduLLcFBmOFQ2MS1VUTg/view?usp=sharing
20160816 01:35:07< tad_> I jsut checked and yes. 
20160816 01:35:11< tad_> I see that now. 
20160816 01:35:29< tad_> How odd. I'm checking to see if the branches are out of sync here.
20160816 01:36:18< tad_> How odd. Seems to not be happening any more.
20160816 01:36:48< tad_> Must have been some error syncing which has cleared.
20160816 01:41:01< vultraz> so, no problem, then?
20160816 01:49:50 * vultraz soundly curses the proprocessor
20160816 01:49:54< vultraz> preprocessor*
20160816 01:50:50< vultraz> there's no way, within the realms of current design, to get a list of scenarios without parsing the campaign config
20160816 01:51:28< celticminstrel> If it's only MP campaigns, maybe it's not too bad?
20160816 01:51:31< celticminstrel> I dunno.
20160816 01:51:38< vultraz> don't forget sp
20160816 01:51:58< celticminstrel> I don't think SP is actually a problem though.
20160816 01:52:13< vultraz> sp campaigns
20160816 01:52:16< vultraz> in mp
20160816 01:52:24< celticminstrel> There's no reason to support that.
20160816 01:53:25< vultraz> the option exists in debug mode
20160816 01:53:53< vultraz> honestly, why would we not want to support co-op campaigns?
20160816 01:54:06< celticminstrel> Because the infrastructure is not in place?
20160816 01:54:26< vultraz> maybe not right now
20160816 01:54:43< celticminstrel> When Wesnoth is in a state where you don't need any special considerations for getting something to work in MP, then maybe.
20160816 01:55:17< vultraz> there shouldn't be, honestly
20160816 01:55:49< vultraz> but whatever
20160816 01:56:06< vultraz> it's still bad to force reloading every time you select a campaign
20160816 01:56:12< vultraz> even if you don't show the loading screen
20160816 01:56:14< vultraz> it takes time
20160816 01:56:25< vultraz> plus, none of these are cached
20160816 01:56:31< vultraz> also bad
20160816 01:57:08< vultraz> but how would one cache these things
20160816 01:57:24< vultraz> you cannot load everything at once when connecting to mp
20160816 01:57:46< vultraz> that's against WML's crappy define= fence paradigm.
20160816 01:57:52< vultraz> and it might take too long
20160816 02:00:13< vultraz> this whole thing with the preprocessor is really the only good argument for splitting the dialogs
20160816 02:00:45< vultraz> as far as I can tell, there is absolutely no workaround
20160816 02:06:27-!- celmin [~celticmin@unaffiliated/celticminstrel] has joined #wesnoth-dev
20160816 02:11:39-!- celmin [~celticmin@unaffiliated/celticminstrel] has quit [Quit: And lo! The minstrel departs, to spread the music to the masses!]
20160816 02:14:20< tad_> Could you add a define= fence to the [scenario] to have the preprocessor bypass the bulk of the scenario? I've been considering if that might help with what looks like memory leaks, to me.
20160816 02:14:40< celticminstrel> There's always a workaround.
20160816 02:15:11< celticminstrel> Maybe you should take a break from the MP create dialog - either let someone else take a look, or mull it over for awhile.
20160816 02:15:37< tad_> Another option would be to preprocess on upload and save a hash and timestamp.
20160816 02:21:24< Aginor> or fix the preprocessor to make it quick
20160816 02:21:33< Aginor> making it a non-issue in that way
20160816 02:23:27< tad_> Trading speed for design only puts off the day of reckoning. It just means people will build bigger content which will be harder to handle when it finally fails.
20160816 02:30:41-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has quit [Quit: Page closed]
20160816 02:31:12< vultraz> Aginor: the preprocessor has gradually been getting faster over the years
20160816 02:31:28< vultraz> Aginor: but i'm guessing there's a limit to just how fast a string parser can go
20160816 02:33:35< vultraz> removing the string preprocessor would be a gigantic task but also make wesnoth 100% better
20160816 02:40:40-!- irker191 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20160816 02:49:56< vultraz> if the data could be guarded behind an actual [if] block or something
20160816 02:50:04< vultraz> and the entire config not reparsed
20160816 02:50:16< vultraz> *pipe dream*
20160816 02:55:04< vultraz> but that would either involve a wml parser connected to lua or something or a wml-like language like fabi is working on
20160816 02:55:06< vultraz> :|
20160816 02:59:26-!- celmin [~celticmin@unaffiliated/celticminstrel] has joined #wesnoth-dev
20160816 03:03:52< celticminstrel> I think you are making assumptions.
20160816 03:04:13< vultraz> celmin: then how might it be doen?
20160816 03:04:13< celticminstrel> Why would supporting [if] in eg [campaign] require either of the two things you mentioned?
20160816 03:04:15< vultraz> done
20160816 03:04:40< celticminstrel> All you need to do is make the part of the engine that interprets [campaign] tags understand that there may be an [if] tag nested.
20160816 03:04:50< vultraz> supporting [if] then would mean all the wml is parsed by the preprocessor all the time
20160816 03:04:55< vultraz> see?
20160816 03:05:04< celticminstrel> No, I do not see.
20160816 03:05:16< celmin> [if] has nothing to do with the preprocessor.
20160816 03:06:45< vultraz> ugh, im explaining myself badly
20160816 03:06:52< vultraz> say we get rid of the preprocessor
20160816 03:07:01< vultraz> we still need some way to parse wml, and some way to guard segments
20160816 03:07:32< celmin> We already have a WML parser.
20160816 03:07:38< celmin> We don't need anything new for that.
20160816 03:08:30< vultraz> ... so what does the preprocessor do, then
20160816 03:08:31< vultraz> macros?
20160816 03:08:48< celmin> It preprocesses. :P
20160816 03:09:17< celmin> You could think of it as a "separate language" from WML.
20160816 03:09:43< celmin> The preprocessor doesn't understand WML syntax. It only understands preprocessor directives, macro inclusions, and angled string literals.
20160816 03:09:59< celmin> (Maybe one or two other things, I'm not entirely sure of the details.)
20160816 03:10:10< vultraz> so wait
20160816 03:10:18< vultraz> is it within the realm of possibility to get rid of it?
20160816 03:10:45< celmin> For 1.16, maybe, though I wouldn't count on it.
20160816 03:11:08< celmin> You can't get rid of it unless you remove the need to use it at all.
20160816 03:11:25< celmin> For example, you would need an all-new syntax for terrain graphics.
20160816 03:12:06< Aginor> vultraz: the pre-processor changes replaces placeholder strings with pre-defined values
20160816 03:12:21 * vultraz ;_;
20160816 03:12:29< celmin> It's basic text replacement, no more, yeah.
20160816 03:12:38< vultraz> who's to say there will even be a 1.16
20160816 03:12:42< celmin> Me?
20160816 03:12:49< Aginor> it's got some smarts, but that's about it
20160816 03:13:00< Aginor> vultraz: are you talking about abandoning the project again?
20160816 03:13:14< vultraz> no, I'm talking about the fact that we have 3 developers :|
20160816 03:13:56< celmin> But look at how active those developers are.
20160816 03:14:17< celmin> Also, I'm pretty sure 3 is a very conservative count.
20160816 03:14:43< celmin> There's probably at least 15-20 people invested in the project.
20160816 03:14:49< celmin> Not counting players.
20160816 03:14:57< celmin> Obviously.
20160816 03:15:03< celmin> (Why did I bother saying that. :P )
20160816 03:15:18< celmin> Even 15-20 might be an underestimate.
20160816 03:15:25< vultraz> if we had 20 players I'd cry
20160816 03:15:29< celmin> Of course, not all of those are C++ programmers.
20160816 03:15:52< celmin> But they're still important.
20160816 03:17:25< vultraz> c++ programmers are the only ones that count in terms of keeping the project alive
20160816 03:17:37< Aginor> vultraz: that's bullshit
20160816 03:17:45< Aginor> 1.12 is mature
20160816 03:17:46< vultraz> how so?
20160816 03:18:02< Aginor> it could more or less go on indefinitively with people developing content and maintaining campaigns
20160816 03:18:04< vultraz> without them, who will fix bugs
20160816 03:18:07< vultraz> who will improve things
20160816 03:18:21< celmin> Without C++ programmers, WML and Lua authors would find workarounds for bugs.
20160816 03:18:56< celmin> There might not be more major releases, admittedly.
20160816 03:18:59-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160816 03:19:10< celmin> There could be minor ones though, even with only WML authors and Lua programmers.
20160816 03:19:32< celmin> (Assuming 1.14 reached a mature state comparable to 1.12.)
20160816 03:19:42< vultraz> that's what I'm aiming for
20160816 03:19:48< Aginor> and then, if someone cares enough to fix particular bugs, they'll probably branch it
20160816 03:19:51< Aginor> or fork
20160816 03:19:54< vultraz> except 1.12 was only stable because of iceiceice
20160816 03:19:59< Aginor> and prepare the next point release
20160816 03:22:02< Aginor> vultraz: so if it wasn't for him 1.12 wouldn't have been released?
20160816 03:22:13< vultraz> it would have
20160816 03:22:23< vultraz> but he fixed many many many bugs
20160816 03:22:25< vultraz> he and gfgtdf
20160816 03:23:30-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 244 seconds]
20160816 03:26:01< celmin> I wonder if Steam has some sort of authentication API that the Wesnoth server could hook into. Like Minecraft for example.
20160816 03:27:44< vultraz> probably connected to your steam account like dota does
20160816 03:28:21< celmin> Hm?
20160816 03:29:55< vultraz> in dota, you don't use a separate account
20160816 03:30:00< vultraz> it's your steam account
20160816 03:30:17< celmin> But we can't quite do that with Wesnoth.
20160816 03:31:30< vultraz> hm?
20160816 03:31:48< celmin> We need something akin to how it links to forum accounts.
20160816 03:32:13< celmin> Maybe Steam users would get names starting with $ or something, I dunno (to avoid conflicts).
20160816 03:32:26< vultraz> conflicts?
20160816 03:32:28< vultraz> with forum names?
20160816 03:33:02< celmin> Yeah.
20160816 03:36:45< Aginor> celmin: I don't know, but it's fully possible that steam is an oauth2 provider
20160816 03:37:05< Aginor> if that's the case, we can integrate with that reasonably easy
20160816 03:37:17< celmin> Aginor: If it is, that'd be perfect. We need something that doesn't require linking with Steams libs, after all.
20160816 03:38:49< vultraz> (maybe)
20160816 03:38:59< vultraz> I'm sure we can find a way around that if we look 
20160816 03:54:07-!- hk238 [~kvirc@t224.ip7.netikka.fi] has joined #wesnoth-dev
20160816 04:47:16-!- irker023 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20160816 04:47:16< irker023> wesnoth: Wedge009 wesnoth:master 63c65d3ac695 / src/gui/dialogs/unit_attack.cpp: Show correct number of attacks when unit has swarm weapon special (bug #24978) https://github.com/wesnoth/wesnoth/commit/63c65d3ac69598bedefa678a37b7ea5f455ed7c0
20160816 04:47:16< irker023> wesnoth: Wedge009 wesnoth:master 5ed288d6359a / changelog players_changelog: Update change logs. https://github.com/wesnoth/wesnoth/commit/5ed288d6359a4d62358b047bed02bf528d32ba77
20160816 04:47:17< irker023> wesnoth: Charles Dang wesnoth:master b4a98cdc84e9 / changelog players_changelog src/gui/dialogs/unit_attack.cpp: Merge pull request #750 from Wedge009/bug_24978_fix https://github.com/wesnoth/wesnoth/commit/b4a98cdc84e9a0206786457be8cc87bfa8208baa
20160816 04:47:48< wedge009> That was quick. Was that the right thing to do?
20160816 04:47:52< wedge009> What I changed, I mean.
20160816 04:48:08< vultraz> I assume you tested it :)
20160816 04:48:28< shadowm> <vultraz> who's to say there will even be a 1.16
20160816 04:48:30< vultraz> and it looked fine
20160816 04:48:33< shadowm> I'd like to ask you to stop doing this.
20160816 04:49:17< wedge009> vultraz: Yes, of course.
20160816 04:49:18< wedge009> Thanks.
20160816 04:54:54-!- Kwandulin [~Miranda@p200300760F35BF31DCC60AA2A3D3806F.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160816 05:07:19-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160816 05:11:19-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 244 seconds]
20160816 05:19:29-!- ancestral [~ancestral@67-4-224-82.mpls.qwest.net] has joined #wesnoth-dev
20160816 05:27:49-!- ancestral_ [~ancestral@67-4-224-82.mpls.qwest.net] has joined #wesnoth-dev
20160816 05:28:24-!- ancestral [~ancestral@67-4-224-82.mpls.qwest.net] has quit [Read error: Connection reset by peer]
20160816 05:28:24-!- ancestral_ is now known as ancestral
20160816 05:39:12< vultraz> good god, the lobby code is a mess
20160816 05:42:15< vultraz> celmin: I propose removing the "other games" player list
20160816 05:42:58-!- travis-ci [~travis-ci@ec2-54-157-210-248.compute-1.amazonaws.com] has joined #wesnoth-dev
20160816 05:42:59< travis-ci> wesnoth/wesnoth#10420 (master - b4a98cd : Charles Dang): The build has errored.
20160816 05:42:59< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/152577355
20160816 05:42:59-!- travis-ci [~travis-ci@ec2-54-157-210-248.compute-1.amazonaws.com] has left #wesnoth-dev []
20160816 05:46:35< celmin> Well, I suppose it could always be put back later anyway.
20160816 05:46:45< celmin> If it's simpler for you to remove it, go ahead?
20160816 06:12:18-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20160816 06:12:18-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Read error: Connection reset by peer]
20160816 06:55:37-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160816 07:00:09-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 260 seconds]
20160816 07:03:21-!- JyrkiVesterinen [~JyrkiVest@87-100-224-31.bb.dnainternet.fi] has joined #wesnoth-dev
20160816 07:17:59< vultraz> JyrkiVesterinen: hello
20160816 07:18:12< JyrkiVesterinen> Hi vultraz
20160816 07:18:18< vultraz> JyrkiVesterinen: I noticed an issue with gui1 dialogs post-your event changes
20160816 07:18:34< vultraz> if you bring up a menu, say, its contents won't be drawn until you move your mouse
20160816 07:18:43< vultraz> (which, I assume, triggers a draw cycle or something)
20160816 07:18:57< vultraz> likewise for gui1 dialogs, where you see a delay before the contents appear
20160816 07:19:10< vultraz> theme items also disappear when they're open
20160816 07:19:24< JyrkiVesterinen> Hmm. I see. I thought that messing with event delivery was risky... :(
20160816 07:20:12< JyrkiVesterinen> I can investigate the problem tomorrow (it's morning where I live). Not today, my head hurts. :(
20160816 07:20:51< vultraz> No rush
20160816 07:21:05< vultraz> It's rather minor, and gui1 dialogs are dropping like flies :P
20160816 07:27:39< vultraz> celmin: I've figured it out
20160816 07:27:45< celmin> ??????
20160816 07:27:54< vultraz> celmin: the listboxes for the player lists in the lobby is an *old implementation*
20160816 07:27:56< vultraz> it's hidden
20160816 07:28:00< vultraz> I shall remove it
20160816 07:28:05< celmin> Oh fun
20160816 07:28:07-!- Kwandulin_2 [~Miranda@p200300760F35BF9CDCC60AA2A3D3806F.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160816 07:28:09< vultraz> it looks like a kind of proto-tree view 
20160816 07:28:25< celmin> I didn't accidentally use it in the low-res, did I?
20160816 07:28:29< celmin> I think I didn't.
20160816 07:28:58< vultraz> no
20160816 07:29:01-!- ChipmunkV [~vova@static-89-94-113-91.axione.abo.bbox.fr] has joined #wesnoth-dev
20160816 07:29:14-!- Kwandulin [~Miranda@p200300760F35BF31DCC60AA2A3D3806F.dip0.t-ipconnect.de] has quit [Ping timeout: 258 seconds]
20160816 07:47:52-!- irker023 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20160816 07:53:23-!- Kwandulin_2 [~Miranda@p200300760F35BF9CDCC60AA2A3D3806F.dip0.t-ipconnect.de] has quit [Ping timeout: 258 seconds]
20160816 07:54:31-!- Kwandulin [~Miranda@p200300760F35BF9CF9FB0B903664D6EA.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160816 07:58:52-!- ChipmunkV [~vova@static-89-94-113-91.axione.abo.bbox.fr] has quit [Quit: ChipmunkV]
20160816 08:03:03-!- VultCave [~chatzilla@124.109.10.167] has joined #wesnoth-dev
20160816 08:04:27-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Ping timeout: 250 seconds]
20160816 08:04:39-!- VultCave is now known as vultraz
20160816 08:06:58< Aginor> vultraz: so can you convert the main gameboard to gui2?
20160816 08:07:04< Aginor> and the help?
20160816 08:07:08< Aginor> and the storyscreen?
20160816 08:07:22< celmin> Main gameboard, probably not...
20160816 08:07:46< celmin> Help will be a huge pain too.
20160816 08:07:53< celmin> Does the storyscreen use GUI1 then?
20160816 08:07:58-!- vultraz [~chatzilla@124.109.10.167] has quit [Read error: Connection reset by peer]
20160816 08:08:01< celmin> For buttons maybe...
20160816 08:08:43-!- vultraz [~chatzilla@124.109.10.167] has joined #wesnoth-dev
20160816 08:08:54-!- vultraz [~chatzilla@124.109.10.167] has quit [Changing host]
20160816 08:08:54-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20160816 08:09:40< Aginor> celmin: it's proto-pre-gui1
20160816 08:09:53< Aginor> it uses the same semantics but lives outside the widget tree
20160816 08:10:01< celmin> You mean proto-pre-gui2?
20160816 08:10:05< Aginor> it wouldn't be hard to move to a gui2 environment
20160816 08:10:09< Aginor> celmin: no, gui1
20160816 08:10:15< Aginor> it appears Old
20160816 08:10:24< celmin> But it uses GUI2's text rendering.
20160816 08:10:29< Aginor> not that I bothered doing a git blame
20160816 08:10:46< celmin> I think the main difficulty in porting it to GUI2 was text fade-in.
20160816 08:11:06-!- VultCave [~chatzilla@124.109.10.167] has joined #wesnoth-dev
20160816 08:12:32< shadowm> I don't really feel attached to the current transition effect these days.
20160816 08:12:42< shadowm> It seems so cheap and pointless.
20160816 08:12:51< Aginor> they are pretty tacky
20160816 08:13:06< VultCave> what transition effect?
20160816 08:13:17< celmin> Well, it doesn't necessarily need to be the same effect, but I think it's good to have some sort of transition effect.
20160816 08:13:23< shadowm> Back when I ported it from direct SDL_ttf wrapper API calls, it was a character-wise effect.
20160816 08:13:46< shadowm> I couldn't figure out how to do it using Pango, so I made it line-wise instead.
20160816 08:13:59-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Ping timeout: 250 seconds]
20160816 08:13:59-!- VultCave is now known as vultraz
20160816 08:14:06< shadowm> An easing alpha blending transition would probably look much better.
20160816 08:14:13< celmin> vultraz: Storyscreen
20160816 08:14:21-!- vultraz [~chatzilla@124.109.10.167] has quit [Changing host]
20160816 08:14:21-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20160816 08:14:23< shadowm> Maybe. Not sure. Not going to try it myself either. I'm trying to not touch Wesnoth's code.
20160816 08:14:59< Aginor> I am still trying to make up my mind
20160816 08:15:10< shadowm> Like an ex-alcoholic trying to not fall off the wagon.
20160816 08:15:15< shadowm> Me, I mean.
20160816 08:15:28< vultraz> the rolling text?
20160816 08:15:30 * Aginor introduces shadowm to slither.io instead
20160816 08:15:44< vultraz> no, you want agar.io
20160816 08:22:01-!- Kwandulin [~Miranda@p200300760F35BF9CF9FB0B903664D6EA.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160816 08:22:03< Aginor> slither is better
20160816 08:22:41-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20160816 08:29:35-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Read error: Connection reset by peer]
20160816 08:29:58-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20160816 08:30:52< zookeeper> what's the problem with the preprocessor WRT LoW/ANL/etc?
20160816 08:32:03-!- ancestral [~ancestral@67-4-224-82.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20160816 08:32:21-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160816 08:33:05-!- Duthlet [~Duthlet@dslb-188-105-138-092.188.105.pools.vodafone-ip.de] has joined #wesnoth-dev
20160816 08:33:37< vultraz> zookeeper: they have define= keys, therefor, when you load their config, the preprocessor runs. This is a problem for the joint mp create/configure dialog, since what used to be a potential loading screen *aftter* create now becomes a loading screen when selecting certain items. I could get around this for mp scenarios by reading from the [multiplayer] tags directly, but that doesn't help...
20160816 08:33:38< vultraz> ...for campaigns, for which I need to load the config to get a list of campaign scenarios for the entry point this. Therefor, it seems I might have to split the dialogs after all.
20160816 08:34:37< vultraz> you see?
20160816 08:36:48< zookeeper> well, then a MP campaign that wants to use entry points simply has to move the ifdef guard into individual scenarios (guarding off the bulk of the WML, but leaving in whatever little the entry point code needs)?
20160816 08:37:30< celmin> But MP campaigns that haven't been updated would still either trigger the loading screen or not properly show their default settings.
20160816 08:37:44< zookeeper> sure. how many UMC MP campaigns are there?
20160816 08:37:48< celmin> No idea.
20160816 08:38:42< zookeeper> 1.13 add-on server says 2.
20160816 08:39:04< celmin> What about 1.12?
20160816 08:40:35< zookeeper> says 13, although i doubt all of those count in this case
20160816 08:40:46< celmin> Why's that?
20160816 08:41:24< zookeeper> just a hunch. anyway, the remaining question is whether any of those use entry points
20160816 08:41:28< zookeeper> is that greppable?
20160816 08:41:46< celmin> Oh, wait, is this not a problem when they don't use entry points?
20160816 08:41:59< zookeeper> well, i'm not sure
20160816 08:42:07< zookeeper> that's the impression i got
20160816 08:42:21< celmin> I'm not sure if entry points are greppable. I think they involve something like "allow_new_game=yes", but I think that can be used in regular MP scenarios too.
20160816 08:43:56-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160816 08:48:24-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 265 seconds]
20160816 08:49:31-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20160816 08:56:13< shadowm> gfgtdf: Have you considered not displaying the debug command alert ("debug command foo was used by playername yadda yadda") in SP mode?
20160816 08:56:51< shadowm> I'm sure there are some people for whom this works as a guilt-tripping tactic, but for a content creator/dev it's just annoying.
20160816 08:57:21< celmin> At least disable it when there are no remote sides
20160816 08:59:13< shadowm> So that those silly people who make AI-only games in the MP server can cheat without observers realizing?
20160816 08:59:16< shadowm> :p
20160816 08:59:55< celmin> No, so I don't have to see it when testing local MP games
20160816 09:01:23< shadowm> Maybe it could be ommitted when you are using debug mode yourself.
20160816 09:02:11< shadowm> i.e. only other players can see it (including people watching replays of your SP game).
20160816 09:02:23-!- Kwandulin [~Miranda@p200300760F35BF9CD4059601E3CC3CA5.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160816 09:04:58-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160816 09:09:37-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 252 seconds]
20160816 09:09:40-!- celticminstrel is now known as celmin|sleep
20160816 09:49:04-!- Sapient [~yourstrul@180.190.93.6] has joined #wesnoth-dev
20160816 09:49:04-!- Sapient [~yourstrul@180.190.93.6] has quit [Changing host]
20160816 09:49:04-!- Sapient [~yourstrul@wesnoth/developer/Sapient] has joined #wesnoth-dev
20160816 09:49:08-!- ChipmunkV [~vova@static-89-94-113-91.axione.abo.bbox.fr] has joined #wesnoth-dev
20160816 10:07:13-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160816 10:17:14-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20160816 10:19:12-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160816 10:21:16< zookeeper> so if one had to convert a lot of ' into ’, is that easily pofixable or otherwise doable in a nice manner?
20160816 10:23:09 * vultraz has no idea
20160816 10:33:36-!- gfgtdf [~chatzilla@x4e363805.dyn.telefonica.de] has joined #wesnoth-dev
20160816 10:34:17-!- Shiki [~Shiki@HSI-KBW-149-172-220-18.hsi13.kabel-badenwuerttemberg.de] has joined #wesnoth-dev
20160816 10:34:20< gfgtdf> 20160816 08:56:13< shadowm> gfgtdf: Have you considered not displaying the debug command alert ("debug command foo was used by playername yadda yadda") in SP mode?
20160816 10:34:54< gfgtdf> well i thinkit mkaes sesne to show them alsi in sp replays to 'sp' seems liek it not the correct way too showdeide whether to show that message
20160816 10:35:37< gfgtdf> shadowm: i think the best solution woudl be just not to show them to the one who issues that command,, .. maybe i'l implement that sometime.
20160816 11:00:39-!- Shiki [~Shiki@HSI-KBW-149-172-220-18.hsi13.kabel-badenwuerttemberg.de] has quit [Remote host closed the connection]
20160816 11:18:14-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20160816 11:23:07-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160816 11:26:19< vultraz> celmin: https://drive.google.com/file/d/0B-mR9s8FduLLamtTSDM1NGJSUDA/view?usp=sharing
20160816 11:26:31< wedge009> zookeeper: What needs to be changed for 'into'?
20160816 11:27:05< zookeeper> wedge009, ''s into ’'s :p
20160816 11:27:15< wedge009> Maybe it's just the background colour, but it reminds me of the buggy 'experimental' lobby.
20160816 11:27:19< wedge009> But looks nice.
20160816 11:27:29< vultraz> wedge009: it IS the experimental lobby :P
20160816 11:27:34< wedge009> Oh, ha ha.
20160816 11:27:36< vultraz> xD
20160816 11:27:43-!- Shiki [~Shiki@HSI-KBW-149-172-220-18.hsi13.kabel-badenwuerttemberg.de] has joined #wesnoth-dev
20160816 11:27:49-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 265 seconds]
20160816 11:27:57< vultraz> I improved the interface before, now I'm improving it more
20160816 11:28:12< wedge009> zookeeper: Oh, you mean quotes and apostrophes into their left/right equivalents.
20160816 11:28:26< zookeeper> wedge009, just apostrophes, but yes
20160816 11:28:52< wedge009> So not so buggy now?
20160816 11:29:12< wedge009> Hope this means we can close all those experimental lobby bug reports afterwards. :D
20160816 11:29:27< vultraz> it still has issues
20160816 11:29:32< wedge009> Progress!
20160816 11:29:32< vultraz> but those are listbox issues
20160816 11:29:49< zookeeper> wedge009, we're supposed to use ’ in text, but everyone such as myself forgets to do so and just uses ' so we have lots of those currently
20160816 11:29:52< vultraz> and they need to be fixed before mp create can be used too
20160816 11:30:36< wedge009> zookeeper: It's worse when they're encoded badly and turn out as gibberish.
20160816 11:30:46< wedge009> Not UTF, probably.
20160816 11:32:00< zookeeper> well i'm sure it is...
20160816 11:32:18< zookeeper> ...have they ever?
20160816 11:36:29< gfgtdf> vultraz: do you currently wokr on the gui2 mpo create tiltescrren siuue ?
20160816 11:38:31< vultraz> gfgtdf: hm?
20160816 11:38:34< vultraz> titlescreen?
20160816 11:38:41< gfgtdf> lloadignscreen
20160816 11:38:59< vultraz> no, I don't know what to do 
20160816 11:39:04< vultraz> besides possibly split the dialog in two
20160816 11:40:35< gfgtdf> vultraz: also is there a reson hwo showin the gui1 dialog after the gui2 one ? It somehow mkes it harder to test whether it cna actuall be used to create a scenario
20160816 11:41:59< vultraz> gfgtdf: no reason, i just didn't think the dialog was ready to actually create a scenario
20160816 11:42:17< vultraz> gfgtdf: and i didn't think the gui2 dialog needed to be in the same loop as the gui1 one
20160816 11:42:27< vultraz> feel free to put the gui1 dialog in an else block
20160816 11:43:13< gfgtdf> vultraz: i think this loop is made so that you get back to the screte screen (instad of the lobby) when you press 'back' from the connect screen
20160816 11:44:13< vultraz> hm true..
20160816 12:17:41-!- Kwandulin [~Miranda@p200300760F35BF9CD4059601E3CC3CA5.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160816 12:36:18-!- Shiki [~Shiki@HSI-KBW-149-172-220-18.hsi13.kabel-badenwuerttemberg.de] has quit [Remote host closed the connection]
20160816 12:57:51-!- irker022 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20160816 12:57:51< irker022> wesnoth: gfgtdf wesnoth:master f264b6cddc71 / src/ (4 files in 2 dirs): gui2 mp create improvements https://github.com/wesnoth/wesnoth/commit/f264b6cddc7183f60001a9552d385bdfb95daddb
20160816 12:57:52< gfgtdf> vultraz: ^
20160816 12:59:01< vultraz> :D
20160816 13:00:20< vultraz> thanks
20160816 13:02:52< vultraz> btw, there're two main listbox issue that needs to be fixed before gui2 mp create and lobby can be used. * child widgets should never shrink smaller than the listbox and * listboxes should optimally not shrink if you replace their contents with smaller contents
20160816 13:02:59-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20160816 13:04:15< vultraz> gfgtdf: confirmed the loading screen no longer shows :D
20160816 13:04:17< vultraz> how did you do it?
20160816 13:04:35< gfgtdf> vultraz: i just used the [multiplayer] taghs withotu reloading the game config
20160816 13:04:41< vultraz> ahh
20160816 13:04:59< gfgtdf> vultraz: when the listbox itslef shirinks that us isuall casues by a window->invalidate_lyout() clal that is not nesessary dont by the listbox
20160816 13:05:22< gfgtdf> vultraz: for exmaple in the mp create screen the croll_label also calls window->invalidate() layout
20160816 13:05:26< vultraz> yeah
20160816 13:05:37< vultraz> that can cause the listbox to shrink :/
20160816 13:06:29< vultraz> that's livable tho, I guess
20160816 13:06:34< vultraz> I can always for a minimum size
20160816 13:06:49< vultraz> more pressing is the toggle panels in the lobby shrinking in width
20160816 13:06:57-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 258 seconds]
20160816 13:06:57-!- wedge010 is now known as wedge009
20160816 13:09:15-!- Kwandulin [~Miranda@p200300760F35BF9C756D6B5013A682B4.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160816 13:11:42< gfgtdf> vultraz:  i just noticed that it currently onl show [options] for the selected [multiplayer]
20160816 13:12:00< gfgtdf> vultraz: but [era] and [modification] can also ontain [options]
20160816 13:12:08< vultraz> yeah I haven't added that yet
20160816 13:18:13-!- Shiki [~Shiki@HSI-KBW-149-172-220-18.hsi13.kabel-badenwuerttemberg.de] has joined #wesnoth-dev
20160816 13:30:19-!- gfgtdf [~chatzilla@x4e363805.dyn.telefonica.de] has quit [Quit: ChatZilla 0.9.92 [Firefox 48.0/20160726073904]]
20160816 13:35:08< wedge009> Aginor: Not expecting you to look at this straight away, but if/when you have a chance: https://gna.org/bugs/?24984. Do you start Wesnoth in full-screen on Linux? I don't think these issues are specific to a type of desktop environment, but I'm wondering if it's something related to SDL2 and/or GUI2.
20160816 13:39:45-!- horrowind [~Icedove@2a02:810a:83c0:404:21b:fcff:fee3:c3ff] has joined #wesnoth-dev
20160816 13:52:11-!- travis-ci [~travis-ci@ec2-54-227-41-166.compute-1.amazonaws.com] has joined #wesnoth-dev
20160816 13:52:12< travis-ci> wesnoth/wesnoth#10421 (master - f264b6c : gfgtdf): The build has errored.
20160816 13:52:12< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/152668755
20160816 13:52:12-!- travis-ci [~travis-ci@ec2-54-227-41-166.compute-1.amazonaws.com] has left #wesnoth-dev []
20160816 14:08:58< irker022> wesnoth: Charles Dang wesnoth:master 0fd26b9dc747 / data/gui/ (widget/unit_preview_pane.cfg window/unit_recruit.cfg): Few small improvements https://github.com/wesnoth/wesnoth/commit/0fd26b9dc747c2f3a426637c866ce8798f7e16a1
20160816 14:09:01< irker022> wesnoth: Charles Dang wesnoth:master a9116a7ed544 / data/gui/widget/toggle_button_icon.cfg: Don't scale toggle button icons if they're smaller than the button size https://github.com/wesnoth/wesnoth/commit/a9116a7ed544545d3e5579084a2eb80fce91e6a3
20160816 14:09:04< irker022> wesnoth: Charles Dang wesnoth:master b58a6de5784f / data/gui/widget/scroll_label_description.cfg: Fixed description_small scroll labels not working https://github.com/wesnoth/wesnoth/commit/b58a6de5784f8d044a55aca8bf4476cd6763c089
20160816 14:09:07< irker022> wesnoth: Charles Dang wesnoth:master d9b6bd7fa162 / data/gui/window/lobby_player_info.cfg: MP Player Info: layout improvements https://github.com/wesnoth/wesnoth/commit/d9b6bd7fa1625aef2192cd7017f69ab06c30455a
20160816 14:09:10< irker022> wesnoth: Charles Dang wesnoth:master 32451f2245f7 / / (4 files in 3 dirs): MP Lobby: layout improvements and cleanup https://github.com/wesnoth/wesnoth/commit/32451f2245f73049ff79c7ad6dda21a7e1692246
20160816 14:23:34-!- Kwandulin [~Miranda@p200300760F35BF9C756D6B5013A682B4.dip0.t-ipconnect.de] has quit [Ping timeout: 258 seconds]
20160816 14:26:54< celmin> vultraz: Screenshot's nice, but needs less space between the checkboxes.
20160816 14:27:12< celmin> gfgtdf: Do your MP create changes affect GUI1 MP create?
20160816 14:32:17< zookeeper> ok, i really need to start being able to compile again.
20160816 14:32:46< zookeeper> old builds aren't functioning anymore
20160816 14:36:49-!- Kwandulin [~Miranda@p200300760F35BF9C756D6B5013A682B4.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160816 14:37:22< zookeeper> is there no one but aquileia (who has been absent for a long time) who can update the dependency archive(s) here? https://wiki.wesnoth.org/CompilingWesnothOnWindows#Visual_Studio
20160816 14:41:53< celmin> So was it not actually using a treebox for rooms before this?
20160816 14:42:35< vultraz> celmin: it was
20160816 14:43:06< vultraz> but it did some weird thing where their labels were set using the values of some labels in the hidden old listbox
20160816 14:43:33< celmin> Weird...
20160816 14:43:54-!- celmin|sleep is now known as celticminstrel
20160816 14:43:59< vultraz> indeed
20160816 14:45:08< vultraz> now it purely uses the listbox
20160816 14:45:12< vultraz> er
20160816 14:45:14< vultraz> tree view
20160816 14:45:16< vultraz> no listboxes
20160816 14:45:54< vultraz> i also thought giving the player lists more space was good, so one can see longer lists
20160816 14:46:12< vultraz> and so they have enough width to fit most usernames without making the column resize
20160816 14:47:45-!- Kwandulin [~Miranda@p200300760F35BF9C756D6B5013A682B4.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160816 14:47:57< celmin> What's with this "edward" guy posting in the Greenlight thread.
20160816 14:48:39< vultraz> blah, he has a colored signature
20160816 14:49:04< celmin> I mean more the fact that he appears to be talking as if he were "part of the team"
20160816 14:49:12< vultraz> that too
20160816 14:49:16< vultraz> he quoted me exactly
20160816 14:49:27< vultraz> and he's flagged..
20160816 14:49:34 * vultraz is unsure what action to take here
20160816 14:49:36< zookeeper> edward who? i see no one like that
20160816 14:49:39 * zookeeper whistles
20160816 14:50:46< vultraz> :P
20160816 14:50:53< zookeeper> (post content was copied, so it was a bot)
20160816 14:51:01< celmin> Ah.
20160816 14:51:14< vultraz> thanks
20160816 14:51:33< celmin> Posted in "remove planning mode" too.
20160816 14:51:39< vultraz> indeed
20160816 14:51:48< celmin> Now that I think about it, that was a copy too.
20160816 14:52:21< vultraz> anyway
20160816 14:52:28< vultraz> i'm calling the lobby layout done now
20160816 14:53:13< vultraz> sorry about the space between checkboxes, but it just looks unbearably ugly to have the filter box stretch halfway across the screen :|
20160816 14:55:26< celmin> I don't see why?
20160816 14:55:43< vultraz> it just really does
20160816 14:57:03< celmin> Still, it looks too small in the screenshot. It could be at least 1.5 - 2 times as large.
20160816 14:57:21< vultraz> that is the default text box size
20160816 14:57:25< celmin> Did you touch the low-res layout?
20160816 14:57:29< vultraz> which could be increased, of course
20160816 14:57:33< vultraz> no, I didn't
20160816 14:57:36< vultraz> but I probably broke it
20160816 14:57:45< celmin> Well, I'll check it out soonish.
20160816 14:58:22< vultraz> just checked, it still works
20160816 14:59:04< zookeeper> loonycyborg, any chance we could have jenkins builds for windows working anytime soon?
20160816 14:59:23< loonycyborg> they're not working?
20160816 14:59:35< zookeeper> well looks like that to me?
20160816 14:59:39< celmin> There were jenkins builds?
20160816 14:59:40< loonycyborg> hmm I can look into it later a bit
20160816 14:59:52-!- gfgtdf [~chatzilla@x4e363805.dyn.telefonica.de] has joined #wesnoth-dev
20160816 15:00:13< vultraz> celmin: small task you can work on that you can probably do in 30 minutes: convert the random map settings dialog to gui2
20160816 15:00:22< vultraz> make use of tfield_bool/integer
20160816 15:00:23< gfgtdf> vultraz: is there some easy way to tes the gui1 lobby? I mean there are usually no people on the 1.13 server so i dont see anything.
20160816 15:00:41< vultraz> gui2 lobby you mean?
20160816 15:00:46< gfgtdf> vultraz: yes
20160816 15:00:57< vultraz> you just need to open other instances of wesnoth and create a game or something
20160816 15:01:19< gfgtdf> vultraz: in the random map dialog the max value of some slidersw is dependent on the values of the other sliders, so it migth be not that easy
20160816 15:01:35< gfgtdf> vultraz: of yourse we coudl also drop that so that the map values stasy constant
20160816 15:01:41< vultraz> if you're looking into the listbox issue, it's immediately evident as soon as any data in it the row changes
20160816 15:02:08< vultraz> gfgtdf: yeah, i'd recommend dropping that. 
20160816 15:02:39< vultraz> (for celmin's reference: the width/height sliders move slightly in tandem with the players slider)
20160816 15:03:10< vultraz> if he wants to keep that, it's really not much effort to implement tho
20160816 15:03:29< vultraz> just add a value_changed signal handler to the players slider and set the w/h sliders accordingly
20160816 15:03:33< gfgtdf> vultraz: the gui2 obby show a horizonal scrollbar when it runs with ~800 with (which is what i usually use when i have multiple wesnoth instances open.)
20160816 15:03:46-!- travis-ci [~travis-ci@ec2-54-157-210-248.compute-1.amazonaws.com] has joined #wesnoth-dev
20160816 15:03:47< travis-ci> wesnoth/wesnoth#10422 (master - 32451f2 : Charles Dang): The build has errored.
20160816 15:03:47< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/152689981
20160816 15:03:47-!- travis-ci [~travis-ci@ec2-54-157-210-248.compute-1.amazonaws.com] has left #wesnoth-dev []
20160816 15:03:49< zookeeper> loonycyborg, great
20160816 15:04:11< vultraz> hm that doesn't seem to happen for me..
20160816 15:04:35< gfgtdf> vultraz: hmm i'll pull form master any try again
20160816 15:05:08< loonycyborg> celmin: http://wesnoth.org:8080/
20160816 15:06:03< gfgtdf> vultraz: and there is still the issue woith the desrciption scroll label on small resolutions (its put the whole text in one line in this case)
20160816 15:06:37< vultraz> gfgtdf: wait, do you mean mp create or the mp lobby has the horizontal scrollbar?
20160816 15:06:39< vultraz> that
20160816 15:06:41< vultraz> s knonw
20160816 15:06:44< vultraz> known
20160816 15:06:48< gfgtdf> lobby has a scrollbar
20160816 15:06:52< vultraz> hm
20160816 15:06:54< gfgtdf> create too
20160816 15:07:10< gfgtdf> on small (horizontal) resolutions that is
20160816 15:07:33< celmin> 800x660 or smaller uses different layout.
20160816 15:07:59< gfgtdf> celmin: you mean lobby or create ?
20160816 15:08:04< celmin> Lobby
20160816 15:08:10< celmin> Haven't implemented that yet for create
20160816 15:11:46 * vultraz thinks he found a bug
20160816 15:12:37< vultraz> weird lag before units attack each other in mp... and console config dumbs
20160816 15:12:54< vultraz> might be from creating the game from the new dialog...?
20160816 15:13:18< gfgtdf> vultraz: what do the dumps say?
20160816 15:13:35< gfgtdf> vultraz: it might alos come from the mote carolo attack prediction changes
20160816 15:14:36< celmin> vultraz: If you think it might be from creating the game from the new dialog, then disable new lobby and try again. :P
20160816 15:14:41< vultraz> lemme try to reproduce
20160816 15:16:27< gfgtdf> vultraz: also th elobby doenst show active modification for games
20160816 15:16:41< vultraz> it's supposed to?
20160816 15:16:48< gfgtdf> vultraz: yes
20160816 15:17:07< gfgtdf> vultraz: hwsin modifications for games is also a problem in the gui1 lobby since if people play with a lot of modifications not all are shown due to not enough soace in one line
20160816 15:17:21< celmin> Tooltips?
20160816 15:18:12< gfgtdf> celmin: that coudl work i guess speciall scine gui2 does already show the full text on tooltips oif a test was clipped
20160816 15:19:34< vultraz> gfgtdf: http://pastebin.com/sNhjbn2c
20160816 15:19:36< vultraz> etc 
20160816 15:20:15< celmin> That's the config diff format… why is that being dumped...
20160816 15:20:19< vultraz> there seems to be a minor lag before every attack begins, but sometime the game freezes completely and this happens..
20160816 15:20:29< gfgtdf> vultraz: hmm it seems liek the server sends you [gamelist_diff] whiel you are in the game
20160816 15:20:34-!- JyrkiVesterinen [~JyrkiVest@87-100-224-31.bb.dnainternet.fi] has quit [Quit: .]
20160816 15:20:56< gfgtdf> vultraz: usually teh servre shouldnt know that you are in a game and not send you those packages i suppose
20160816 15:21:20< celmin> Huh, so the network protocol uses the diff format. I didn't know that.
20160816 15:21:58< gfgtdf> vultraz: whats alo rather strange is that i saw you on the playerist as in 'in other games' but i didnt saw you game on he gamelist 
20160816 15:22:13< vultraz> hm
20160816 15:26:52< celmin> I have no idea what all those little icons mean.
20160816 15:27:11< gfgtdf> celmin: on 800x600 the game show s ascrollbar at the bottom, but one on german translations
20160816 15:27:27< gfgtdf> celmin: mostlilely somce of the texts is longer in german
20160816 15:27:38< celmin> Which text?
20160816 15:27:41< celmin> Buttons?
20160816 15:27:56< vultraz> the little icons need to be redone
20160816 15:28:29< gfgtdf> celmin: dont know a lt of texts are longer, but not sure which is the important one
20160816 15:29:13< gfgtdf> celmin: also when for exmaple run on 800x100 it use the 'large' layout, i it'd be better if it used the sall layout there too
20160816 15:29:19< gfgtdf> 800x1000*
20160816 15:29:28< celmin> Game entry doesn't take full width.
20160816 15:30:08< celmin> Huh? I thought it would've used small layout there since one dimension is 800...
20160816 15:30:10< vultraz> yes, that's an annoying listbox bug
20160816 15:30:29< vultraz> somehow, child widgets (the toggle panel, in this case) are allowed to shrink
20160816 15:41:26-!- Shiki [~Shiki@HSI-KBW-149-172-220-18.hsi13.kabel-badenwuerttemberg.de] has quit [Remote host closed the connection]
20160816 15:47:47-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160816 15:55:14-!- Sapient [~yourstrul@wesnoth/developer/Sapient] has quit [Quit: Have fun, cya!]
20160816 16:02:56-!- ancestral [~ancestral@67-4-224-82.mpls.qwest.net] has joined #wesnoth-dev
20160816 16:04:27-!- JyrkiVesterinen [~JyrkiVest@87-100-250-87.bb.dnainternet.fi] has joined #wesnoth-dev
20160816 16:10:50-!- mjs-de [~mjs-de@x4e3190fe.dyn.telefonica.de] has joined #wesnoth-dev
20160816 16:19:37-!- ancestral [~ancestral@67-4-224-82.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20160816 16:25:59< gfgtdf> vultraz: ^
20160816 16:26:04< irker022> wesnoth: gfgtdf wesnoth:master af5897172782 / src/gui/widgets/listbox.cpp: fix listbox items to small after tlistbox::set_row_shown https://github.com/wesnoth/wesnoth/commit/af58971727828f2045cf8e61fe088b9dccc01bb2
20160816 16:29:51-!- Shiki [~Shiki@HSI-KBW-149-172-220-18.hsi13.kabel-badenwuerttemberg.de] has joined #wesnoth-dev
20160816 16:32:49-!- ancestral [~ancestral@67-4-224-82.mpls.qwest.net] has joined #wesnoth-dev
20160816 17:20:48-!- travis-ci [~travis-ci@ec2-54-81-250-138.compute-1.amazonaws.com] has joined #wesnoth-dev
20160816 17:20:49< travis-ci> wesnoth/wesnoth#10423 (master - af58971 : gfgtdf): The build has errored.
20160816 17:20:49< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/152731066
20160816 17:20:49-!- travis-ci [~travis-ci@ec2-54-81-250-138.compute-1.amazonaws.com] has left #wesnoth-dev []
20160816 17:30:30-!- Kwandulin [~Miranda@p200300760F35BF9CBC4EB186147F7676.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160816 17:32:36-!- Greg-Bog_ [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160816 17:36:34-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 260 seconds]
20160816 17:38:53-!- boucman_work [~boucman@2a02-8428-034f-f800-9e32-0c7c-b391-6223.rev.sfr.net] has joined #wesnoth-dev
20160816 17:38:54-!- boucman [~rosen@2a02-8428-034f-f800-6451-74a2-180b-7330.rev.sfr.net] has joined #wesnoth-dev
20160816 17:38:54-!- boucman [~rosen@2a02-8428-034f-f800-6451-74a2-180b-7330.rev.sfr.net] has quit [Changing host]
20160816 17:38:54-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20160816 17:53:33-!- hk238 [~kvirc@t224.ip7.netikka.fi] has quit [Quit: http://www.kvirc.net/ 4.9.1 Aria]
20160816 17:55:40-!- Kwandulin [~Miranda@p200300760F35BF9CBC4EB186147F7676.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160816 17:57:20-!- Greg-Bog_ [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160816 17:58:03-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160816 18:01:54-!- ancestral [~ancestral@67-4-224-82.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20160816 18:30:23-!- Shiki [~Shiki@HSI-KBW-149-172-220-18.hsi13.kabel-badenwuerttemberg.de] has quit [Remote host closed the connection]
20160816 18:31:08-!- Shiki [~Shiki@HSI-KBW-149-172-220-18.hsi13.kabel-badenwuerttemberg.de] has joined #wesnoth-dev
20160816 19:03:33-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160816 19:05:12-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160816 19:06:07-!- Kwandulin [~Miranda@p200300760F35BF9C1140AAB99EDFF524.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160816 19:07:37< aeth> Is there any reason why the campaign macro defines for the extra advancements causes OOS (at least in 1.12)?
20160816 19:08:46< celmin> Probably.
20160816 19:09:08< celmin> I guess the not-hosts don't honour the defines, or something.
20160816 19:09:28-!- Shiki_ [~Shiki@HSI-KBW-149-172-220-18.hsi13.kabel-badenwuerttemberg.de] has joined #wesnoth-dev
20160816 19:09:40< celmin> I suspect only the scenario gets transmitted across the network.
20160816 19:09:52-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 244 seconds]
20160816 19:11:38-!- gfgtdf [~chatzilla@x4e363805.dyn.telefonica.de] has quit [Read error: Connection reset by peer]
20160816 19:14:25-!- Shiki [~Shiki@HSI-KBW-149-172-220-18.hsi13.kabel-badenwuerttemberg.de] has quit [Quit: Verlassend]
20160816 19:21:59-!- iceiceice [~chris@cpe-24-161-53-115.hvc.res.rr.com] has joined #wesnoth-dev
20160816 19:21:59-!- iceiceice [~chris@cpe-24-161-53-115.hvc.res.rr.com] has quit [Changing host]
20160816 19:21:59-!- iceiceice [~chris@unaffiliated/iceiceice] has joined #wesnoth-dev
20160816 19:23:54-!- Spixi [~chatzilla@pD9549364.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160816 19:24:05-!- Spixi [~chatzilla@pD9549364.dip0.t-ipconnect.de] has quit [Client Quit]
20160816 19:26:19-!- irker022 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20160816 19:31:49-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160816 19:33:34 * zookeeper wonders where to copypaste the new x attack stats symbol from
20160816 19:35:49< zookeeper> from wikipedia, apparently
20160816 19:42:14< zookeeper> there was only two files in which the attack notation is used in text and i can change those from - to x whenever, but i guess i might as well wait since there's been a strange silence regarding that symbol change.
20160816 19:49:30-!- mjs-de [~mjs-de@x4e3190fe.dyn.telefonica.de] has quit [Remote host closed the connection]
20160816 19:50:21-!- Shiki_ [~Shiki@HSI-KBW-149-172-220-18.hsi13.kabel-badenwuerttemberg.de] has quit [Remote host closed the connection]
20160816 19:52:55-!- Ravana_ [~Ravana@unaffiliated/ravana/x-2327071] has quit [Disconnected by services]
20160816 19:53:19-!- Ravana__ [~Ravana@12-44-46-176.dyn.estpak.ee] has joined #wesnoth-dev
20160816 19:53:31-!- Ravana__ is now known as Ravana_
20160816 19:53:33-!- Ravana_ [~Ravana@12-44-46-176.dyn.estpak.ee] has quit [Changing host]
20160816 19:53:33-!- Ravana_ [~Ravana@unaffiliated/ravana/x-2327071] has joined #wesnoth-dev
20160816 19:56:04-!- hay207__ [~hay207@41.34.13.213] has joined #wesnoth-dev
20160816 20:03:44-!- Shiki [~Shiki@HSI-KBW-149-172-220-18.hsi13.kabel-badenwuerttemberg.de] has joined #wesnoth-dev
20160816 20:04:06-!- Shiki [~Shiki@HSI-KBW-149-172-220-18.hsi13.kabel-badenwuerttemberg.de] has quit [Client Quit]
20160816 20:48:43-!- JyrkiVesterinen [~JyrkiVest@87-100-250-87.bb.dnainternet.fi] has quit [Quit: .]
20160816 20:59:57< zookeeper> whoever implemented SCALE_INTO: the wiki doesn't say whether it prioritizes width or height.
20160816 21:00:03-!- horrowind [~Icedove@2a02:810a:83c0:404:21b:fcff:fee3:c3ff] has quit [Quit: horrowind]
20160816 21:06:36< iceiceice> did i implement that?
20160816 21:06:38< iceiceice> i don't remember
20160816 21:06:53< iceiceice> if i did i dont remember the answer anyways
20160816 21:07:44< zookeeper> right :p
20160816 21:08:01< celmin> I implemented it, and I don't remember...
20160816 21:08:08< zookeeper> it's kind of essential knowledge if one wants to use it
20160816 21:08:21< celmin> Let me check.
20160816 21:09:46< celmin> I'm not sure if I can say that it prioritizes width or height.
20160816 21:10:31-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160816 21:10:40< celmin> ratio = min(requested_width / original_width, requested_height / original_height)
20160816 21:10:53< celmin> The multiply original width and height by the ratio.
20160816 21:11:42< zookeeper> so in plain english, how does it work?
20160816 21:13:09< iceiceice> i guess, it stretches it as little as possible in order that it will fit in the requested box?
20160816 21:13:15< celmin> Thinking about it, I wonder if it should've been max rather than min…?
20160816 21:13:24< celmin> Yeah, I think what iceiceice said was the idea.
20160816 21:13:30< iceiceice> and if aspect change is required, there will be a gap along horizontal / vertical
20160816 21:13:42< celmin> It's supposed to change size as little as possible.
20160816 21:13:44< zookeeper> right, that'd make sense
20160816 21:13:54-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160816 21:13:59< zookeeper> so it's "scale up/down just enough that the source image fits into the specified box"
20160816 21:14:15< celmin> Yeah, that was the intent, at least.
20160816 21:14:45< zookeeper> sure, that's a good thing to have. i'd very much suggest making the description match it, though.
20160816 21:14:53< celmin> Huh?
20160816 21:15:12< zookeeper> what?
20160816 21:15:31< celmin> You mean the description on the wiki?
20160816 21:15:34< zookeeper> yes
20160816 21:15:52-!- Kwandulin [~Miranda@p200300760F35BF9C1140AAB99EDFF524.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160816 21:15:58< celmin> Maybe I should also do some tests with scaling into non-square destination boxes...
20160816 21:17:01< celmin> From the code it looks like it only makes the correct decision when scaling up, and would choose the smaller option when scaling down...
20160816 21:17:15< celmin> I'll edit the wiki with the intent, anyway.
20160816 21:17:24-!- minzbonbon is now known as miniplane
20160816 21:18:04-!- ancestral [~ancestral@67-4-224-82.mpls.qwest.net] has joined #wesnoth-dev
20160816 21:18:34< zookeeper> great
20160816 21:34:58-!- ChipmunkV [~vova@static-89-94-113-91.axione.abo.bbox.fr] has quit [Quit: ChipmunkV]
20160816 21:35:55-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has joined #wesnoth-dev
20160816 21:36:15< tad_> Is zookeeper up atm?
20160816 21:36:22< zookeeper> yes
20160816 21:36:40< tad_> Need guidance on TSG objectives for S02.
20160816 21:36:46-!- irker499 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20160816 21:36:46< irker499> wesnoth: ln-zookeeper wesnoth:master d6225b891ae3 / data/core/sounds/ (mud-fist-miss.ogg mud-fist.ogg): Removed residual metadata tags https://github.com/wesnoth/wesnoth/commit/d6225b891ae38497fc2984dc3702710353091bd2
20160816 21:36:46< irker499> wesnoth: ln-zookeeper wesnoth:master 25c31ed72026 / / (17 files in 8 dirs): Merge branch 'master' of https://github.com/wesnoth/wesnoth https://github.com/wesnoth/wesnoth/commit/25c31ed72026b38769f86f108e556b076cf24368
20160816 21:37:00< zookeeper> yeah?
20160816 21:37:15< tad_> They read 'end of ninth turn' but implementation is actually 10th turn. Should I fix [objectives] or turn events?
20160816 21:39:09< zookeeper> the event, i suppose. just changing the "turn 10" event to "side 2 turn 9" would do it, right?
20160816 21:39:44< zookeeper> i mean it's still effectively the same
20160816 21:39:56< zookeeper> you lose if you let turn 9 pass without taking the city
20160816 21:40:09< tad_> It says end of 9th, name=turn 9 end fixes that. At present it does not defeat until turn 10 end
20160816 21:40:44< zookeeper> eh? name=turn 10 will trigger at the start of turn 10
20160816 21:40:47< tad_> side 2 turn 9 start would do it, too. simply means not watching side 2/3 move before you lose.
20160816 21:42:24< zookeeper> yeah, and not having to watch that might be nice because it's just wasted time
20160816 21:42:31< zookeeper> aside from that i don't see any problem
20160816 21:42:59< zookeeper> well, the extra hint does say "You must move Deoran here by the end of the next turn" whereas that should probably be "this turn"
20160816 21:43:35< tad_> I was playing with it and it seemed wrong. Maybe I'm confusing what it says and what it does. ... Already have a change for that line. 
20160816 21:44:23< tad_> What I was thinking was once I got the implemenation and words matching, I'd do something to get the turns-left on [objectives] to match, too.
20160816 21:44:30-!- Appleman1234 [~Appleman1@KD036012018092.au-net.ne.jp] has quit [Ping timeout: 244 seconds]
20160816 21:45:00< tad_> OK. So it should end after side 1 ends their movement on turn 9. And events and words should be adjusted as needed to fit.
20160816 21:45:58< tad_> Thanks. I think I can see how to proceed, now.
20160816 21:46:25< zookeeper> i'm just saying that as far as i see, they did match already. you'd lose after turn 9, if you hadn't taken the city. it's just that making the player watch the AI turn is unnecessary
20160816 21:47:04< zookeeper> anyway, since you're here, there's one general thing i'd suggest:
20160816 21:47:12< tad_> OK?
20160816 21:48:56< zookeeper> at least in the previous PR's you had often made all sorts of WML-side workarounds for various glitches, like the li'sar object unit selection bug and such. as a general rule i'd very much discourage those kind of things, because these are still development releases and if there's engine bugs that create problems, it's really not necessary to try to patch them up in WML.
20160816 21:50:05< zookeeper> maybe it's not always obvious what's an engine bug and what's not, but if in doubt you can always ask here
20160816 21:52:06< tad_> Ah. OK. My goal was to get it working with the minimum fuss/interruption for the core team. But, as I'm getting more 'into' how things work, I'm getting more confidence on reaching into the Lua and fixing it 'right'. That said, noted: in future I will try to avoid workarounds for things I can't fix and seek help.
20160816 21:52:08< zookeeper> well, maybe "often" is an exaggeration, since i can't off the top of my head name anything but the selection bug and some of the original recall error suppression.
20160816 21:54:14< tad_> I think it was three times, but one of them wasn't flagged as a workaround. 'use advisor' beat me up getting the advisor to hide but the problem went away recently and I simplified the code back to how it should have worked all along.
20160816 21:55:49< zookeeper> i'm sure there are some problems that might be so hard to fix that we can't expect them to be fixed in a timely manner, but usually if you run into some kind of visual glitch or seemingly unnecessary error/warning or whatever, they'll probably get fixed relatively quickly as soon as the right people hear of it
20160816 21:57:27< tad_> My goal, coming in, was to not interrupt the work on the C++ side since their time is best spent getting the bigger issues fixed. What I've been learning is most of the issues I've had can be fixed in the Lua, where I have a lot more confidence.
20160816 21:58:25< tad_> There have been some visual glitches which I could neither fix nor work around. But most of those have recently be fixed (apperently inadvertently).
20160816 21:58:41< zookeeper> sure. it's just that letting c++ people know where they've botched something up isn't interruption, it's helping them do their job :p
20160816 21:59:43< celmin> Use advisor was the one that I accidentally fixed?
20160816 21:59:59< celmin> That was most likely due to a carry when constructing fake unit IDs?
20160816 22:00:20< zookeeper> i mean, those kind of issues are usually indicative of someone having accidentally introduced a bug or otherwise broken something
20160816 22:00:21< tad_> Yep.
20160816 22:01:18< tad_> Anyway .. the only visual glitch I still see has to do with multi-tile units. Gryphons and Gryphon Riders in HttT leave wing-tips behind while moving under move_unit_fake.
20160816 22:01:45< tad_> There is an older PR which might be related to that, but it appears stagnated.
20160816 22:01:53< zookeeper> for me i think that happens even with their standing animations
20160816 22:02:00< zookeeper> or at least i think i saw that recently
20160816 22:02:55< tad_> I may not notice it when they're standing because the artifacts seem to clear when the animation finishes. And that's a lot longer under move_unit_fake and a LONG move, which is what HttT has in 2 or 3 places/.
20160816 22:04:41< tad_> For Gryphons, to see it, you need to entice side 2 to attack the gryphons but avoid them completely yourself. At present (pending my PR) that means letting side 2 kill Mother Gryphon and then killing side 2 leader .. to trigger the fly-by departure sequence.
20160816 22:05:32< tad_> The glitch stands out best though when you get the eggs then later get the Gryphon Riders. That's a long sequence and the contrasting background is darker, so the wing-tips really show.
20160816 22:05:35< zookeeper> Aginor, ^ aware of the wingtips?
20160816 22:06:31< tad_> PR 318 appears as if it may be related to the issue.
20160816 22:06:52< Aginor> zookeeper: no
20160816 22:07:07< Aginor> hmm
20160816 22:07:22< zookeeper> IIRC the wingtips brush beyond the surrounding hexes so that's probably where it happens
20160816 22:07:23< tad_> Well, I'm thinking the zoom change leaving artifacts may be related to the animation leaving artifacts.
20160816 22:07:35< Aginor> they're probably not undrawing properly, or the tiles they overlap aren't flagged as dirty
20160816 22:08:11< zookeeper> tad_, anyway, do you have any particularly notable plans for TSG?
20160816 22:08:15< tad_> In my old windows programmer way of thinking, it's probably just need to invalidate the surrounding hex so they redraw .. what you said :)
20160816 22:08:30< Aginor> tad_: yeah
20160816 22:08:41< Aginor> tad_: all of that code needs quite a bit of TLC
20160816 22:10:15< tad_> I'm fine-toothing TSG and my big plans for it are to properly fix choose_level and next_level. There do not appear to be any TODOs. The code is messy and I have to keep stepping back from a complete rewrite. But it looks right now like my current PR has most of the errors covered. So it will probably be small things like getting the words to match the implementation as in S02
20160816 22:10:43-!- Appleman1234 [~Appleman1@KD036012016147.au-net.ne.jp] has joined #wesnoth-dev
20160816 22:12:26< zookeeper> tad_, okay, cool. the main thing the campaign really really needs is redesign of some of the last scenarios. return to kerlath and vengeance are more or less abysmal.
20160816 22:13:11< zookeeper> the former because it's so meh and duh story-wise, the latter because of shoddy gameplay
20160816 22:13:17< tad_> Yeah .. they beat me up every time I play test. Just do d#$%ed hard for an introductory campaign 
20160816 22:13:28< celmin> The last scenario is a bit bizarre. Not as bad as (what I've heard of) the last DiD, but…
20160816 22:14:21< celmin> Normally when faced with an enemy, you're expected to kill them. Granted, the objective states otherwise, but from what I recall, every enemy you kill makes it harder for you.
20160816 22:14:34< tad_> Oh yeah.
20160816 22:14:45< zookeeper> P.S. also, please don't try to rewrite the long march :p
20160816 22:14:46< celmin> Normally killing an enemy would make it easier.
20160816 22:14:48< tad_> I've learned for the elven ending it's best to run-and-hide.
20160816 22:15:11< celmin> I'm not saying it should be changed (unlike DiD, that should be changed), but maybe it could be clarified better.
20160816 22:15:34< celmin> Make it more clear to the player that avoiding kills is a good idea.
20160816 22:15:41< tad_> Long march? Oh .. yeah .. you said that already.
20160816 22:15:51< celmin> (Is that the name of the final scenario?)
20160816 22:16:15< zookeeper> i'm protective of that gem of beautiful WML
20160816 22:16:24< tad_> Long march in the run through the forest, under fog, with the outlaw doing pathfinding in WML.
20160816 22:16:49-!- boucman_work [~boucman@2a02-8428-034f-f800-9e32-0c7c-b391-6223.rev.sfr.net] has quit [Read error: Connection reset by peer]
20160816 22:16:52< celmin> Is this the scenario before heading underground?
20160816 22:16:53< tad_> Well, the only reason that gem of WML works is it's load is spread over every turn.
20160816 22:17:27< zookeeper> eh, the pathfinding isn't heavy
20160816 22:17:34< tad_> I tried a similar approach in HttT when I added snow to the Dwarved Doors and it took several minutes. Lua was way .. Way .. WAY faster.
20160816 22:18:01< celmin> Was that pathfinding?
20160816 22:18:12< tad_> It's a lot of WML and takes a lot of thinking to understand.
20160816 22:18:13< tad_> IT'
20160816 22:18:33< tad_> Not really pathfinding. But he's exposing the fog you'd expose next turn if you moved him first.
20160816 22:19:04< celmin> …okay…?
20160816 22:19:22< zookeeper> sure it's a lot to understand. which is why i prefer to not have to learn _another_ implementation of the same logic.
20160816 22:19:44< vultraz> gfgtdf: huge thanks for fixing that :D
20160816 22:19:52< tad_> Which you said a few weeks ago, so I'm tip-toeing around it :P
20160816 22:20:00< zookeeper> good, good :>
20160816 22:20:49< tad_> If that area is buggy and needs fixin' I'll let you know. But, as I recall, it works and doesn't present any bugs.
20160816 22:20:56-!- Duthlet [~Duthlet@dslb-188-105-138-092.188.105.pools.vodafone-ip.de] has quit [Quit: leaving]
20160816 22:21:42< tad_> My only issue with it is the mass of black you see if you remove fog at the end .. all those units in each other's way down south. I am SO glad they don't all come after you or it's be impossible to make it through!
20160816 22:22:55< vultraz> only one small listbox bug left
20160816 22:23:19< vultraz> it doesn't preserve position well
20160816 22:24:01< vultraz> celmin: I think we can enable the gui2 lobby
20160816 22:24:24< celmin> Is MP Create functional for every possible type of game?
20160816 22:24:31 * zookeeper thought there was already a fix to limit the total undead numbers so there won't be hundreds of them
20160816 22:24:50< celmin> GUI2 MP Create, I mean.
20160816 22:24:55< tad_> voltraz: on lists and such .. when you get the layouts working would it be possible to move the focus as you page down? I find it strange to page down, find what I need and find I should have used arrows instead of PgDn because I have to go through the list again.
20160816 22:24:56< vultraz> not yet, but I said lobby, not mp create
20160816 22:25:09< celmin> Well, they're currently linked to the same preference, after all.
20160816 22:25:26< vultraz> one can simply move the lobby out of the conditional :P
20160816 22:25:39< celmin> And since no-one has done load tests as far as I know, making the new lobby unconditional seems a bit hasty.
20160816 22:25:39< vultraz> tad_: hmmmm
20160816 22:25:58< vultraz> celmin: indeed, we need to test it
20160816 22:26:53< zookeeper> tad_, vultraz, i'm pretty sure i find it awkward if the selected item changes if i just casually scroll up and down a list
20160816 22:27:06< celmin> Maybe someone could run 100 or so instances of the host plugin from the MP test: https://github.com/wesnoth/wesnoth/blob/master/utils/travis/mp_test_executor.sh
20160816 22:27:35< celmin> About how many games are usually on the lobby anyway?
20160816 22:27:43< celmin> Maybe more than 100 is better.
20160816 22:27:57< celmin> Would also be better if they're not all identical, as would be the case when using the plugin.
20160816 22:28:19< vultraz> I've never seen more than 20 or so
20160816 22:28:25< tad_> And run-to-crash might be good.
20160816 22:28:32< celmin> Okay, then 100 seems like a good number to me.
20160816 22:28:34< vultraz> anyway
20160816 22:28:49< vultraz> I want it enabled for 1.13.6, so the community can provide feedback and testing
20160816 22:28:49< celmin> I suspect my computer wouldn't be able to handle it though.
20160816 22:28:50< tad_> If 100 runs, try 200 ...
20160816 22:28:56< celmin> Sure?
20160816 22:29:42< zookeeper> tad_, to what kind of value do you think the number of the undead spawns should be capped? as in, only spawn new ones if there's less than X currently?
20160816 22:29:45< celmin> If someone connects a hoard of bots, then what needs to be done is for people to join the server with the new lobby and check out performance.
20160816 22:30:10< celmin> I wonder if it's actually possible to play a game against a client running the host plugin...
20160816 22:30:20-!- gfgtdf [~chatzilla@x4e363805.dyn.telefonica.de] has joined #wesnoth-dev
20160816 22:30:49< tad_> zookeeper: that's the problem, I don't know. Personally, I'd suggest a dynamic cap based upon how many you have, how many are in position to be a nuisance or have been killed
20160816 22:31:08< vultraz> tad_: listboxes do handle pageup/pagedown
20160816 22:31:09-!- travis-ci [~travis-ci@ec2-54-81-250-138.compute-1.amazonaws.com] has joined #wesnoth-dev
20160816 22:31:10< travis-ci> wesnoth/wesnoth#10424 (master - 25c31ed : ln-zookeeper): The build has errored.
20160816 22:31:10< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/152810605
20160816 22:31:10-!- travis-ci [~travis-ci@ec2-54-81-250-138.compute-1.amazonaws.com] has left #wesnoth-dev []
20160816 22:31:25< tad_> It seems to me the sheer mass of them is an attempt to work around the AI not being smart enough to get 'em focused on kicking the player's butt.
20160816 22:32:26< zookeeper> tad_, well i was just thinking on preventing them from becoming a performance problem, so not something that would need to depend on anything
20160816 22:32:57< tad_> Oh, yeah, I do notice performance dropping on my memory-restricted VM.
20160816 22:33:20< celmin> Oh, BTW vultraz, there's a problem with the text field in the new lobby.
20160816 22:33:28< vultraz> oh?
20160816 22:33:30< vultraz> what is it
20160816 22:33:37< celmin> If you click anywhere else, it's deactivated, but really it should only be deactivated when you click the filter box.
20160816 22:33:55< celmin> ie, it should always have input focus, unless the filter box has focus.
20160816 22:34:07< tad_> zookeeper: I'll play with it some, when I get to that point. But I'm not sure there is a good solution. Too few and the player can ignore them. Too many and performance drops.
20160816 22:34:22< iceiceice> celmin, that's interesting
20160816 22:34:24< celmin> (Actually, this probably applies to the Lua console too, though maybe it's already the case, I can't remember.)
20160816 22:34:31< iceiceice> i think you would want different behavior if you are on a phone
20160816 22:34:41< iceiceice> b/c when a textbox is highlighted usually the keyboard shows up
20160816 22:34:50< tad_> zookeeper: maybe their AI needs to be more agressive.
20160816 22:34:56< celmin> iceiceice: Ah, I see your point.
20160816 22:35:04< iceiceice> but maybe on a computer that's the right behavior
20160816 22:35:09< iceiceice> i didn't think about it before
20160816 22:35:44< zookeeper> tad_, well the player can't ignore 40 units even if that's the ceiling, and i doubt it makes a difference whether it's 40 or 80 (since the player is very far already when 40 is hit, and thus new spawns wouldn't ever get anywhere near them)
20160816 22:35:46< celmin> Maybe what should happen is that, if the filter box is not focused, any keypress that is not a hotkey activates the chat field?
20160816 22:36:11< tad_> zookeeper: I've noticed some places where the AI losses focus on the task at hand. HttT Princess and AToTB Guarded Castle, for instance.
20160816 22:36:16< iceiceice> that sounds like a good solution
20160816 22:36:44< iceiceice> maybe the "dialog background" can forward key presses it recieves to the chatbox?
20160816 22:36:59< zookeeper> tad_, i can add a unit cap right away and tweak the value later, if that doesn't interfere with your stuff
20160816 22:37:06< celmin> Maybe?
20160816 22:37:33-!- ancestral [~ancestral@67-4-224-82.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20160816 22:37:49< tad_> zookeeper: maybe a two-fold approach .. cap the number .. and have them appear 12 or so hex away from the player instead of having to slog it through the forest 
20160816 22:37:56< vultraz> i have found the place where the listbox scrolling goes wonky
20160816 22:38:31< tad_> zookeeper: I try to track master and sync every morning so I can fix merge conflicts as they appear.
20160816 22:38:52< zookeeper> tad_, in principle maybe, but that'd require proper rebalancing
20160816 22:38:53< tad_> zookeeper: so, go ahead and change what you want and I'll adapt.
20160816 22:38:57< zookeeper> okay
20160816 22:40:45< tad_> zookeeper: istm that capping should be done with an area-search .. "If there are N undead within 12 hex of (center,bottom) then break" or something like that
20160816 22:41:53< tad_> zookeeper: The performance hit is probably the AI trying to pathfind a unit and finding it can't because all possible hex are occupied.
20160816 22:42:20< tad_> zookeeper: during move-to-target
20160816 22:43:40< tad_> zookeeper: If so, I dread the day the move-to-target sorts the mass outer-first and can find paths for ALL of them!
20160816 22:43:56< zookeeper> when looking at a replay from the forums, the AI has a respectable 81 units :P
20160816 22:46:04< tad_> One time, for grins, I tried leaving a couple tanks and mages down south to kill 'em off as they appeared. It was quickly apparent it was hopeless and running away was the best plan.
20160816 22:48:05-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Remote host closed the connection]
20160816 22:48:58< zookeeper> humm. does [have_unit] count=1-50 [/have_unit] pass if there's >50 units?
20160816 22:49:15< tad_> No.
20160816 22:49:25< tad_> count=51-99999 would
20160816 22:50:01< zookeeper> you sure? the wiki description is vague on that account
20160816 22:50:21< tad_> As I understand it, the exact count must be in the range.
20160816 22:50:33< zookeeper> well, that's good
20160816 22:51:07< tad_> so [not][have_unit]count=51-9999 might be what you want
20160816 22:52:08-!- ancestral [~ancestral@67-4-224-82.mpls.qwest.net] has joined #wesnoth-dev
20160816 22:57:41< zookeeper> oops, almost made a mistake...
20160816 22:58:01 * zookeeper forgot that undead and still-ambushing elves are on the same side
20160816 22:58:09< tad_> I wonder if limited the undead moght free up more during move_to-target and force revisting balance anyway
20160816 22:59:43< irker499> wesnoth: Charles Dang wesnoth:master 1209a2fb5b26 / src/gui/widgets/scrollbar_container.cpp: Slightly more sensible scrollbar positioning on content resize https://github.com/wesnoth/wesnoth/commit/1209a2fb5b261fa40635a3871bcd831487a6e4c3
20160816 23:00:10< vultraz> Still needs a proper "fix".
20160816 23:00:16< vultraz> the list should never jump around
20160816 23:00:42< aeth> I wonder if there's a library Wesnoth could switch to to make the networked play a bit better
20160816 23:01:06< vultraz> what, besides how we just switched away from SDL_net?
20160816 23:01:14< aeth> Just switched away?
20160816 23:01:18< aeth> Okay, I didn't catch that
20160816 23:01:21< aeth> I thought you were still using it
20160816 23:01:59< vultraz> no, we use boost asio now.
20160816 23:02:15< tad_> What is the issue you're hoping to solve with networked play?
20160816 23:02:19< vultraz> campaignd still uses sdl_net but loonycyborg already has a pr for that
20160816 23:04:10< aeth> tad_: the MP campaign code is a bit glitchy, at least in 1.12
20160816 23:04:26< aeth> e.g. if player ownership changes in scenario 2, then in scenario 3 there's OOS, or at least there was for me last night
20160816 23:05:14< aeth> also the whole thing about the defines not being broadcast or something, causing OOS when you want to e.g. have wolves (not wolf riders) have promotions through the campaign macro for it
20160816 23:05:30< zookeeper> that doesn't really have anything to do with what network library is used...
20160816 23:05:33< aeth> and there are a few other things. I haven't gotten a test game with someone else in 1.13 to see what is still there
20160816 23:05:42< aeth> zookeeper: well, the right library makes it easier to do the right thing
20160816 23:05:53< zookeeper> not in this case, i'm pretty sure
20160816 23:06:13< aeth> ok, sorry. Most of my experience thus far in engine code has been graphics
20160816 23:06:27< tad_> Sounds so. Changing the networking would make sense if there's congestion problems, or the server is lagging. But that sounds like a bug elsewhere
20160816 23:06:31< zookeeper> if the logic is broken it doesn't matter how shiny tubes you push it through :P
20160816 23:07:15< aeth> It's possible that there's a higher level library that would help out Wesnoth's networking
20160816 23:08:16< Aginor> I don't think it's the networking libraries that are the problem of the game having different state across all machines in the network
20160816 23:08:52< tad_> As I said, depends on the issue. If you want more than a few hundred people connected at once (say 250,000) you need a vastly different library than if you are fine with the limit.
20160816 23:09:21< aeth> depends on how well the Steam launch goes :-p
20160816 23:09:27< Aginor> even if we had a super-shiny library that could push entire objects from the memory of one host to another, it'd be a bad idea from a security point of view
20160816 23:09:57< Aginor> tad_: I don't think the networking libraries we use would be the main problems in that scenario either ;)
20160816 23:10:19< celmin> The networking library has nothing to do with Wesnoth's MP problems.
20160816 23:10:28< Aginor> I'd expect the cpu/memory to fall over long before we hit the limits of the networking libraries
20160816 23:10:57< Aginor> if the library allows us to so select() or something similar and allows us to shuffle data at line speed, it does the job
20160816 23:11:06< Aginor> the rest are engine problems
20160816 23:11:20< aeth> Aginor: I've already reached the limit of Wesnoth's CPU usage in my 400x200 scenario with more than 1500 units
20160816 23:11:48< celmin> I'm not sure if a centralized server is what Wesnoth needs,...
20160816 23:12:01< aeth> client side though
20160816 23:12:13< aeth> celmin: centralized servers generally seem to work better in gaming because otherwise it's hard to avoid cheating
20160816 23:12:15< irker499> wesnoth: ln-zookeeper wesnoth:master 402627d85169 / data/campaigns/The_South_Guard/scenarios/06b_The_Long_March.cfg: Roughly cap the number of undead spawns to avoid performance issues https://github.com/wesnoth/wesnoth/commit/402627d851695bd155044914d35becf20ea6815d
20160816 23:12:17< irker499> wesnoth: ln-zookeeper wesnoth:master 6382660fd349 / src/gui/widgets/scrollbar_container.cpp: Merge branch 'master' of https://github.com/wesnoth/wesnoth https://github.com/wesnoth/wesnoth/commit/6382660fd349dcb156a201d9bcd244bb0eb7eeba
20160816 23:12:22< celmin> Well, you could say that part of the problem is that all the work is being done client-side.
20160816 23:12:27< celmin> And on every client.
20160816 23:12:49< aeth> celmin: actually, that's a good way to do it if you can get away with it
20160816 23:12:51< aeth> keeps it in sync
20160816 23:12:57< aeth> if AAA FPSes could do that they probably would too
20160816 23:13:02< aeth> It's just there's too much going on with those games
20160816 23:13:08< aeth> and definitely too much in e.g. an MMO
20160816 23:13:09< celmin> I don't see how that's a good way.
20160816 23:13:29< celmin> Isn't it better for only one machine to do the work, and the rest only get the results?
20160816 23:13:55< aeth> that's probably the next best way
20160816 23:14:04< aeth> which is still impractical for some games
20160816 23:14:13< Aginor> celmin: if you can get away with it, you don't want your centralised bottleneck do more work than it absolutely have to
20160816 23:14:30< Aginor> or you're going to need more central servers, driving up cost
20160816 23:14:34< celmin> Well, even if Wesnoth's basic networking model is unchanged, there's also the problem that it doesn't transmit everything it really needs.
20160816 23:14:59< Aginor> ergo, it's the engine and not the library
20160816 23:15:07< celmin> Yeah, it's the engine, of course.
20160816 23:15:27< aeth> celmin: a problem with the engine as a whole is that there's this odd mix of Lua and WML
20160816 23:15:44< aeth> and sometimes you write your WML in Lua and sometimes you have to call WML from Lua because Lua can't do it etc.
20160816 23:15:56< aeth> it does show in some places in MP.
20160816 23:16:30< aeth> If I make my Lua literal WML strings loaded via macros, I don't need to require a download. If I do use Lua the way it's supposed to I require a download. Same Lua, same actions, different result.
20160816 23:16:41< tad_> I've not looked at what the central server maintains. My impression it it has a copy of the game state, but leaves the work to each client. That said, I see no reason a modest server could not hold hundreds, if not thousands, of on-going games.
20160816 23:17:15< aeth> tad_: would that mean that clients would no longer have to sit through possibly 10+ minutes of state syncing if they disconnect?
20160816 23:17:24< gfgtdf> aeth: 1.13 will give the player an optionto automatically download addons if needed so requiring download shouldn't be an issue liek in .12.
20160816 23:17:26< aeth> that would be the main advantage to switching imo
20160816 23:18:01< aeth> gfgtdf: not as much of an issue, but still an issue. My add-on is technically pre-release. Even if it's not an issue, I don't really want it on the add-on server because it's so buggy and incomplete.
20160816 23:18:07< tad_> aeth: depends on the design of the server. As I said, I've not looked at how it's being done and what's on the wire.
20160816 23:18:09< aeth> s/it's/downloading is/
20160816 23:18:12< celmin> Lua can do anything that WML can do though.
20160816 23:18:23< aeth> celmin: but a lot of the time that's through buggy WML interfaces
20160816 23:18:27< celmin> True.
20160816 23:18:28< aeth> s/buggy/hacky/
20160816 23:18:34< celmin> I'm trying to change that a bit.
20160816 23:18:47< tad_> celmin: and faster. syncing is the issue
20160816 23:19:00< aeth> e.g. wesnoth.add_modification(unit, "object", { T["effect"] { ...
20160816 23:19:05< aeth> very messy way to add an object
20160816 23:19:18< aeth> even when an object is the right thing, that kind of encourages doing the wrong thing with the other dozen ways to modify a unit
20160816 23:19:21< aeth> because it's so messy.
20160816 23:19:43< aeth> everything in filter.lua and most in fire.lua are pretty messy in my add-on
20160816 23:19:49< aeth> That I need to wesnoth.fire to do some things is messy
20160816 23:19:54< celmin> By the way, does anyone want to rewrite [modify_unit] to use the Lua proxy methods instead of storing and unstoring the units?
20160816 23:21:30< gfgtdf> aeth: i dont realyl se anything meyy about that way to add an object
20160816 23:22:14< aeth> Imo all of the ScenarioWML that's possible should be written in Lua using an API exposed to the Lua
20160816 23:22:29< aeth> gfgtdf: consider a hypothetical wesnoth.add_object(unit, ...)
20160816 23:22:37< aeth> gets rid of a lot of boilerplate and could be implemented better
20160816 23:22:51< aeth> Imo, WML should be calling Lua, not the other way around
20160816 23:23:01< celmin> Yes.
20160816 23:23:12< celmin> Like I said, I was trying to fix that a bit.
20160816 23:23:33< celmin> I don't think add_object is right though, because there's three types of modifications with no significant difference between them.
20160816 23:23:54< celmin> Anyway, when adding a modification, you absolutely need WML, because it's stored in the unit.
20160816 23:24:09< tad_> I think it should not make a difference. WML->Lua or Lua->WML should both be possible and which the author uses, both should get you to the same point.
20160816 23:24:14< celmin> Admittedly the add_modification function could theoretically be made to construct that WML rather than taking it raw.
20160816 23:24:56< aeth> tad_: the problem is WML in Lua is a messy sort of only-in-Wesnoth Lua programming
20160816 23:25:03< aeth> which makes it harder to pick up
20160816 23:25:15< aeth> which imposes a bit of a barrier to entry for people wanting to use Lua in Wesnoth
20160816 23:25:28< celmin> I personally think that it should not be necessary for Lua authors to call functions in wesnoth.wml_actions, wesnoth.wml_conditionals, or wesnoth.effects, unless they're implementing one in terms of another or something.
20160816 23:25:48< aeth> notice how I always do T["effect"] etc in my style instead of the equivalent T.effect that all the documentation uses. This is because (1) it looks like a WML tag sort of and (2) you have to do T["or"] etc if it's a Lua built in
20160816 23:25:49< tad_> In my experience, every embedded Lua as a similar problem.
20160816 23:25:52< aeth> T.or is an error
20160816 23:26:02< aeth> tad_: well, yeah, every game is coded horribly, with few exceptions if any
20160816 23:26:15< aeth> but usually they're rushed, Wesnoth doesn't have as much of an excuse
20160816 23:26:17< celmin> So aeth, what would be the arguments to that hypothetical wesnoth.add_object?
20160816 23:26:40< tad_> ACtually, this is the first game engine I've worked with. My Lua has been mainly playing with Apache.
20160816 23:26:51< celmin> Other than the unit, of course.
20160816 23:27:02< celmin> …Lua with Apache?
20160816 23:27:10< tad_> modlua.
20160816 23:27:18< aeth> celmin: easiest thing would probably be just to flatten the syntax a bit
20160816 23:27:36< celmin> So, for example, take a list of WML tables that would go inside [effect]?
20160816 23:27:40< aeth> so e.g. wesnoth.add_object(unit, {apply_to = attack, ...})
20160816 23:27:57< celmin> That seems like something I could arrange.
20160816 23:28:00< aeth> i.e. you skip "object", { T["effect"] and keep the rest
20160816 23:28:01< gfgtdf> well you can add ahelper function if you realyl want that, 
20160816 23:28:18< aeth> gfgtdf: well yeah, it's the easiest solution
20160816 23:28:31< gfgtdf> but since for exampel teh roposed function can onyl add one effect, not have an id to delete the obejc tetc, i'd doubt it be used often
20160816 23:28:53< celmin> gfgtdf: Nah, it'd probably be variadic or something.
20160816 23:30:06< celmin> Technically I guess you can remove objects based on their identity, huh… by filtering on the effect tags...
20160816 23:30:24< celmin> I think the only important key besides id in [object] is duration.
20160816 23:30:29-!- ancestral [~ancestral@67-4-224-82.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20160816 23:30:58< celmin> Not quite sure if id is actually that important, but duration has a definite meaning in [object] even when in [modifications].
20160816 23:33:00< celmin> In [trait], there are text keys which are relevant, I guess.
20160816 23:33:07< tad_> I'm outta here. Going to get back to TSG.
20160816 23:33:14< gfgtdf> celmin: hmm yes but often you want the durion eigher forever to to be ended at a cusotm time not covered by 'duration', so i usually rather rather add some cutom key like 'id' or similar ad then manually use u:remvoe_modification with wml filter from an event.
20160816 23:33:14-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has quit [Quit: Page closed]
20160816 23:33:21< celmin> [advancement] also has some relevant keys.
20160816 23:34:21< gfgtdf> celmin: the [advancement] key are mostly shown during that dialog and have no effect one it is added
20160816 23:34:26< gfgtdf> once*
20160816 23:34:53< celmin> gfgtdf: Note that, if you use remove_modification using a filter containing only duration, the WML filter is not run at all, and modifications with that duration are removed.
20160816 23:35:26< celmin> I'm pretty sure some of the [advancement] keys are relevant when determining which advancements you can take on level-up?
20160816 23:35:27< gfgtdf> celmin: hmm that sonds like a useful feature.
20160816 23:35:40< gfgtdf> celmin: only 'id' afaik
20160816 23:35:51< celmin> Maybe.
20160816 23:35:57< celmin> You're probably right, I guess.
20160816 23:36:10< gfgtdf> celmin: think like 'max_times' are takes form the 'new' advancemnt surng that dialog, not form the old one with that id that the unit already have
20160816 23:36:13< celmin> exclude_amla and strict_amla are only relevant for the advancement being taken.
20160816 23:36:54< gfgtdf> hmm gettign assertion failures from gui2 mp create
20160816 23:48:00< vultraz> :D
20160816 23:48:07< vultraz> D: *
20160816 23:48:11< vultraz> what are they?
20160816 23:48:40< gfgtdf> vultraz: when adding treeview nodes
20160816 23:49:14< vultraz> :|
20160816 23:49:17< vultraz> dammit
20160816 23:49:42< gfgtdf> vultraz: i still am not 100% sure how to reprodce
20160816 23:50:33< celmin> Okay, so on my GUI todo list is… treeview inspector, MP lobby tweaks (maybe), MP create low-res version, and maybe that small dialog you mentioned… what was it again?
20160816 23:51:12< vultraz> random map settings
20160816 23:51:20< vultraz> it's in default_map_generator::user_config
20160816 23:51:24< celmin> Oh, right.
20160816 23:51:33< vultraz> should bevery simple
20160816 23:51:35< celmin> Isn't there a GUI2 version of this in the editor already?
20160816 23:51:40< vultraz> i recommend you work on that first
20160816 23:51:45< vultraz> no, that's generator selection
20160816 23:52:07< celmin> I know there's a way to generate random maps in the editor.
20160816 23:52:12< vultraz> yes
20160816 23:52:17< celmin> Though, I suppose it might be using the same GUI1 dialog.
20160816 23:52:58-!- travis-ci [~travis-ci@ec2-54-157-210-248.compute-1.amazonaws.com] has joined #wesnoth-dev
20160816 23:52:59< travis-ci> wesnoth/wesnoth#10425 (master - 1209a2f : Charles Dang): The build has errored.
20160816 23:52:59< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/152829027
20160816 23:52:59-!- travis-ci [~travis-ci@ec2-54-157-210-248.compute-1.amazonaws.com] has left #wesnoth-dev []
20160816 23:53:04< vultraz> there's a gui2 dialog to select the generator and a gui1 for the the settings
20160816 23:54:02< vultraz> it's a very simple dialog
20160816 23:54:07< vultraz> should take you 30 mins tops
20160816 23:57:18-!- Greg-Bog_ [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160816 23:57:18< vultraz> and what do you mean treeview inspector?
20160816 23:58:01< celmin> I'd like to use a treeview instead of a listbox pair in the inspector.
20160816 23:59:34< vultraz> ah
20160816 23:59:36< vultraz> sure
--- Log closed Wed Aug 17 00:00:44 2016