--- Log opened Sat Feb 13 00:00:40 2016
20160213 00:14:42< vultraz> shadowm: here's an idea... what if we turn the addon manager ui into a full-window layout like the titlescreen or lobby
20160213 00:15:26< Aginor> shadowm: thanks for your input
20160213 00:15:44< vultraz> on consideration, I don't think the main window would have room for the description dialog, height-wise, without the list becoming short It would look weird to the left/right, though.
20160213 00:17:04< vultraz> however, by making it a borderless dialog we have a lot of space to put stuff in
20160213 00:17:33< vultraz> including space for future stuff like, say, Featured Addon
20160213 00:17:54< vultraz> (again, drawing a bit of inspiration from dota 2's addon layout here)
20160213 00:23:55< gfgtdf> vultraz: i agreww with full windows addon manager
20160213 00:25:40< gfgtdf> vultraz: also see this bugreport http://gna.org/bugs/?22570
20160213 00:25:53-!- Appleman1234 [~Appleman1@KD119104014069.au-net.ne.jp] has quit [Ping timeout: 276 seconds]
20160213 00:26:17< vultraz> hmmm ok
20160213 00:26:29< gfgtdf> vultraz: this basically suggests to add a panel that downs the details of the addons
20160213 00:26:37< gfgtdf> vultraz: simiar to hot its done in campaigsn meanu
20160213 00:26:44< vultraz> ya
20160213 00:26:49< gfgtdf> how*
20160213 00:27:32< gfgtdf> or the loadgame menu
20160213 00:37:22-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160213 00:42:20< shadowm> 21:14:44 <vultraz> shadowm: here's an idea... what if we turn the addon manager ui into a full-window layout like the titlescreen or lobby
20160213 00:42:57< shadowm> I proposed this for the Help browser and the Add-ons Manager itself before, the idea didn't seem to garner much interest.
20160213 00:43:20< vultraz> ah, the Help browser!
20160213 00:43:23< vultraz> I didn't even think of that
20160213 00:43:29< vultraz> I will certainly do that
20160213 00:43:44< shadowm> Back when it was at least a little bit more useful to ask other developers.
20160213 00:44:02< vultraz> I'll probably have a layout prototype for the Manager at window size by tomorrow
20160213 00:44:09< shadowm> You know, when people actually felt strongly about things and wouldn't have let me push IftU into mainline.
20160213 00:45:59< shadowm> Now I believe what mordante wanted based on the old GUI proposals from circa 1.0 was to have the descriptions be handled like what I made you do with Advanced Preferences.
20160213 00:46:48< shadowm> I'm not opposed to that but I feel it will not be very usable since descriptions tend to have an arbitrary length nowadays.
20160213 00:47:15< vultraz> that is how it's currently handled
20160213 00:47:23< vultraz> but it's a bad design and ill change it
20160213 00:47:27< shadowm> Ah yes here: https://wiki.wesnoth.org/GUI#Campaign_download_Dialog
20160213 00:47:38< shadowm> It's not bad design, vultraz ugh.
20160213 00:47:44< vultraz> Bad design for *this*
20160213 00:47:46< shadowm> It's just suboptimal for us.
20160213 00:48:11< shadowm> If descriptions were single paragraphs no more than 3 lines long then it'd be just fine.
20160213 00:48:25< vultraz> (I'd still like to get some kind of graphic for tab bars but we need an actual widget for that)
20160213 00:48:40< shadowm> No we really don't.
20160213 00:48:53< shadowm> It'd just be a matter of reskinning the horizontal listbox and toggle panels.
20160213 00:49:14< shadowm> For the tab bar.
20160213 00:49:32< shadowm> Not for the tab container, which can be a reskinned stacked_widget.
20160213 00:49:46< vultraz> right
20160213 00:50:12< shadowm> The thing is I don't feel it's important to worry about it.
20160213 00:50:17< vultraz> As for the mockups on those pages, those look pretty nice, actually. But you're right that it wouldn't do for what we have today
20160213 00:50:24< shadowm> You use Chrome, so you can see that in its Preferences page it doesn't even use tabs.
20160213 00:51:16< shadowm> The tab concept appears to be slowly drifting away, becoming relegated to an MDI implementation detail.
20160213 00:51:39< vultraz> anyway, I'm off to buy a lightswitch 
20160213 00:51:41< shadowm> People tend to use lists as tabs just as often these days.
20160213 00:54:56< shadowm> But of course for things like these two people's opinions are far from sufficient.
20160213 00:55:46< shadowm> Especially given that both of them are _coders_.
20160213 01:00:25-!- travis-ci [~travis-ci@ec2-54-196-239-78.compute-1.amazonaws.com] has joined #wesnoth-dev
20160213 01:00:26< travis-ci> wesnoth/wesnoth#8446 (guifixes - ea649ec : Andreas Löf): The build is still failing.
20160213 01:00:26< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/108913937
20160213 01:00:26-!- travis-ci [~travis-ci@ec2-54-196-239-78.compute-1.amazonaws.com] has left #wesnoth-dev []
20160213 01:11:38-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160213 01:18:16-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 250 seconds]
20160213 01:21:57-!- Appleman1234 [~Appleman1@KD119104011105.au-net.ne.jp] has joined #wesnoth-dev
20160213 02:00:36-!- travis-ci [~travis-ci@ec2-54-196-239-78.compute-1.amazonaws.com] has joined #wesnoth-dev
20160213 02:00:37< travis-ci> wesnoth/wesnoth#8448 (master - 0f47daf : ln-zookeeper): The build has errored.
20160213 02:00:37< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/108917231
20160213 02:00:37-!- travis-ci [~travis-ci@ec2-54-196-239-78.compute-1.amazonaws.com] has left #wesnoth-dev []
20160213 02:00:47-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Ping timeout: 276 seconds]
20160213 02:09:07-!- gfgtdf_ [~chatzilla@f050180205.adsl.alicedsl.de] has joined #wesnoth-dev
20160213 02:11:08-!- gfgtdf [~chatzilla@f054140019.adsl.alicedsl.de] has quit [Ping timeout: 250 seconds]
20160213 02:11:17-!- gfgtdf_ is now known as gfgtdf
20160213 02:26:38-!- gfgtdf [~chatzilla@f050180205.adsl.alicedsl.de] has quit [Quit: ChatZilla 0.9.92 [Firefox 44.0.2/20160210153822]]
20160213 02:28:01< shadowm> vultraz: http://i.imgur.com/qAGrAjs.png
20160213 02:28:18< shadowm> I know you don't read the logs, I'm just putting this here so I can easily repaste it later.
20160213 02:32:50-!- prkc [~prkc@51B6E6A0.dsl.pool.telekom.hu] has quit [Quit: Leaving]
20160213 02:59:56-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160213 03:01:58-!- neverEnough [~nEnough@host76-232-dynamic.21-79-r.retail.telecomitalia.it] has quit [Quit: zZzZ]
20160213 03:04:33-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 240 seconds]
20160213 03:07:55-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20160213 03:16:21-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20160213 03:27:32-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20160213 03:31:48< shadowm> 23:28:02 <shadowm> vultraz: http://i.imgur.com/qAGrAjs.png
20160213 03:36:53-!- Robertdebrus [6322841b@gateway/web/freenode/ip.99.34.132.27] has joined #wesnoth-dev
20160213 03:45:07-!- Robertdebrus [6322841b@gateway/web/freenode/ip.99.34.132.27] has quit [Quit: Page closed]
20160213 03:50:32-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has quit [Quit: And lo! The computer falls into a deep sleep, to awake again some other day!]
20160213 04:19:53-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Ping timeout: 276 seconds]
20160213 04:22:16-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: So long and thanks for all the fish.]
20160213 04:47:54-!- irker031 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20160213 04:48:20-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160213 04:52:46-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 250 seconds]
20160213 05:40:11-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20160213 05:42:45-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160213 05:47:02-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 256 seconds]
20160213 06:36:44-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160213 06:41:26-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 256 seconds]
20160213 06:45:49-!- Kwandulin [~Miranda@p200300760F0BC5E8583E7BF0A0476587.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160213 07:05:35-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has joined #wesnoth-dev
20160213 07:21:02-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20160213 07:22:10-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20160213 07:30:52-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160213 07:35:06-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 240 seconds]
20160213 08:06:13< ancestral> Dis good: https://sourceforge.net/blog/sourceforge-acquisition-and-future-plans/
20160213 08:06:27< ancestral> ‘Our first order of business was to terminate the “DevShare” program.’
20160213 08:06:43< Aginor> very much so
20160213 08:06:45< ancestral> ‘The DevShare program delivered installer bundles as part of the download for participating projects. We want to restore our reputation as a trusted home for open source software, and this was a clear first step towards that.’
20160213 08:07:50< ancestral> They’re smart to make changes now. GitHub is dealing with the dear-github fallout
20160213 08:15:33-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20160213 08:15:41-!- noy [~Noy@wesnoth/developer/noy] has quit [Ping timeout: 240 seconds]
20160213 08:16:39-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20160213 08:16:58< shadowm> Yeah I mean SF.net got handed over by a non-software-related company to another non-software-related company. Things can only possibly improve, right?
20160213 08:24:58-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160213 08:27:10-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has joined #wesnoth-dev
20160213 08:29:50-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 272 seconds]
20160213 08:43:47< vultraz> shadowm: would you be against removing Size from the main list of addons and keeping it in the description when selected?
20160213 08:43:54< shadowm> 00:31:49 <shadowm> 23:28:02 <shadowm> vultraz: http://i.imgur.com/qAGrAjs.png
20160213 08:44:12< shadowm> Very against.
20160213 08:44:29< vultraz> Why are you linking me that
20160213 08:44:37< shadowm> 05:43:54 <shadowm> 00:31:49 <shadowm> 23:28:02 <shadowm> vultraz: http://i.imgur.com/qAGrAjs.png
20160213 08:44:48< shadowm> Because I like doing that.
20160213 08:45:11< shadowm> After all, we all know that I'm an expert native English speaker with a degree on UI text writing, as well as a doctorate on UI design.
20160213 08:45:22< shadowm> vultraz I'm asking you for comments, come on.
20160213 08:45:54< vultraz> Hmmm
20160213 08:46:05-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20160213 08:46:42< shadowm> There is a purpose to all of the columns that there are on the Add-ons Manager, btw. Don't remove any of them.
20160213 08:47:01-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20160213 08:47:09< shadowm> I didn't spend literally weeks deciding on what to keep/remove only for someone else to do that again.
20160213 08:47:28< vultraz> Guess I need to add Type because that doesn't exist here yet
20160213 08:47:28< shadowm> (Yes, I actually spent several weeks deciding that, when I did the last redesign of the dialog.)
20160213 08:47:47< shadowm> Oh yeah, I already told you that mordante did his own thing.
20160213 08:48:04< shadowm> You are supposed to reconcile it with my work on the production version.
20160213 08:48:19< vultraz> I am
20160213 08:48:58< shadowm> Anyway, I'm still waiting for feedback on the patch implied by the screenshot.
20160213 08:49:13< shadowm> Do say if I need to elaborate further on my design decisions.
20160213 08:49:26< vultraz> The text seems a bit...
20160213 08:49:28< vultraz> I'm not sure
20160213 08:49:36< shadowm> iceiceice used tooltips on the column header labels.
20160213 08:50:01< shadowm> But: 1) tooltips don't work with [label]; 2) that's the most non-obvious approach you could possibly use, no-one will ever see the tooltips even if they worked.
20160213 08:51:02< shadowm> I don't think we should be afraid of in-dialog explanations.
20160213 08:51:10< vultraz> It seems a little, how should I say... impersonal? 
20160213 08:51:12< vultraz> Technical?
20160213 08:51:27< vultraz> But I can't think of any better text
20160213 08:51:49< vultraz> I think get rid of "pop-ups"
20160213 08:52:08< shadowm> I'd rather not, 'notification' is too vague ("sound notification").
20160213 08:52:37< shadowm> Also, that's not the solution if you chose the right words to describe your issue with it.
20160213 08:53:02< shadowm> What you want is probably a more profound change in tone and/or structure.
20160213 08:53:58< shadowm> It was a bit more direct originally.
20160213 08:55:10< shadowm> "You may choose to be alerted of events in multiplayer games by playing sounds and (if the game window is inactive or minimized) displaying notification pop-ups on your desktop. You may also choose whether to receive these alerts while in the lobby."
20160213 08:55:23< shadowm> (Not the original, I just wrote that.)
20160213 08:56:43< shadowm> "You may choose to be alerted of events in multiplayer games or in the multiplayer lobby by playing sounds and (if the game window is inactive or minimized) displaying notification pop-ups on your desktop."
20160213 08:58:31< vultraz> (off topic, but we should probably expand the desktop popups to SP as well)
20160213 08:58:38< vultraz> That seems a bit better
20160213 08:58:47< vultraz> I still dislike "pop-ups", though
20160213 08:58:50< shadowm> Uh, maybe not.
20160213 08:58:53< vultraz> pop ups have bad connotation 
20160213 08:59:28< vultraz> "desktop notifications" is a pretty standard term, Ibelieve
20160213 08:59:55< shadowm> I really don't feel we need notifications in SP but I'm too lazy to elaborate right now.
20160213 09:00:18< shadowm> Something something nature of afk in solo games vs. playing against humans.
20160213 09:00:27< shadowm> Something something pause/don't end turn.
20160213 09:00:37< shadowm> Something something reload autosave or save.
20160213 09:00:49< shadowm> Something something long AI turns = poor scenario design.
20160213 09:01:03< shadowm> (Hi NR.)
20160213 09:02:11< vultraz> 	"You may choose a sound alert or (if the game window is inactive or minimized) desktop notification of events in multiplayer games. You may also choose whether to receive these alerts while in the lobby."
20160213 09:02:43< shadowm> No, you can't choose "a sound alert".
20160213 09:02:58< shadowm> According to iceiceice and the removed tooltip text, you can, but that's not the dialog's goal.
20160213 09:03:18< shadowm> You can choose whether to play a sound alert, but you can't choose (in the dialog) what sound alert to use.
20160213 09:04:29< shadowm> "Choose how you wish to be notified of events in multiplayer games or in the multiplayer lobby; either by playing sound alerts or (if the game window is inactive or minimized) displaying desktop notifications."
20160213 09:04:42< vultraz> I like
20160213 09:05:53< shadowm> I'm just a bit concerned it might not be obvious that you can't choose which notification type to use in the lobby; you can only choose whether the same notification mechanisms to the left will be used in the lobby or not.
20160213 09:06:58< shadowm> Alsoo, thoughts about removing the colons from the column headers?
20160213 09:07:17< vultraz> Oh, that's what changed 
20160213 09:07:18< vultraz> I like
20160213 09:08:00< shadowm> Well, I also changed the checkboxes' alignment on the columns. Previously they were left-aligned, now they are centered.
20160213 09:08:14< vultraz> Looks good
20160213 09:08:42< shadowm> I also added some extra spacing above the dialog control buttons (Defaults, Close) because otherwise the Defaults button seemed to belong to the Turn changed row somehow.
20160213 09:08:42< vultraz> Maybe consider disabling each row's In Lobby checkbox if both other boxes are unselected
20160213 09:09:15< shadowm> Not really, that'd be a bit inconvenient if you are in the middle of tweaking settings.
20160213 09:10:02< shadowm> It means you have to go through the columns in order even if you already have the mouse over the lobby column.
20160213 09:11:27< shadowm> e.g. if you have A[] B[] C[] cleared, your mouse is over cell C, and you want to check B and C, you shouldn't have to check B before C.
20160213 09:11:43< vultraz> ah
20160213 09:11:57< shadowm> I'm not sure if I'm explaining myself properly here. The idea is to not increase mouse travel distances when not strictly required.
20160213 09:12:19< vultraz> I understand
20160213 09:12:56-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20160213 09:13:24< shadowm> The extra spacing bit above the dialog controls is something I'm tempted to make a convention across all of our UI.
20160213 09:14:46< shadowm> It's actually very common that you get text of some sort right above the controls, making the whole thing look unbalanced and weird.
20160213 09:15:26< vultraz> that's reasonable
20160213 09:15:49< shadowm> This has to be the most common case: https://dl.dropboxusercontent.com/u/21371130/Screenshots/zzzzz5.png
20160213 09:19:06-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160213 09:23:34-!- irker054 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20160213 09:23:34< irker054> wesnoth: Ignacio R. Morelle wesnoth:master bd8066313ce0 / data/gui/default/window/mp_alerts_options.cfg: gui2/tmp_alerts_options: Substantial dialog design cleanup https://github.com/wesnoth/wesnoth/commit/bd8066313ce063d1c9d929b571745cdee47e37d1
20160213 09:24:04-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 256 seconds]
20160213 09:25:33< shadowm> vultraz: Also, do you still want this? https://dl.dropboxusercontent.com/u/21371130/Screenshots/zzzzz6.png
20160213 09:25:53< vultraz> Hmmm
20160213 09:25:59< vultraz> well it does mean an extra click...
20160213 09:26:12< vultraz> but otoh it looks better
20160213 09:26:20< shadowm> Well, if our buttons weren't stupidly size-restricted!
20160213 09:26:54< shadowm> (Or alternatively, if I could emulate LordBob's button design without spending a whole month on it.)
20160213 09:27:21< vultraz> we need support for SVG rendering for buttons
20160213 09:27:38< shadowm> OTOH I could use two standalone toggle buttons and use wesnoth.set_dialog_callback to trigger the choice.
20160213 09:28:01< shadowm> But then I'd still need to skin them so it's obvious that clicking on them dismisses the dialog.
20160213 09:28:32< shadowm> I'll tell you this much, our current buttons would be just as hard to do using vector graphics.
20160213 09:28:45< shadowm> They look very hand-painted to me.
20160213 09:28:55< vultraz> You'd have to speak to LB
20160213 09:29:00< vultraz> but it's something to consider
20160213 09:29:01< shadowm> Mostly because of that noise effect they have.
20160213 09:29:11< vultraz> since this WILL become a problem more and more
20160213 09:29:21< shadowm> I won't talk to him about SVGs, no.
20160213 09:29:24< vultraz> And unless we want to turn into Windows 10...
20160213 09:29:33< vultraz> why not?
20160213 09:29:35< shadowm> Why would I do that when we don't have support for SVG files?
20160213 09:29:43< vultraz> we need to add them
20160213 09:29:45< vultraz> s/them/it
20160213 09:30:01< shadowm> Besides, if he's a professional artists something tells me he probably uses Adobe Illustrator instead.
20160213 09:30:15< shadowm> And I suspect Inkscape doesn't (fully) support their format. :p
20160213 09:31:02< vultraz> And I counter that Illustrator probably has an export option for that
20160213 09:31:26< shadowm> And SVG probably doesn't support whatever proprietary filters AI uses.
20160213 09:31:43< shadowm> *SVG or your SVG renderer of choice
20160213 09:32:11< shadowm> That's also why .psds are really useless in practice, since not everyone can afford Adobe Photoshop or is running on a platform where it'll run.
20160213 09:32:31< shadowm> There's a limit to what people can achieve through reverse-engineering in finite time.
20160213 09:34:03< shadowm> Oh and anyway there's a problem here with GUI2 not allowing user-defined widget definitions right now.
20160213 09:34:20< vultraz> True
20160213 09:34:31< shadowm> Which means that I'd have to make buttons with Konrad and Li'sar's sprites on them and make them part of the core GUI2 configuration.
20160213 09:34:47< vultraz> Something tells me mordante never meant GUI2 to be used by addons/campaigns like this
20160213 09:35:02< shadowm> I told you he didn't mean for the Lua bindings to exist.
20160213 09:35:40< shadowm> Specifically not before the GUI2 API was stable.
20160213 09:36:07< shadowm> He did mean for GUI2 to be themeable by add-ons, however, which would have involved user-defined widget definitions of course.
20160213 09:36:34-!- horrowind [~Icedove@2a02:810a:8380:834:21b:fcff:fee3:c3ff] has joined #wesnoth-dev
20160213 09:37:18< shadowm> That aspect of the framework just didn't get much priority since, again, not a stable API, and the core functionality was and still is unfinished.
20160213 09:38:22< vultraz> I think we're going to have to do a bug refactor of gui2 soon
20160213 09:38:25< shadowm> Doing otherwise would be the equivalent of working our ass off on support for fancy items before even deciding on your campaign's gameplay design.
20160213 09:38:29< vultraz> big
20160213 09:40:33< shadowm> So the bottomline is, you've learned a lot since the last time we talked about the Tutorial character selection dialog.
20160213 09:40:48< shadowm> Back then I had about the exact same thing and you wanted it to be committed immediately.
20160213 09:41:05< shadowm> Now you finally saw the same flaw that's been preventing me from working on it.
20160213 09:41:34< vultraz> shadowm: https://www.dropbox.com/s/qpuevwln0utjwhm/GUI2%20addons%20list%20FULL%20WINDOW%20staging.PNG?dl=0
20160213 09:42:22< shadowm> ... Sigh.
20160213 09:42:49< shadowm> Rule #1 of Wesnoth UI design club: NEVER start working at full HD resolutions.
20160213 09:43:12< shadowm> `-r 800x480 -w` in the command line.
20160213 09:43:30< shadowm> What looks good at 1080p will probably look like ass even on 1280x800.
20160213 09:43:37< vultraz> (as for the Tutorial dialog, I'd say leave it out for now, since I don't think that extra click is worth it)
20160213 09:43:48< shadowm> Most common symptoms: unexpected scrollbars.
20160213 09:44:21< shadowm> Scrollbars on text labels. Scrollbars on whole widget regions. Scrollbars around buttons.
20160213 09:45:30< shadowm> Instead of "No description available", take AtS' description, paste it thrice into the box, and use that as your test case.
20160213 09:45:55< shadowm> Or just go to the closest Lorem Ipsum generator you can find with Google and paste three paragraphs of that.
20160213 09:46:16< shadowm> AND USE LOW RESOLUTIONS.
20160213 09:46:36< shadowm> I can't emphasize that enough because IRC doesn't have a universal standard for text size specifications.
20160213 09:47:05< shadowm> Also, never forget to consider translations.
20160213 09:47:30< vultraz> wesnoth crashes if you use "-w", just saying
20160213 09:47:46< shadowm> It might sound absurd, but assuming that a short word ("yes") can be translated to something very long ("fjdakslfjaklgjaklgjaklgjalkjdflkadsjfasdklfjasdklj") helps.
20160213 09:48:52< vultraz> alright, work is needed on the layout
20160213 09:48:56< shadowm> (Don't take that example literally, because core words like "yes" and "no" are obviously going to be universally short for psycholinguistic reasons.)
20160213 09:49:03< vultraz> or else the entire window gets scrollbars 
20160213 09:49:14< shadowm> (The same obviously doesn't apply to nouns or uncommon words.)
20160213 09:49:44< vultraz> Right
20160213 09:50:15< vultraz> But I want to hear what you have to say on the layout in general
20160213 09:50:38< vultraz> filter options to the right? to the bottom? description to the right? to the left?
20160213 09:50:39< shadowm> "SP/MP campaigns" in Spanish is "Campaña normal/multijugador", for convenient reference, because we don't get univerally-accepted non-ambiguous abbreviations for single/multiplayer.
20160213 09:51:29< shadowm> Filter options on a separate dialog box OR filter options on an area that can be toggled on/off (off by default).
20160213 09:51:37< vultraz> (also as you can see, I have discovered the magic of [draw])
20160213 09:51:49< shadowm> Otherwise this looks like a dialog straight from the 80s.
20160213 09:52:50< shadowm> The installation status filter is an exception though. I'd ideally see that at the top next to the filter box.
20160213 09:53:03< vultraz> (also remind me to ditch that text box for Wesbsite. We have links now)
20160213 09:53:05< shadowm> It's the *most* commonly used filter option.
20160213 09:53:30< shadowm> No, keep that textbox there.
20160213 09:53:43< shadowm> I'll tell you what to do when new information becomes available.
20160213 09:54:10< vultraz> Alright
20160213 09:54:39< shadowm> (The most important bit being that desktop::open() is only supported on three platforms and it's not reliable on two of them.)
20160213 09:56:00< shadowm> Also, while we are discussing priorities, still no fix for the listbox header scrolling nonsense?
20160213 09:56:48< shadowm> Because I don't know exactly how more important that would be compared to this task you came up with, but I suspect it's a lot.
20160213 09:56:55< shadowm> *much more
20160213 09:57:13< shadowm> Especially since it's a prerequisite for the dialog to work as intended.
20160213 09:57:51< vultraz> It's a lot more complicated than I though
20160213 09:58:04< vultraz> It would be much easier without the repeating buttons, possibly
20160213 09:58:43< shadowm> Programming has never been about taking the easiest path.
20160213 09:59:22< shadowm> Otherwise, we wouldn't even have CPUs.
20160213 09:59:43< vultraz> I'd appreciate some help on the issue
20160213 09:59:50< vultraz> I can't deal with All The Gui2 Things
20160213 10:00:07< shadowm> You seemed more sure you could do it at the time.
20160213 10:00:44< shadowm> https://dl.dropboxusercontent.com/u/21371130/Screenshots/zzzzz7.png
20160213 10:01:04< shadowm> I probably don't need to scroll down to show what the rest looks like.
20160213 10:01:33< vultraz> Yes I know that's something else >_>
20160213 10:01:43< shadowm> Also, exactly how does Wesnoth crash for you when using -w? It's the shorthand form of --windowed (guess what it does).
20160213 10:02:00< shadowm> Oh.
20160213 10:02:03< shadowm> Huh.
20160213 10:02:26< shadowm>     wesnoth: /usr/include/boost/smart_ptr/scoped_ptr.hpp:99: T* boost::scoped_ptr<T>::operator->() const [with T = sdl::twindow]: Assertion `px != 0' failed.
20160213 10:02:33< shadowm> This way, I guess.
20160213 10:02:34< vultraz> yes, that
20160213 10:02:53< shadowm> Someone did a bad thing, I can tell that much.
20160213 10:03:04< shadowm> I'll get back to you when I get an -O0 build.
20160213 10:07:54< vultraz> I think I'll move most of the filter options to the right
20160213 10:08:05< vultraz> else there's this big square of unused space
20160213 10:10:27< vultraz> suddenly I'm getting gui2 mp lobby flashbacks
20160213 10:11:49< shadowm> Aginor: So since this involves the SDL 2 work, I'm going to suggest you try running Wesnoth with -w (shorthand of --windowed) in the command-line, with a debugger attached.
20160213 10:12:35< shadowm> Aginor: Some early initialization code wants to set a flag in CVideo to disable fullscreen mode, and CVideo now does that by calling a method on an object that does not exist at the time.
20160213 10:12:50< vultraz> shadowm: that's my code :|
20160213 10:12:53< vultraz> I think
20160213 10:13:12-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160213 10:13:20< vultraz> at least I was working on our sdl2 window handling
20160213 10:13:35< vultraz> so it's possible I broke something...
20160213 10:14:10< shadowm> vultraz: Is this your code? https://gist.githubusercontent.com/shikadilord/25bf08fd44de2cb0bb34/raw/e3c9b1949722ff0046ebd472bc1db936b78b6b97/gistfile1.txt
20160213 10:14:27< shadowm> (Stack frame #5 in particular.)
20160213 10:14:50< vultraz> Yes
20160213 10:15:00< shadowm> Aginor: vultraz tells me it's his code, not yours.
20160213 10:15:39< shadowm> vultraz: Figure out how to defer the operation (preferably within CVideo) until the object in question is constructed.
20160213 10:16:50< shadowm> It shouldn't be that hard, but I'm not going to give you any clues because, again, I don't like dealing in actual code because that makes any given problem my own task to solve.
20160213 10:17:31< shadowm> Why within CVideo? Because it should not be the initialization's code responsibility to know that CVideo relies on other objects that haven't been constructed at that time.
20160213 10:17:35-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 240 seconds]
20160213 10:17:43< shadowm> *initialization code in wesnoth.cpp's
20160213 10:18:42< shadowm> Yes, that's a clue, but it's abstract enough that it doesn't count as actual code.
20160213 10:20:18< vultraz> Alright, I have a fix
20160213 10:23:04< shadowm> Oh great, tlistbox itself is a tscrollbar_container.
20160213 10:23:15-!- travis-ci [~travis-ci@ec2-54-167-163-133.compute-1.amazonaws.com] has joined #wesnoth-dev
20160213 10:23:16< travis-ci> wesnoth/wesnoth#8450 (master - bd80663 : Ignacio R. Morelle): The build passed.
20160213 10:23:16< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/108976690
20160213 10:23:16-!- travis-ci [~travis-ci@ec2-54-167-163-133.compute-1.amazonaws.com] has left #wesnoth-dev []
20160213 10:23:22< vultraz> shadowm: -f is for fullscreen, right?
20160213 10:23:41< shadowm> Yes.
20160213 10:24:07< shadowm> First thought: this is impossible without refactoring tlistbox, regardless of whether the scrollbar buttons are kept or not.
20160213 10:25:01< shadowm> If you made any additional discoveries about tlistbox's relationship with the scrollbars, do share them before I start running in circles like a headless chicken.
20160213 10:27:43< vultraz> shadowm: the listbox uses content_grid() to find its child widgets. I believe that is the grid with id "content_grid_". Now, I have a suspicion that grid is the same grid scrollbars scroll. But since we want only the vertical scrollbar to not scroll the header, we have a problem.
20160213 10:28:41< vultraz> Or I could be totally wrong
20160213 10:28:58< vultraz> Either way, I got to a point where the listbox contents (not the header) wouldn't draw at all.
20160213 10:29:01< shadowm> Actually, scrap that first thought of mine, it's invalidated by one crucial detail.
20160213 10:29:22< shadowm> The scrollbars are part of the listbox grid, and they clearly don't scroll themselves.
20160213 10:29:41< shadowm> (That'd be pretty bizarre and break some laws of physics and UI design.)
20160213 10:30:48< shadowm> Ohhhh, tscrollbar_container is the one that refers to _content_grid.
20160213 10:31:09< irker054> wesnoth: Charles Dang wesnoth:master 8d6ca11241d0 / src/video.cpp: Check that window is initialized when setting fullscreen https://github.com/wesnoth/wesnoth/commit/8d6ca11241d00deacb6e474bcc90b89e959429b4
20160213 10:31:23< vultraz> shadowm, Aginor ^
20160213 10:32:26< vultraz> the function will fall back on preference flag setting only if window isn't initialized yet
20160213 10:33:02< vultraz> Fixes crashes with both -f and -w
20160213 10:33:09< shadowm> Good, good.
20160213 10:33:50< shadowm> Okay, there are several tscrollbar_container subclasses, hm.
20160213 10:33:56< vultraz> I'm not sure why you were telling me to look at wesnoth.cpp when I knew the caller was in game_launcher.cpp
20160213 10:34:20< shadowm> I didn't tell you to look at wesnoth.cpp.
20160213 10:34:38< shadowm> I did misidentify the caller, though.
20160213 10:35:01< shadowm> The point was that you shouldn't have to change (or even look at) the caller, anyway.
20160213 10:35:52< vultraz> That is true
20160213 10:37:14< shadowm> I really want to refactor our initialization code after SDL 1.2 support is finally cast into oblivion. It's kind of a fragmented mess.
20160213 10:37:51< shadowm> There's also an important feature that I'm planning to add that'd be much simpler to implement if SDL 1.2 is taken out of the equation.
20160213 10:38:02< vultraz> The window initialization part was horrible before I got to it :P
20160213 10:38:06< vultraz> What feature is this?
20160213 10:39:25< shadowm> Display message boxes if Wesnoth fails to initialize early.
20160213 10:40:17< shadowm> The Windows-specific log file setup code does that, but I want to implement it as a more generic mechanism across all platforms and SDL 2 seems to have just the thing I need.
20160213 10:42:26< shadowm> Okay, let me see if I can understand the listbox build logic.
20160213 10:47:58< shadowm> tscrollbar_container sets up the scrollbars, then creates a spacer (!), replaces _content_grid with said spacer.
20160213 10:48:35< shadowm> _content_grid is reparented to the tscrollbar_container (but where is it in the grid layout now that it's been extracted????).
20160213 10:49:25< shadowm> Because reparenting seems to have no effect on the layout.
20160213 10:50:21< shadowm> Then, before returning to the subclass finalize(), we call the subclass ourselves!
20160213 10:50:50< shadowm> Except tlistbox doesn't override the method called, so that does nothing, okay.
20160213 10:51:14< shadowm> So I'm still unsure where the crap the extracted _content_grid goes to.
20160213 10:51:43< vultraz> Now you see why this is more complicated than one would think
20160213 10:52:24< shadowm> Oh wait, it doesn't matter that it went to limbo, we still have the pointer.
20160213 10:52:32< shadowm> I forgot this is a class member.
20160213 10:53:18< shadowm> So it's in limbo, but tlistbox can still access it via the inherited getter.
20160213 10:54:03< shadowm> I guess tscrollbar_container only provides shared functionality but leaves it up to its subclasses to actually rescue the _content_grid from layout limbo?
20160213 10:54:47< shadowm> (Note to self: _content_grid contains _list_grid.)
20160213 10:55:12< shadowm> Waaait.
20160213 10:55:39< shadowm> That means that at this particular point the whole of the listbox is a spacer and the actual contents, scrollbars and everything, are in limbo.
20160213 10:55:56< vultraz> What's the config key with the campaign's type?
20160213 10:56:26< shadowm> src/addon/info.hpp for all your campaignd [campaign]-related inquiries.
20160213 10:56:42< vultraz> thanks
20160213 10:57:03< shadowm> Okay, so, next we swap the _header_grid (which is in _content_grid which is in limbo) with whatever header->build() creates.
20160213 10:57:20< shadowm> I mean _header_grid/_footer_grid were placeholders to begin with anyway, so that's expected.
20160213 10:57:39< shadowm> Blah blah blah header footer set-up stuff.
20160213 10:57:53< shadowm> Seems like footers are actually a bit crippled right now, they don't get sorting functionality.
20160213 10:58:01< shadowm> Then again, I guess that's a post-mordante addition so it figures.
20160213 10:58:25-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has joined #wesnoth-dev
20160213 10:58:53< shadowm> Now we swap the _list_grid that's part of _content_grid (which is still in limbo??) with the stuff returned by the generator, which I assume builds the actual list-of-items grid.
20160213 10:59:19< vultraz> Improved layout https://www.dropbox.com/s/jsjk0936h2gfdfv/GUI2%20addons%20list%20FULL%20WINDOW%20staging%202.PNG?dl=0
20160213 10:59:28< shadowm> And my limbo theory is incorrect, because we can't possibly return with the entire grid existing in a parallel universe.
20160213 11:00:24< shadowm> I don't get it, tscrollbar_container quite plainly replaces it with a _spacer_.
20160213 11:00:26< vultraz> (I did change the description scroll label to the default defintion. No need for that special bg anymore)
20160213 11:00:47-!- travis-ci [~travis-ci@ec2-54-159-254-240.compute-1.amazonaws.com] has joined #wesnoth-dev
20160213 11:00:48< travis-ci> wesnoth/wesnoth#8451 (master - 8d6ca11 : Charles Dang): The build has errored.
20160213 11:00:48< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/108982787
20160213 11:00:48-!- travis-ci [~travis-ci@ec2-54-159-254-240.compute-1.amazonaws.com] has left #wesnoth-dev []
20160213 11:01:02< shadowm> Maybe I missed finalize_subclass() implementors...
20160213 11:01:21< shadowm> Nope, only implementor is tscroll_label.
20160213 11:01:40< shadowm> 08:00:26 <vultraz> (I did change the description scroll label to the default defintion. No need for that special bg anymore)
20160213 11:01:51< shadowm> Make it special again.
20160213 11:02:01< shadowm> It's supposed to be a stand-in for a multiline textbox.
20160213 11:02:19< shadowm> I refuse to look at the screenshot until it's low-res.
20160213 11:02:29< vultraz> :|
20160213 11:04:32< vultraz> Well it seems like the mp lobby can do it so I guess i can too 
20160213 11:06:18< shadowm> I really don't get how this works at all.
20160213 11:07:20-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160213 11:10:39< shadowm> I don't even see what's the tangible effect of reparenting widgets.
20160213 11:12:04< shadowm> Other than that being an important variable in the visibility query.
20160213 11:12:05-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 260 seconds]
20160213 11:12:59< vultraz> gahhh how the hell do I make this work on a tiny @()*@#$() resolution
20160213 11:13:13< vultraz> there's no way
20160213 11:13:14< shadowm> Hm, something something tgrid::place().
20160213 11:13:38< vultraz> I'd have to shrink the options panel to one column....
20160213 11:13:42< shadowm> I may have to look into the debug output.
20160213 11:20:16< shadowm> Reason #59 I hate C++: I need to recompile 50 files because I added a protected method I only use in one of them.
20160213 11:21:13< shadowm> Um, generators aren't grids.
20160213 11:25:28< shadowm> Hm.
20160213 11:25:45< shadowm> Very interesting.
20160213 11:30:35-!- mjs-de [~mjs-de@x4db6bf8f.dyn.telefonica.de] has joined #wesnoth-dev
20160213 11:46:37< vultraz> tfw dialog crashes at 800x600
20160213 11:46:40< vultraz> DAMN YOU, GUI2
20160213 11:52:01< vultraz> I'll deal with the size stuff later
20160213 11:54:42-!- Kwandulin [~Miranda@p200300760F0BC5E8583E7BF0A0476587.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160213 11:55:45 * vultraz figures out how to use addon_info
20160213 12:01:27-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160213 12:06:34< vultraz> listbox bug :|
20160213 12:06:35-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 276 seconds]
20160213 12:08:49< vultraz> gfgtdf: for when you're around https://www.dropbox.com/s/jsjk0936h2gfdfv/GUI2%20addons%20list%20FULL%20WINDOW%20staging%202.PNG?dl=0
20160213 12:09:04< vultraz> (not tghe bug just the layout)
20160213 12:09:06< vultraz> the*
20160213 12:15:02-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Ping timeout: 276 seconds]
20160213 12:39:44-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160213 12:51:37-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160213 12:52:13-!- louis94 [~~louis94@13.149-243-81.adsl-dyn.isp.belgacom.be] has joined #wesnoth-dev
20160213 12:56:34-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 256 seconds]
20160213 12:59:26-!- louis94 [~~louis94@13.149-243-81.adsl-dyn.isp.belgacom.be] has quit [Ping timeout: 240 seconds]
20160213 13:02:22-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20160213 13:04:09-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has quit [Quit: Connection closed for inactivity]
20160213 13:31:19-!- irker054 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20160213 13:37:53-!- Kwandulin [~Miranda@p200300760F0BC5114122825F0007FCA3.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160213 13:41:03-!- gfgtdf [~chatzilla@f050180205.adsl.alicedsl.de] has joined #wesnoth-dev
20160213 13:45:04< gfgtdf> shadowm: well, that is actually the main reason to hate c++
20160213 13:45:40-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160213 13:45:43< gfgtdf> vultraz: hmm ooks fine to me, but i wonder how it behaves if the size of the desciption changes?
20160213 13:46:19< gfgtdf> vultraz: i also wonder whther its really that iseful to have all the sort options visible all the time.
20160213 13:48:47< gfgtdf> vultraz: i think it would be good if the sort options could be collapsed/hidden similar to a splitter. but faik we have no widet for that
20160213 13:50:04-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 252 seconds]
20160213 14:05:30-!- louis94 [~~louis94@13.149-243-81.adsl-dyn.isp.belgacom.be] has joined #wesnoth-dev
20160213 14:39:41-!- neverEnough [~nEnough@host76-232-dynamic.21-79-r.retail.telecomitalia.it] has joined #wesnoth-dev
20160213 14:39:54-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160213 14:44:46-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 272 seconds]
20160213 14:45:31-!- vultraz [~chatzilla@124.109.10.167] has joined #wesnoth-dev
20160213 14:45:31-!- vultraz [~chatzilla@124.109.10.167] has quit [Changing host]
20160213 14:45:31-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20160213 14:55:38-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has joined #wesnoth-dev
20160213 14:58:12< vultraz> 'lo, celticminstrel 
20160213 14:58:35< vultraz> I come bearing more gui2 things https://www.dropbox.com/s/jsjk0936h2gfdfv/GUI2%20addons%20list%20FULL%20WINDOW%20staging%202.PNG?dl=0
20160213 14:59:44< vultraz> gfgtdf: best I can do is maybe a toggle button that sets some to visible/hidden
20160213 14:59:53< vultraz> or maybe tree view but maybe not
20160213 15:00:26-!- irker006 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20160213 15:00:26< irker006> wesnoth: loonycyborg wesnoth:asio_wesnothd cea8fb1dc9fe / src/server/ (10 files): Clean up some commented out code and dead code https://github.com/wesnoth/wesnoth/commit/cea8fb1dc9fecf91b27ce33485cd92da364e7385
20160213 15:02:04< gfgtdf> vultraz: i somehow doubt that treeviews can handle this usecase.
20160213 15:02:18< vultraz> yeah probably
20160213 15:02:38< vultraz> i'm currently seeing if i can make the 'new' gui2 lists work 
20160213 15:02:41< vultraz> but
20160213 15:02:45< vultraz> they crash wesnoth w/o error >_>
20160213 15:02:58< gfgtdf> vultraz: hmm i think the 
20160213 15:03:23< gfgtdf> problem is that the 'new' gui2 lists assume that every dialog canhandle them
20160213 15:03:43< gfgtdf> (i assume you mean teh one that are enables by the EXPEREMENTAL_LISTBOX macro?)
20160213 15:04:00< vultraz> gfgtdf: no i mean it crashes on "initializing user interface" at startup. and yes
20160213 15:04:25< gfgtdf> vultraz: and it gives no error on stderr.txt ?
20160213 15:04:42< vultraz> ah wait it does
20160213 15:06:05< gfgtdf> vultraz: i think that is we want to have the 'new' listoxes our parroach shoudl be to have the 2 listboxes at the same time (tlist and tlistbox) so that we dont need to write each dialog to support both listboxes
20160213 15:06:21< gfgtdf> vultraz: instead of replacing the old lstbox i mean
20160213 15:06:26< vultraz> hm yes
20160213 15:06:33< gfgtdf> vultraz: if the new listox works, we can still replace it adfterwards
20160213 15:06:37< vultraz> right
20160213 15:07:13< gfgtdf> vultraz: do you know what are teh aadvantages of the new listboxes ?
20160213 15:07:20< vultraz> no
20160213 15:07:31< vultraz> that's why I'd like to look at a working instance
20160213 15:09:29< gfgtdf> vultraz: iirc the add-ons manager used the 'new' listoxes at soem time but i might have broken that when ia dded sort support to them
20160213 15:09:43< vultraz> yeah...
20160213 15:09:49< gfgtdf> to the addon-on manager dialog i men
20160213 15:09:51< gfgtdf> mean
20160213 15:10:51< gfgtdf> vultraz: bu iir it dodnt really work well also before, often the listbox item panel would overlap over teh list and impossible to remove from there
20160213 15:13:19< celticminstrel> Have I mentioned that I can't see dropbox images?
20160213 15:13:30< vultraz> ughh
20160213 15:14:04< gfgtdf> celticminstrel: maybe you have to enable javascript ?
20160213 15:14:32< celticminstrel> No, more likely it's cookies.
20160213 15:16:02< gfgtdf> celticminstrel: so you have javacript enables for dropbox page ?
20160213 15:17:11< vultraz> celticminstrel: https://drive.google.com/file/d/0B-mR9s8FduLLcDVjQzBPak1WSk0/view?usp=sharing
20160213 15:19:00< vultraz> gfgtdf: wow i can't even get the addons dialog to show up using the new lists and --new-widgets. seems like find_widget calls fail. Ugh
20160213 15:19:15< vultraz> maybe we should just keep fixing the current listbox and ditch this
20160213 15:23:39< celticminstrel> The "addons will be sorted by" section seems kind of redundant.
20160213 15:24:14< vultraz> yeah, I don't really like that section
20160213 15:24:32< vultraz> It's basically...what sorting do you want when you're not sorting :P
20160213 15:24:50< celticminstrel> Though it does have two extra options.
20160213 15:25:35< celticminstrel> I don't think there should be an option for an unsorted list.
20160213 15:25:55< celticminstrel> Sorting by last updated would be good though, so maybe a column for should be added.
20160213 15:26:40< celticminstrel> I wonder, could we replace _ with spaces in the names?
20160213 15:27:13< celticminstrel> Why are no checkboxes checked?
20160213 15:27:33< vultraz> Because this is a layout prototype and largely nonfunctional
20160213 15:27:49< vultraz> You can see the Lorem Upsum and the fake staging^ labels :P
20160213 15:27:54< celticminstrel> Cancel should probably be changed to Exit, since the Install button doesn't (or shouldn't?) close the dialog.
20160213 15:27:54< vultraz> Ipsum
20160213 15:29:04< celticminstrel> Are personal mods still relegated to the bottom?
20160213 15:30:22< vultraz> right now they're nowhere
20160213 15:30:37< celticminstrel> I wonder if the checkboxes would work as a single column if the radiobuttons were taken out.
20160213 15:31:10< celticminstrel> And filter could then be moved up so that it's above the listbox, which could be widened to make room for another column or two.
20160213 15:31:31< vultraz> I don't really think it needs more columns
20160213 15:31:39< celticminstrel> Size should probably be shown with units.
20160213 15:31:51< vultraz> yes, I know
20160213 15:32:06< vultraz> again, this is a proof of concept layout test
20160213 15:32:14< celticminstrel> Maybe it doesn't need more columns, but I think sorting by last update time would be useful, and the best way to do that is add that as a column, in my opinion.
20160213 15:33:19< celticminstrel> Sorting by version actually seems utterly useless.
20160213 15:33:52< celticminstrel> Perhaps that could be replaced with last update time.
20160213 15:33:52-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160213 15:33:57< vultraz> the header is disabled, actually
20160213 15:34:05< vultraz> so you can't sort by version
20160213 15:34:21< celticminstrel> Clearly I'm not the only one who thought it was useless!
20160213 15:34:31< vultraz> :P
20160213 15:34:37< vultraz> it's also impossible
20160213 15:34:45< celticminstrel> Alright, let's see if it's missing anything...
20160213 15:35:00< celticminstrel> Filter, check.
20160213 15:35:25< celticminstrel> Name/version/author/size/dl/type cols, check.
20160213 15:35:43< celticminstrel> Filter by type, check.
20160213 15:35:59< celticminstrel> I guess the combobox is all/installed/upgradable/not installed?
20160213 15:37:04-!- prkc [~prkc@51B6E6A0.dsl.pool.telekom.hu] has joined #wesnoth-dev
20160213 15:37:27< celticminstrel> Description pane has type/version/author/last update/status/size/dl/first upload/deps/trans/website. Yeah, that seems to be everything the old one has.
20160213 15:37:45< celticminstrel> That combobox could have additional options pertaining to locally authored addons.
20160213 15:38:38-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 256 seconds]
20160213 15:40:06< vultraz> "Not uploaded"?
20160213 15:40:39< celticminstrel> Sure, and "Authored" which would include any that have a server.pbl or whatever that file's called.
20160213 15:41:58< celticminstrel> Also, currently uploaded add-ons are displayed twice, once with a "Delete" option and once with a "Publish" option. Might be nice to instead show an extra button next to "Install" (and change Install to Upload. Not sure if that's already done.)
20160213 15:42:14< vultraz> it's not
20160213 15:42:16< vultraz> i would do that
20160213 15:42:40< celticminstrel> I feel like there's insufficient space for the dependencies and translations fields.
20160213 15:42:55< celticminstrel> The old dialog had them below website.
20160213 15:43:07< celticminstrel> There'd probably be enough space if they can use the entire width of the dialog.
20160213 15:43:41< celticminstrel> But it looks like the listbox and description share a column while the options are a second column - rather than the listbox and options sharing a row with the description being a subsequent row. Does that make sense?
20160213 15:44:13< vultraz> yes
20160213 15:44:25< vultraz> you think description should extend all the way across?
20160213 15:45:01< celticminstrel> If dependencies and translations are going to be where they currently are, yes. If they were moved back underneath Website, I don't think it would matter.
20160213 15:46:10< celticminstrel> An alternate option would be to move Last Updated and First Uploaded to where Dependencies and Translations currently are, and move Deps and Trans below Website.
20160213 15:46:24< celticminstrel> Basically, Deps and Trans potentially need a lot of space. That's what I'm worried about.
20160213 15:47:41< vultraz> you raise a good point
20160213 15:47:55< celticminstrel> ...I just thought of a cool idea. Place a "goto" type button next to each dependency in the dependency list, which when clicked selects the addon in the list (or alternatively, shows its description in place of whichever is currently selected).
20160213 15:48:49< vultraz> or just make it so you can click the dependency?
20160213 15:48:55< celticminstrel> Something.
20160213 15:49:10< celticminstrel> If you made it look like a hyperlink, that would work too.
20160213 15:49:47< celticminstrel> In another world, Add-on Types could be a combobox, but gfgtdf's version of combobox won't be able to do that.
20160213 15:50:00< vultraz> it could be done
20160213 15:50:11< vultraz> with some tweaking
20160213 15:50:40< celticminstrel> Probably not worth it unless we really, really need more space in the dialog.
20160213 15:51:52< celticminstrel> Looking at the columns, I guess there's enough wasted space there that an extra column could likely be slotted in even without making it any wider... though the window in general is kind of overly wide. How does it look in 800x600?
20160213 15:52:05< vultraz> it doesn't
20160213 15:52:12< vultraz> it crashes bc of insufficient width :P
20160213 15:52:13< vultraz> must fix :P
20160213 15:52:17< celticminstrel> I see.
20160213 15:52:58< celticminstrel> As a simple quickfix you could add horizontal scrolling to the listbox, though I don't think that's an ideal situation.
20160213 15:53:09< celticminstrel> ...that's assuming the listbox supports something like that, too.
20160213 15:53:27< vultraz> it does
20160213 15:53:59< celticminstrel> What do you think about removing the version column?
20160213 15:54:40< vultraz> can't
20160213 15:54:53< vultraz> shadowm wants his layout in that respect preserved
20160213 15:55:17< celticminstrel> Okay, but like I mentioned earlier, the version could be moved into the name column.
20160213 15:55:53< celticminstrel> I guess it's not important though.
20160213 15:56:01< celticminstrel> Just a thought.
20160213 15:56:21< celticminstrel> I do think showing the version in the listbox is important.
20160213 15:57:23< celticminstrel> Well, that's all I have to say on this.
20160213 15:58:54< vultraz> ok, thanks
20160213 16:18:46-!- Kwandulin [~Miranda@p200300760F0BC5114122825F0007FCA3.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160213 16:22:41-!- louis94 [~~louis94@13.149-243-81.adsl-dyn.isp.belgacom.be] has quit [Ping timeout: 276 seconds]
20160213 16:27:59-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160213 16:28:35-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20160213 16:33:04-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 272 seconds]
20160213 16:49:16-!- Alduin_ [~Alduin@p200300864458E300E29467FFFE0EE8C4.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160213 16:52:26-!- Kwandulin [~Miranda@p200300760F0BC51171FF05B61EAE054B.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160213 16:59:00-!- louis94 [~~louis94@13.149-243-81.adsl-dyn.isp.belgacom.be] has joined #wesnoth-dev
20160213 17:03:22-!- Alduin_ [~Alduin@p200300864458E300E29467FFFE0EE8C4.dip0.t-ipconnect.de] has quit [Ping timeout: 250 seconds]
20160213 17:10:02< gfgtdf> vultraz: we should change the code to something like if(width < desired_width) { system("shotdown")} :p
20160213 17:10:16< vultraz> what...?
20160213 17:11:09 * celticminstrel assumes that's a joke.
20160213 17:12:11< gfgtdf> vultraz: that was sarcasic and meant that we should realy fix the 'wesnoth crashes on isufficent space' issue.
20160213 17:12:29< vultraz> oh
20160213 17:12:31< vultraz> yea we should
20160213 17:14:27< celticminstrel> In other words, we should excise the use of assertions in GUI2.
20160213 17:15:05< celticminstrel> Well, there might be a few that can stay. Situations that are logically impossible, like "the window is null".
20160213 17:15:37< celticminstrel> But still, they shouldn't be used for things like insufficient screen space. Just clip it into the available space somehow.
20160213 17:15:54< celticminstrel> Even overlapping elements would be better than a crash.
20160213 17:16:00-!- Alduin_ [~Alduin@2001:4ca0:4fff:4::a] has joined #wesnoth-dev
20160213 17:16:29< vultraz> maybe you could remove some of them?
20160213 17:16:39< celticminstrel> Some elements?
20160213 17:16:53< celticminstrel> That's probably a bad idea, but still better than crashing.
20160213 17:17:11< vultraz> i mean asserts
20160213 17:17:23< celticminstrel> Automatically instating a scroll pane in the event of insufficient space would probably be the best solution.
20160213 17:17:59< celticminstrel> Though random scroll panes showing up in dialogs would probably be a little annoying.
20160213 17:21:49< vultraz> celticminstrel: ok wow. 545 asserts in 92 files in gui2 
20160213 17:22:08-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160213 17:22:19< celticminstrel> Wow indeed. o.o
20160213 17:23:49< vultraz> you have your work cut out for you :P
20160213 17:24:21< celticminstrel> I'd much prefer not to touch GUI2 if I can avoid it. >_> Apart from maybe constructing a new dialog here and there.
20160213 17:24:41< celticminstrel> But I suppose I might try something eventually if gfgtdf and Aginor don't get to it first.
20160213 17:24:50< vultraz> celticminstrel: quick OT... what does 'explicit' do again?
20160213 17:24:59< vultraz> taddon_list has an explicit constructor 
20160213 17:25:08< celticminstrel> That's for constructors (and, in C++11, conversion operators).
20160213 17:25:35< celticminstrel> Any constructor that could potentially be called with exactly one argument (including due to defaulted parameters) qualifies to be declared "explicit".
20160213 17:26:00< celticminstrel> If you don't declare it explicit, it's considered to define an implicit conversion between the argument type and the class type that it's declared in.
20160213 17:26:29< celticminstrel> Which, most of the type, actually isn't what you want.
20160213 17:26:31< celticminstrel> ^time
20160213 17:27:01-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 276 seconds]
20160213 17:27:26-!- horrowind [~Icedove@2a02:810a:8380:834:21b:fcff:fee3:c3ff] has quit [Quit: horrowind]
20160213 17:34:14-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160213 17:35:26< gfgtdf> y i wonder why they didnt make explicit the default
20160213 17:36:34< celticminstrel> So do a lot of people.
20160213 17:36:46< celticminstrel> But it's kinda too late to change now.
20160213 18:00:24-!- irker006 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20160213 18:16:14-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160213 18:21:38-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 276 seconds]
20160213 18:34:54-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160213 18:39:08-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 250 seconds]
20160213 19:10:59-!- travis-ci [~travis-ci@ec2-54-81-197-37.compute-1.amazonaws.com] has joined #wesnoth-dev
20160213 19:11:00< travis-ci> wesnoth/wesnoth#8460 (asio_wesnothd - cea8fb1 : loonycyborg): The build has errored.
20160213 19:11:00< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/109011870
20160213 19:11:00-!- travis-ci [~travis-ci@ec2-54-81-197-37.compute-1.amazonaws.com] has left #wesnoth-dev []
20160213 19:12:12-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160213 19:15:48-!- louis94 [~~louis94@13.149-243-81.adsl-dyn.isp.belgacom.be] has quit [Ping timeout: 248 seconds]
20160213 19:17:31-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 276 seconds]
20160213 19:19:56< vultraz> celticminstrel: is there a simple way to measure a string length and insert a '\n' character after the lest space character found before the provided length?
20160213 19:20:10< celticminstrel> Yes.
20160213 19:20:26< celticminstrel> But first, what do you actually mean by "length"? There are two distinct concepts here.
20160213 19:21:46< vultraz> Number of characters
20160213 19:22:15< celticminstrel> That's super-easy, though I wonder if this is really the correct length measure for your use-case.
20160213 19:22:46< celticminstrel> I think find_last_of might do what you're saying.
20160213 19:27:04< zookeeper> not a very successful attempt at a prototype logo animation: https://dl.dropboxusercontent.com/u/63964618/wesnoth/logo_hexed.gif just saying that if we ever want (and could support) something even remotely along those lines, i ought to be able to provide an animation.
20160213 19:27:55< zookeeper> (i had always wanted the thankfully already gone wipe animation replaced with some kind of pixelly or hex-based transition)
20160213 19:28:23< vultraz> zookeeper: to be honest, I don't really like it
20160213 19:28:28< vultraz> zookeeper: it feels very '2005'
20160213 19:29:33< zookeeper> that's a strange description :P
20160213 19:30:09< vultraz> 2005 meaning like something from one of the pre-iphone flash games you'd play on the internet
20160213 19:30:39< vultraz> it doesn't feel...elegant
20160213 19:32:24< zookeeper> well, i believe elegance in this case is a matter of implementation details
20160213 19:32:55 * neverEnough wonders, do apple users code too? O_O
20160213 19:34:00< vultraz> zookeeper: perhaps
20160213 19:37:14< celticminstrel> So vultraz, are you trying to implement word-wrapping?
20160213 19:37:31< vultraz> celticminstrel: no, I remembered there's characters_per_line :P
20160213 19:39:29< celticminstrel> ?
20160213 19:39:41< celticminstrel> I really have no idea what you're doing.
20160213 19:39:53< celticminstrel> But that still sounds wrong.
20160213 19:40:11< celticminstrel> Potentially.
20160213 19:40:20< vultraz> gui2 key
20160213 19:40:22< vultraz> for labels
20160213 19:40:44< celticminstrel> Huh?
20160213 19:41:11< vultraz> gui2 labels have a characters_per_line key
20160213 19:41:33< celticminstrel> Why the heck do they have something like that?
20160213 19:42:02< zookeeper> Aginor, oh btw. i now have an sdl2 build and all, and while ö is still the physical key for user command, it does now correctly display it as ö in the hotkeys list. just saying, since i complained about that before.
20160213 19:42:08< vultraz> it's actually kinda like a word wrap function
20160213 19:42:47< gfgtdf> vultraz: note that this is note reallte related to the actual number of characters
20160213 19:43:02< vultraz> gfgtdf: yeah it's size
20160213 19:43:14< vultraz> gfgtdf: i figured that out the other day with the character width difference
20160213 19:44:28< celticminstrel> Oh, so the key is misnamed?
20160213 19:44:36< gfgtdf> celticminstrel: no not really
20160213 19:45:06< gfgtdf> celticminstrel: it more like max_width is calculates by 'characters_per_line'*'average size of a character'
20160213 19:45:29< gfgtdf> celticminstrel: where average size of a character is unrelated to the characters used in that specific text
20160213 19:45:40-!- trewe [~trewe@bl20-32-64.dsl.telepac.pt] has joined #wesnoth-dev
20160213 19:47:57< celticminstrel> Sounds like it's still doing word wrapping wrong. Oh well?
20160213 19:48:20< zookeeper> urgh, i checked and i've actually said that i can live with deprecation of MENU_IMG_TXT... because that's awful.
20160213 19:48:24< celticminstrel> The proper way to word-wrap is to use the routines provided by the text layouter to calculate the length in pixels and compare that to the available space.
20160213 19:48:36< vultraz> zookeeper: ?
20160213 19:48:38< celticminstrel> zookeeper: Huh?
20160213 19:50:21< vultraz> celticminstrel: it's actually handled by ttext::recalculate()
20160213 19:54:16< zookeeper> vultraz, what's this?
20160213 19:54:16< zookeeper> error gui/layout: ttext::font::ttext::set_markup_helper text '<big>Victory:</big>
20160213 19:54:32< vultraz> context, please
20160213 19:56:43< zookeeper> vultraz, http://pastebin.com/dwDgKapb
20160213 19:59:48< vultraz> zookeeper: I don't understand
20160213 20:00:44< zookeeper> whole error: http://pastebin.com/DcD5Zv5a
20160213 20:01:19< zookeeper> if that message is correct, it has cut off the last > which breaks the syntax
20160213 20:01:27< vultraz> zookeeper: I believe it's because </small> is inside the quotes
20160213 20:02:24< zookeeper> well of course it's inside the quotes of the string it's part of
20160213 20:03:20< gfgtdf> zookeeper: does that happen in all wesnoth versions ?
20160213 20:03:21< vultraz> no it's not
20160213 20:03:26< vultraz> it's describing the string
20160213 20:03:32< vultraz> oh wait
20160213 20:03:36< vultraz> ehh??
20160213 20:04:26< zookeeper> gfgtdf, 1.13
20160213 20:04:27< vultraz> now im confused
20160213 20:06:29-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160213 20:09:27< vultraz> zookeeper: could you help out the guy with pr 596
20160213 20:10:04< zookeeper> you mean could i help him with how to use git?
20160213 20:10:22< zookeeper> because i think you know the answer to that
20160213 20:10:50< vultraz> yes. also explain about indent, because somehow he messed it up :|
20160213 20:11:02-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 252 seconds]
20160213 20:11:12< zookeeper> umm. no, i can't explain to him how to rebase or whatever.
20160213 20:11:29< vultraz> celticminstrel maybe?
20160213 20:11:57< celticminstrel> Maybe what?
20160213 20:12:33< vultraz> could you explain to him how to rebase?
20160213 20:12:46< vultraz> I'm not very good at the actual git commands
20160213 20:12:50< vultraz> I'm spoiled by TGit
20160213 20:12:52< celticminstrel> Link?
20160213 20:13:08< vultraz> https://github.com/wesnoth/wesnoth/pull/596
20160213 20:13:20< zookeeper> also tell him to squash
20160213 20:18:41< zookeeper> "error scripting/lua: The &image=col1=col2 syntax is deprecated, use new DescriptionWML instead." <- but doesn't say where i'm purportedly using it
20160213 20:20:01< gfgtdf> vultraz: i wonder whether we can change teh MENU_IMG_TXT macro into something like #define MENU_IMG_TXT IMAGE TEXT  {TEXT} image = {IMAGE} #enddef ?
20160213 20:20:13< zookeeper> gfgtdf, yeah i'm actually testing that
20160213 20:20:35< zookeeper> looks easy enough, only "warning wml: [option] has both label= and message=, ignoring the latter" gets in the way and that ought to be easy to change
20160213 20:21:28< zookeeper> of course that might cause problems when used in [campaign]
20160213 20:25:18-!- Alduin_ [~Alduin@2001:4ca0:4fff:4::a] has quit [Ping timeout: 250 seconds]
20160213 20:25:55< gfgtdf> zookeeper: campaigsn accepts differetn keys than message ?
20160213 20:26:45< zookeeper> description, image
20160213 20:26:46< vultraz> no
20160213 20:27:16< gfgtdf> hmm ok it seems like difficilfy_descriptions is the old format, y that migth give problems
20160213 20:28:29< vultraz> gfgtdf: yes, but that key is deprecated
20160213 20:28:31< vultraz> don't use it
20160213 20:30:57< vultraz> Someone remind me to add a standard table of hex colors to font.cpp at some point
20160213 20:33:10< vultraz> celticminstrel: c++ equivalent of lua's string.format()?
20160213 20:37:12< gfgtdf> vultraz: you coudl eigher strignstrems, or boost::format
20160213 20:45:55< zookeeper> vultraz, so i'm looking at message.lua and i see no reason why it couldn't just split the old-style string into label,description,image,default?
20160213 20:46:19< zookeeper> looks like it wouldn't require more than basic string splitting
20160213 20:46:44< vultraz> zookeeper: I want people to abandon the syntax as fast as possible
20160213 20:46:48< vultraz> the old syntax*
20160213 20:47:26< zookeeper> yeah and you know no one else here appreciates that
20160213 20:48:31< vultraz> *if* you can split the raw string into the new format *without* using the old marked_up_text class, then fine
20160213 20:48:40< vultraz> emphasis on *without*
20160213 20:49:05< vultraz> ah, tlegacy_menu_item is what it's called
20160213 20:49:32< zookeeper> as i said, i'm talking about the lua code
20160213 20:49:53< vultraz> well as for the lua, celticminstrel would have to be consulted
20160213 20:50:05< vultraz> there's also some c++ backend that makes use of that class
20160213 20:54:06-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20160213 20:54:06-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Remote host closed the connection]
20160213 21:00:25-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160213 21:05:02-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 256 seconds]
20160213 21:12:13< celticminstrel> [Feb 13@3:33:10pm] vultraz: celticminstrel: c++ equivalent of lua's string.format()?
20160213 21:12:14< celticminstrel> Probably nothing, but I'm not too familiar with the Lua here. How does it work?
20160213 21:13:11< celticminstrel> [Feb 13@3:20:01pm] gfgtdf: vultraz: i wonder whether we can change teh MENU_IMG_TXT macro into something like #define MENU_IMG_TXT IMAGE TEXT  {TEXT} image = {IMAGE} #enddef ?
20160213 21:13:13< celticminstrel> [Feb 13@3:20:13pm] zookeeper: gfgtdf, yeah i'm actually testing that
20160213 21:13:13< celticminstrel> I doubt this would work.
20160213 21:13:41< vultraz> string.format("puts %s together", "string) would print "puts string together"
20160213 21:13:43< vultraz> IIRC
20160213 21:13:48< vultraz> I've been away from lua for a few months
20160213 21:14:06< celticminstrel> The change could be made eventually, after MENU_IMG_TXT is ready to be removed, but that's a different story.
20160213 21:14:18< zookeeper> works just fine for me, in option message context: http://pastebin.com/Tn1YChdg
20160213 21:14:27< celticminstrel> Basically I think it's impossible for MENU_IMG_TXT to work both in the new and old syntax, which is what it would have to do.
20160213 21:15:09< celticminstrel> If I'm wrong, then go ahead and change it, I guess?
20160213 21:15:22< vultraz> why the ""?
20160213 21:15:30< zookeeper> to make message=""
20160213 21:15:30< celticminstrel> vultraz: C++ has nothing like that. There's the unsafe sprintf and safe variants of it, but normally you'd use stringstreams instead.
20160213 21:16:49< celticminstrel> zookeeper: Does that cause a deprecated message if used in [option], then?
20160213 21:17:16< celticminstrel> Also, the other problem is that it wouldn't work in [campaign].
20160213 21:17:24-!- ancestral [~ancestral@209.181.254.220] has joined #wesnoth-dev
20160213 21:17:36< zookeeper> yes, and yes
20160213 21:17:39< zookeeper> err
20160213 21:17:41< zookeeper> no, and yes
20160213 21:18:04< zookeeper> deprecation warning doesn't come from the macro, but from the syntax
20160213 21:18:25< celticminstrel> So the presence of an empty message= doesn't raise the warning?
20160213 21:18:38< zookeeper> correct. it just raises the warning i said it does.
20160213 21:19:08< celticminstrel> vultraz: ABout PR 596. Is it the case that we want this merged?
20160213 21:19:22< vultraz> I intended to do that
20160213 21:19:39< vultraz> even if zookeeper does his thing, the fewer uses of it the better
20160213 21:19:53< gfgtdf> vultraz: please dont merge it unless are issues with it are fixed
20160213 21:20:04< gfgtdf> vultraz: all
20160213 21:20:08< vultraz> gfgtdf: ...obviously
20160213 21:20:09< celticminstrel> gfgtdf: I was considering merging and squashing it manually, if we want it.
20160213 21:20:40< gfgtdf> vultraz: specially meaning the whitepases and teh squashing
20160213 21:20:49< vultraz> yes. I know
20160213 21:21:13< celticminstrel> Are there still whitespace issues in the current head of that PR?
20160213 21:21:32< gfgtdf> vultraz: i actualy wonder whether we can mabe just have a scropt in in python wml update tools to port code to the new attributes
20160213 21:22:11< gfgtdf> celticminstrel: yes, just llok at the prs diff
20160213 21:22:24< celticminstrel> I'm looking. Haven't seen anything obvious yet.
20160213 21:23:18< gfgtdf> celticminstrel: hmm but the files ANL_worker_options.cfg (which is over 50% of the diff) is full of such changes
20160213 21:23:28< celticminstrel> Haven't gotten to that file yet.
20160213 21:23:54< celticminstrel> Okay, ANL_research_options seems to have whitespace changes too.
20160213 21:24:36< gfgtdf> celticminstrel: well i dont know fore sure of they are all bad, but they are such unrelated to the isses that are fixed by thic ommits (the syntax change in [option])
20160213 21:24:40< vultraz> celticminstrel, gfgtdf: small update https://drive.google.com/file/d/0B-mR9s8FduLLMlFyT2ZwTng3ZG8/view?usp=sharing
20160213 21:25:11< gfgtdf> vultraz: 'Beim Anzeigen des Bildes ist ein Problem aufgetreten.'
20160213 21:25:14< celticminstrel> I don't see anything different.
20160213 21:25:50< vultraz> celticminstrel: I've gotten the list filled out
20160213 21:25:54< celticminstrel> Except the removal of "staging^" stuff.
20160213 21:26:03< celticminstrel> There was stuff in the list before, too.
20160213 21:26:09< vultraz> but the wrong stuff
20160213 21:26:15< celticminstrel> Oh okay.
20160213 21:26:20< vultraz> bad size labels
20160213 21:26:22< vultraz> etc
20160213 21:28:05-!- Kwandulin [~Miranda@p200300760F0BC51171FF05B61EAE054B.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160213 21:28:14< gfgtdf> zookeeper: i wonder why ANL has these 'wmlindent: start ignoring' in it ?
20160213 21:28:30< vultraz> gfgtdf: pango markup
20160213 21:28:35< vultraz> wmlindent doesn't like it
20160213 21:28:35< zookeeper> celticminstrel, so, as i said i don't think it looks hard to fix message.lua:144-> to just convert from the old syntax without needing deprecation warnings. you wouldn't want to do that, would you?
20160213 21:29:42-!- ancestral [~ancestral@209.181.254.220] has quit [Quit: i go nstuf kthxbai]
20160213 21:30:51< celticminstrel> The combined diff for ANL_worker_options is impossible to follow because he seems to have indented a giant block of code. That may not be a problem if the indentation was in a separate commit, though...
20160213 21:31:22< celticminstrel> The main thing I dislike about this PR is that his commits imply that he did the work primarily through the github interface.
20160213 21:32:48< gfgtdf> im trying to port the gotky preferences dialog to gui2 but its harder than i thought mainly becasue of the 'press the key (esc to abort)' mainly becasue in when using window.connect_signal i see no way to know the scancode of the key.
20160213 21:33:27< gfgtdf> celticminstrel: well when you know ahty you are doing and when you only want to change one file then i see no disadvantages of using github webinterface.
20160213 21:35:20-!- ancestral [~ancestral@209.181.254.220] has joined #wesnoth-dev
20160213 21:36:53< celticminstrel> gfgtdf: Yeah, if it really affects only a single file, sure. I've done that on occasion (I think even in Wesnoth - RELEASE_NOTES for example).
20160213 21:37:17< celticminstrel> But in this case it looks like he edited several files in sequence using github's interface.
20160213 21:38:19< gfgtdf> celticminstrel: y in this case using github webinterfacse was sureley not a good way to do it.
20160213 21:38:43< celticminstrel> Looking at this, I'm suspecting wmlindent might actually have a bug.
20160213 21:39:28< gfgtdf> celticminstrel: why ?
20160213 21:40:29< celticminstrel> gfgtdf: https://github.com/wesnoth/wesnoth/pull/596#discussion-diff-52833783
20160213 21:41:21< gfgtdf> celticminstrel: hmm mabye its the 'wmlindent doesn't like pango markup' issue that vultraz mentioned above?
20160213 21:41:37< celticminstrel> You think that's related?
20160213 21:41:47< celticminstrel> I was assuming it just didn't properly ignore indentation within quotes.
20160213 21:43:04< gfgtdf> celticminstrel: i dont really know anything about wmlindent., but since pango markup and wmlindent in involved ,vultraz notice about pango markup and wmlindent was the first thing that came in my mind.
20160213 21:43:27< celticminstrel> I dunno.
20160213 21:43:47< gfgtdf> celticminstrel: maybe vultraz knows.
20160213 21:49:57< zookeeper> vultraz, celticminstrel, so... it'd be great if you let me know whether either of you is going to make message able to handle the old syntax as well.
20160213 21:51:32-!- Roniga [6f4549ac@gateway/web/freenode/ip.111.69.73.172] has joined #wesnoth-dev
20160213 21:51:52< celticminstrel> I'm pretty sure [option] handles the old syntax already?
20160213 21:53:50< Roniga> I'm inviting the author of PR 596 to irc
20160213 21:53:52< Roniga> <--- Aginor
20160213 21:54:32-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160213 21:55:20< zookeeper> celticminstrel, ah, true, it does. it just detects the old syntax and outputs a deprecation warning.
20160213 21:56:10-!- Roniga [6f4549ac@gateway/web/freenode/ip.111.69.73.172] has quit [Client Quit]
20160213 21:56:50< zookeeper> so it's basically just a matter of removing deprecation warnings and keeping the old code.
20160213 21:57:47< vultraz> what old code?
20160213 21:58:04< vultraz> also, don't remove the deprecation warnings
20160213 21:58:07< celticminstrel> Why are we removing deprecation warnings?
20160213 21:58:09< vultraz> this *is* deprecated
20160213 21:58:13-!- ancestral [~ancestral@209.181.254.220] has quit [Quit: i go nstuf kthxbai]
20160213 21:58:13< celticminstrel> They're there as a reminder to update code.
20160213 21:58:56-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 240 seconds]
20160213 22:00:23< zookeeper> because when there's clearly no good technical reason to remove support for that old syntax which is very widely used, then it should not be. my bad for not insisting on that before, but i thought there was a good reason.
20160213 22:00:43< celticminstrel> Well, we're not removing support for it.
20160213 22:00:53< vultraz> We will be 
20160213 22:00:55< celticminstrel> I suppose we could suppress the deprecation warnings until 1.14 or something.
20160213 22:01:01< vultraz> It's ugly and shouldn't be around
20160213 22:01:15< vultraz> It's even called *legacy* 
20160213 22:01:38< zookeeper> shadowm, lend me your squeaky mallet
20160213 22:01:38< vultraz> notes in the code said a new syntax should be devided
20160213 22:01:46< vultraz> we've done so!
20160213 22:02:42< vultraz> plus it may be widely used, but how many people actually understand how it works
20160213 22:02:56< vultraz> it makes maintenance hard 
20160213 22:03:05< celticminstrel> I don't really have any problem with putting off the deprecation aspect of it.
20160213 22:03:06< vultraz> plus *you* greenlit the warnings
20160213 22:03:40< zookeeper> if necessary, we can put all the old code you hate in a file called here_be_dragons_dont_look.cpp and not need to worry about it like almost ever.
20160213 22:04:09< celticminstrel> XD
20160213 22:04:11< vultraz> :|
20160213 22:07:40< zookeeper> the old code for this stuff is... limited to old_markup.*pp? those files are <100 lines of easy code each. there's no maintenance problem.
20160213 22:08:04< vultraz> yes, that has to hang around for backwards compatibility
20160213 22:08:11< vultraz> but I'm referring to the WML syntax
20160213 22:08:29< vultraz> well, the not!wml syntax in this case
20160213 22:12:06< celticminstrel> Is old_markup.?pp still used for the old handling of this? I can't remember.
20160213 22:12:28< vultraz> yes
20160213 22:13:43< zookeeper> what i don't understand is how the heck does this work: https://github.com/wesnoth/wesnoth/blob/master/src/scripting/lua_gui2.cpp#L268-L276
20160213 22:13:57< zookeeper> how does short_opt end up with meaningful contents?
20160213 22:14:22< zookeeper> oh, the luaW_totstring call does it, i presume
20160213 22:14:27< zookeeper> instead of testing for... something
20160213 22:15:12< zookeeper> i still don't see how it knows what to stuff into it
20160213 22:16:56-!- Robertdebrus [6322841b@gateway/web/freenode/ip.99.34.132.27] has joined #wesnoth-dev
20160213 22:17:48-!- trewe [~trewe@bl20-32-64.dsl.telepac.pt] has quit [Quit: quit]
20160213 22:24:45-!- gfgtdf [~chatzilla@f050180205.adsl.alicedsl.de] has quit [Quit: ChatZilla 0.9.92 [Firefox 44.0.2/20160210153822]]
20160213 22:27:25< vultraz> ugh
20160213 22:27:45< vultraz> annoying, this is
20160213 22:28:05< vultraz> gui2 layouts, that is
20160213 22:28:22< celticminstrel> zookeeper: Yeah, luaW_totstring stores its result in it.
20160213 22:28:53< vultraz> GUI2 sucks
20160213 22:29:05< celticminstrel> Based on what's at index -1 on the Lua stack.
20160213 22:29:25< zookeeper> vultraz, so i presumed you were talking about the objectives bug when you said that <vultraz> now im confused
20160213 22:29:33< zookeeper> was that right?
20160213 22:29:34< vultraz> yes
20160213 22:29:42< zookeeper> ok, so i presume you'll look at it sooner or later?
20160213 22:30:33< vultraz> possibly
20160213 22:31:06< zookeeper> well that sounds like it's not your fault, meaning i could pester someone else about it if i knew who
20160213 22:36:47< vultraz> perhaps that would be best
20160213 22:37:35< vultraz> why why why did mordante never implement widget dimension settings >_>
20160213 22:39:11< vultraz> oh, and of course I come across the damn height bug 
20160213 22:39:17< vultraz> t(-___-t)
20160213 22:39:31< vultraz> i need a break
20160213 22:45:30-!- louis94 [~~louis94@13.149-243-81.adsl-dyn.isp.belgacom.be] has joined #wesnoth-dev
20160213 22:48:39-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160213 22:53:42-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 272 seconds]
20160213 22:55:17< zookeeper> anyway. lua_gui2.cpp apparently only prints the deprecation warning to the log, which is all good and dandy. the user-visible warning comes solely from https://github.com/wesnoth/wesnoth/blob/master/data/lua/wml/message.lua#L143-L146 so i'm just suggesting removing those lines.
20160213 22:55:29< zookeeper> shadowm, ^ please agree that this is the way
20160213 22:56:40< shadowm> The way to do what? I'm too busy to check the logs right now.
20160213 22:57:15< shadowm> Also, I believe that whole file is celticminstrel's baby.
20160213 22:57:43< vultraz> shadowm: he wants to remove user-facing deprecation warnings for the old descriptionwml syntax
20160213 22:57:56< vultraz> I disagree strongly
20160213 22:58:11< shadowm> I don't even know the new syntax, don't ask me for an opinion.
20160213 22:59:06< shadowm> I do generally defer to zookeeper's opinion on WML API stuff because he's been doing this for much longer than any of us current developers (save our translation manager).
20160213 22:59:24< vultraz> The new syntax is three wml keys
20160213 22:59:26< zookeeper> well you don't need to know the new syntax to appreciate how widely used the old syntax is (in the form of MENU_IMG_TXT*)
20160213 22:59:36< vultraz> much simpler, cleaner, and easier for wml coders to understand
20160213 23:00:03< vultraz> And its widespread use is exactly why we should phase it out :|
20160213 23:00:31< shadowm> WML is very much like Perl (Lua's presence only increases the truth in this analogy) so I'm not opposed to there being two ways to do the same thing.
20160213 23:01:36< shadowm> All involved parties should carefully consider the strengths and weaknesses of their respective proposals, though.
20160213 23:01:56< zookeeper> oh and all this is now in context of [option] message= only, not [campaign] difficulty_description
20160213 23:02:08< shadowm> Also keep in mind what I said earlier this year or maybe last year (god why do I always have to repeat myself).
20160213 23:02:24< shadowm> "blah blah blah can't afford alienating existing UMC maintainers blah blah blah"
20160213 23:02:29< zookeeper> ^^^^^^
20160213 23:02:32< vultraz> zookeeper: you agreed to the deprecation warnings for campaign difficulties - this is the same syntax. why should one be deprecated but not the other?
20160213 23:02:32< shadowm> I'm paraphrasing, really.
20160213 23:02:49< shadowm> It was a far more elaborate statement than that, but again, busy.
20160213 23:04:18< zookeeper> vultraz, i'm pretty sure i've never said it should be deprecated anywhere. but deprecating it for campaign difficulties is obviously a much less problematic change.
20160213 23:04:45< vultraz> zookeeper: MENU_IMG_TXT2 is used for the old campaign difficulty descriptions. So it's the same kind of change from a user's POV
20160213 23:05:15-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20160213 23:05:29< zookeeper> it is and it isn't
20160213 23:05:36< zookeeper> what are you getting at?
20160213 23:06:02< vultraz> That it will confuse people if they see deprecation warnings related to one use of the macro but not the other
20160213 23:06:26< zookeeper> right
20160213 23:06:41< vultraz> But I concede I may be thinking about that from a technical standpoint
20160213 23:07:12< zookeeper> however, even currently they won't see a deprecation warning from the macro. although arguably you'd want to do that as well, i'm sure :p
20160213 23:08:53< vultraz> A good reason to repurpose the macro is a bit less code has to be changed, and the warning shouldn't occur. A point against is that it will immediately break [campaign] difficulty_description
20160213 23:09:04< vultraz> (I think, at least)
20160213 23:09:21< zookeeper> so what's your problem with keeping the old syntax for _both_? like, none of those usual silly reasons of technical purism that no one else really cares about, but how does support for the old syntax get in your way personally?
20160213 23:09:32-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 272 seconds]
20160213 23:09:32-!- wedge010 is now known as wedge009
20160213 23:09:46< vultraz> it doesn't get in my way
20160213 23:09:57< vultraz> It's just that it's extremely confusing to newcomers
20160213 23:10:12< zookeeper> so you want to remove it to help UMC authors?
20160213 23:10:16< vultraz> The new syntax makes the functions a lot more obvious
20160213 23:10:18< zookeeper> like, make their life easier?
20160213 23:10:18< vultraz> Yes
20160213 23:10:26< zookeeper> well, i'm glad to hear that
20160213 23:10:29< celticminstrel> That file's not entirely my doing; gfgtdf and vultraz in particular also had some input.
20160213 23:10:58< celticminstrel> I have no particular objection to removing the deprecation warning at the moment.
20160213 23:11:09< vultraz> when I first started learning WML I assumed MENU_IMG_TXT was a holy magic macro that made images appear in options
20160213 23:12:46< vultraz> WML's thing, for better or for worse, is that it's human readable
20160213 23:13:41< zookeeper> i realize that there is no method of divine scrying with which we can confirm with absolute certainty whether it's removing or keeping the support for the old syntax which will help UMC authors more, but i can guarantee that if there was, i'd bet my head that "keeping" is the answer by a gargantuan margin.
20160213 23:14:31< vultraz> you may argue that the macro is readable, but I'd prefer the use of wml keys, clean
20160213 23:15:12< vultraz> plus from a technical standpoint, it would be easier if we ever revamped our scripting language to have these oddities disposed of
20160213 23:15:20< vultraz> (that is one thing I agree with fabi one)
20160213 23:15:38< zookeeper> no one argues that the macro is more readable or even good
20160213 23:15:47< zookeeper> i'm sure you know that
20160213 23:17:25< vultraz> right
20160213 23:19:01< celticminstrel> We haven't dropped the support for the old syntax.
20160213 23:19:17< celticminstrel> Possibly thanks to me insisting while vultraz was working on it?
20160213 23:19:45< celticminstrel> You're only complaining that the warning is shown to users, which I don't particularly mind dropping.
20160213 23:19:57< celticminstrel> Or changing it to a log message.
20160213 23:20:13-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: mattsc]
20160213 23:20:38< celticminstrel> Like most of the other warnings in that file.
20160213 23:21:50< celticminstrel> log(text, "warning") in other words
20160213 23:22:02< zookeeper> yes, i only care about the user-facing warning
20160213 23:23:22< loonycyborg> https://github.com/boostorg/iostreams/pull/23#event-549638365 <- nice, they actually merged my PR.
20160213 23:23:24-!- irker199 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20160213 23:23:24< irker199> wesnoth: Celtic Minstrel wesnoth:master d28365730b00 / data/lua/wml/message.lua: Demote [option] deprecated message to log https://github.com/wesnoth/wesnoth/commit/d28365730b004d852afa3371be347514cdb098ba
20160213 23:23:36< loonycyborg> that'll make building boost on windows easier for me :P
20160213 23:26:18-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20160213 23:28:17< zookeeper> celticminstrel, i'm still seeing it though?
20160213 23:29:41< celticminstrel> Maybe I need it to be at debug level... :|
20160213 23:29:56< celticminstrel> You don't have debug mode active in-game, do you?
20160213 23:30:09< zookeeper> nope
20160213 23:30:21< celticminstrel> Only asking because I think more errors are put to chat area when it's active.
20160213 23:30:45< celticminstrel> So basically, my change didn't change anything?
20160213 23:32:35< zookeeper> as far as i can see, correct
20160213 23:33:03< zookeeper> i dunno if the fact that i'm running a debug build affects what domains are displayed by default
20160213 23:33:12< celticminstrel> Could do.
20160213 23:34:25< celticminstrel> Do you have access to the debug commands?
20160213 23:34:38< celticminstrel> eg, right-click -> create unit
20160213 23:35:35< zookeeper> nope
20160213 23:35:57< shadowm> No, it doesn't.
20160213 23:38:41< celticminstrel> I assume you never get messages like "No speaker found for message" (which is a log message emitted if the speaker doesn't exist, for example when they already died).
20160213 23:38:50-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160213 23:42:05< irker199> wesnoth: Celtic Minstrel wesnoth:master 1873b3824347 / data/lua/wml/message.lua: Demote [option] deprecated message even further https://github.com/wesnoth/wesnoth/commit/1873b382434732b764124a170b1b13b444747e7d
20160213 23:42:49< zookeeper> yeah i don't recall ever seeing that
20160213 23:42:57-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 240 seconds]
20160213 23:43:15< celticminstrel> Then this commit should stop the message showing up in chat.
20160213 23:43:15< zookeeper> it's gone now
20160213 23:49:00-!- gfgtdf [~chatzilla@f050180205.adsl.alicedsl.de] has joined #wesnoth-dev
--- Log closed Sun Feb 14 00:00:11 2016