--- Log opened Wed Nov 11 00:00:17 2009
--- Day changed Wed Nov 11 2009
20091111 00:00:17< fendrin> The new recruit/recall dialogue displays the unit that is selected like in a wml message the speaker. But what shall it display before one unit is selected? The unit that is the recaller/recruiter, the leader? Or just nothing or the wesnoth narrator icon or a still to made image?
20091111 00:01:17< Soliton> leader seems like a good idea.
20091111 00:02:39< fendrin> Okay, that is it then.
20091111 00:03:16< Ivanovic> i'd say narrator
20091111 00:03:35< Ivanovic> since the leader unit type might also be recruitable
20091111 00:04:39< fendrin> That case would save me some work. Easier to code.
20091111 00:05:33-!- noy [n=Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20091111 00:06:24< Ivanovic> when converting the recruit/recall dialog don't forget the "create unit" dialog (it probably uses the same backend, looks at least rather similar)
20091111 00:07:24< Ivanovic> heading off to bed now, n8
20091111 00:07:29< fendrin> Ivanovic: The create unit in debug mode is the only one that is already a gui2 dialog.
20091111 00:07:37< Ivanovic> ahh, okay
20091111 00:08:07< fendrin> But I can remove it and base the feature on top of the new wml_message class.
20091111 00:08:29< Soliton> we display the player name when recruiting now so we might as well show his unit. though why not just select the first unit like we do atm?
20091111 00:10:04< fendrin> Selecting the first unit is what the code does at the moment. Keeping that is fine as well. No idea why I found that behaviour strange. 
20091111 00:11:00< fendrin> Displaying the unit in the unit_box to the right of the gui works well. 
20091111 00:11:40< fendrin> But the jump for the eyes from the list to the unit box might be too huge.
20091111 00:11:59< Soliton> as long as the unit_box is there and visible.
20091111 00:12:36< fendrin> Soliton: A theme that removes that box but allows recruiting?
20091111 00:13:49< Soliton> what does a theme have to do with recruiting?
20091111 00:14:44< fendrin> the theme can remove the unit box und the campaign designer can use it in a scenario where recruiting is allowed, was that the usecase you worry about?
20091111 00:15:36< Soliton> i didn't have a specific use case in mind but sure that'd be one possibility.
20091111 00:16:20< fendrin> I can't think of another case where the unit box isn't visible but the player could need to recall or recruit.
20091111 00:18:18< Soliton> the point is that it seems like a bad idea to depend on some external GUI element in a dialog unless you are really sure that GUI element has to always be there and visible whenever you show the dialog.
20091111 00:19:00< fendrin> I never saw the unit box missing.
20091111 00:20:29< fendrin> By the way, if a unit has modified stats like better resistance, how can I see that? It can't, can it?
20091111 00:21:13< Soliton> mouse over the HP.
20091111 00:23:45< fendrin> I see. The values displayed in the help are still the units base values.
20091111 00:24:44< Soliton> sure, it's the help for the unit type.
20091111 00:26:32< fendrin> Okay, what about a extra dialoge that gives more informations about a unit than the help and the unit_box. It can be triggered from the recruit/recall dialoge and leaves the new dialogue small.
20091111 00:30:05< Soliton> assuming by extra information you mean information pertaining to the single unit instead of the unit type that'd be nice. no idea why that would be useful in the recruit dialog though.
20091111 00:31:02-!- teaser [n=tsr@h-37-115.A254.priv.bahnhof.se] has joined #WESNOTH-DEV
20091111 00:31:50-!- SonIcco [n=SonIcco@pD9512B68.dip0.t-ipconnect.de] has joined #wesnoth-dev
20091111 00:32:24< fendrin> Soliton: Right, it's only to solve the information gape in case the unit box is not visible. All type related information can be seen be pressing the profile button so it's only the recalling dialogue that needs some more powerfull slave dialogue.
20091111 01:01:54-!- noy [n=Noy@wesnoth/developer/noy] has quit ["GO, GET TO THE CHOPPAH!!!"]
20091111 01:02:39-!- grzywacz [n=grzywacz@chello089074205085.chello.pl] has joined #wesnoth-dev
20091111 01:11:48-!- boucman [n=rosen@wesnoth/developer/boucman] has quit ["Leaving."]
20091111 01:15:04-!- Tigge_ [n=tigge@c-5ba372d5.015-28-73746f34.cust.bredbandsbolaget.se] has joined #wesnoth-dev
20091111 01:20:00-!- Tigge [n=tigge@c-5ba372d5.015-28-73746f34.cust.bredbandsbolaget.se] has quit [Read error: 104 (Connection reset by peer)]
20091111 01:20:28-!- Tigge_ is now known as Tigge
20091111 01:25:55-!- mjs-de [n=mjs-de@vpw.wh.uni-dortmund.de] has quit [Remote closed the connection]
20091111 01:26:15-!- teaser [n=tsr@h-37-115.A254.priv.bahnhof.se] has quit ["Ex-Chat"]
20091111 01:30:45-!- SonIcco [n=SonIcco@pD9512B68.dip0.t-ipconnect.de] has quit [Remote closed the connection]
20091111 01:31:49-!- Chusslove [n=Chusslov@brsg-d9bef911.pool.mediaWays.net] has quit [Read error: 110 (Connection timed out)]
20091111 01:33:22-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has joined #wesnoth-dev
20091111 01:38:26-!- alink [n=alink@wesnoth/developer/alink] has quit [Remote closed the connection]
20091111 01:38:59-!- Chusslove [n=Chusslov@brsg-d9bef206.pool.mediaWays.net] has joined #wesnoth-dev
20091111 01:54:31-!- Deusite [n=Tom@92-238-120-246.cable.ubr26.newt.blueyonder.co.uk] has left #wesnoth-dev []
20091111 02:34:27-!- loonycyborg [n=sergey@wesnoth/developer/loonycyborg] has quit ["Zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz"]
20091111 02:39:09-!- stikonas [n=and@wesnoth/translator/stikonas] has quit [Remote closed the connection]
20091111 02:44:52-!- noy [n=Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20091111 02:46:04< noy> why do people continually suggest "new ways to distribute wesnoth"
20091111 02:47:43< shadowmaster> well, of course using 7z archives may be save more space than bzip2 tarballs - oh, User's Forum you mean, right.
20091111 02:47:54< noy> yeah
20091111 02:48:02< Zarel> o_O
20091111 02:48:20< shadowmaster> it sounds like a proposal for some commercial software vendor.
20091111 02:48:25< noy> it does
20091111 02:48:30< Zarel> "New ways" as in "Hey guys, .tar.bz2 has better compression ratio than .tar.gz!"
20091111 02:48:52< Zarel> Actually, the first thing I thought of was binary diffs. ;)
20091111 02:49:01< noy> and while I understand the Iphone issue brings up some conflicting issues, its the only way to distribute onto the Iphone
20091111 02:49:06< noy> that platform
20091111 02:50:27-!- ardesh [n=ardesh@port-92-195-6-248.dynamic.qsc.de] has quit [Read error: 110 (Connection timed out)]
20091111 02:50:46-!- ardesh [n=ardesh@port-92-195-68-100.dynamic.qsc.de] has joined #wesnoth-dev
20091111 02:50:49< Zarel> Hmm. Wesnoth on Impulse is fine, though.
20091111 02:50:56< noy> Uh I hate that like "I can facilitate it for you" 
20091111 02:51:03< noy> line*
20091111 02:51:07< Zarel> It's like a package management system, except for Windows. And with paid stuff on it, too.
20091111 02:51:30< noy> meh, whats wrong with our current distribution method
20091111 02:52:10< Zarel> Nothing. Just because, say, you offer the source on your website doesn't mean Ubuntu shouldn't put it in their repos.
20091111 02:52:12< Zarel> Same idea here
20091111 02:52:40-!- elynia [n=shyde@wesnoth/umc-dev/misc/elynia] has joined #wesnoth-dev
20091111 02:52:50< Aethaeryn> I agree with Zarel on the overall thing. I agree with noy that that "I can facilitate it for you" line isn't good.
20091111 02:53:03< Aethaeryn> If anyone does, it would be Ivanovic and/or the Windows packager.
20091111 02:53:22< Zarel> Though I agree it reeks of pretension, I've never bothered to be jugdmental about vocabulary, but to each their own.
20091111 02:53:35< shadowmaster> and I don't see what we'd get in exchange.
20091111 02:53:35< noy> Oh I agree, but at the same time to be honest I find the current distribution method gives us a manageable exposure 
20091111 02:53:47< shadowmaster> let's see: trolls? no thanks.
20091111 02:53:51< noy> Aethaeryn: or the secretary of wesnoth possibly?
20091111 02:54:00< Zarel> :P All so negative.
20091111 02:54:22< Zarel> It's not like they even need to ask. Wesnoth is GPL; you can't exactly stop anyone from redistributing.
20091111 02:54:25< Aethaeryn> noy: Well, maybe the initial message, but I'd assume the packager would handle the packaging. Like there's a Fedora packager, an Ubuntu packager, etc.
20091111 02:54:53< noy> Zarel: of course. 
20091111 02:55:21< Zarel> There's an Ubuntu packager?
20091111 02:55:30< Aethaeryn>  /s/Ubuntu/Debian
20091111 02:55:39< Aethaeryn> :P
20091111 02:55:46< Zarel> Now you're thinking with portals!
20091111 02:55:58< shadowmaster> there are Ubuntu packagers too AFAIK.
20091111 02:56:01< Zarel> s/portals/Debian and Ubuntu politics/
20091111 02:56:05< shadowmaster> they are just less involved with mainline.
20091111 02:56:28< Zarel> Aethaeryn: Also, your slashes are in the wrong places. ;)
20091111 02:56:55< shadowmaster> Aethaeryn doesn't know sed or Perl.
20091111 02:57:13< Aethaeryn> [s]Ubuntu[/s] Debian. ;)
20091111 02:57:21< Aethaeryn> phpBB for you
20091111 02:57:27< shadowmaster> Aethaeryn: I made that tag.
20091111 02:57:35< grzywacz> …
20091111 02:57:46< shadowmaster> grzywacz, hey!
20091111 02:57:50< Zarel> Well, um, I made more tags on my forum, so there!
20091111 02:58:04< shadowmaster> it's also called bbCode, not phpBB
20091111 02:58:12< grzywacz> shadowmaster, hey...
20091111 02:58:16< shadowmaster> <span style="text-decoration: line-through">{TEXT}</span>
20091111 02:58:30< shadowmaster> because IIRC the current XHTML DTD doesn't have the <strike> element.
20091111 02:58:42< Zarel> It has the <del> element, though.
20091111 03:01:01< Zarel> Incidentally, I learned s/something/something else/ in IRC way before I learned about regexes. Oh, the wonder of IRC.
20091111 03:01:55< shadowmaster> same here
20091111 03:02:42< Aethaeryn> shadowmaster: I know it's bbCode, but that's standard across more forums, most of which don't have [s]
20091111 03:04:27< shadowmaster> http://www.bbcode.org/reference.php
20091111 03:05:08< shadowmaster> it's so extensible that different software implementations or sites may use different codes and expansions
20091111 03:07:51< Aethaeryn> ok
20091111 03:07:55< Aethaeryn> Nevermind, then
20091111 03:09:17< shadowmaster> we also have some extra bbcodes which people don't even know they exist.
20091111 03:09:23< Aethaeryn> [wiki]
20091111 03:09:35< Espreon> Hmmmm...
20091111 03:09:42< shadowmaster> Aethaeryn: no, that one is publicly known
20091111 03:09:59< shadowmaster> I'm looking at the other bbCodes I added and deliberately set to not be displayed in the post editor.
20091111 03:10:08< Aethaeryn> Oh.
20091111 03:10:32< Aethaeryn> shadowmaster: What makes you think you'd get trolls, btw?
20091111 03:10:34< Zarel> Warzone has [rev], [s], [ticket], [wiki]
20091111 03:10:51< Aethaeryn> Obviously, any influx of users means some will be n00bs.
20091111 03:11:00< shadowmaster> Zarel: [ticket] is for your bug/issue tracker I guess?
20091111 03:11:01< Aethaeryn> But some will also be content creators, some will be ladder masters, etc.
20091111 03:11:03< Zarel> Yep.
20091111 03:11:28< shadowmaster> and [rev] is some link to a vcs commit viewed through some web interface?
20091111 03:11:35< Zarel> Trac for both.
20091111 03:11:47< Zarel> And the wiki, too. :P
20091111 03:12:00< Espreon> shadowmaster; whoever can deal with this: If I try to add the wiki's search plugin to my browser (Firefox 3.5.5), it says that it cannot download it.
20091111 03:12:24< shadowmaster> windows xp sp3 booting.
20091111 03:12:45< shadowmaster> (only operating system here that has FF 3.5.x)
20091111 03:13:27< Aethaeryn> < noy> Oh I agree, but at the same time to be honest I find the current distribution method gives us a manageable exposure <- Is that a no?
20091111 03:14:00< noy> that basically is a no
20091111 03:15:09< Aethaeryn> ok.
20091111 03:15:22< shadowmaster> Espreon: I'm seeing it.
20091111 03:15:31< Aethaeryn> I guess having it on Linux package managers is more controlled since it only is < 3% of the market.
20091111 03:15:39< Espreon> shadowmaster: Grand. :)
20091111 03:15:52< shadowmaster> but I cannot fix it.
20091111 03:16:02< Espreon> Damn it. :(
20091111 03:16:07< Espreon> But who can fix it?
20091111 03:16:08< shadowmaster> I have no idea what's wrong with it. Some MediaWiki silliness.
20091111 03:16:11< noy> Aethaeryn: we really don't need huge exposure like other games. We get high levels of contributions which we can manage. 
20091111 03:16:12< Espreon> I see...
20091111 03:17:25< shadowmaster> Espreon: let me try something.
20091111 03:18:31< shadowmaster> um, nope. No idea.
20091111 03:18:45< Aethaeryn> noy: What do you mean by "high level of contributions"?
20091111 03:19:11< noy> We get really good exposure and draw great talent
20091111 03:19:16< noy> as is
20091111 03:39:02< noy> Aethaeryn: thanks for the post
20091111 03:39:06< noy> I also replied
20091111 03:39:38< Aethaeryn> noy: Yeah, I didn't want to reply with a no because I'm not a dev.
20091111 03:39:55< Aethaeryn> But I did want to not lead him into thinking he should do it himself.
20091111 03:40:07< noy> oh thats for sure
20091111 03:40:57< noy> ugh, its kinda a piss off when someone suggests something like that, particularly that individual
20091111 03:41:13< noy> I mean who egged him on
20091111 04:04:26-!- Zarel [n=Zarel@warzone2100/developer/Zarel] has quit ["This computer has gone to sleep"]
20091111 04:05:48-!- shadowmaster is now known as SpinachMaster
20091111 04:09:19-!- SpinachMaster is now known as shadowmaster
20091111 04:09:37-!- elynia is now known as PinkUnicorn
20091111 04:28:13-!- PinkUnicorn is now known as elynia
20091111 04:29:22-!- elynia [n=shyde@wesnoth/umc-dev/misc/elynia] has quit ["nyu"]
20091111 04:35:47-!- Ivanovic_ [n=ivanovic@dtmd-4db2b22a.pool.mediaWays.net] has joined #wesnoth-dev
20091111 04:36:33< shadowmaster> wesbot: tell me sapient's last host.
20091111 04:36:51< shadowmaster> oh wait, it must be cloaked. never mind.
20091111 04:43:53< Aethaeryn> Why is $unit.name being used in strings?
20091111 04:44:22< shadowmaster> unit.name is translated.
20091111 04:44:34< Aethaeryn> Is it?
20091111 04:44:49< shadowmaster> it's the name displayed to users, not the internal name (unit.id).
20091111 04:45:12< Aethaeryn> So if the unit name is nominative (it is), and the sentence it is in demands it as an indirect object, it automatically converts it to dative in Latin?
20091111 04:45:17< Aethaeryn> Wow, what magic that is.
20091111 04:45:36< shadowmaster> well, yeah, that fails.
20091111 04:45:56< shadowmaster> there's a forum post about why using variable expansion in translatable strings may be a problem for certain languages.
20091111 04:46:09< shadowmaster> good that e.g. Spanish doesn't suffer from that.
20091111 04:46:27< Espreon> :)
20091111 04:46:46< Aethaeryn> Yes, but it's bad practice for a reason. In Latin, word order does not matter. So "Bob goes to the store." and The store goes to Bob.
20091111 04:47:11< Aethaeryn> Both could mean the same thing, if the endings are the same between the sentences.
20091111 04:48:44< Aethaeryn> In other words, I have to translate "$unit.name" as "hic actor" (literally "this unit"), thus killing any benefit $unit.name might have had.
20091111 04:48:53-!- Blueblaze [n=nick@adsl-99-158-46-90.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20091111 04:50:45< Espreon> Then a goal for 1.9/1.10 shold be to purge these uses of variables.
20091111 04:51:02< Espreon> ... if realistically possible...
20091111 04:51:55-!- Ivanovic [n=ivanovic@wesnoth/developer/ivanovic] has quit [Read error: 113 (No route to host)]
20091111 04:53:45-!- Ivanovic_ is now known as Ivanovic
20091111 05:25:21-!- Espreon [n=espreon@wesnoth/developer/espreon] has quit ["WRYYYYYYYYYYYYYYYYYYYY!"]
20091111 05:28:33-!- noy [n=Noy@wesnoth/developer/noy] has quit []
20091111 05:31:36-!- Zarel [n=Zarel@nat-portal-160-94-47-16.uofm.wireless.umn.edu] has joined #wesnoth-dev
20091111 05:54:32-!- Blueblaze [n=nick@adsl-99-158-46-90.dsl.hstntx.sbcglobal.net] has quit [Remote closed the connection]
20091111 06:10:44-!- Chusslove [n=Chusslov@brsg-d9bef206.pool.mediaWays.net] has quit [Read error: 110 (Connection timed out)]
20091111 06:17:38-!- Chusslove [n=Chusslov@brsg-d9bee374.pool.mediaWays.net] has joined #wesnoth-dev
20091111 06:21:15-!- wesbot changed the topic of #wesnoth-dev to: string/feature freeze active! | 69 bugs, 246 feature requests, 10 patches | logs: http://irclogs.wesnoth.org/ | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com | http://imagebin.org
20091111 06:49:44-!- lizard_r [n=Miranda@wesnoth/umc-dev/developer/lizard] has joined #wesnoth-dev
20091111 06:50:03-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has quit []
20091111 06:52:17-!- lizard_r [n=Miranda@wesnoth/umc-dev/developer/lizard] has quit [Client Quit]
20091111 07:48:54< Aethaeryn> esr: http://forums.wesnoth.org/viewtopic.php?f=11&t=27738 <- someone stealing Wesnoth images in a game in which they're selling a "full version" for $30
20091111 07:49:33< Aethaeryn> Since you know more about licensing than anyone else, perhaps you could give your two cents.
20091111 07:50:27< esr> I'll try to find time to reply.
20091111 07:51:45< Aethaeryn> Okay, ty.
20091111 07:52:32< Aethaeryn> I hate it when people put 0 effort in something and try to make money off of it
20091111 07:58:57-!- zookeeper [n=l@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20091111 08:05:30-!- Vetinari [n=lukjad00@unaffiliated/lukjad007] has quit [Read error: 110 (Connection timed out)]
20091111 08:05:56-!- lukjad007 [n=lukjadOO@unaffiliated/lukjad007] has quit [Read error: 110 (Connection timed out)]
20091111 08:07:52-!- silene [n=plouf@83.197.60.28] has joined #wesnoth-dev
20091111 08:14:11-!- Netsplit leguin.freenode.net <-> irc.freenode.net quits: _teddy, Aethaeryn
20091111 08:15:52-!- Netsplit over, joins: Aethaeryn
20091111 08:23:11-!- noy [n=Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20091111 08:45:02-!- EdB [n=edb@79.88.117.18] has joined #wesnoth-dev
20091111 09:07:28-!- Aethaeryn [n=Michael@wesnoth/umc-dev/developer/aethaeryn] has quit ["Exit."]
20091111 09:35:47-!- Zarel [n=Zarel@warzone2100/developer/Zarel] has quit ["This computer has gone to sleep"]
20091111 09:37:47-!- boucman [n=rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20091111 09:37:59< boucman> hey all
20091111 09:47:40< fendrin> hi boucman
20091111 09:48:27-!- Zarel [n=Zarel@nat-portal-160-94-47-16.uofm.wireless.umn.edu] has joined #wesnoth-dev
20091111 10:04:30-!- silene [n=plouf@wesnoth/developer/silene] has quit ["Leaving."]
20091111 10:10:59-!- dtiger [n=dtiger@fe26.hc.ru] has joined #wesnoth-dev
20091111 10:20:27-!- noy [n=Noy@wesnoth/developer/noy] has quit []
20091111 10:22:56-!- Crab_ [i=crab@wesnoth/developer/crab] has joined #wesnoth-dev
20091111 10:23:52< boucman> hey Crab_
20091111 10:23:57< Crab_> hi boucman
20091111 10:24:07< boucman> the AI seems to stabilize at 25/66/8.5
20091111 10:24:23< boucman> so definitely a huge improvement
20091111 10:24:43< Crab_> yes
20091111 10:25:04< Zarel> ?
20091111 10:25:32< boucman> Zarel: we are testing some AI improvements, and one of them seems really good
20091111 10:25:43< Zarel> Can you tell me what the numbers mean?
20091111 10:25:50< Crab_> boucman: alink's patch is for candidate action, but it can be ported to default_ai easily (now it is 2-line change)
20091111 10:25:56< Crab_> Zarel: see http://wesnoth.terraninfo.net/index.php?dim1=ai_config_me&latest_svn_only=yes
20091111 10:26:24< Zarel> What's the win percentage against?
20091111 10:26:31< boucman> the other AI
20091111 10:26:39< Zarel> Oh, I see.
20091111 10:26:45< Crab_> Zarel: we are testing 'rca ai' vs 'rca ai in which targeting phase is patched by alink'.
20091111 10:26:49< boucman> Crab_: how far are we of making RCA the default AI
20091111 10:27:22< Crab_> boucman: should be doable by simple switch in default ai config. I can test the result.
20091111 10:28:04< boucman> Crab_: if you consider it a good idea...
20091111 10:28:40< Crab_> I think that there's one minor thing that need to be fixed (goto phase, which I haven't implemented yet - it's used in a couple of scenarios)
20091111 10:28:49< Crab_> yes, it is worth trying to do
20091111 10:29:02< boucman> Crab_: doable before 1.8 ?
20091111 10:29:12< Crab_> of course
20091111 10:29:18< Crab_> goto phase is ~30 lines of code
20091111 10:29:45< boucman> cool
20091111 10:30:49< Crab_> the main advantages of rca-based ai are absence of certain classes of bugs and easier extension (it's a lot easier to add a new candidate action via [modify_ai] than to plug in something into default_ai - since to plug in something into default ai, it is needed to put it either *before* or *after* the entire ai (with the exception of recruitment) )
20091111 10:32:01< Crab_> and, with rca-based ai, it's possible to add a new candidate action with a exact priority wanted, and, it'll be dynamic - i.e. if the new candidate action is 'attack using poisoners', it will fire not only once, but each time such an opportunity arises
20091111 10:32:05< boucman> Crab_: yes, the question is more about RCA being ready
20091111 10:32:59< Crab_> well, it is ready with the exception of goto_phase, and is on par with default ai in efficiency terms.
20091111 10:33:54< boucman> ok, so it's good for porting once we have goto phase
20091111 10:35:24< Crab_> yes
20091111 10:39:15< Crab_> and, later, I will also need to optimize rca's speed - it can be done if we set an prior upper bound on each candidate action score.
20091111 10:40:19< Crab_> then, it will be possible to skip the evaluation of those candidate actions which 'certainly will not be selected for execution' on this evaluation run.
20091111 10:41:05< boucman> hmm
20091111 10:41:11< boucman> not sure how you envision that...
20091111 10:41:30< Crab_> i.e. if goto_phase evaluated to 200, and all other candidate actions have less 'maximum score' than 200, then it is not needed to evaluate other candidate actions on this run.
20091111 10:41:42< boucman> you need to evaluate to know you're not going to select...
20091111 10:41:54< boucman> aaah
20091111 10:42:27< boucman> add an optional "maximum score" attribute to CA to avoid calculating them
20091111 10:42:29< boucman> makes sense
20091111 10:42:30< Crab_> yes
20091111 10:43:12< Crab_> this will make rca speed even 'slightly better' than default_ai speed.
20091111 10:43:37< Crab_> (rca avoids some recomputations due to its structure)
20091111 10:43:56< Crab_> and, currently, phases scores are static, so it'll work for all of them.
20091111 10:45:14< boucman> sounds good
20091111 10:47:14-!- Zarel [n=Zarel@warzone2100/developer/Zarel] has quit ["This computer has gone to sleep"]
20091111 11:03:33-!- ardesh [n=ardesh@port-92-195-68-100.dynamic.qsc.de] has quit [Read error: 60 (Operation timed out)]
20091111 11:10:59-!- ardesh [n=ardesh@port-92-195-68-100.dynamic.qsc.de] has joined #wesnoth-dev
20091111 11:34:14-!- ilor [n=ilor@wesnoth/developer/ilor] has joined #wesnoth-dev
20091111 11:37:19< CIA-55> crab * r39677 /trunk/src/ai/testing/ (ca.cpp ca.hpp): implemented candidate action: 'testing_ai_default::goto_phase'
20091111 11:44:49-!- lukjad007 [n=lukjadOO@unaffiliated/lukjad007] has joined #wesnoth-dev
20091111 12:09:18-!- loonycyborg [n=sergey@79.139.138.110] has joined #wesnoth-dev
20091111 12:12:10< dtiger> since last few 1.7.x versions AI leader never attacks at his turn. Only recruits or just stay out.
20091111 12:12:23< boucman> Crab_: ^^^
20091111 12:12:50< boucman> I think this had already been reported and might have been fixed in 1.7.8, did you check with yesterday's release ?
20091111 12:13:12< dtiger> yes, 1.7.8 too
20091111 12:13:14< Crab_> dtiger: can you give a savegame or how-to-reproduce ?
20091111 12:13:41< Crab_> ( for example, via PM to Crab on wesnoth forums )
20091111 12:13:50< boucman> or dl.free.fr
20091111 12:14:35< dtiger> no problem to reproduce - just start multiplayer with AI and move your units to AI leader :)
20091111 12:16:07< Crab_> ok, will try
20091111 12:24:58-!- fendrin [n=fabi@wesnoth/developer/fendrin] has quit [Remote closed the connection]
20091111 12:26:58-!- mjs-de [n=mjs-de@vpw.wh.uni-dortmund.de] has joined #wesnoth-dev
20091111 12:31:07< Crab_> ok, tried. default ai has the bug, rca ai has not.
20091111 12:32:32< dtiger> rca ai? where can i get it?
20091111 12:33:22< Crab_> dtiger: run wesnoth with -d flag
20091111 12:33:31< dtiger> ok
20091111 12:33:59< Crab_> try 'RCA ai (experimental targeting)'
20091111 12:35:24< Crab_> (it's mp-only)
20091111 12:35:43< Crab_> (since it's not enabled in campaign scenarios)
20091111 12:36:02< boucman> Crab_: ok, so either make rca default for 1.8 or fix in default ai
20091111 12:36:13< Crab_> or both :)
20091111 12:36:43< dtiger> ok, started. Where can I read description of all available AIs?
20091111 12:37:16< boucman> there is no such description, that's why it's debug only :)
20091111 12:37:21< Crab_> 'default ai' is default, 'rca ai' is slightly modernized default
20091111 12:37:30< dtiger> i see :)
20091111 12:37:32< dtiger> thanks
20091111 12:38:10< Crab_> 'RCA ai (experimental targeting)' contains a small bugfix, and wins against other AIs 2/3 of the time (on large enough maps)
20091111 12:39:51< boucman> Crab_: we won't keep the current default AI once RCA has been promoted as default, will we ?
20091111 12:41:00< Crab_> once we're certain that all the 'features' are accounted for, and if the speed is good enough, there'll be no reason to keep it (since rca is based on the forked code of default)
20091111 12:41:45< boucman> ok, so we'll have an "old ai" and a "default ai" available to players in 1.8 ?
20091111 12:43:45< Crab_> boucman: we'll see. optimally, only the 'rca ai as default'
20091111 12:43:53< Crab_> ok, I've tracked down the bug
20091111 12:43:57< boucman> k
20091111 12:45:50< Crab_> new_turn; consider_combat -> true; combat phase: not fighting with leader because recruiting is more preferable;  combat phase: returns false; consider_combat -> false; recruit; skip combat phase in later do_move evaluations; no attack with leader even after recruitment.
20091111 12:46:53< Crab_> no wonder rca is not affected - it is not vulnerable to such classes of bugs because there's no need for recruiting_preferred_ and consider_combat_ hacks - turn sequence is truly dynamic
20091111 12:47:45< Crab_> now searching for a way to fix the bug without introducing an infinite loop....
20091111 12:51:21< Crab_> ok, compiling a candidate fix..
20091111 12:53:03-!- ardesh [n=ardesh@port-92-195-68-100.dynamic.qsc.de] has quit [Connection timed out]
20091111 13:12:19< CIA-55> crab * r39678 /trunk/src/ai/default/ai.cpp: fixed a bug with ai_default leader sometimes not attacking even if it is advantageous to do so
20091111 13:12:21< Crab_> dtiger: ^ fixed. thanks for reporting.
20091111 13:12:26< CIA-55> crab * r39679 /trunk/ (5 files in 3 dirs): important speed optimization of candidate action evaluation loop - ending candidate action evaluation loop if current score is greater-or-equal than the upper bound of score for remaining candidate actions
20091111 13:12:58< Crab_> boucman: ^ done. shall recompile on batch testing server to test the game duration change.
20091111 13:13:13< boucman> k
20091111 13:13:35< boucman> does it really work as you described in the commitlog ?
20091111 13:13:44< boucman> i.e stop the loop when allremaining are lower
20091111 13:13:55< Crab_> yes. I tested it locally.
20091111 13:14:06< boucman> or does it work by skipping any that has a max value lower than the current score ?
20091111 13:14:19< Crab_> it sorts them by max_score DESC
20091111 13:14:54< Crab_> so, if next CA has max_score<=current_score, we break out of the loop;
20091111 13:15:13< boucman> ok, I didn't know about the sorting
20091111 13:15:25< Crab_> tiny implementation detail
20091111 13:19:16< Crab_> final results of previous test run: http://aitest.wesnoth.org/index.php?dim1=ai_config_me&filter_svn_release=39651
20091111 13:21:17-!- yamekidzu-it [n=rainy-it@host146-75-dynamic.16-79-r.retail.telecomitalia.it] has joined #wesnoth-dev
20091111 13:23:30< boucman> ok, so 8.3 second per turn....
20091111 13:26:04< Crab_> it would be interesting to see how well "rca+alink's targeting patch+dk's recruitment" will fare against default_ai...
20091111 13:26:20< boucman> hehe
20091111 13:26:39< boucman> do you have dk's recruitement in a working state to test that ?
20091111 13:29:24< Crab_> all I know is that DKs recruitment works ok on some maps.
20091111 13:29:32-!- EdB [n=edb@79.88.117.18] has quit [Remote closed the connection]
20091111 13:29:33< Crab_> and it's easy enough to integrate it into any ai.
20091111 13:30:05< boucman> dk is connected, we could ping him now if you want
20091111 13:30:23< Crab_> no,it's easy enough to pick out a map where it works
20091111 13:30:59< boucman> Crab_: how do we know what mp dk's recruitment work on ? would it be possible to have a test for it somehow ?
20091111 13:31:48< Crab_> well, it either works or not (errors), it's evident on ai turn 1, and can be easily seen by launching the map in MP.
20091111 13:32:11< boucman> hmm
20091111 13:32:13< boucman> ok
20091111 13:32:35< boucman> so we can test it on any ai but not really integrate into default AI yet
20091111 13:32:43< Crab_> (I guess a NULL creeps in during some kind of map analysis on some maps, but I'm not sure exactly)
20091111 13:32:48< Crab_> yes
20091111 13:32:58< boucman> deekay: around ?
20091111 13:33:27< Crab_> and, {AI_SIMPLE_FORMULA_AI_EXPERIMENTAL_RECRUITMENT} macro can be used to enable it for any side in any scenario. in fact, i use it in LoW 2
20091111 13:36:31-!- stikonas [n=and@wesnoth/translator/stikonas] has joined #wesnoth-dev
20091111 13:40:11< deekay> boucman: Sort of
20091111 13:40:54< boucman> deekay: do you konw what maps your recruitement formula works on and which it doesnt, 
20091111 13:41:16< boucman> the mid/long term goal is to integrate it to the mainline AI, if possible for 1.8
20091111 13:41:40< deekay> boucman: Last time when I checked there was a problem with maps that have some statues (can't remember wheter I fixed that or not) but there was alos some more serious problem
20091111 13:42:12< boucman> ok
20091111 13:42:22< deekay> boucman: I don't know how it is right now, but Foruma AI variables were not saved to a savefile
20091111 13:42:34< deekay> Thus breaking my recruitment upon reloading
20091111 13:42:41< boucman> foruma ?
20091111 13:43:05< deekay> formula
20091111 13:43:11< deekay> typo :)
20091111 13:43:19< boucman> :)
20091111 13:43:42< deekay> I don't have too much time lately, work + unit are consuming my time greately
20091111 13:43:43< boucman> Crab_: did you know about that ?
20091111 13:44:23< deekay> I think I mentioned it some time ago, but it is easy to forget that (I myself forgot till now when you mention recruitment)
20091111 13:45:20< Crab_> no, I don't know about the exact issues. 
20091111 13:45:46< Crab_> save/loading of formula_ai works via src/ai/composite/engine_fai.cpp
20091111 13:46:05< deekay> Formula AI variables are issue IIRC
20091111 13:46:43< deekay> Last time what we did lack was saving to/loading form config while saving/loading the game
20091111 13:46:47< Crab_> fai_engine has, in constructor : formula_ai_(context,cfg.child_or_empty("formula_ai")   and, in to_config,    cfg.add_child("formula_ai",formula_ai_.to_config())
20091111 13:47:53< Crab_> and formula_ai::to_config has no proper handling in save/load
20091111 13:48:49< Crab_> deekay: where are those variables stored ?
20091111 13:49:12< Crab_> game_logic::map_formula_callable vars_ ?
20091111 13:50:07< deekay> Crab_: Yes
20091111 13:51:23< Crab_> then it should be fairly easy to serialize/deserialize them (there's sample code in unit.cpp )
20091111 13:52:14< Crab_> (I can add such serialization/deserialization)
20091111 13:55:30< deekay> Crab_: Yes, code from unit.cpp is definiately what we need, if you can, please do so. :)
20091111 13:55:38< Crab_> ok
20091111 13:57:27< deekay> I am currently leader of the project group at my uni (we are supposed to write e-bay-like portal in like 1.5 month), have to participate in 4 other projects at my uni plus I'm working as a contractor... that does not help with free time, especially when I want to have some for myself...
20091111 13:57:48< deekay> So I can't guarantee I that will be really helpful now :/
20091111 13:57:59< boucman> deekay: that's fine, we all have ups and downs with wesnoth...
20091111 13:58:20< boucman> but if we can get as mucha s possible in place for when you come back, that's good
20091111 13:58:27< deekay> I do hope about some more time after december..
20091111 14:02:23< Crab_> (test run #2 started, testing speed optimization of r39679)
20091111 14:05:28-!- lukjad007 [n=lukjadOO@unaffiliated/lukjad007] has quit [Read error: 110 (Connection timed out)]
20091111 14:06:38-!- _teddy [n=fedor76@ppp-78-24-25-8-bras0.istra.ru] has joined #wesnoth-dev
20091111 14:28:54< boucman> Crab_: it's only 11 games, (so not really significant) but turn duration is aprox 5s now
20091111 14:30:08< Crab_> that's around -40%
20091111 14:32:15< Crab_> I'll try replacing default_ai by rca ai and playtesting one of the campaigns...
20091111 14:40:11< boucman> Crab_: should we state on the 1.8beta1 release thread that LoW is a showcase of the new AI and that we are particularly interested in feedback ?
20091111 14:40:35< Crab_> not yet.
20091111 14:42:52< boucman> you sure ? it's not good enough for beta testing ?
20091111 14:43:11< boucman> if the next release is the final, we won't have any more chance for testing???
20091111 14:44:26< Crab_> sure. it's not 'broad enough' for beta testing, there's still work to do. and, I doubt that the next release will be final one.
20091111 14:46:10< boucman> k
20091111 14:47:56-!- yamekidzu-it [n=rainy-it@host146-75-dynamic.16-79-r.retail.telecomitalia.it] has quit [Read error: 113 (No route to host)]
20091111 15:16:32< boucman> Crab_: when looking at the AI testing stats, i'm a bit suprised that both AI have exactly the same average time per turn... is the time per turn simply calculated by dividing the time of the match by the number of turn, or is it the time taken by that particular AI computed only with the time for that particular AI ?
20091111 15:18:09< Crab_> it is 'game duration'
20091111 15:18:27< Crab_> so, time per turn is 'dividing the time of the match by the number of turn'
20091111 15:19:14< boucman> ok
20091111 15:41:09-!- dtiger [n=dtiger@fe26.hc.ru] has quit ["Konversation terminated!"]
20091111 15:59:02-!- Sapient [n=patrickp@wesnoth/developer/sapient] has joined #wesnoth-dev
20091111 16:03:26-!- Crab_ [i=crab@wesnoth/developer/crab] has quit ["Leaving."]
20091111 16:18:08-!- crimson_penguin [n=ben@64.201.60.211] has joined #wesnoth-dev
20091111 16:24:11-!- fendrin [n=fabi@88-134-116-206-dynip.superkabel.de] has joined #wesnoth-dev
20091111 16:27:45< fendrin> hello
20091111 16:31:43-!- dtiger [n=dtiger@dynamic-vpdn-91-149-132-222.telecom.by] has joined #wesnoth-dev
20091111 16:32:44< fendrin> Sapient: You suggested to have more leaders that can recruit per side?
20091111 16:33:17< Sapient> no, I just pointed out that it's a fairly frequent request
20091111 16:33:26< Sapient> in the WML Workshop
20091111 16:33:46< fendrin> It would be nice to have recruit_lists for each as well.
20091111 16:33:53< Sapient> so if the feature were added, it would be used
20091111 16:34:15< fendrin> I will make sure that the part of the code I redo now is prepared for it.
20091111 16:34:31< fendrin> But I have a problem with the user interface. 
20091111 16:35:20< fendrin> What if a player has more than one leader in a keep hex. Which one's recruit dialog shall be displayed?
20091111 16:35:47< boucman> when right clicking in an empty hex ?
20091111 16:35:51< boucman> who cares..
20091111 16:36:13< boucman> the list would be a merge of all leader's list, but we don't really care which one recruits in case both provide a unit
20091111 16:37:12< fendrin> When right clicking in an empty hex the recruit_lists of all leaders that are in a keep in that castle can be merged.
20091111 16:37:45< fendrin> But what if the player opens the dialogue per hotkey usage.
20091111 16:37:49< fendrin> ?
20091111 16:38:18< boucman> hmm
20091111 16:38:27-!- Blueblaze [n=nick@adsl-99-158-46-90.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20091111 16:38:35< boucman> take a random/the first leader in a keep, I guess
20091111 16:38:43< boucman> there is no "good" way to solve that afaict
20091111 16:39:19< fendrin> I could display a dialogue that let's the user choice the recruiting leader.
20091111 16:48:48-!- noy [n=Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20091111 16:52:50-!- Zarel [n=Zarel@nat-portal-160-94-47-16.uofm.wireless.umn.edu] has joined #wesnoth-dev
20091111 16:56:53-!- Zarel [n=Zarel@warzone2100/developer/Zarel] has quit [Client Quit]
20091111 17:00:36-!- noy [n=Noy@wesnoth/developer/noy] has quit ["GO, GET TO THE CHOPPAH!!!"]
20091111 17:05:01-!- zookeeper [n=l@wesnoth/developer/zookeeper] has quit [Read error: 145 (Connection timed out)]
20091111 17:06:42-!- zookeeper [n=l@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20091111 17:08:18-!- Zarel [n=Zarel@nat-portal-160-94-47-16.uofm.wireless.umn.edu] has joined #wesnoth-dev
20091111 17:17:48< grzywacz> There's something wrong with description display: http://units.wesnoth.org/trunk/pl_PL/Spectre.html
20091111 17:25:15-!- zookeeper2 [n=l@88-148-251-223.bb.dnainternet.fi] has joined #wesnoth-dev
20091111 17:26:41-!- yamokidzu-it [n=rainy-it@host240-67-dynamic.24-79-r.retail.telecomitalia.it] has joined #wesnoth-dev
20091111 17:26:43-!- zookeeper [n=l@wesnoth/developer/zookeeper] has quit [Read error: 145 (Connection timed out)]
20091111 17:34:05< AI0867> grzywacz: the backslash or something else?
20091111 17:35:06< AI0867> great, html injection
20091111 17:35:24< AI0867> is \" the proper way to escape in PO files?
20091111 17:37:03< grzywacz> I think so.
20091111 17:38:29< Ivanovic> hi
20091111 17:41:19-!- Aethaeryn [n=Michael@wesnoth/umc-dev/developer/aethaeryn] has joined #Wesnoth-dev
20091111 17:41:46< fendrin> Can't find the code section that scrolls to a speaking unit. It's cursed.
20091111 17:42:56-!- yamokidzu-it [n=rainy-it@host240-67-dynamic.24-79-r.retail.telecomitalia.it] has quit [Connection timed out]
20091111 17:42:59< Aethaeryn> Zarel: You're right. tar.bz2 compresses better than tar.gz. (I just reread logs on the Impulse discussion from last night; I must have missed that comment.)
20091111 17:43:11< Zarel> LOL
20091111 17:43:32< Aethaeryn> Yes, that's the line that stuck with me. :P
20091111 17:43:39< Zarel> It's true, but it's also less supported. If you wanted ridiculously high compression ratios and don't care about support, RAR and 7Z work fine
20091111 17:44:07< Aethaeryn> It's true that its compression is better though: I've used tar.bz2 every time I cared about space. Every time I cared about support, .zip.
20091111 17:44:34< Aethaeryn> RAR I have issues with. Support ain't that great in the FOSS universe with it.
20091111 17:45:26< Aethaeryn> 7z, never tried it tbh.
20091111 17:46:13< Aethaeryn> Still, when we're talking about save files, wouldn't .tar.bz2 be better? I know from experience that it's better on plain text than .tar.gz, and if you have to install a dependency to make it, anyway...
20091111 17:47:27-!- dtiger [n=dtiger@dynamic-vpdn-91-149-132-222.telecom.by] has quit [Remote closed the connection]
20091111 17:48:13-!- dtiger [n=dtiger@dynamic-vpdn-91-149-132-222.telecom.by] has joined #wesnoth-dev
20091111 17:51:18< Ivanovic> Aethaeryn: why tar?
20091111 17:51:23< Ivanovic> tar is only needed to bundle things...
20091111 17:51:40< Aethaeryn> Ivanovic: Good catch. :P
20091111 17:51:54< Ivanovic> though for normal savegames the gzip compression is perfectly fine and enough, no need to make it even more difficult for some windows machines
20091111 17:51:55< Ivanovic> ;)
20091111 17:52:05< Aethaeryn> I've never noticed the savegames were just .gz
20091111 17:52:18< Aethaeryn> I'm so used to treating ".tar.gz" as if it were ".targz"
20091111 17:53:31< Aethaeryn> And bah, being cross-platform with Windows makes Wesnoth inferior in terms of efficiency to being written solely for Linux.
20091111 17:53:39< Aethaeryn> Why can't Microsoft go bankrupt already? ;)
20091111 17:54:37< Ivanovic> boucman, crab_: we could continue shipping both ais calling the current "default ai" something like "stupid ai"
20091111 17:54:58< boucman> Ivanovic: hmm
20091111 17:55:01< Ivanovic> no, it does not make it inferior in case of speed and the likes
20091111 17:55:19< boucman> if it's in 1.8 we need to keep it until 1.10
20091111 17:55:38< Ivanovic> the only thing is that on windows some stuff is *a lot* slower (eg loading/saving)
20091111 17:55:46< Ivanovic> that is at least for some users this is the case
20091111 17:55:50< Ivanovic> boucman: uhm, why?
20091111 17:56:21< boucman> because (IIRC) a scenario can specify the AI it wants to use 
20091111 17:56:36< boucman> so for compatibility throughout the 1.8 serie, we can't remove any AI
20091111 17:56:53< Ivanovic> boucman: then we can't remove it now anyway (and not change the ID either!)
20091111 17:57:05< Ivanovic> since we are now in compatibility mode due to the restart of the addon server!
20091111 17:57:13-!- Zarel [n=Zarel@warzone2100/developer/Zarel] has quit ["This computer has gone to sleep"]
20091111 17:57:20< boucman> well, currently there is only the default ai
20091111 17:57:23< Ivanovic> but once 1.8 is out we can directly drop the old ai in trunk
20091111 17:57:34< boucman> (the other ones are available only in debug mode, and thus not in MP
20091111 17:57:58< Ivanovic> i still don't see the argument what leaving the "old ai" in 1.8 has to do with 1.10
20091111 17:58:12< Ivanovic> we can remove it while we are in 1.9.x times, there is no problem
20091111 17:58:23< Ivanovic> sure, branches/1.8 still has that ai, but who cares?
20091111 17:58:32< boucman> Ivanovic: that's what Imeant, sorry for unclarity
20091111 17:58:58< Ivanovic> and we are free to label all bug reports for "old ai" as "won't fix"
20091111 17:59:00< Ivanovic> ;)
20091111 18:09:29-!- zookeeper2 is now known as zookeeper
20091111 18:21:15-!- wesbot changed the topic of #wesnoth-dev to: string/feature freeze active! | 68 bugs, 246 feature requests, 10 patches | logs: http://irclogs.wesnoth.org/ | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com | http://imagebin.org
20091111 18:28:56-!- yann [n=dwitch@nan92-1-81-57-214-146.fbx.proxad.net] has quit [Read error: 110 (Connection timed out)]
20091111 18:41:31< Aethaeryn> Ivanovic: did you get my translation email?
20091111 18:41:46< Ivanovic> yeah, had no time so far to commit it, about to do it
20091111 18:42:00< Aethaeryn> ok
20091111 18:43:03< Aethaeryn> It's ok, I translated 19% of wesnoth-manual while waiting
20091111 18:45:27< Aethaeryn> If you want me to send that in too that's okay. If you're already commiting it that's okay too.
20091111 18:45:38< Aethaeryn> No rush to complete it since 1.8 is still a ways off
20091111 18:47:19< CIA-55> ivanovic * r39680 /trunk/ (6 files in 5 dirs): updated Latin translation
20091111 18:47:47-!- yann [n=dwitch@nan92-1-81-57-214-146.fbx.proxad.net] has joined #wesnoth-dev
20091111 18:47:54< Ivanovic> the manual only makes a difference if at least 80% are done
20091111 18:48:06< Ivanovic> unles taht is the case the manual will not be created (the same for the manpages)
20091111 18:48:44< Aethaeryn> ah, well, most of what I was doing was transferring already translated phrases over.
20091111 18:48:49< Aethaeryn> caught two typos that way.
20091111 18:53:45< Aethaeryn> Ivanovic: do you mean 80% of core or 80% of total, btw?
20091111 18:53:49< Aethaeryn> that's a big difference in 80%
20091111 18:58:07< Aethaeryn> A difference between 22 meeting the qualifications and 13
20091111 18:59:22< Ivanovic> 80% of "wesnoth-manual"
20091111 18:59:30< Ivanovic> the same for the manpages
20091111 19:01:50< Aethaeryn> oh
20091111 19:03:29-!- SonIcco [n=SonIcco@pD9512DD1.dip0.t-ipconnect.de] has joined #wesnoth-dev
20091111 19:09:30< CIA-55> ivanovic * r39681 /trunk/po/ (11 files in 11 dirs): updated Polish translation
20091111 19:09:45< CIA-55> fendrin * r39682 /branches/fendrin_gui_stuff/: Branch for the work on the gui2 recall/recruit dialogue.
20091111 19:11:57< CIA-55> ivanovic * r39683 /trunk/doc/man/Makefile.am: since nonexisting manpages should be properly ignored these days, just list all the ones also available in the LINGUAS file, so that all manpages will be installed at make install time
20091111 19:13:57-!- ilor_ [n=ilor@wesnoth/developer/ilor] has joined #wesnoth-dev
20091111 19:15:44< CIA-55> fendrin * r39684 /branches/fendrin_gui_stuff/ (11 files in 5 dirs): Initial commit for the new gui2 recruit/recall dialogue. Working prototype.
20091111 19:17:08-!- Zarel [n=Zarel@nat-portal-160-94-47-16.uofm.wireless.umn.edu] has joined #wesnoth-dev
20091111 19:17:46< Zarel> Aethaeryn: I was talking about source when I said .tar.bz2, so I'm innocent. ;)
20091111 19:31:20-!- ilor [n=ilor@wesnoth/developer/ilor] has quit [Read error: 110 (Connection timed out)]
20091111 19:37:30< Aethaeryn> Maybe within a week, Latin will hit 50% core packages translated.
20091111 19:38:21-!- silene [n=plouf@AMarseille-251-1-54-28.w83-197.abo.wanadoo.fr] has joined #wesnoth-dev
20091111 19:42:03-!- mordante [n=mordante@wesnoth/developer/mordante] has joined #wesnoth-dev
20091111 19:42:16< mordante> servux
20091111 19:42:18< mordante> servus*
20091111 19:46:16< fendrin> hi mordante
20091111 19:46:55< mordante> Aethaeryn, you can use Wesnoth to unpack .gz files... of course command line and Windows is not the greatest combination ;-)
20091111 19:46:59< mordante> hi fendrin 
20091111 20:02:44-!- Blueblaze [n=nick@adsl-99-158-46-90.dsl.hstntx.sbcglobal.net] has quit [Remote closed the connection]
20091111 20:08:26-!- Netsplit leguin.freenode.net <-> irc.freenode.net quits: Rhonda, shadowmaster, boucman, stikonas, SonIcco, yann, crimson_penguin, zookeeper, Aethaeryn, silene,  (+23 more, use /NETSPLIT to show all of them)
20091111 20:18:50-!- fendrin [n=fabi@wesnoth/developer/fendrin] has quit [Remote closed the connection]
20091111 20:19:12-!- fendrin [n=fabi@88-134-116-206-dynip.superkabel.de] has joined #wesnoth-dev
20091111 20:37:23-!- Bob_The_Mighty [n=chatzill@cpc4-brig15-0-0-cust904.3-3.cable.virginmedia.com] has joined #wesnoth-dev
20091111 20:38:40< Bob_The_Mighty> hello sapient
20091111 20:38:58< Sapient> o Bob_The_Mighty
20091111 20:39:06< Sapient> hello
20091111 20:39:32< Bob_The_Mighty> hi, i notice there is a beta version out for 1.8 - many wml changes?
20091111 20:39:38< Sapient> looks like we have a big netsplit
20091111 20:39:51< Bob_The_Mighty> huh?
20091111 20:40:07< Sapient> netsplit is when IRC servers are split apart
20091111 20:40:23< Sapient> so a lot of users are missing
20091111 20:40:28< Sapient> from the room
20091111 20:41:15< Sapient> Bob_The_Mighty: there are a lot of changes from 1.6 to 1.8, yes
20091111 20:41:18< Bob_The_Mighty> The new lobby screen seems a bit confusing, but has there been drastic wml changes between 1.6 and 1.8
20091111 20:41:46< Sapient> but I can't think of many that would cause backwards compatibility issues
20091111 20:42:13< Bob_The_Mighty> so i can go upload some stuff right now?
20091111 20:42:15< Sapient> if anything, there might be some nice elegant behavior broken by the drop-in LUA replacements
20091111 20:42:37< Sapient> you should definitely test your stuff to see that it works first
20091111 20:42:51< Sapient> but I don't *think* anything that worked in 1.6 will be broken
20091111 20:42:56< Bob_The_Mighty> there is no one to help me test it :(
20091111 20:43:15< Sapient> in any case, you can run wmllint over it
20091111 20:43:21< Sapient> just make a backup first
20091111 20:43:38< Sapient> wmllint is designed so that you don't have to upgrade content by hand
20091111 20:44:00< Bob_The_Mighty> i have used it before, but i always get how every time - no gui yet?
20091111 20:44:04< Sapient> Bob_The_Mighty: well, if you're lazy just upload it and blame me then :p
20091111 20:45:44< Bob_The_Mighty> that's what i wanted to hear, thanks!
20091111 20:48:34-!- Netsplit over, joins: mordante, silene, Zarel, ilor_, SonIcco, yann, dtiger, Aethaeryn, zookeeper, crimson_penguin (+12 more)
20091111 20:48:34-!- happygrue [n=George@wesnoth/developer/wintermute] has joined #wesnoth-dev
20091111 20:48:34-!- Netsplit over, joins: CIA-55, kimrhh, isaac, shadowmaster, shikadibot, Rhonda
20091111 20:48:34-!- Soliton [n=Soliton@wesnoth/developer/soliton] has joined #wesnoth-dev
20091111 20:48:34-!- Netsplit over, joins: wesbot, ettin, freim
20091111 20:49:23< mordante> fendrin, I've an idea what your problem is, but your branch doesn't build I miss gui/dialogs/unit_input.hpp
20091111 20:50:00< Aethaeryn> Bob_The_Mighty!
20091111 20:50:10< Bob_The_Mighty> Hi Aeth
20091111 20:52:06< Aethaeryn> Hi.
20091111 20:58:53-!- Crab_ [i=crab@wesnoth/developer/crab] has joined #wesnoth-dev
20091111 20:59:14< Sapient> esr: does wmllint change all the old WML font markup from @ and ~ and so on to Pango markup?
20091111 20:59:41< Sapient> I know there was some discussion of this in the mailing list a while back
20091111 20:59:49< Aethaeryn> I think it does.
20091111 20:59:56< Sapient> and I got overruled in my desire to maintain backwards compatibility
20091111 21:00:28< Crab_> boucman: so, it stayed 5s per ai turn.
20091111 21:00:29< esr> Sapient: Utvdoes, yes.
20091111 21:00:35< esr> Sorry.
20091111 21:00:43< boucman> Crab_: yes, definitively an improvement...
20091111 21:00:44< esr> It does, yes.
20091111 21:00:56< Sapient> Bob_The_Mighty: you need to run wmllint then
20091111 21:01:01< boucman> so, rca is good for default now ? :)
20091111 21:01:08< Bob_The_Mighty> is this wiki statement still true: At the mement, Pango markup only works in message= attributes within [message] tags. More attributes within other tags will be enabled in the future.   ??
20091111 21:01:36< Soliton> i'm pretty sure it works in a lot more now.
20091111 21:01:52< Soliton> objectives for example.
20091111 21:02:56< Crab_> boucman: I'll convert LoW to use it, and see how it will look like.
20091111 21:03:06< boucman> k
20091111 21:04:57-!- lizard_r [n=Miranda@wesnoth/umc-dev/developer/lizard] has joined #wesnoth-dev
20091111 21:07:35< Aethaeryn> Pango works in labels.
20091111 21:07:45< Aethaeryn> I just tested out making labels in a dummy game with things like <tt> and <s>
20091111 21:16:18< CIA-55> fendrin * r39685 /branches/fendrin_gui_stuff/src/ (game_events.cpp gui/dialogs/wml_message.cpp): Removed an unused class.
20091111 21:16:22< fendrin> mordante: please try the new revision
20091111 21:17:10< fendrin> mordante: If you want to do a test start low scenario seven. It calls the right dialogue that doesn't scroll right.
20091111 21:23:22< silene> Bob_The_Mighty: pango markup should work in anything accessible from wml; if it doesn't, please fill a bug-report
20091111 21:24:01< Bob_The_Mighty> okay
20091111 21:24:31< mordante> fendrin, are you sure you also committed all config files?
20091111 21:25:01< mordante> silene, Bob_The_Mighty it shouldn't work at all places, campaign titles are an exception 
20091111 21:25:07< fendrin> mordante: what are you missing?
20091111 21:25:24< Aethaeryn> mordante: why are campaign titles an exception?
20091111 21:25:46< mordante> fendrin, not sure but looks like a window definition
20091111 21:25:58< mordante> Aethaeryn, we don't want red bold items on the addon server
20091111 21:26:10< fendrin> mordante: the unit_input dialogue class was only a first experiment. It is totaly obsolete.
20091111 21:26:12< silene> mordante: i know; what i meant by "accessible from wml" is accessible from inside a scenario; campaign titles are not
20091111 21:26:25< mordante> silene, ok
20091111 21:27:49< mordante> fendrin, do you have the config files of that experiment still in your checkout?
20091111 21:28:07< fendrin> I have
20091111 21:28:46< mordante> I don't and I think settings.[c|h]pp still needs them 
20091111 21:29:25< fendrin> mordante: fixing 
20091111 21:29:44< Bob_The_Mighty> wmllint is telling me that these units are not known: recruit="Cavalryman,Horseman,Spearman,Fencer,Heavy Infantryman,Bowman,Mage"
20091111 21:30:46< Soliton> you didn't include the core unit files in that wmllint run then.
20091111 21:31:06-!- Deusite [n=Tom@92-238-120-246.cable.ubr26.newt.blueyonder.co.uk] has joined #wesnoth-dev
20091111 21:31:38-!- silene [n=plouf@wesnoth/developer/silene] has quit ["Leaving."]
20091111 21:31:40< Bob_The_Mighty> oh, so i can ignore it
20091111 21:31:49< Soliton> yep.
20091111 21:39:55-!- Netsplit leguin.freenode.net <-> irc.freenode.net quits: yann
20091111 21:40:02< fendrin> mordante: It compiles fine here. Do you have the latest revision?
20091111 21:40:12-!- Netsplit over, joins: yann
20091111 21:40:24< mordante> fendrin, it compiles fine here as well but fails to run
20091111 21:43:04< Bob_The_Mighty> funny, wmllint changed [unstone] to [petrify] but did not change the closing tag [/unstone]
20091111 21:43:12-!- Blueblaze [n=nick@adsl-99-158-46-90.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20091111 21:43:50< fendrin> mordante: I see. missed that one.
20091111 21:44:10< mordante> fendrin, I made a fix, shall I commit it?
20091111 21:44:23< fendrin> yes, please
20091111 21:44:49< mordante> will do after looking at the scroll problem
20091111 21:46:24< fendrin> can you reproduce it?
20091111 21:47:45< mordante> yes I can, I'll try to have a closer look tomorrow
20091111 21:47:59< mordante> I think it's an engine "feature"
20091111 21:48:41< fendrin> I just don't get where the code is that does that jumping. 
20091111 21:48:59< CIA-55> mordante * r39686 /branches/fendrin_gui_stuff/src/gui/widgets/ (settings.cpp settings.hpp): Remove some obsolete code.
20091111 21:48:59< mordante> I don't think it jumps ;-)
20091111 21:49:17< mordante> I think it's the rerendering of the background
20091111 21:49:33< mordante> not sure however so need to test more
20091111 21:50:12< mordante> I'm off night
20091111 21:50:14< fendrin> bye
20091111 21:50:27-!- mordante [n=mordante@wesnoth/developer/mordante] has quit ["Leaving"]
20091111 21:52:49-!- Zarel [n=Zarel@warzone2100/developer/Zarel] has quit ["This computer has gone to sleep"]
20091111 22:00:27< Bob_The_Mighty> if a race has no plural name it causes wml errors..
20091111 22:04:24< boucman> Bob_The_Mighty: open a bug so we don't forget it...
20091111 22:05:01< Bob_The_Mighty> ok
20091111 22:11:01-!- dtiger [n=dtiger@dynamic-vpdn-91-149-132-222.telecom.by] has quit [Remote closed the connection]
20091111 22:17:07< CIA-55> crab * r39687 /trunk/data/core/macros/ai.cfg: additional changes to AI macros
20091111 22:17:17< CIA-55> crab * r39688 /trunk/data/campaigns/Legend_of_Wesmere/scenarios/01_The_Uprooting.cfg: LoW 1: new AI syntax
20091111 22:17:23< CIA-55> crab * r39689 /trunk/data/ai/aliases/stable_singleplayer.cfg: RCA (with alink's targeting patch) -> default stable singleplayer AI for scenarios which are converted to new ai syntax
20091111 22:31:11-!- zookeeper [n=l@wesnoth/developer/zookeeper] has quit []
20091111 22:35:18< Bob_The_Mighty> what does this mean?
20091111 22:35:19< Bob_The_Mighty> 20091111 21:18:20 error gui/event: distributor mouse motion [mp_method_selection]: Failed to remove hover timer.
20091111 22:35:21< Bob_The_Mighty> 20091111 21:18:20 error gui/event: distributor mouse motion [mp_method_selection]: show hover tooltip bailing out, no hover widget.
20091111 22:37:00< boucman> Bob_The_Mighty: tha would be for mordante, but apprently he's not around atm
20091111 22:38:50-!- Zarel [n=Zarel@nat-portal-160-94-47-16.uofm.wireless.umn.edu] has joined #wesnoth-dev
20091111 22:43:03< Crab_> fendrin: around /
20091111 22:43:04< Crab_> ?
20091111 22:47:30< CIA-55> crab * r39690 /trunk/data/campaigns/Legend_of_Wesmere/scenarios/04_Elvish_Treasury.cfg: LoW 4: new AI syntax
20091111 22:47:35< CIA-55> crab * r39691 /trunk/data/campaigns/Legend_of_Wesmere/scenarios/05_Saurian_Treasury.cfg: LoW 5: new AI syntax
20091111 22:47:40< CIA-55> crab * r39692 /trunk/data/campaigns/Legend_of_Wesmere/scenarios/06_Acquaintance_in_Need.cfg: LoW 6: new AI syntax
20091111 22:47:46< CIA-55> crab * r39693 /trunk/data/campaigns/Legend_of_Wesmere/scenarios/07_Elves_last_stand.cfg: LoW 7: new AI syntax
20091111 22:52:03< CIA-55> fendrin * r39694 /branches/fendrin_gui_stuff/ (2 files in 2 dirs): unit list defined
20091111 22:52:54< Bob_The_Mighty> can someone explain to me what changes have been made in the area of custom terrains since 1.6? 
20091111 22:53:10< CIA-55> soliton * r39695 /trunk/src/server/room.cpp: removed unsused include
20091111 22:53:13< Crab_> fendrin: I see that LoW 3 now includes a 'full map' from LoW 7, albeit part of it is unaccessible. it also includes LoW 7-specific labels. is this intended ?
20091111 22:53:21< CIA-55> soliton * r39696 /trunk/src/server/server.cpp: remove leaving users from ghost_players_ as well
20091111 22:53:25< CIA-55> soliton * r39697 /trunk/src/server/server.cpp: avoid a lookup in players_ when not in debug mode
20091111 22:53:29< CIA-55> soliton * r39698 /trunk/src/server/server.cpp: properly start with the next user when looking for clones
20091111 22:53:37< CIA-55> soliton * r39699 /trunk/po/wesnoth/en_GB.po: removed translation notes from translated strings
20091111 22:54:30< fendrin> Crab_: Hello crab. I switched on the labels, they should fit since it is the same map. There has nothing changed with the scenario in relation to the map. 
20091111 22:54:32-!- ilor__ [n=ilor@aud220.neoplus.adsl.tpnet.pl] has joined #wesnoth-dev
20091111 22:54:56< fendrin> The strange looking border is a new bug that must have been introduced lately.
20091111 22:54:57< Crab_> 'Keep of FOO' ?
20091111 22:55:16< CIA-55> soliton * r39700 /trunk/data/multiplayer/scenarios/2p_Howling_Ghost_Badlands.cfg: fixed team names
20091111 22:55:29< Crab_> fendrin: there's keep labels for enemy leaders which are not there yet, and the location of Kalenz's keep is slightly different from LoW 7
20091111 22:55:32< fendrin> The keep of labels are useless anyway. 
20091111 22:55:47< fendrin> They can be removed.
20091111 22:56:15< fendrin> Crab_: Can you do that please? I don't have a trunk checkout around atm.
20091111 22:57:13< Crab_> ok, I'll try (they're not useless in LoW 7, btw ) And, another question: how easy you want  'LoW on easy' to be ?
20091111 22:57:59< Crab_> (I'm asking because I'm going to re-tune the ai of 1st arc (1-7) of LoW)
20091111 22:58:10< fendrin> It's too easy at the moment I guess?
20091111 22:58:43< Crab_> apart from LoW 2, yes
20091111 22:59:27< fendrin> It can increase significantly.
20091111 23:00:41-!- grzywacz [n=grzywacz@wesnoth/developer/grzywacz] has quit [Remote closed the connection]
20091111 23:01:06-!- Appleman1234 [n=Appleman@131.181.101.200] has joined #wesnoth-dev
20091111 23:01:31-!- ilor_ [n=ilor@wesnoth/developer/ilor] has quit [Read error: 60 (Operation timed out)]
20091111 23:24:22-!- Espreon [n=espreon@wesnoth/developer/espreon] has joined #wesnoth-dev
20091111 23:29:41-!- lukjad007 [n=lukjadOO@unaffiliated/lukjad007] has joined #wesnoth-dev
20091111 23:34:28-!- Espreon [n=espreon@wesnoth/developer/espreon] has quit [Read error: 54 (Connection reset by peer)]
20091111 23:48:07-!- Espreon [n=espreon@wesnoth/developer/espreon] has joined #wesnoth-dev
20091111 23:48:45-!- lizard_r [n=Miranda@wesnoth/umc-dev/developer/lizard] has quit ["Saurian Augur - I'll heal you by 4 hp if you leave next to me"]
20091111 23:52:28-!- boucman [n=rosen@wesnoth/developer/boucman] has quit ["Leaving."]
20091111 23:54:26-!- Deusite [n=Tom@92-238-120-246.cable.ubr26.newt.blueyonder.co.uk] has left #wesnoth-dev []
20091111 23:56:52-!- Bob_The_Mighty [n=chatzill@cpc4-brig15-0-0-cust904.3-3.cable.virginmedia.com] has quit ["ChatZilla 0.9.85 [Firefox 3.0.15/2009101601]"]
--- Log closed Thu Nov 12 00:00:51 2009