--- Log opened Sun Dec 27 00:00:00 2009
20091227 00:05:22-!- alink [n=alink@wesnoth/developer/alink] has quit [Remote closed the connection]
20091227 00:11:47-!- stikonas [n=and@wesnoth/translator/stikonas] has quit [Read error: 60 (Operation timed out)]
20091227 00:24:23-!- stikonas [n=and@wesnoth/translator/stikonas] has joined #wesnoth-dev
20091227 00:28:49-!- ilor [n=user@wesnoth/developer/ilor] has quit [Read error: 60 (Operation timed out)]
20091227 00:29:07-!- ilor [n=user@wesnoth/developer/ilor] has joined #wesnoth-dev
20091227 00:42:34-!- aditya [n=aditya@59.95.14.171] has joined #wesnoth-dev
20091227 00:44:03-!- zookeeper [n=l@wesnoth/developer/zookeeper] has quit []
20091227 01:08:41-!- aditya [n=aditya@59.95.14.171] has quit [Read error: 110 (Connection timed out)]
20091227 01:09:41-!- mordocai [n=mordocai@66.119.9.243] has joined #wesnoth-dev
20091227 01:27:08-!- mordocai [n=mordocai@66.119.9.243] has quit ["Leaving"]
20091227 01:36:51-!- stikonas_ [n=and@wesnoth/translator/stikonas] has joined #wesnoth-dev
20091227 01:44:01-!- stikonas [n=and@wesnoth/translator/stikonas] has quit [Read error: 60 (Operation timed out)]
20091227 01:45:25-!- aditya [n=aditya@59.95.9.167] has joined #wesnoth-dev
20091227 02:12:20-!- Noyga [n=lame-z@wesnoth/developer/noyga] has quit [Remote closed the connection]
20091227 02:13:21-!- stikonas_ [n=and@wesnoth/translator/stikonas] has quit [Read error: 60 (Operation timed out)]
20091227 02:15:46-!- Netsplit kubrick.freenode.net <-> irc.freenode.net quits: Tigge
20091227 02:19:50-!- Chusslove [n=Chusslov@brsg-d9bee89c.pool.mediaWays.net] has quit [Read error: 110 (Connection timed out)]
20091227 02:21:17-!- Netsplit kubrick.freenode.net <-> irc.freenode.net quits: erl
20091227 02:21:49-!- Netsplit over, joins: erl
20091227 02:25:24-!- ilor [n=user@wesnoth/developer/ilor] has quit [Read error: 60 (Operation timed out)]
20091227 02:25:32-!- Chusslove [n=Chusslov@brsg-d9bef391.pool.mediaWays.net] has joined #wesnoth-dev
20091227 02:31:23-!- shadowm_bluecore [n=ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20091227 02:32:24-!- Tigge [n=tigge@c-5ba372d5.015-28-73746f34.cust.bredbandsbolaget.se] has joined #wesnoth-dev
20091227 02:46:25-!- loonycyborg [n=sergey@wesnoth/developer/loonycyborg] has quit ["Zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz"]
20091227 02:46:27-!- aditya [n=aditya@59.95.9.167] has quit [Read error: 110 (Connection timed out)]
20091227 02:46:59-!- loonybot [n=loonybot@wesnoth/bot/loonybot] has quit [Remote closed the connection]
20091227 02:53:31-!- aditya [n=aditya@59.95.16.27] has joined #wesnoth-dev
20091227 02:55:42-!- Netsplit kubrick.freenode.net <-> irc.freenode.net quits: Tigge
20091227 03:01:57-!- deekay [n=dk@wesnoth/developer/dragonking] has quit []
20091227 03:10:40-!- Netsplit over, joins: Tigge
20091227 03:14:13-!- ardesh [n=ardesh@port-92-195-232-64.dynamic.qsc.de] has quit [Read error: 110 (Connection timed out)]
20091227 03:14:34-!- ardesh [n=ardesh@port-92-195-132-190.dynamic.qsc.de] has joined #wesnoth-dev
20091227 03:15:01-!- Netsplit kubrick.freenode.net <-> irc.freenode.net quits: Ivanovic, Vetinari
20091227 03:17:58-!- Ivanovic [n=ivanovic@wesnoth/developer/ivanovic] has joined #wesnoth-dev
20091227 03:18:04-!- Netsplit over, joins: Vetinari
20091227 03:20:32-!- noy [n=Noy@wesnoth/developer/noy] has quit []
20091227 03:22:40-!- aditya [n=aditya@59.95.16.27] has quit [Read error: 110 (Connection timed out)]
20091227 03:26:53-!- Zarel [n=Zarel@warzone2100/developer/Zarel] has quit ["Leaving"]
20091227 03:29:27-!- shadowm_bluecore is now known as Pasta
20091227 03:29:30-!- crimson_pinvin [n=ben@unused-66-225-190-136.vianet.ca] has joined #wesnoth-dev
20091227 03:40:41-!- Appleman1234 [n=Appleman@CPE-124-191-178-150.oxqn1.cha.bigpond.net.au] has joined #wesnoth-dev
20091227 03:46:28-!- Pasta is now known as ShikadiLord
20091227 03:49:47-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has quit [No route to host]
20091227 03:49:47-!- crimson_pinvin is now known as crimson_penguin
20091227 04:03:10-!- ShikadiLord [n=ignacio@wesnoth/developer/shadowmaster] has quit ["This TCP socket is closing in a few milliseconds!"]
20091227 04:32:37-!- aditya [n=aditya@59.95.14.57] has joined #wesnoth-dev
20091227 04:32:44-!- Ivanovic_ [n=ivanovic@dtmd-4db2e102.pool.mediaWays.net] has joined #wesnoth-dev
20091227 04:43:52-!- crimson_pinvin [n=ben@unused-66-225-190-179.vianet.ca] has joined #wesnoth-dev
20091227 04:50:20-!- Ivanovic [n=ivanovic@wesnoth/developer/ivanovic] has quit [Read error: 110 (Connection timed out)]
20091227 04:50:46-!- Ivanovic_ is now known as Ivanovic
20091227 05:02:40-!- Appleman1234 [n=Appleman@CPE-124-191-178-150.oxqn1.cha.bigpond.net.au] has quit ["Leaving"]
20091227 05:03:51-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has quit [Read error: 113 (No route to host)]
20091227 05:03:54-!- crimson_pinvin is now known as crimson_penguin
20091227 05:48:19-!- esr [n=chatzill@wesnoth/developer/esr] has quit ["ChatZilla 0.9.86 [Firefox 3.5.5/20091109134913]"]
20091227 05:59:44-!- FreeAsABeer [n=croiv@219.142.242.194] has joined #wesnoth-dev
20091227 06:01:30-!- FreeAsABeer [n=croiv@219.142.242.194] has left #wesnoth-dev []
20091227 06:01:35-!- aditya [n=aditya@59.95.14.57] has quit [Read error: 54 (Connection reset by peer)]
20091227 06:13:04-!- crimson_pinvin [n=ben@wesnoth/developer/crimsonpenguin] has joined #wesnoth-dev
20091227 06:14:29-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has quit [Read error: 60 (Operation timed out)]
20091227 06:14:29-!- crimson_pinvin is now known as crimson_penguin
20091227 06:16:37-!- esr [n=chatzill@static-71-162-243-5.phlapa.fios.verizon.net] has joined #wesnoth-dev
20091227 06:20:54-!- aditya [n=aditya@59.95.28.163] has joined #wesnoth-dev
20091227 06:21:16-!- wesbot changed the topic of #wesnoth-dev to: string/feature freeze active! | 67 bugs, 246 feature requests, 9 patches | logs: http://irclogs.wesnoth.org/ | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com | http://imagebin.org
20091227 07:09:06-!- _teddy [n=fedor76@ppp-78-24-27-29-bras0.istra.ru] has quit [Read error: 104 (Connection reset by peer)]
20091227 08:06:02-!- mordante [n=mordante@wesnoth/developer/mordante] has joined #wesnoth-dev
20091227 08:06:15< mordante> servus
20091227 08:09:22-!- boucman [n=rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20091227 08:12:54-!- boucman [n=rosen@wesnoth/developer/boucman] has left #wesnoth-dev []
20091227 08:13:05-!- boucman [n=rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20091227 08:19:26-!- Appleman1234 [n=Appleman@CPE-124-191-178-150.oxqn1.cha.bigpond.net.au] has joined #wesnoth-dev
20091227 08:35:55-!- Blueblaze [n=nick@adsl-76-202-22-180.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20091227 08:50:03-!- silene [n=plouf@AMarseille-251-1-26-141.w83-113.abo.wanadoo.fr] has joined #wesnoth-dev
20091227 09:01:36-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has quit ["poof! 3AM is too late"]
20091227 09:08:36-!- dtiger [n=dtiger@dynamic-vpdn-91-149-132-17.telecom.by] has joined #wesnoth-dev
20091227 09:55:32-!- Blueblaze [n=nick@adsl-76-202-22-180.dsl.hstntx.sbcglobal.net] has quit [Read error: 104 (Connection reset by peer)]
20091227 10:13:08-!- stikonas [n=and@ctv-79-132-179-139.vinita.lt] has joined #wesnoth-dev
20091227 10:19:03-!- deekay [n=dk@wesnoth/developer/dragonking] has joined #wesnoth-dev
20091227 10:33:22< Ivanovic> moin
20091227 10:39:34< mordante> i Ivanovic 
20091227 10:39:36< mordante> hi*
20091227 10:42:08< CIA-28> mordante * r40416 /trunk/src/gui/widgets/ (5 files): 
20091227 10:42:08< CIA-28> Rename tgenerator_::get_item() to item().
20091227 10:42:08< CIA-28> Make the name more consistent with other similair functions. Also make
20091227 10:42:08< CIA-28> the const version public, no reason to hide it when the non-const
20091227 10:42:08< CIA-28> version is public.
20091227 10:42:11< CIA-28> mordante * r40417 /trunk/src/gui/widgets/window.cpp: 
20091227 10:42:11< CIA-28> Improve the drawing against UB.
20091227 10:42:14< CIA-28> When the filter removes all items to be drawn, it called back() on an
20091227 10:42:16< CIA-28> empty vector. Now test before the call.
20091227 10:42:18< CIA-28> mordante * r40418 /trunk/src/gui/widgets/ (widget.cpp widget.hpp): Don't try to draw items without a size.
20091227 10:44:50-!- EdB [n=edb@60.12.95-79.rev.gaoland.net] has joined #wesnoth-dev
20091227 10:56:38< Ivanovic> hmm, it would be lovely to have a libsdl version which is not affected by the "clicking does not work" before releasing 1.8...
20091227 10:57:39< mordante> indeed, still not sure whether the hal or sdl is to blame however
20091227 10:57:52-!- zookeeper [n=l@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20091227 10:58:15< Ivanovic> mordante: just wrote another mail to the ml also pointing to the bugtracker
20091227 10:58:34< Ivanovic> it seems to be some interaction between DE/WM and sdl that only occurs in windowed mode
20091227 10:58:43< Ivanovic> something where the mouseclicks get really screwed
20091227 10:58:55 * mordante starts to read
20091227 10:59:19< Ivanovic> (the gentoo bugtracker links to the arch bugtracker where they say that holding right mouse button and then leftclicking somehow seems to work to get a leftclick)
20091227 11:00:08 * mordante is surprised what kind of work-aronds people try 
20091227 11:00:20< boucman> Ivanovic: that sounds a lot like a falsh bug I had in ubuntu, 
20091227 11:00:51< Ivanovic> boucman: please have a look at the thread on the dev-ml, maybe you got a good idea what it could be
20091227 11:01:26< Ivanovic> in general i fear that it is some weird sdl bug (and we have to somehow make sam able to reproduce it, or the chances for having it fixed are *really* slim...)
20091227 11:01:28< boucman> Ivanovic: is that the wesnoth dev ml ?
20091227 11:01:34< Ivanovic> jupp
20091227 11:01:49< Ivanovic> thread: [Wesnoth-dev] Wesnoth has (many) problems with libsdl-1.2.14
20091227 11:01:56< boucman> havn't received any mail about that yet, apprently
20091227 11:02:08< Ivanovic> sent another update some mins ago
20091227 11:02:15< Ivanovic> in general this thread is "old" by now
20091227 11:02:18< Ivanovic> ;)
20091227 11:02:49< boucman> ok, got the thread
20091227 11:03:44< Ivanovic> good
20091227 11:03:44< mordante> Ivanovic, are you able to reproduce the problem?
20091227 11:03:52< Ivanovic> not constantly
20091227 11:03:58< Ivanovic> i got the problem *sometimes*
20091227 11:04:37< mordante> I wonder whether it's gui1 or gui2 or both
20091227 11:04:45< Ivanovic> no idea
20091227 11:04:52< boucman> could you try launching wesnoth with  export GDK_NATIVE_WINDOWS=1
20091227 11:05:06< mordante> me neither
20091227 11:05:08< boucman> that's the fix for a bug with exactly the same symptom in flash
20091227 11:05:11< Ivanovic> since many of the people reporting it have the prob in the main menu i'd guess that it is at least gui1
20091227 11:05:14< boucman> it's a long shot, but worth a try
20091227 11:05:51< Ivanovic> boucman: freakwent2 in #wesnoth has this prob, too, maybe he can test...
20091227 11:05:57-!- EdB [n=edb@60.12.95-79.rev.gaoland.net] has quit [Remote closed the connection]
20091227 11:44:19< Ivanovic> AI0867, boucman, Chusslove, deekay, esr, fendrin, mordante, shadowmaster, silene, Soliton, zookeeper, everyone else who cares: what do you think of releasing 1.7.11 in about one hour?
20091227 11:44:28< Ivanovic> any blockers speaking against doing so?
20091227 11:45:16-!- Noyga [n=lame-z@wesnoth/developer/noyga] has joined #wesnoth-dev
20091227 11:45:27< fendrin> Ivanovic: no, go ahead
20091227 11:45:34< Chusslove> Ivanovic: Don't know if you should run update-po4a. I've committed manual POs and images for Serbian, but it's not really important that it appears in 1.7.11.
20091227 11:47:10< Ivanovic> Chusslove: i always run the update commands right before releasing (as part of the "final" pot-update)
20091227 11:47:56< Chusslove> Great.
20091227 11:48:42< Ivanovic> that is: i tend to include those commands in "normal" pot-updates, too
20091227 11:48:57< Ivanovic> just takes ages to finish the manuals, which is why i don't run the commands too often...
20091227 11:49:02< Chusslove> Right.
20091227 11:49:40< boucman> fine here
20091227 11:49:54< Chusslove> Ah, uh, forgot about this http://pastebin.com/m57aba9c5 -- who should give a go for CSS style update for the manual?
20091227 11:50:31< Chusslove> This one does not change anything in existing manuals, just covers more Docbook markup, if used (I used it in translation).
20091227 11:50:36< Ivanovic> Chusslove: uhm, in theory it is ettins area of work, though he is inactive
20091227 11:51:00< Ivanovic> Chusslove: so i'd guess it is a case of: just commit it yourself if it really works!
20091227 11:51:08< Chusslove> Can I commit it then? I guarantee it doesn't break (or even change) anything.
20091227 11:51:18< Chusslove> Great.
20091227 11:53:45< CIA-28> caslav_ilic * r40419 /trunk/ (changelog doc/manual/styles/manual.css): Additional styles for GUI markup in manuals.
20091227 11:55:07< silene> Ivanovic: no blocker that i know of
20091227 12:11:44< Ivanovic> mordante: your commit stack is empty?
20091227 12:11:55< Ivanovic> or is there anything i should wait for on your end before getting 1.7.11 out?
20091227 12:12:36< stikonas> 1.7.11 = beta4?
20091227 12:13:04< Ivanovic> jupp
20091227 12:13:33< stikonas> I can probably send small translation update
20091227 12:16:37< stikonas> Ivanovic: sorry, the wrong file
20091227 12:16:37-!- aditya [n=aditya@59.95.28.163] has quit [Read error: 104 (Connection reset by peer)]
20091227 12:16:42< stikonas> will resend in a minute
20091227 12:19:03-!- ilor [n=user@wesnoth/developer/ilor] has joined #wesnoth-dev
20091227 12:21:16-!- wesbot changed the topic of #wesnoth-dev to: string/feature freeze active! | 66 bugs, 246 feature requests, 9 patches | logs: http://irclogs.wesnoth.org/ | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com | http://imagebin.org
20091227 12:22:51< CIA-28> ivanovic * r40420 /trunk/ (12 files in 11 dirs): updated Lithuanian translation
20091227 12:23:01< Ivanovic> ilor: ready for 1.7.11?
20091227 12:28:18< Ivanovic> running the pot-update now
20091227 12:30:20< fendrin> silene: Do you have some time for me?
20091227 12:31:12< silene> fendrin: not much, food for lunch is almost ready; but say anyway, i will look at it this afternoon
20091227 12:31:26< ilor> Ivanovic: haven't done much sadly 
20091227 12:31:33< Ivanovic> mkay
20091227 12:32:52< fendrin> silene: http://wesnoth.pastebin.com/d41ce22b1 This is my fixed heuristic. Can't see anything wrong with it but new_h is never < h. 
20091227 12:34:06-!- aditya [n=aditya@59.95.4.60] has joined #wesnoth-dev
20091227 12:37:10< silene> fendrin: at first glance, i don't see anything wrong with it; are your teleports correct?
20091227 12:37:22< silene> (i'm off btw)
20091227 12:37:56< fendrin> silene: Yes, everything looks fine in the debugger, the nodes are in place.
20091227 12:38:33< Ivanovic> oh, you'll love me for this pot-update!
20091227 12:42:03< CIA-28> ivanovic * r40421 /trunk/ (865 files in 28 dirs): 
20091227 12:42:03< CIA-28> pot-update (reference update only, no string changes/additions)
20091227 12:42:03< CIA-28> regenerated doc files
20091227 12:42:21< stikonas> 16 files changed, 3754 insertions(+), 3771 deletions(-)
20091227 12:43:36< stikonas> Ivanovic: ^^ that is not entirely correct. Some strings were removed
20091227 12:43:57< Ivanovic> oh, it is correct
20091227 12:44:12< Ivanovic> no string was *changed* and no string was *added*
20091227 12:44:20< Ivanovic> i don't mention removal at all...
20091227 12:44:21< Ivanovic> ;)
20091227 12:44:47-!- Noyga [n=lame-z@wesnoth/developer/noyga] has quit [Remote closed the connection]
20091227 12:46:04-!- loonybot [n=loonybot@ppp79-139-139-192.pppoe.spdop.ru] has joined #wesnoth-dev
20091227 12:46:46-!- loonycyborg [n=sergey@ppp79-139-139-192.pppoe.spdop.ru] has joined #wesnoth-dev
20091227 12:50:04< Ivanovic> loonycyborg: ready for 1.7.11?
20091227 12:50:21< loonycyborg> Yes.
20091227 13:10:18< mordante> Ivanovic, nothing left to commit for the next release
20091227 13:10:30< Ivanovic> mkay
20091227 13:10:45< Ivanovic> i am currently running optipng on the images, once it is done and commited i will package 1.7.11
20091227 13:15:48-!- stikonas_ [n=and@wesnoth/translator/stikonas] has joined #wesnoth-dev
20091227 13:17:31-!- stikonas [n=and@wesnoth/translator/stikonas] has quit [Connection reset by peer]
20091227 13:18:06< stikonas_> Ivanovic: I think that I found a serious bug in Two Brothers
20091227 13:18:18< Ivanovic> what is the bug?
20091227 13:18:20< stikonas_> it seems, that there is no Bjarn in scenario 3
20091227 13:18:27< stikonas_> you have to free him
20091227 13:18:37< stikonas_> the scenario is still winnable though
20091227 13:19:02< stikonas_> if you go to the right place
20091227 13:19:20-!- stikonas_ is now known as stikonas
20091227 13:21:23< Ivanovic> zookeeper: can you please have a look at two brothers?
20091227 13:21:40< Ivanovic> at the end of scenario2 i got this message in the logs when brena joins the group:
20091227 13:21:42< Ivanovic> 20091227 13:20:47 error engine: Couldn't find function for wml tag: modifications
20091227 13:22:03< stikonas> indeed, I haven't seen Brena too
20091227 13:22:17< stikonas> So at least 2 characters are missing
20091227 13:23:34-!- EdB [n=edb@60.12.95-79.rev.gaoland.net] has joined #wesnoth-dev
20091227 13:23:47< Ivanovic> brena was working
20091227 13:23:57< Ivanovic> okay, what exactly was changed in the engine?
20091227 13:24:12< Ivanovic> i now get this when bjarn should be placed in the dungeon:
20091227 13:24:14< Ivanovic> 20091227 13:23:27 error engine: wrong side in [unit] tag - not a number
20091227 13:25:57< CIA-28> ivanovic * r40422 /trunk/data/ (18 files in 16 dirs): 
20091227 13:25:57< CIA-28> ran utils/wesnoth-optipng
20091227 13:25:57< CIA-28> Overall statistics (only for files with a smaller recompressed size):
20091227 13:25:57< CIA-28> Original size: 3539 KiB on 18 files
20091227 13:25:58< CIA-28> Optimized size: 3461 KiB
20091227 13:26:00< CIA-28> Total saving: 77 KiB = 2% decrease
20091227 13:28:58< Ivanovic> Chusslove: please run util/wesnoth-optipng on translated images before commiting them
20091227 13:29:24< Ivanovic> Chusslove: at the top of the file you find a list of the progs you need to be able to do so (basically optipng and advcomp)
20091227 13:29:51< Ivanovic> Chusslove: you can just put the images into one folder and call the optipng script from within that folder so that you don't run it on *all* of wesnoth which takes a while
20091227 13:32:07< Chusslove> Ivanovic: I was using stock optipng so far, but I see it's not enough. Will look into what's needed for wesnoth-optipng.
20091227 13:33:28< Ivanovic> just run the script, it is *really* a nice thing (setting niceness and number of threads to use automatically, stuff like this)
20091227 13:41:20< CIA-28> ivanovic * r40423 /trunk/data/campaigns/Two_Brothers/scenarios/3_Guarded_Castle.cfg: make sure that bjarn gets placed correctly...
20091227 13:45:06< Ivanovic> zookeeper: how to add the strong trait to a unit generated this way?
20091227 13:45:13< Ivanovic> {NAMED_LOYAL_UNIT 1 (Iron Mauler) 8 1 Brena (_ "Brena")}
20091227 13:49:41-!- aditya [n=aditya@59.95.4.60] has quit [Nick collision from services.]
20091227 13:49:48-!- adibis [n=aditya@59.95.55.84] has joined #wesnoth-dev
20091227 13:51:00< fendrin> silene: Don't bother anymore, I think I found the bug.
20091227 13:54:31< fendrin> silene: It's working now like expected. Of course I broke the village silvermage teleportation and the new code is slow but I can't tell you how happy I am.
20091227 13:55:21< zookeeper> Ivanovic, looking...
20091227 13:58:18< zookeeper> Ivanovic, i guess [+unit]    [modifications]        {TRAIT_LOYAL}        {TRAIT_STRONG}    [/modifications][/unit]
20091227 13:58:31< zookeeper> (right after the macro call)
20091227 13:58:44< silene> fendrin: note that your heuristic function doesn't support consecutive teleports, so the computed paths will be wrong if they are allowed to be long enough (and they will be allowed to, since A* is commonly used to get the distance between any two points of the map, irrespective of unit moves)
20091227 13:59:24< zookeeper> but better to just use the NAMED_UNIT macro instead..
20091227 13:59:47< fendrin> silene: Uh, that is right. But I can't see a solution that wouldn't end in divergenz.
20091227 14:00:12< fendrin> s/divergenz/divergency
20091227 14:00:45< zookeeper> Ivanovic, the current code for placing brena should work just fine...but i think it should just make him strong, not loyal,strong
20091227 14:01:04< silene> fendrin: fivergence of what?
20091227 14:01:10< zookeeper> err, wait, nevermind that.
20091227 14:02:29< zookeeper> i'll fix brena
20091227 14:03:42< fendrin> silene: The heuristic. If I did understand you right. The code how it is now will only allow to use one teleport at a time. If I try to fix the heuristic to use more than one I will have to check for every teleport again and again.
20091227 14:04:21< CIA-28> zookeeper * r40424 /trunk/data/campaigns/Two_Brothers/scenarios/2_The_Chase.cfg: Placed Brena with the NAMED_UNIT macro instead, to make assignment of the strong trait less ackward.
20091227 14:05:28< silene> fendrin: yes, that's what i explained last time; the idea is to precompute a matrix that gives a distance (either heuristic or optimal) between any two teleport points (even if they are unrelated); then you just use a variant of the old heuristic function  (or of yours)
20091227 14:06:20< zookeeper> ok, so why is arne missing the loyal overlay? i thought they were supposed to be used consistently in mainline now.
20091227 14:06:31< fendrin> silene: If you give me a rough draw how that can be implemented I will do it.
20091227 14:06:54 * stikonas understood why I did not see Brema. I was playing on Hard difficulty
20091227 14:06:57< zookeeper> any search for places to use the overlay in should have brought up arne, since he has {TRAIT_LOYAL}...
20091227 14:07:09< zookeeper> eh, s/arne/bjarn
20091227 14:07:10< silene> fendrin: is the idea clear for the heuristic function at least?
20091227 14:07:46< fendrin> silene: Sorry, no. 
20091227 14:08:34< fendrin> silene: I wasn't fit enough to follow all the conversation last time we talked. But I guess I have learned much since then.
20091227 14:09:49< silene> fendrin: the old heuristic function and yours behave this way: for any two related teleport points, they compute a lower bound the distance to one, the distance to the other, they assume the distance between them is one, so the sum of these three values is a lower bound on the total distance; do you agree?
20091227 14:10:08< zookeeper> right, looks like no one actually made loyal units use the loyal overlay, just added it to the loyal unit macros.
20091227 14:11:10< zookeeper> if my grep is correct, there's over 200 loyal units missing the loyal overlay
20091227 14:12:55< fendrin> silene: Right, that is the idea as I understand it as well.
20091227 14:12:56< silene> zookeeper: nice, i think that if we had purposely wanted to confuse the user, we wouldn't have done better :-)
20091227 14:14:41< zookeeper> a-ha, i missed checking for IS_LOYAL...looks like not too many cases left after all
20091227 14:14:56< zookeeper> which is...nice
20091227 14:15:13< silene> fendrin: now imagine that the function considers even unrelated teleport points and everything else is left unchanged; the heuristic function becomes correct for multiteleport; unfortunately it is not a good heuristic function, since it assumes the distance between any two teleport points is 1; do you still agree?
20091227 14:15:33< zookeeper> and of course bjard doesn't have the overlay since he already has the hero overlay
20091227 14:16:20< fendrin> silene: What is a unrelated teleport?
20091227 14:17:11< fendrin> zookeeper: I don't consider that as a problem. Most (all?) heroes are loyal.
20091227 14:17:31< zookeeper> fendrin, who does?
20091227 14:18:11< silene> fendrin: two locations where units are teleported, even if there is no direct teleport between them; for instance, if your map has two tunnels, the four entries of the tunnels are all teleport points, but they are not all related
20091227 14:19:30< fendrin> zookeeper: No one then. I have overrated your statement.
20091227 14:19:40< zookeeper> roger :)
20091227 14:20:49< fendrin> silene: My tunnels are one way tunnels for now. But I got the idea. The heuristic must consider that a unit can use two independent tunnels or more to reach a goal.
20091227 14:23:18< silene> fendrin: yes, so the heuristic function has to compute a lower bound on the distance between two teleport points (which may be the entrances of two independent tunnels); if the function says that this lower bound is 1, it is a globally correct heuristic function; okay?
20091227 14:23:41< fendrin> yes
20091227 14:25:24< silene> fendrin: so, as a first step, you could slightly modify your heuristic function and its data structures so that it behaves like that; as a second step, we will later see how to improve the lower bound
20091227 14:26:12< fendrin> silene: Right, but I have no idea how to archieve that.
20091227 14:28:40-!- stikonas_ [n=and@ctv-79-132-179-139.vinita.lt] has joined #wesnoth-dev
20091227 14:29:47< silene> fendrin: for instance, just put all the teleport points in an array, and iterate over this array "twice": one time for the distance to the starting point of the path (new_srch) and one for the distance to the ending point (new_dsth); the code will be almost identical to your current one, except that new_h will be computed outside the loops instead of inside
20091227 14:29:49-!- alink [n=alink@wesnoth/developer/alink] has joined #wesnoth-dev
20091227 14:30:24< CIA-28> alink * r40425 /trunk/changelog: update changelog
20091227 14:31:28< fendrin> silene: Ah, I understand. Should be doable.
20091227 14:33:18-!- stikonas_ [n=and@wesnoth/translator/stikonas] has quit [Client Quit]
20091227 14:33:25-!- stikonas_ [n=and@ctv-79-132-179-139.vinita.lt] has joined #wesnoth-dev
20091227 14:34:05-!- stikonas [n=and@wesnoth/translator/stikonas] has quit [Read error: 60 (Operation timed out)]
20091227 14:37:58-!- stikonas_ is now known as stikonas
20091227 15:00:04-!- Netsplit kubrick.freenode.net <-> irc.freenode.net quits: Tigge
20091227 15:00:05-!- silene [n=plouf@wesnoth/developer/silene] has quit [Read error: 104 (Connection reset by peer)]
20091227 15:01:03-!- Netsplit over, joins: Tigge
20091227 15:03:57-!- silene [n=plouf@wesnoth/developer/silene] has joined #wesnoth-dev
20091227 15:04:09-!- alink [n=alink@wesnoth/developer/alink] has quit [Remote closed the connection]
20091227 15:09:17-!- kimrhh [n=kimhh@exherbo/developer/kimrhh] has left #wesnoth-dev []
20091227 15:10:28< fendrin> silene: But how does this solve the problem? The algo will compute heuristics between tunnels that are not connected in any way. 
20091227 15:11:18-!- Netsplit kubrick.freenode.net <-> irc.freenode.net quits: Tigge
20091227 15:11:31-!- Netsplit over, joins: Tigge
20091227 15:15:47< shadowmaster> Ivanovic: I think you are crazy.
20091227 15:16:02 * fendrin seconds shadowmaster.
20091227 15:16:03< Ivanovic> shadowmaster: huh, context?
20091227 15:17:01< shadowmaster> "blah blah 1.7.11 in about one hour" ;)
20091227 15:17:24< shadowmaster> I didn't see that coming.
20091227 15:18:38< Ivanovic> ah, okay
20091227 15:18:58< Ivanovic> but do you have any reason that speaks against 1.7.11?
20091227 15:19:07< shadowmaster> nope
20091227 15:21:28< Ivanovic> you see, perfect timing on my side, noone objecting
20091227 15:21:29< Ivanovic> ;)
20091227 15:21:34< loonycyborg> Considering that an hour had already passed, that was a rather overzealous statement :P
20091227 15:21:39< CIA-28> ivanovic * r40426 /trunk/ (changelog configure.ac players_changelog src/wesconfig.h): bump version to 1.7.11-1.8beta4
20091227 15:22:23< Ivanovic> loonycyborg: i had no time since i had to eat breakfast and get rid of the christmas tree
20091227 15:22:25< Ivanovic> ;)
20091227 15:22:45< Ivanovic> silene, mordante: you know that there are *many* warnings about missed mouse events?
20091227 15:23:08< Ivanovic> at least when using libsdl 1.2.14 that is
20091227 15:23:19< shadowmaster> same with 1.2.13, seems harmless
20091227 15:23:44< mordante> Ivanovic, I know there are some harmless mouse warnings which dwell low at my todo list
20091227 15:23:48< Ivanovic> 20091227 15:23:35 warning gui/event: distributor mouse button left [transient_message]: SDL left button up. The mouse button is already up, we missed an event.
20091227 15:24:18< silene> fendrin: it solves the problem in the sense that the heuristic function computes a lower bound; "lower bound" is the important point, if it isn't, the paths are not correct; a heuristic function that always return 1 would be correct (though it would be terribly inefficient, since A* would become a simple Dijkstra)
20091227 15:24:40< silene> fendrin: your previous function was not correct, since the result was not a lower bound when multiple teleports were involved
20091227 15:25:33< mordante> Ivanovic, and this problem causes Wesnoth to sometimes miss events, so it's a real warning (but happens very seldom and has nothing to do with SDL 1.2.14)
20091227 15:25:47< Ivanovic> very seldom is wrong
20091227 15:25:59< Ivanovic> (that is regarding the occurence of the warning)
20091227 15:26:11< Ivanovic> it happens with every ingame dialog
20091227 15:26:18< Ivanovic> (dialog as in chars talking to each other)
20091227 15:26:44< fendrin> silene: I see. That is not a fix but a hack to provide a heuristic that is good enough.
20091227 15:27:06< Ivanovic> fendrin, esr: for whom of you are the warnings in DM?
20091227 15:28:13< Ivanovic> http://wesnoth.pastebin.com/m575937fb
20091227 15:28:29< Ivanovic> those are the warnings i get with the first scenario of DM (the story only one)
20091227 15:28:32 * loonycyborg gets tons of 'broken markup' warnings on campaign screen due to svn version of TDH
20091227 15:28:44< silene> fendrin: i'm not sure to understand; your function was broke, so this is a fix (and the returned values are not even worse when there is only one teleport involved, so it's quite good)
20091227 15:29:05< Ivanovic> OUTSCH
20091227 15:29:19< Ivanovic> the campaign sceptre of fire can not be started
20091227 15:29:26< Ivanovic> 20091227 15:28:51 error general: Unbekanntes Szenario: '1_A_Bargain_is_Struck'
20091227 15:29:33< Ivanovic> "unknown scenario"
20091227 15:29:45< Ivanovic> argh, outsch
20091227 15:29:59< Ivanovic> error on my end
20091227 15:30:39< loonycyborg> Seems to work for me.
20091227 15:32:10< fendrin> Ivanovic: both of us.
20091227 15:32:39< Ivanovic> message in utbs: 20091227 15:32:19 warning filesystem: Illegal path 'terrain/../scenery/fire1.png' replaced by 'scenery/fire1.png'
20091227 15:33:17< fendrin> Ivanovic: These are warnings, so I am going to fix them after today's release?
20091227 15:33:45< Ivanovic> fine
20091227 15:35:31< shadowmaster> harmless tooo...
20091227 15:35:39-!- Ivanovic changed the topic of #wesnoth-dev to: releasing 1.7.11 | string/feature freeze active! | 66 bugs, 246 feature requests, 9 patches | logs: http://irclogs.wesnoth.org/ | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com | http://imagebin.org
20091227 15:36:13< shadowmaster> I don't see any need to fix that kind of warnings, since the only valid fix is duplicating image files in different paths, which is *damn annoying*
20091227 15:36:40< shadowmaster> I'd rather demote the warning to info.
20091227 15:41:46< Ivanovic> sounds good to me...
20091227 15:47:05-!- Blarumyrran [n=Blarumyr@81-20-159-197.levira.ee] has joined #wesnoth-dev
20091227 15:51:46-!- ilor [n=user@wesnoth/developer/ilor] has quit [Remote closed the connection]
20091227 15:52:00-!- ilor [n=user@wesnoth/developer/ilor] has joined #wesnoth-dev
20091227 15:53:32< fendrin> shadowmaster: Ivanovic: That means I take no action in the case of the warnings?
20091227 15:54:18 * shadowmaster is wondering if he's supposed to demote it now?
20091227 15:54:52< Ivanovic> i think those terrain images are clearly not a case of "wrong path" but a case of "wrong detection"
20091227 15:55:16< Ivanovic> fendrin: the move event stuff probably warrants a closer look
20091227 15:56:35< fendrin> Okay
20091227 16:02:02< fendrin> silene: is this http://wesnoth.pastebin.com/d2a5109d4 what you have in mind?
20091227 16:04:51< CIA-28> ivanovic * r40427 /tags/1.7.11/: tagged 1.7.11-1.8beta4 (using r40426)
20091227 16:07:14< fendrin> Ivanovic: The move_unit fake warnings are from DM: A night in the swamp?
20091227 16:07:43< Ivanovic> fendrin: they are from starting the campaign in the main screen and then just clicking through the very first scenario
20091227 16:07:51< fendrin> Ivanovic: thanks
20091227 16:07:55< Ivanovic> (that is the "text only" stuff)
20091227 16:08:13-!- loonycyborg [n=sergey@wesnoth/developer/loonycyborg] has quit ["KVIrc 3.4.2 Shiny http://www.kvirc.net/"]
20091227 16:08:16 * shadowmaster wishes trunk would compile as fast as the 1.0 branch does!
20091227 16:08:33< Ivanovic> hehehe
20091227 16:08:45-!- loonybot [n=loonybot@wesnoth/bot/loonybot] has quit [Remote closed the connection]
20091227 16:08:56 * Ivanovic wishes that the tarball was as small as with 1.0.x
20091227 16:09:03< Ivanovic> then uploading would not take me some 45mins...
20091227 16:09:44< fendrin> Ivanovic: why not build the tarball on the wesnoth server?
20091227 16:09:54< Ivanovic> fendrin: because i got to test it, too
20091227 16:10:14< Ivanovic> fendrin: so either i got to upload it to wesnoth.org or download to test if everything works
20091227 16:10:15< shadowmaster> it would also take some important CPU time...
20091227 16:10:36< Ivanovic> and i don't want to use bzip2 -9 on the mp server in some "rush hour time" when i can prevent it
20091227 16:10:38< Ivanovic> ;)
20091227 16:11:04< shadowmaster> also, compiling the AI code tends to cause Linux to throw everything to swap here even if I have 1.2 GB of RAM free.
20091227 16:11:35< Ivanovic> wow, the 1.0 tarball really just had 37.4MB?!?
20091227 16:11:36< shadowmaster> while we have... 89 MB free?
20091227 16:11:50< Ivanovic> a normal xdelta these days is about 2 to 10MB already...
20091227 16:11:57< shadowmaster> I hope free -m is just lying ;)
20091227 16:12:47< shadowmaster> (yes, I'm aware of wesnothd's requirements for the official instances...)
20091227 16:13:00< shadowmaster> (I'm just kidding and half asleep if you can't tell)
20091227 16:15:01< fendrin> Ivanovic: The warning is just lying. There is a route between the 2 locations.
20091227 16:15:07-!- Blarumyrram [n=Blarumyr@81-20-159-197.levira.ee] has joined #wesnoth-dev
20091227 16:15:12< Ivanovic> lovely
20091227 16:15:37< fendrin> Since there are no enemies on this map it can't be blocked by that either.
20091227 16:15:59-!- Netsplit kubrick.freenode.net <-> irc.freenode.net quits: esr, Blarumyrran, AnMaster, Appleman1234, knotwork, deekay
20091227 16:16:04 * shadowmaster is awaiting Ivanovic 's post-release bump
20091227 16:17:34< CIA-28> ivanovic * r40428 /trunk/ (changelog configure.ac players_changelog src/wesconfig.h): post release version bump to 1.7.11+svn
20091227 16:20:29-!- Ivanovic changed the topic of #wesnoth-dev to: released 1.7.11, announcing "soon" | string/feature freeze active! | 66 bugs, 246 feature requests, 9 patches | logs: http://irclogs.wesnoth.org/ | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com | http://imagebin.org
20091227 16:24:45-!- Blarumyrran [n=Blarumyr@81-20-159-197.levira.ee] has joined #wesnoth-dev
20091227 16:24:45-!- deekay [n=dk@wesnoth/developer/dragonking] has joined #wesnoth-dev
20091227 16:24:45-!- Appleman1234 [n=Appleman@CPE-124-191-178-150.oxqn1.cha.bigpond.net.au] has joined #wesnoth-dev
20091227 16:24:45-!- esr [n=chatzill@wesnoth/developer/esr] has joined #wesnoth-dev
20091227 16:24:45-!- knotwork [n=markm@142.177.234.40] has joined #wesnoth-dev
20091227 16:25:46-!- AnMaster [n=AnMaster@unaffiliated/anmaster] has joined #wesnoth-dev
20091227 16:30:30< CIA-28> shadowmaster * r40429 /trunk/src/filesystem.cpp: 
20091227 16:30:30< CIA-28> Demote a filesystem interface warning to the info log level because it
20091227 16:30:30< CIA-28> gets too noisy with certain add-ons using terrain graphics from a
20091227 16:30:30< CIA-28> upper-level directory (e.g. scenery/) and is completely harmless
20091227 16:30:30< CIA-28> otherwise.
20091227 16:34:09-!- shadowm_bluecore [n=ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20091227 16:34:53< mordante> shadowmaster, shouldn't those addons be fixed instead of hiding the warning?
20091227 16:34:59< shadowmaster> no
20091227 16:35:09< shadowmaster> read above when I mentioned "valid fix"
20091227 16:35:34< Ivanovic> mordante: we are not talking about addons
20091227 16:35:48< Ivanovic> we are talking about using scenery objects in mainline as terrains
20091227 16:36:30< mordante> Ivanovic, the commit entry says add-ons ;-)
20091227 16:36:43< shadowmaster> UtBS and THoT are, strictly speaking, mainline add-ons
20091227 16:36:49< Ivanovic> mordante: and shadowmaster was/is wrong with the message
20091227 16:37:07< mordante> Ivanovic, ok
20091227 16:37:17< shadowmaster> some mainline macros also rely on this convenient mechanism to store images in scenery/ so that nobody tries to find a brazier in terrain/, because braziers are not terrains
20091227 16:37:24< mordante> IMO UtbS and THoT are part of mainline ;-)
20091227 16:37:28< shadowmaster> those macros are used e.g. in THoT and IftU.
20091227 16:39:54 * shadowmaster notes that mainline doesn't require UtBS and THoT to work
20091227 16:40:44-!- Netsplit kubrick.freenode.net <-> irc.freenode.net quits: esr, Blarumyrran, knotwork, Appleman1234, deekay
20091227 16:41:09< mordante> true but then a lot more can be tagged addon ;-) but at least the commit message is now clear
20091227 16:41:09< shadowmaster> (except the test scenario, probably... which I suppose makes running wesnoth -t on Debian without installing UtBS or the -all metapackage a bit nasty
20091227 16:41:09< shadowmaster> )
20091227 16:41:20-!- loonybot [n=loonybot@ppp79-139-139-192.pppoe.spdop.ru] has joined #wesnoth-dev
20091227 16:41:20-!- v [n=Blarumyr@81-20-159-197.levira.ee] has joined #wesnoth-dev
20091227 16:41:32 * shadowmaster notes that people enjoy writing even more fuzzy commit messages here nowadays!
20091227 16:41:41-!- v is now known as Blarumyrran
20091227 16:42:08-!- loonycyborg [n=sergey@ppp79-139-139-192.pppoe.spdop.ru] has joined #wesnoth-dev
20091227 16:43:54 * shadowmaster can't wait to see http://forums.wesnoth.org/viewtopic.php?p=400333#p400333 become a spam post
20091227 16:45:00< boucman> shadowmaster: wher's the link or the add in it ?
20091227 16:45:08< shadowmaster> there's no one *yet*
20091227 16:45:10< boucman> (pure curiosity, it's obviously spammy)
20091227 16:45:34< shadowmaster> they have done this in the past: http://shadowm.rewound.net/blog/?p=24
20091227 16:45:51-!- Netsplit over, joins: deekay, Appleman1234, esr, knotwork
20091227 16:49:15< Ivanovic> off for some time
20091227 16:50:41-!- lukjad007 is now known as Superman
20091227 16:51:10-!- Superman is now known as Guest86944
20091227 16:52:04-!- Guest86944 is now known as lukjad007
20091227 17:10:34-!- Blarumyrram [n=Blarumyr@81-20-159-197.levira.ee] has quit [Read error: 110 (Connection timed out)]
20091227 17:14:27-!- Appleman1234 [n=Appleman@CPE-124-191-178-150.oxqn1.cha.bigpond.net.au] has quit [Connection timed out]
20091227 17:15:17< silene> fendrin: in the first loop, you are overwriting new_srch each time; instead you should take the minimum of all the values; in the second pair of loops, you don't have to compute new_h each time, you just have to compute the minimum of new_dsth, and move the computation of new_h outside
20091227 17:15:54< silene> fendrin: also, you should think of a better datastructure, yours is not adapted to village teleport
20091227 17:17:15< fendrin> silene: What problems do you see with village teleport?
20091227 17:17:22-!- loonybot [n=loonybot@wesnoth/bot/loonybot] has quit [Read error: 54 (Connection reset by peer)]
20091227 17:17:56< fendrin> silene: A village is just a tunnel that has every other village as target. Should do well.
20091227 17:18:38< silene> fendrin: no, it won't do well; think about it, if there are 20 villages (not so big a value), you will have a total of 400 targets; the slow down will be impressive
20091227 17:20:36< fendrin> silene: That is why I wanted to handle the both teleport types in different code parts. The village teleportation like it was before, the tunnels like the new code does it.
20091227 17:25:12< esr> zookeeper: ping?
20091227 17:25:25< zookeeper> esr, yeah?
20091227 17:26:00< esr> Couple small bugs you should look at before release...
20091227 17:26:14< esr> https://gna.org/bugs/?14429
20091227 17:26:27< esr> Is that one easily fixable?
20091227 17:26:46< shadowmaster> esr: btw, 1.7.11 has just been released
20091227 17:26:55< shadowmaster> you may want to go and close fixed bugs
20091227 17:27:09< esr> Well, he should fix it anyway! :-)
20091227 17:27:17< esr> I'll go close bugs.
20091227 17:28:52< zookeeper> esr, maybe, but i don't know how exactly.
20091227 17:29:39< esr> Who can fix terrain-transition problems? Is there someone else I should assign it to?
20091227 17:31:37< shadowmaster> no one that's not MIA at least
20091227 17:31:45< esr> Crap.
20091227 17:33:00 * boucman was thinking mog, but we havn't seen him in ages...
20091227 17:33:21< shadowmaster> yeah, mgoe is MIA as far as IRC and forum posting goes, but he's occassionally seen by phpBB
20091227 17:33:25< shadowmaster> by me too.
20091227 17:34:59< zookeeper> well, i'll take a quick another look right now
20091227 17:35:04< zookeeper> no promises
20091227 17:36:48< esr> 14 bug reports, 1 FR and 2 patches closed.  Performance not so good this cycle. 
20091227 17:37:33< shadowmaster> maybe those were really important bugs (e.g. lobby resizing, which is in the RELEASE_NOTES this time)
20091227 17:37:41< shadowmaster> (as fixed, that is)
20091227 17:38:10< shadowmaster> as the old saying goes: quality is often more important than quantity
20091227 17:39:07< boucman> Ivanovic: did you get a chance to test the env variable I mentionned earlier ?
20091227 17:40:34< mordante> I don't think all progress can be measured in closed bugs ;-)
20091227 17:40:53< esr> zookeeper: Also,  #14926 and #14927, both in the tutorial, are yours.
20091227 17:41:16< esr> mordante: Yo own about lf of the remaining ones, I think.
20091227 17:41:35< esr> s/lf/half/
20091227 17:42:56< zookeeper> esr, the tutorial isn't really my cup of tea.
20091227 17:43:13< mordante> esr, indeed and a lot of them are postponed since they all require the new gui
20091227 17:43:18< esr> zookeeper: Who owns it, then?
20091227 17:44:11< shadowmaster> The guardian of the Void does ;)
20091227 17:44:24< zookeeper> esr, i already talked about the double-movement bug with someone...probably crab
20091227 17:44:35< fendrin> guardian of the void sounds like an alias for shadowmaster ;-)
20091227 17:44:45< zookeeper> not sure if someone actually committed the fix to it alread
20091227 17:45:07< shadowmaster> fendrin: nah, it's an IftU/AtS thing
20091227 17:45:31< esr> Ivanovic: I think there is a strong case for a string-freeze exception to fix bug #14765: Negative finishing gold is not added.  That's a bug that gets right in the user's face.
20091227 17:46:42< loonycyborg> Ivanovic: 1.7.11 windows release is uploaded.
20091227 17:47:24< esr> zookeeper: If the tutorial double-move bug is tied to one of crab's issues, please leave a comment to that effect so I will know to chase him about it.  Or just assign it to him.
20091227 17:47:39< zookeeper> esr, or just set it to 0? that's perfectly clear to the user
20091227 17:48:10-!- ilor [n=user@wesnoth/developer/ilor] has quit [Remote closed the connection]
20091227 17:48:23-!- ilor [n=user@wesnoth/developer/ilor] has joined #wesnoth-dev
20091227 17:48:58< esr> zookeeper: But night have a significant effect on blalance in longer campaigns.  Not a good can of worms to open this late in the major-release cycle, IMO. 
20091227 17:50:04< zookeeper> esr, no, i meant set it to 0 so the user sees ""You will start the next scenario with 0 on top of the minimum defined starting gold."
20091227 17:50:14< fendrin> esr: setting it to 0 is the thing that happens actualy if you carry over negative gold. So no change in balancing.
20091227 17:50:28< esr> Oh, I see.
20091227 17:50:57< shadowmaster> I think I just heard the name "Jessene" int he news.
20091227 17:51:14< esr> I guess that's reasonable.  fendrin, will you dpo that and annotate the bug, please?
20091227 17:51:18< shadowmaster> or some mangled Spanish variant.
20091227 17:51:57< esr> shadowmaster: I checked art one point, it's a very rare variant.
20091227 17:54:22< fendrin> esr: dpo?
20091227 17:55:07< shadowmaster> I guess that's "do" ;)
20091227 17:55:24< fendrin> esr: I can try but I am not familiar with that part of the c++ code.
20091227 17:56:31< shadowmaster> not sure if I'm recalling wrong, but euschn and yogibear have had to do with the gamestate/gamestatus and savegame/replay code this year.
20091227 17:57:59< zookeeper> great choice of a thread to post in: http://www.wesnoth.org/forum/viewtopic.php?p=400340#p400340
20091227 17:58:25< fendrin> I guess it maybe a simple if (gold < 0) { gold = 0} thing.
20091227 18:00:38< shadowmaster> zookeeper, Ivanovic : maybe you want to split it?
20091227 18:00:53 * shadowmaster is playing the lazy unless someone finds an outstanding spambot to play with.
20091227 18:02:27-!- _teddy [n=fedor76@ppp-78-24-27-29-bras0.istra.ru] has joined #wesnoth-dev
20091227 18:05:28< fendrin> esr: The bug https://gna.org/bugs/index.php?12978 seems to be better assigned to crab, if he hasn't fixed that already. I guess dave isn't going to fix bugs in wesnoth anymore.
20091227 18:08:26< esr> shadowmaster: It wiould be a good thing i fendrin got more familiar with the C++ level.  Goddess knows I don't want to be...had quite enough of that when I emergency-fixed the AI.
20091227 18:09:44 * shadowmaster throws a template class using multi-inheritance, virtual and pure virtual methods at esr
20091227 18:09:55-!- Noyga [n=lame-z@wesnoth/developer/noyga] has joined #wesnoth-dev
20091227 18:10:05< shadowmaster> the AI always looked like some sort of swamp to me
20091227 18:10:23 * esr destroys it with a +5 Vorpal Fish of Doom.
20091227 18:10:31< shadowmaster> game.cpp is also a Swamp of Dread, actually.
20091227 18:11:06< esr> It isn't quite as bad.  I've had to do fixes in that, too.
20091227 18:12:41< esr> The difference is this: in game.cpp,  the code is a typical overcomplicated C++ mess. In the AI, it's a typical overcomplicated C++ mess *and* it's algorithmically tricky.
20091227 18:12:55< fendrin> esr: I am doing my best in getting familiar with the c++ heart of wesnoth. I have visited the wml parser/tokenizer/pathfinder/event handler/hotkeys/gamedisplay sections already.
20091227 18:13:51< esr> fendrin: That's a part I don't know well.  I've mainly looked at the old GUI, the upper levels of the game engine and the AI.
20091227 18:14:38 * fendrin could need a mentor that looks at all the evil hacks waiting on his harddisc.
20091227 18:14:50< shadowmaster> the WML parser, pathfinder and event handler components are very tricky so you'd better handle those with care; overdose of that can't be good
20091227 18:15:11< shadowmaster> the tokenizer is brain-dead simple, so I wonder if you meant something else :P
20091227 18:15:57< esr> The WML parser isn't so bad -- I managed to add a feature to the peeprocessor once.  The pathfinder is pretty nasty.
20091227 18:15:59< shadowmaster> (the pathfinder also doubles as something I could probably not even try to understand without reading a book on the A* algorithm)
20091227 18:16:55< shadowmaster> s/probably/definitely/
20091227 18:16:56< esr> I think the pathdfinder would be a goot target for being replaced by lua code.  That would make someone a good SOC project.
20091227 18:19:52< fendrin> The pathfinder isn't that bad. It's not commented but I managed to understand most parts of it with the help of silene and alink. It doesn't need a rewritting in my opinion.
20091227 18:20:06< mordante> just curious, why would a rewrite in lua make it better?
20091227 18:20:40< mordante> fendrin, I hope you then added comment so the next person can figure it out without help ;-)
20091227 18:20:55< fendrin> mordante: I did but they are in german.
20091227 18:21:16-!- wesbot changed the topic of #wesnoth-dev to: released 1.7.11, announcing "soon" | string/feature freeze active! | 65 bugs, 246 feature requests, 9 patches | logs: http://irclogs.wesnoth.org/ | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com | http://imagebin.org
20091227 18:21:46< mordante> fendrin, would be great if you translate them and commit them
20091227 18:22:14< fendrin> mordante: Can't commit, what I did to the pathfinder is far away from being in a clean state.
20091227 18:22:24< zookeeper> ok, i got a fix for the chasm<->castle thing...
20091227 18:22:49< mordante> fendrin, it's also oke to commit it later, but I just love more comment in the code
20091227 18:23:13< fendrin> mordante: but the new recruit/recall thing and the cursor stuff is nearly there.
20091227 18:23:34< mordante> still wish I had time to look at it :-(
20091227 18:29:14< fendrin> mordante: I should have commited it when it was ready a month ago. It would be stable until now and be ready for 1.8.
20091227 18:29:53< fendrin> mordante: In fact, I am using it without major problems since then.
20091227 18:30:06< shadowmaster> have you got more testers?
20091227 18:30:55< fendrin> shadowmaster: No not realy.
20091227 18:31:18< shadowmaster> would be nice that you got more testers before deploying on trunk
20091227 18:32:04< fendrin> shadowmaster: Sure, but the matter is already settled. Ivanovic spoke a might word about it. (Not sure if the term might word is known to non german speakers)
20091227 18:32:42< shadowmaster> is that like "word of power"?
20091227 18:32:43< fendrin> s/non/known
20091227 18:32:50< fendrin> shadowmaster: Yes, I guess.
20091227 18:33:45< shadowmaster> for a moment I thought it would be nice if the 1.8 branch existed before 1.8 was out so trunk could be used for unstable stuff already.
20091227 18:34:06< shadowmaster> then I remembered that I don't want to get distracted by mainline bugs and FRs. ;)
20091227 18:34:19< fendrin> shadowmaster: That is something I would like to have for a long time.
20091227 18:37:12< mordante> fendrin, I like to have a look at the code before committing it to trunk
20091227 18:37:37< mordante> and 1.8 is badly delayed, mainly due to the lobby problems being discovered that late
20091227 18:37:51< fendrin> mordante: Sure, I won't commit it without your green light.
20091227 18:38:00 * shadowmaster is attempting to fetch all branches with git-svn
20091227 18:38:24< mordante> ok
20091227 18:41:10< shadowmaster> shikadibot: log HEAD
20091227 18:41:12< shikadibot> Revision 40429 (shadowmaster, 2009-12-27 15:30:16 +0000 (Sun, 27 Dec 2009)):
20091227 18:41:15< shikadibot> Demote a filesystem interface warning to the info log level because it
20091227 18:41:18< shikadibot> gets too noisy with certain add-ons using terrain graphics from a
20091227 18:41:20< shikadibot> upper-level directory (e.g. scenery/) and is completely harmless
20091227 18:41:23< shikadibot> otherwise.
20091227 18:41:25< shikadibot> Web interface URL: http://svn.gna.org/viewcvs/wesnoth?view=rev&rev=40429
20091227 18:41:29< shadowmaster> gaa, I'm still at r38951
20091227 18:41:35< fendrin> shadowmaster: If you are going to try the gui recall/recruit stuff let me clean that branch first. I messed it up with my attempts to bring the frogatto wml schema validator in.
20091227 18:42:36< shadowmaster> I don't think I could fetch that branch yet; I'll need more bandwidth if I want to avoid spending days with the laptop on this cahir
20091227 18:43:51-!- Noyga [n=lame-z@wesnoth/developer/noyga] has left #wesnoth-dev ["Quitte"]
20091227 18:48:19-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has joined #wesnoth-dev
20091227 18:48:44< crimson_penguin> Ivanovic: ...so my internet's been down since yesterday
20091227 18:48:57 * crimson_penguin is on dialup
20091227 18:50:29< shadowmaster> crimson_penguin: that sucks :(
20091227 18:50:47< crimson_penguin> it being a holiday sunday, I doubt they'll fix it until tomorrow
20091227 18:52:46< fendrin> mordante: I am going to takeover bug https://gna.org/bugs/index.php?11276 if you don't mind.
20091227 18:55:02< mordante> fendrin, no no problem
20091227 18:56:48< crimson_penguin> man, I hate Windows, and MSVC++
20091227 18:57:17< crimson_penguin> the first problem was that you have to include windows.h before you include gl/gl.h (why doesn't it include it itself?!)
20091227 18:57:59< crimson_penguin> so I just made VC++ force-include windows.h on everything, so as to not modify the code... but now boost::asio doesn't work if windows.h is included before it...
20091227 18:58:19< fendrin> esr: https://gna.org/bugs/index.php?13655 seems more important to normal to me. Maybe silene wants to take care of it.
20091227 18:58:23< crimson_penguin> so there's 1 file that includes gl/gl.h, and then boost/asio.h
20091227 18:58:31< fendrin> s/to/than
20091227 18:58:56< CIA-28> zookeeper * r40430 /trunk/data/core/ (terrain-graphics/base.cfg terrain-graphics.cfg): Fixed bug #14429 by tweaking KEEP_BASE a bit and using it instead of TERRAIN_BASE for castle and keep floors.
20091227 18:59:10< zookeeper> esr, ^
20091227 19:02:00< shadowmaster> successfully fetched all branches, yays \o/
20091227 19:09:29< silene> fendrin: no, this is hardly in my domain; the bug is not related to lua, it is related to the --multiplayer option: games started this way are not started by the standard engine; several bugs have already been reported about this (e.g. issues with eras); if --multiplayer games were fixed so that they are played by the same engine than single and multiplayer games, all these bugs would vanish
20091227 19:10:47< fendrin> silene: Is this only when using --multiplayer from the commandline or is it also when selecting multiplayer from ingame?
20091227 19:11:06< silene> fendrin: no, only when using the command line
20091227 19:18:29-!- ilor [n=user@wesnoth/developer/ilor] has quit [Read error: 60 (Operation timed out)]
20091227 19:19:44-!- ilor [n=user@wesnoth/developer/ilor] has joined #wesnoth-dev
20091227 19:29:32-!- loonycyborg [n=sergey@wesnoth/developer/loonycyborg] has quit ["KVIrc 3.4.2 Shiny http://www.kvirc.net/"]
20091227 19:44:49< Ivanovic> re
20091227 19:46:29< crimson_penguin> hi Ivanovic 
20091227 19:50:20< CIA-28> fendrin * r40431 /trunk/src/playsingle_controller.cpp: 
20091227 19:50:20< CIA-28> Set the amount of gold carried over to 0 in the carryover dialog if it is negative.
20091227 19:50:20< CIA-28> That fixes bug #14765 without a string frezze.
20091227 19:50:20< CIA-28> I am not sure if the bug report should be closed or kept open for a stringchange after the release.
20091227 19:50:22< fendrin> esr: ^
20091227 19:56:16< mordante> ilor, how easy would it be to switch to diff the user and the game list in the lobby
20091227 19:56:52< ilor> mordante: as in don't clear but use remove/add?
20091227 19:57:20< mordante> yes
20091227 19:58:41< ilor> well, easy, but will need to track a bunch more stuff, so might take a while
20091227 19:59:32< mordante> well it should do some speeding up as well
20091227 20:00:19< mordante> I tested with some other changes, but that didn't help...
20091227 20:00:30< mordante> and then I noticed I still used the debug containers of gcc for testing... so need to retest later 
20091227 20:00:40< ilor> I'll get on it, maybe I'll start with the player list
20091227 20:00:52< mordante> thanks
20091227 20:00:53< ilor> since it's easier to see that being slow when up-downing on the gamelist
20091227 20:01:14< fendrin> silene: What is your opinion: seperate code for tunnels/village teleport or search for a better data structure?
20091227 20:05:08< Ivanovic> crimson_penguin: i hope your internet will be working as usual again soon
20091227 20:05:28< Ivanovic> crimson_penguin: if it does already work again on monday there should not be too much of a problem with the release
20091227 20:05:37< Ivanovic> then i will most likely just delay by a day and that's it
20091227 20:05:43< Ivanovic> esr: btw feel free to close bug reports
20091227 20:05:57< crimson_penguin> well, I should be able to use my brother's internet if nothing else
20091227 20:06:04< crimson_penguin> possibly even from here
20091227 20:09:35< silene> fendrin: i would go for a better data structure; i don't think it has to be that different: just add a pointer before the set, so that all the villages point to the same set; and introduce another datastructure (e.g. a vector) that contains all the destinations without any duplicate and that is used just by the heuristic; this second set can be precomputed from the first one
20091227 20:10:09< fendrin> silene: Okay, thank you for the help.
20091227 20:22:26-!- shadowm_bluecore [n=ignacio@wesnoth/developer/shadowmaster] has quit ["Lost terminal"]
20091227 20:33:48-!- zookeeper2 [n=l@88-148-251-223.bb.dnainternet.fi] has joined #wesnoth-dev
20091227 20:35:58< shadowmaster> that stevejoe guy is creeping me out
20091227 20:54:23-!- cib0 [n=cib@p5DC43116.dip.t-dialin.net] has joined #wesnoth-dev
20091227 20:56:07-!- ilor [n=user@wesnoth/developer/ilor] has quit [Remote closed the connection]
20091227 20:56:22-!- ilor [n=user@wesnoth/developer/ilor] has joined #wesnoth-dev
20091227 20:57:08-!- zookeeper [n=l@wesnoth/developer/zookeeper] has quit [Read error: 110 (Connection timed out)]
20091227 20:58:52< cib0> are any forum moderators here? could you please look at http://www.wesnoth.org/forum/viewtopic.php?f=15&t=26415&start=75? i am not sure what the standard procedure there is
20091227 21:04:38-!- dtiger [n=dtiger@dynamic-vpdn-91-149-132-17.telecom.by] has quit [Remote closed the connection]
20091227 21:26:44-!- ilor_ [n=user@wesnoth/developer/ilor] has joined #wesnoth-dev
20091227 21:30:33-!- loonybot [n=loonybot@ppp79-139-139-192.pppoe.spdop.ru] has joined #wesnoth-dev
20091227 21:31:21-!- loonycyborg [n=sergey@ppp79-139-139-192.pppoe.spdop.ru] has joined #wesnoth-dev
20091227 21:39:10-!- ilor [n=user@wesnoth/developer/ilor] has quit [Read error: 110 (Connection timed out)]
20091227 21:40:04-!- silene1 [n=plouf@AMarseille-251-1-67-231.w83-205.abo.wanadoo.fr] has joined #wesnoth-dev
20091227 21:40:16-!- EL_Caballero [n=saul@201.243.27.90] has joined #wesnoth-dev
20091227 21:40:33< EL_Caballero> Hello
20091227 21:40:54< EL_Caballero> is anybody there?
20091227 21:41:06< loonycyborg> Yes. Why?
20091227 21:41:38< EL_Caballero> ok...  i am trying to change a little the Conquest of Europe...
20091227 21:42:00< EL_Caballero> I know that the map can be changed with the editor map.
20091227 21:42:15< EL_Caballero> But... i do not know how change the .cfg file. 
20091227 21:42:25< EL_Caballero> is it did manually?
20091227 21:43:01< EL_Caballero> do you undestand me?
20091227 21:43:02< loonycyborg> Almost everything except map making is done by editing wml.
20091227 21:43:36< loonycyborg> There are specific tools to aid for particular tasks though.
20091227 21:44:00< EL_Caballero> ok. I want to fix/change the configuration of Conquest Europe.
20091227 21:44:03< EL_Caballero> Exactly.
20091227 21:44:12< EL_Caballero> do you know that tools?
20091227 21:44:53< EL_Caballero> i do not believe that it is do manually.
20091227 21:44:56< fendrin> EL_Caballero: There is an editor called emacs (not quite the easiest one) but there is a plugin for it that does tab completion for wml code.
20091227 21:45:29< loonycyborg> That's probably should be done manually. I only know of trackplacer to make tracks on maps and an animation editor among specific tools.
20091227 21:46:18< zookeeper2> ...just use a text editor
20091227 21:46:26< zookeeper2> if you don't know what a text editor is, use notepad
20091227 21:46:34-!- zookeeper2 is now known as zookeeper
20091227 21:46:50< EL_Caballero> hmmm... ok ok
20091227 21:47:16< EL_Caballero> thanks Fendrin... i will search something about wml.
20091227 21:47:28< EL_Caballero> thanks Zookeeper :).
20091227 21:47:58< fendrin> EL_Caballero: You won't find something usefull for wml outside the wesnoth project. That language is wesnoth specific.
20091227 21:48:14< EL_Caballero> hmmm... ok
20091227 21:49:05< fendrin> EL_Caballero: http://wiki.wesnoth.org/ReferenceWML
20091227 21:49:05< EL_Caballero> so... i will read detailly the file.
20091227 21:49:14< EL_Caballero> wow, excellent :)
20091227 21:49:32-!- EdB [n=edb@60.12.95-79.rev.gaoland.net] has quit [Remote closed the connection]
20091227 21:50:12< fendrin> EL_Caballero: Good luck with your project :-)
20091227 21:50:42< EL_Caballero> thanks fendrin :-D. 
20091227 21:51:14< EL_Caballero> if it will be sucessfully, i will post it in the forum :D
20091227 21:54:26< fendrin> EL_Caballero: You can upload it to the add-on server as well. That way you will reach more poeple.
20091227 21:55:33-!- silene [n=plouf@wesnoth/developer/silene] has quit [Read error: 110 (Connection timed out)]
20091227 22:08:33< CIA-28> mordante * r40432 /trunk/src/gui/widgets/debug.cpp: Fix a compilation error.
20091227 22:13:50< mordante> I'm off night
20091227 22:14:28-!- mordante [n=mordante@wesnoth/developer/mordante] has quit ["Leaving"]
20091227 22:16:46-!- ancestral [n=ancestra@mobile-166-137-136-107.mycingular.net] has joined #wesnoth-dev
20091227 22:17:28-!- ancestral [n=ancestra@mobile-166-137-136-107.mycingular.net] has quit [Client Quit]
20091227 22:18:34-!- giusef [n=giusef@unaffiliated/giusef] has joined #wesnoth-dev
20091227 22:27:52< esr> fendrin: Problem with bug #14765 solved.
20091227 22:30:01-!- crimson_pinvin [n=ben@64.201.60.211] has joined #wesnoth-dev
20091227 22:30:50-!- Blarumyrran [n=Blarumyr@81-20-159-197.levira.ee] has quit [Read error: 54 (Connection reset by peer)]
20091227 22:32:04< fendrin> esr: good solution.
20091227 22:32:56< esr> fendrin: Is WOV scenario 2 playable at this point?  Can I test it?
20091227 22:38:23< fendrin> esr: no it isn't and you will need to use my c++ code base which isn't in an own svn branch right now.
20091227 22:39:05< esr> So we're not going to be able to even test it before 1.9?
20091227 22:39:17< fendrin> esr: Not with svn trunk.
20091227 22:39:39< esr> Ouch.
20091227 22:40:12< fendrin> Not that bad, I am going to make it work during the week and branch wesnoth once more for the new pathfinding code.
20091227 22:46:32-!- Crab_ [n=Crab_@wesnoth/developer/crab] has joined #wesnoth-dev
20091227 22:47:09< crimson_pinvin> turns out the problem was ice on the antenna; I'm back on semi-fast internet now
20091227 22:48:59< fendrin> crimson_pinvin: May I ask where you live?
20091227 22:49:11< crimson_pinvin> Canada
20091227 22:49:22< fendrin> crimson_pinvin: The british or the french part?
20091227 22:49:37< crimson_pinvin> British :P Ontario
20091227 22:50:02< crimson_pinvin> I live in the middle of nowhere though; not in a city or even a town
20091227 22:50:27< fendrin> crimson_pinvin: I assumed that, if your internet is going via an antenna.
20091227 22:50:48< Ivanovic> mordante: clicking makes no difference between GUI1 and GUI2
20091227 22:50:49< Crab_> hmm, so, 1.7.11 is out here..
20091227 22:50:56< Ivanovic> mordante: cf the #wesnoth logs
20091227 22:50:58-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has quit [Nick collision from services.]
20091227 22:50:58-!- crimson_pinvin is now known as crimson_penguin
20091227 22:50:59< fendrin> hi Crab_ :-)
20091227 22:51:02< Crab_> hi fendrin
20091227 22:55:44< Crab_> Ivanovic: do you know the cause of the bug that you addressed in r40423 ?
20091227 22:56:03-!- zookeeper [n=l@wesnoth/developer/zookeeper] has quit []
20091227 22:57:23< CIA-28> fendrin * r40433 /branches/fendrin_pathfind/: Branch for the pathfinding code that enables custom teleporters.
20091227 22:58:28< Ivanovic> Crab_: no idea
20091227 22:58:42< Ivanovic> Crab_: i just know that stikonas pointed out that the unit ain't there anymore
20091227 23:00:10< Crab_> ok. the old code ( in unit.cpp) assumed that 'if no side is given in create [unit] in [event] tag, then the side is default, 1. my rewrite of the team construction/adding units in src/game_events.cpp does a more strict checking and assumes 'no side given - error'.
20091227 23:00:22-!- EL_Caballero [n=saul@201.243.27.90] has quit ["Leaving"]
20091227 23:01:18< Crab_> therefore, the question is, what should be done, (a) keep the old behavior or (b) keep the more strict checking for possible errors (as the log message is quite clear about 'what went wrong')
20091227 23:02:28< silene1> Crab_: lots of wml action tags assume that when no side is given, side is 1; there is no reason [unit] behaves differently
20091227 23:02:33-!- silene1 is now known as silene
20091227 23:03:05< Ivanovic> Crab_: at least post a warning about this
20091227 23:03:06< Crab_> silene: ok. then I'll patch game_events.cpp it to assume that.
20091227 23:03:22< Ivanovic> since it clearly does break compatibility and probably several other places are affected, too
20091227 23:03:32< Crab_> Ivanovic: i.e. 'use default=1 and log a warning' ?
20091227 23:03:41< Ivanovic> jupp
20091227 23:03:51< Crab_> (yes, this is unintended side effect)
20091227 23:05:00< silene> no, please don't put a warning; or put for all the wml tags; again, there is no reason [unit] behaves differently from the other tags
20091227 23:05:05< Crab_> ok
20091227 23:08:11< Ivanovic> silene: i'd say that every entry should be clear, so a warning for all tags that just assume side1 is meant is probably in place
20091227 23:09:16< Ivanovic> so yeah, probably *many* warnings will appear soon
20091227 23:09:18< Ivanovic> ;)
20091227 23:10:20< Ivanovic> the reasoning behind this is that it might help with getting wml a little clearer and later on makes things maybe a little easier to eventually port to multiplayer stuff (mp campaigns and all the likes)
20091227 23:12:01< silene> Ivanovic: so, you mean a real warning? one that appears on screen?
20091227 23:12:17< Ivanovic> silene: for the moment not
20091227 23:12:20< Ivanovic> but later on: sure
20091227 23:12:26< silene> then what's the point.
20091227 23:12:28< silene> ?
20091227 23:12:29< Crab_> Ivanovic: maybe patch wmllint instead ?
20091227 23:12:56< Ivanovic> if we directly add it at the moment it will be *lots* of work to update all of mainline to get rid of the warnings there...
20091227 23:14:19< Ivanovic> silene: the point is that we can "soon" fix things anyway and are not forced to hurry
20091227 23:14:44< silene> Ivanovic: what do you mean "fix"? you don't fix something that is not broken
20091227 23:15:09< Ivanovic> s/fix/improve
20091227 23:15:16< stikonas> maybe postpone to 1.9
20091227 23:16:06< stikonas> otherwise 1.8 will never be ready :)
20091227 23:16:12< Crab_> stikonas: it should be tied with our efforts to produce a 'wml schema'
20091227 23:16:32-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has quit [Read error: 113 (No route to host)]
20091227 23:16:33< Crab_> stikonas: as this is clearly a schema issue - 'is an attribute required or not?'
20091227 23:17:13< Crab_> Ivanovic: how do you playtest campaigns before releasing ? 'do they start' or 'does each scenario loads if we :cl' ?
20091227 23:17:24< Ivanovic> just a plain "do they start"
20091227 23:17:31< Crab_> ok
20091227 23:17:34< Ivanovic> since not all campaigns can really be done using :cl anyway
20091227 23:18:08< Ivanovic> (beside this i would probably take >1h to test all campaigns with :cl, plain starting already takes a while...)
20091227 23:18:57 * Ivanovic would love to have some "start all campaigns installed" method
20091227 23:19:15-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has joined #wesnoth-dev
20091227 23:19:23< Ivanovic> maybe even combined with "go through all scenarios"
20091227 23:19:25< Ivanovic> ;)
20091227 23:19:28< shadowmaster> that would be so utterly useless.
20091227 23:19:30< silene> Crab_: i don't believe in wml schemas; especially not now that wml tags can be modified on the fly (and some already are in mainline, so mainline wouldn't even validate against a wml schema)
20091227 23:20:41< shadowmaster> heh, sounds like my argument some days ago.
20091227 23:20:44< Crab_> silene: yes, but even a partial solution (e.g wmllint that we have now) is better than nothing. so, imo, the ability to statically check wml should be improved
20091227 23:21:49< Crab_> silene: btw, what might be the best way to allow modification (from scenario files) of unit lists at scenario start, *before* those units are parsed by game engine to place units/leaders on map ?
20091227 23:22:20< Crab_> silene: currently, 'placing initial units' is done even before 'teams' vector is fully constructed
20091227 23:22:48< silene> Crab_: i'm not sure to understand; just place the units inside a prestart event, so it will happen later
20091227 23:25:16< Crab_> even 'prestart' is sometimes too late and requires additional code to fix things up. Example: in previous scenarios, player 1 has heroes A and B. We need B to play for side 3 in this scenario. we need B to have the right XP/items/etc - as B was previously 'owned' by player. Also, we need a new B to be created if we just :cl to this scenario.
20091227 23:25:59< Crab_> (this sometimes happens in LoW, in SotBE)
20091227 23:27:04< silene> how can it be too late? the screen is still black at the time of the prestart event
20091227 23:28:37< CIA-28> fendrin * r40434 /branches/fendrin_pathfind/src/pathfind.hpp: changed the declaration of the a_star_search method to take the new custom teleporters as an argument.
20091227 23:29:37< Crab_> yes, the screen is still black, but the team construction has already taken place, so it is not possible to write things in simple declarative style i.e. 'transfer B to side 3 and let team construction handle any placement/duplication issues', we need to do things like 'if side 1 has B, then find whether side 3 has B, and if side3's B is on table, replace him, otherwise, replace B on recall list of side 3'
20091227 23:30:03-!- Blueblaze [n=nick@adsl-76-202-22-180.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20091227 23:30:39< Crab_> also, there might be sides with save_id's that should be hidden and not present in this scenario but which can contribute units to the scenario's starting positions.
20091227 23:30:40< CIA-28> fendrin * r40435 /branches/fendrin_pathfind/src/astarsearch.cpp: implement custom teleporters in the a_star_search algo/method.
20091227 23:32:05< Crab_> so, if there was a way to mess with the units stored in 'persistent sides' which are present in the start-of-scenario save, it would allow more easier manipulation of units which are stored in the campaign.
20091227 23:33:05< silene> Crab_: are you saying that you want to add a wml tag that allows access to persistent sides? i don't mind
20091227 23:33:21< Crab_> silene: yes. and to allow such wml to be executed *before* teams are constructed
20091227 23:33:38< silene> Crab_: no way
20091227 23:34:18< Crab_> silene: maybe split team construction in two stages ? e.g. 1) create all teams, don't place leader,units  2) fire a special event 3) place leader, units ?
20091227 23:34:22< esr> fendrin: You're aware that I made a map for scenario 4?
20091227 23:35:09< fendrin> esr: I have had a look on it yesterday, it's already nice.
20091227 23:35:18< silene> Crab_: i would prefer: let's get rid of team construction, and let wml take care of whatever it wants to take care
20091227 23:36:30< esr> fendrin: I'll do the modified Sunrise Island map for scenario 5 next.
20091227 23:36:53< Crab_> silene: we cannot get rid of team construction per se, but we can recode it in WML or lua (providing a default implementation which will fire wml events to allow pluggability). 
20091227 23:37:37< fendrin> esr: I am still not a help with the maps. My lacking of playing skills with drakes is a real problem when it comes to mapmaking.
20091227 23:38:13-!- Espreon [n=espreon@wesnoth/developer/espreon] has joined #wesnoth-dev
20091227 23:38:40< silene> Crab_: why can't we get rid of it? the only thing that matters is that std::vector<team> is filled at scenario start, everything else can be done from wml
20091227 23:38:46< esr> fendrin: That's OK, I can do maps.  Except I'm not sure I can do a good one for 2 because I don't understand what you're doing with the teleporters.
20091227 23:40:55< fendrin> esr: That is easy. There are two maps. An underground one and an overground one. I am going to merge this map into a single one. They are separeted to ease the editing. Both maps have at the same locations the connections between them. On the overground map it's they are  cave tiles, on the underground map the entrances are marked by the "lit" overlay.
20091227 23:41:10< Crab_> silene: well, we need at least a default implementation of 'find out the correct qty of gold and handle gold carryover,set team objectives,  If the game state specifies additional units that can be recruited by the player, add them, place leader, place units (handling issues with duplicated units)'. language in which this is written doesn't matter, of course, but we need to provide a 'default implementation' and allow it to be extended from WML.
20091227 23:43:00< silene> Crab_: why do we need that? how many scenarios do you think there are that don't already set objectives in their (pre)start events? why would gold carryover be handled at scenario start? why can't the leader be placed from an event? and so on
20091227 23:43:36< silene> i really don't understand what your default implementation is supposed to do
20091227 23:44:29< Crab_> silene: you can see either the old version of the code, or the new version of the code ( class team_builde ), in gamestatus.cpp
20091227 23:45:05< Crab_> basically, getting even 'place leader' right is quite hard forrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr
20091227 23:45:11< Crab_> (sorry, bug)
20091227 23:45:58< Crab_> ..is quite hard for wml creator - e.g. many mainline campaigns, such as HttT break when we use :cl to skip scenarios
20091227 23:46:47< boucman> we should have an "emergency win" event in WML that would be called on :cl
20091227 23:47:17< Crab_> boucman: this is already possible, but it isn't enough
20091227 23:47:29-!- cib0 [n=cib@p5DC43116.dip.t-dialin.net] has quit [Remote closed the connection]
20091227 23:47:40< Crab_> boucman: there are other issues such as 'loading old mid-campaign start-of-scenario saves'
20091227 23:48:46< Crab_> boucman: e.g if we add a new 'required' character to the campaign, we need to make sure he's created if he doesn't exist to allow people who completed half-of-a-campaign earlier, to continue
20091227 23:50:26< Crab_> for example, LoW previously used a system which depended on an 'earlier' 'betrayal' scenario to set things right for 'final' scenario, which, of course, broke badly wrt old saves ('each scenario set things up for itself' is a lot more robust implementation)
20091227 23:50:45< silene> Crab_: it's already possible to do that from wml; i don't see the point of making the c++ even uglier that it already is (and no, your recent commits did not improve the readability of it)
20091227 23:51:57< Crab_> silene: main aim of my recent commits was not 'readability improvements', but to get rid of some old hacks, and to allow [unit] tags in [side] and [event] to behave the same, removing code duplication
20091227 23:53:46< silene> Crab_: removing code duplication? is that a joke? your commit added 200 lines of code
20091227 23:54:44-!- stikonas_ [n=and@wesnoth/translator/stikonas] has joined #wesnoth-dev
20091227 23:54:44-!- stikonas_ [n=and@wesnoth/translator/stikonas] has quit [Read error: 54 (Connection reset by peer)]
20091227 23:55:13< Crab_> silene: no, I'm about add_unit functionality which was previously duplicated and behaved differently in [side][unit], [side]for_leader_directly_in_side and [event][unit] tags
20091227 23:55:43< Crab_> silene: but, the original question is 'to allow each scenario properly set up for itself, we need WML access to all the persistent sides before things that c++ code which places units  is activated. what is the best way to do that?'
20091227 23:56:32< silene> my answer has not changed; we get rid of c++ placing units, so that scenarios can place them whichever way they want whenever they want
20091227 23:58:10< Crab_> silene: my comment to that answer was 'then, we need to provide a WML/lua implementation of 'placing initial units', to avoid the need to reimplement the wheel for scenario creators'
20091227 23:59:06< silene> but it already exists! what's wrong with the current [unit] tag in the [event] tag; as you just said, it now behaves exactly the same way then the [unit] tag in the [side] tag!
--- Log closed Mon Dec 28 00:00:09 2009