--- Log opened Sun Oct 11 00:00:21 2009
20091011 00:00:30< shadowmaster> Espreon: do you hav all build dependencies installed?
20091011 00:00:37< Espreon> ... Yes.
20091011 00:00:42< shadowmaster> in this case, some libfribidi
20091011 00:00:57< Espreon> I have fribidi...
20091011 00:01:04< shadowmaster> I read that.
20091011 00:01:20< shadowmaster> let me check what's up with font.cpp
20091011 00:02:00-!- SonIcco [n=SonIcco@pD9510BF9.dip0.t-ipconnect.de] has quit [Remote closed the connection]
20091011 00:02:39< shadowmaster> Espreon: that include was added on August 1st
20091011 00:02:55< shadowmaster> so um, it should have worked all this time.
20091011 00:03:38< Espreon> Yeah...
20091011 00:04:04< shadowmaster> Espreon: did you perform some upgrade of libfribidi lately?
20091011 00:04:21< Espreon> Nope.
20091011 00:04:54< shadowmaster> are you using scons, cmake or autotools?
20091011 00:05:33< Espreon> scons.
20091011 00:05:46< shadowmaster> okay, did you try scons --config=force just in case?
20091011 00:06:36< Espreon> I shall do that...
20091011 00:15:27< CIA-13> ilor * r39235 /branches/1.6/ (changelog src/editor2/editor_controller.cpp): fix bug #14266 (port r39103 from trunk)
20091011 00:15:43< ilor> Espreon: ^, finally
20091011 00:15:50-!- Blueblaze [n=nick@adsl-99-171-161-30.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20091011 00:16:10< Espreon> Heh...
20091011 00:19:56 * shadowmaster notes that trunk is perfectly healthy and compiles successfully
20091011 00:20:07-!- Blueblaze [n=nick@adsl-99-171-161-30.dsl.hstntx.sbcglobal.net] has quit [Remote closed the connection]
20091011 00:21:36< loonycyborg> Espreon: Is there /usr/include/fribidi.h file?
20091011 00:21:56< shadowmaster> mine is in  /usr/include/fribidi/fribidi.h
20091011 00:22:11< shadowmaster> and yes, scons bothers to pass -I/usr/include/fribidi to the compiler here
20091011 00:22:27< Arc> oh yea i discovered what the difference was between me getting squares and others getting weird glyphs - i didnt have a recent version of fribidi installed, once i upgraded i got weird glyphs too
20091011 00:22:29< Espreon> Yes, it exists.
20091011 00:22:40< loonycyborg> That's because it uses pkg-config.
20091011 00:23:28< loonycyborg> Espreon: What does running 'pkg-config --libs --cflags fribidi' give?
20091011 00:24:30< Espreon> -I/usr/include/fribidi  -lfribidi
20091011 00:27:10< loonycyborg> What's the command-line just before the error?
20091011 00:28:21< Espreon> I do not remember; the information is probably lost by now. Let us see how this compilation goes... OK?
20091011 00:29:05< shadowmaster> you can save yourself some time by not replying until your build gets to that point.
20091011 00:29:26< Espreon> ...
20091011 00:29:30< shadowmaster> also, some terminal emulators (e.g. Konsole) have this nice thing called "infinite scrollback"
20091011 00:29:49< shadowmaster> screen can also be set to have a ridiculously big scrollback
20091011 00:36:02-!- Netsplit pratchett.freenode.net <-> irc.freenode.net quits: esr, Ivanovic, Doppp, Aethaeryn
20091011 00:36:46-!- Netsplit over, joins: Ivanovic, Aethaeryn, esr, Doppp
20091011 00:36:46-!- MikeJB [n=Michael@69.251.9.23] has joined #Wesnoth-dev
20091011 00:37:42-!- MikeJB [n=Michael@69.251.9.23] has left #Wesnoth-dev []
20091011 00:45:43< Aethaeryn> Espreon: Did you solve your compile problem?
20091011 00:45:56< Espreon> Still compiling...
20091011 00:46:11< shadowmaster> thought you had a more powerful machine
20091011 00:46:24< Espreon> And done...
20091011 00:46:27< CIA-13> alink * r39236 /trunk/src/ai/testing/ca_testing_move_to_targets.cpp: 
20091011 00:46:27< CIA-13> Commit experimental bugfix for defaut AI in RCA AI (exerimental targeting)
20091011 00:46:27< CIA-13> Fix a very old bug making the AI rates targets as if they were all reachable in 1 turn.
20091011 00:46:27< CIA-13> AI should now prefer targets reachable in 2 turns against one needing 3 (if same value).
20091011 00:46:27< CIA-13> Note that it only affects targets needing more than 1 turn to reach.
20091011 00:46:30< CIA-13> This also make "complex targetting" works and should make the AI pick the best unit
20091011 00:46:31< Espreon> Now to install it...
20091011 00:46:32< CIA-13> to reach the best target (probably begin by sending scouts grab undefended villages)
20091011 00:46:39< shadowmaster> Espreon: so it works?
20091011 00:46:48< Espreon> Yes.
20091011 00:46:58< shadowmaster> and why install it? if you are tracking trunk it tends to be more useful to run it from the source tree
20091011 00:47:06< shadowmaster> (since installing take ages for some, e.g. me)
20091011 00:47:27< loonycyborg> Install is just waste of disk space.
20091011 00:47:33< alink> Crab_ : ^one-liner committed, seems to work as expected but i am currently evaluating its usefulness
20091011 00:47:46 * Espreon wants his disk space to be wasted.
20091011 00:47:48< loonycyborg> Just run it as './wesnoth' from the checkout dir.
20091011 00:47:51 * Espreon 's system just works fine.
20091011 00:48:00< Espreon> *works just fine
20091011 00:48:22< Crab_> alink: ok. I can launch an ai-vs-ai test in ~20h, to see if there any effect. which MP map do you suggest ?
20091011 00:48:48< alink> a big map is needed
20091011 00:49:03-!- Espreon [n=espreon@wesnoth/developer/espreon] has quit ["WRYYYYYYYYYYYYYYYYYYYY!"]
20091011 00:49:28< Crab_> 2p_Cynsaun_Battlefield is ok ?
20091011 00:49:32< alink> as mentionned it affect comparaison of targets needing several turns to reach
20091011 00:50:14< alink> Crab_: give me one minute, I check it that map print a lot of my debug code about this
20091011 00:52:11< alink> Crab_: yes seems good
20091011 00:52:33< Crab_> ok
20091011 00:53:07< alink> hum that bugfix will really hurt one of my recent optimization, but I think i can adapt it if needed.
20091011 00:54:15-!- Espreon [n=espreon@wesnoth/developer/espreon] has joined #wesnoth-dev
20091011 00:59:58< alink> We shoud give a chatterbot feature to the different AIs and send them to the MP servers to see which one is best against humans ;-p
20091011 01:00:41< alink> unless someone already do that.
20091011 01:01:07< Crab_> alink: well, it's too early for that - the ai must have a good enough chance to deal with humans before that (unless we fiddle with ssrng to favor AIs :p , of course) :)
20091011 01:01:45< Aethaeryn> The self-aware AI that eliminates mankind will originate from Wesnoth... :P
20091011 01:02:01-!- Zarel [n=Zarel@warzone2100/developer/Zarel] has quit ["This computer has gone to sleep"]
20091011 01:02:04< Crab_> alink: there are plans to add a chatterbox feature, but, a good way to intercept/analyze some game events is required to make it useful.
20091011 01:02:43< Crab_> e.g., allow rules like "say something from "mha-ha-ha-ha" list if a certain enemy unit misses 4/4 shots"
20091011 01:03:50-!- krasicki [n=kvirc@public15523.xdsl.centertel.pl] has joined #wesnoth-dev
20091011 01:04:00-!- krasicki [n=kvirc@public15523.xdsl.centertel.pl] has quit [Read error: 54 (Connection reset by peer)]
20091011 01:04:14-!- ilor [n=user@wesnoth/developer/ilor] has quit [Read error: 113 (No route to host)]
20091011 01:06:09-!- Aethaery1 [n=Michael@wesnoth/umc-dev/developer/aethaeryn] has joined #Wesnoth-dev
20091011 01:06:12< alink> alternatively, create some public challenge against experimental AI (invisibly switching the AI used) and see which one is the best
20091011 01:07:25-!- Aethaeryn [n=Michael@wesnoth/umc-dev/developer/aethaeryn] has quit [Nick collision from services.]
20091011 01:07:25< alink> mmh human learning AI flaws will maybe hurt the statical analysis of that
20091011 01:07:27-!- Aethaery1 is now known as Aethaeryn
20091011 01:07:35< alink> *statistical
20091011 01:09:13< Crab_> well, the current ai frameworks have huge problems with ZoC. until this is solved, there are well-known flaws in ai thinking that can be exploited.
20091011 01:09:47-!- Nayela [n=Nayela@cpc3-lich6-0-0-cust673.brhm.cable.ntl.com] has quit [Remote closed the connection]
20091011 01:09:58< alink> yeah but finding the less bad AI is already an info
20091011 01:10:19< Crab_> yes
20091011 01:11:29< alink> I mentionned AI flaw, because if same humans play against AI1, learns its flaws, then play against AI2 (which have the same flaws). AI2 may lose more regardless of its other improvements
20091011 01:12:13< Crab_> this particular bias is eliminated by having humans play vs AIx, where x is either 1 or 2.
20091011 01:12:53< Crab_> unless the human will be able to guess which AI he's facing and switch to a suitable counter-strategy :)
20091011 01:13:20-!- ilor [n=user@wesnoth/developer/ilor] has joined #wesnoth-dev
20091011 01:13:38< alink> yeah but it's a problem for coding incremental improvement of the AI
20091011 01:14:52< alink> unless you find fresh testers each time you have a new code needing testing
20091011 01:14:58< Crab_> :)
20091011 01:19:49-!- Espreon [n=espreon@wesnoth/developer/espreon] has quit [Remote closed the connection]
20091011 01:20:23 * alink just realizes now the recursive complexity of writing an AI knowing that it will play against itself
20091011 01:21:38< Crab_> if the AI has position evaluation function, it can be reformulated in a less-complex way, using game theory
20091011 01:22:42< Crab_> ok, time to sleep for me. night.
20091011 01:22:48< alink> yeah i suppose that it's a classical game theory problem, but i never thought about it with the write-such-AI-for-wesnoth angle
20091011 01:22:52< alink> Crab_: gn
20091011 01:22:55-!- loonycyborg [n=sergey@wesnoth/developer/loonycyborg] has quit ["Zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz"]
20091011 01:23:30< Crab_> well, one of the ai approaches I'm (albeit slowly) coding now, has position analyzer. 
20091011 01:23:39-!- Crab_ [i=crab@wesnoth/developer/crab] has quit ["Leaving."]
20091011 01:23:43-!- loonybot [n=loonybot@wesnoth/bot/loonybot] has quit [Remote closed the connection]
20091011 01:31:39-!- ilor [n=user@wesnoth/developer/ilor] has quit [Read error: 113 (No route to host)]
20091011 01:32:08-!- stikonas [n=and@wesnoth/translator/stikonas] has quit [Remote closed the connection]
20091011 01:33:39-!- stikonas [n=and@bcm-131-111-216-119.girton.cam.ac.uk] has joined #wesnoth-dev
20091011 01:34:39-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has joined #wesnoth-dev
20091011 01:35:07< alink> aah good the fixed AI won 3 games on 3 against old AI :-)
20091011 01:35:21< alink> so at least it's not a breaking it
20091011 01:35:37< alink> *not breaking it
20091011 01:35:38-!- Espreon [n=espreon@wesnoth/developer/espreon] has joined #wesnoth-dev
20091011 01:37:56-!- stikonas [n=and@wesnoth/translator/stikonas] has quit [Read error: 104 (Connection reset by peer)]
20091011 01:39:54< Arc> don't make it too hard :-)
20091011 01:40:03-!- stikonas [n=and@bcm-131-111-216-119.girton.cam.ac.uk] has joined #wesnoth-dev
20091011 01:40:04< Arc> some of us rely on the AI not being the brightest in the world
20091011 01:40:26< Arc> like how it always sends its troops after a single lone scout that can play chicken while the powerful advace comes in from the other side
20091011 01:40:46< Arc> "ooh look, he's weak and alone!"
20091011 01:42:17< Espreon> Indeed.
20091011 01:43:00< Aethaeryn> Or the 41/42 XP vanguard.
20091011 01:43:17< Aethaeryn> "Oh look, he's about to level, let's ignore him and whatever he's blocking with his ZoC"
20091011 01:43:56-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has quit []
20091011 01:45:12-!- stikonas [n=and@wesnoth/translator/stikonas] has quit [Remote closed the connection]
20091011 01:45:38< alink> don't worry, this change only make the AI go for the nearest target instead of ignoring target distance
20091011 01:46:10< alink> I suspect it will manly improve its village capturing
20091011 01:46:18< alink> *mainly
20091011 01:46:22< Aethaeryn> Yes... Manly village capturing.
20091011 01:46:49-!- stikonas [n=and@bcm-131-111-216-119.girton.cam.ac.uk] has joined #wesnoth-dev
20091011 01:47:30< Arc> only capture masculine villages
20091011 01:47:34< Arc> leave the women and children alone
20091011 01:47:58< alink> I recently noticed that in WML villages had "male_names"
20091011 01:48:05< alink> only male_names
20091011 01:49:38< alink> in names.cfg
20091011 01:51:21-!- boucman [n=rosen@wesnoth/developer/boucman] has quit ["Leaving."]
20091011 02:04:50< Aethaeryn> < alink> I recently noticed that in WML villages had "male_names" <- There you go, then. Proof of manly village capturing.
20091011 02:05:47< alink> or the origin of my lapsus ;-)
20091011 02:14:53-!- alink [n=alink@wesnoth/developer/alink] has quit [Remote closed the connection]
20091011 02:19:31-!- ardesh [n=ardesh@port-92-206-14-177.dynamic.qsc.de] has quit [Read error: 60 (Operation timed out)]
20091011 02:20:03-!- ardesh [n=ardesh@port-92-206-60-157.dynamic.qsc.de] has joined #wesnoth-dev
20091011 02:20:27-!- Zarel [n=Zarel@warzone2100/developer/Zarel] has joined #wesnoth-dev
20091011 03:36:14-!- Zarel [n=Zarel@warzone2100/developer/Zarel] has quit ["This computer has gone to sleep"]
20091011 03:37:26-!- Zarel [n=Zarel@warzone2100/developer/Zarel] has joined #wesnoth-dev
20091011 03:40:10-!- Blueblaze [n=nick@adsl-99-171-161-30.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20091011 03:41:41< shadowmaster> }}}}}}}723
20091011 03:54:07< Espreon> It appears that double-clicking on a campaign's entry in the campaign selection menu no longer starts it, is this intentional?
20091011 03:54:48< shadowmaster> works for me.
20091011 03:55:13< Espreon> Odd...
20091011 03:55:38-!- DDR [n=chatzill@66.183.125.196] has quit [Read error: 60 (Operation timed out)]
20091011 03:56:52-!- DDR [n=chatzill@66.183.125.196] has joined #wesnoth-dev
20091011 04:01:27-!- stikonas [n=and@wesnoth/translator/stikonas] has quit [Remote closed the connection]
20091011 04:02:54< Espreon> shadowmaster: It works for mainline campaigns, but not usermade campaigns.
20091011 04:03:13< shadowmaster> it works for UMC.
20091011 04:03:35< shadowmaster> it doesn't work for one UMC.
20091011 04:03:45< shadowmaster> why me?
20091011 04:03:58< Espreon> LOL, wut?
20091011 04:04:45< shadowmaster> not for NR either.
20091011 04:04:48< shadowmaster> um. weird.
20091011 04:05:01< Espreon> It seems that if the cursor is on the text, it won't work.
20091011 04:05:02< shadowmaster> it seems to be completely random.
20091011 04:05:25< Espreon> No, it doesn't work if the cursor is on the text.
20091011 04:06:33< Espreon> I guess that I shall file a bug.
20091011 04:13:31< Espreon> mordante: https://gna.org/bugs/index.php?14497
20091011 04:13:38< Espreon> Enjoy.
20091011 04:23:51-!- Blueblaze [n=nick@adsl-99-171-161-30.dsl.hstntx.sbcglobal.net] has quit [Remote closed the connection]
20091011 04:45:30-!- Chusslove [n=Chusslov@217.190.237.186] has quit [Read error: 145 (Connection timed out)]
20091011 04:51:59-!- Ivanovic_ [n=ivanovic@dtmd-4db2acdc.pool.mediaWays.net] has joined #wesnoth-dev
20091011 05:01:07-!- Chusslove [n=Chusslov@217.190.225.115] has joined #wesnoth-dev
20091011 05:08:20-!- Ivanovic [n=ivanovic@wesnoth/developer/ivanovic] has quit [Read error: 110 (Connection timed out)]
20091011 05:09:56-!- Blueblaze [n=nick@adsl-99-171-161-30.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20091011 05:09:57-!- Ivanovic_ is now known as Ivanovic
20091011 05:18:18-!- Zarel_ [n=Zarel@warzone2100/developer/Zarel] has joined #wesnoth-dev
20091011 05:23:14-!- Zarel is now known as Zarel__
20091011 05:23:17-!- Zarel_ is now known as Zarel
20091011 05:23:21-!- Zarel__ is now known as Zarel_
20091011 05:38:04-!- Zarel [n=Zarel@warzone2100/developer/Zarel] has quit [Read error: 54 (Connection reset by peer)]
20091011 05:39:11-!- grzywacz [n=grzywacz@wesnoth/developer/grzywacz] has quit [Remote closed the connection]
20091011 05:42:37-!- ancestral [n=ancestra@97-116-105-75.mpls.qwest.net] has joined #wesnoth-dev
20091011 06:15:07-!- Aethaeryn is now known as Aethaeddynyn
20091011 07:12:31-!- mordante [n=mordante@wesnoth/developer/mordante] has joined #wesnoth-dev
20091011 07:12:44< mordante> servus
20091011 07:15:36-!- Aethaeddynyn [n=Michael@wesnoth/umc-dev/developer/aethaeryn] has quit ["meh"]
20091011 07:18:02< Espreon> Hello mordante.
20091011 07:18:09< mordante> hi Espreon 
20091011 07:18:27< Espreon> mordante: Have you read the logs?
20091011 07:18:58< mordante> Espreon, reading it
20091011 07:19:04< Espreon> OK. :)
20091011 07:24:02 * mordante hands Espreon the RELEASE_NOTES
20091011 07:24:55< Espreon> Ah, I see...
20091011 07:25:56< mordante> and indeed it doesn't work when the mouse is on the text label
20091011 07:26:31< mordante> ancestral, didn't you also have problems with some glyphs not shown in menus?
20091011 07:27:15< mordante> Arc, did you remove the screenshot for https://gna.org/bugs/index.php?14487?
20091011 07:27:40< ancestral> mordante: I did but was fixed awhile ago. I instead have some issues with languages not substituting properly
20091011 07:28:37< mordante> you mean when changing languages
20091011 07:32:17< Arc> mordante: no i did not
20091011 07:33:11< Arc> the screenshots are still up on http://media.apogean.org/img/wesnoth-1.png and http://media.apogean.org/img/wesnoth-2.png
20091011 07:33:30< Arc> note that upgrading fribidi got me weird glyphs instead of squares
20091011 07:34:43< mordante> guess then gna lost them
20091011 07:35:32< ancestral> mordante: correct
20091011 07:35:37< mordante> I already had those weird glyphs but assumed it was the same problem ;-)
20091011 07:36:26< mordante> ancestral, ok, since we also have glyph problems with en@shaw
20091011 07:36:34< mordante> thought it might be related
20091011 07:37:42< ancestral> I haven't gotten a chance to test this, since I would need to build
20091011 07:39:52< Arc> the problem is with any 32-bit utf-8
20091011 07:40:14< Arc> i already tracked it down, you're rendering the characters as ucs2 characters one 16-bit chunk at a time, which is whats breaking it
20091011 07:40:30< Arc> it cant render half the glyph's bits
20091011 07:41:04< ancestral> oops, was referring to https://gna.org/bugs/?func=detailitem&item_id=13878 I haven't gotten a chance to test since this fix
20091011 07:41:08< mordante> yes I read that in the bug report
20091011 07:41:16< Arc> ah
20091011 07:41:42< mordante> ancestral, silene said that the fix wouldn't help, but can't hurt to test
20091011 07:42:13< ancestral> Anyone else running OS X that's on right now?
20091011 07:43:03< mordante> Arc, but reimplementing the event handler caused a few regressions which I like to fix first
20091011 07:45:52-!- silene [n=plouf@wesnoth/developer/silene] has joined #wesnoth-dev
20091011 07:48:07< Arc> oh no rush
20091011 07:49:06< Arc> we'd like to play through some of tese campaigns im translating without having to guess what things say, thats all :-)  its still playable so long as you know where things are
20091011 07:49:24< Arc> all the story dialog, character names, etc are all properly rendered
20091011 07:50:37< Arc> as of tomorrow when Ivanovic submits what i sent him tonight, en@shaw will be more complete than esperonto
20091011 07:56:08< mordante> I understand why you like to have it fixed, and I hope to find some time for it soon
20091011 07:56:33-!- Bellerophon_true [n=chatzill@89.142.149.100] has quit ["ChatZilla 0.9.85 [Firefox 3.5.3/20090824101458]"]
20091011 07:56:41< mordante> so almost at 25% translated, nice
20091011 08:15:32-!- ancestral [n=ancestra@97-116-105-75.mpls.qwest.net] has quit ["And that’s the end of THAT chapter."]
20091011 08:33:14< CIA-13> jetryl * r39237 /trunk/data/core/ (23 files in 3 dirs): Added a new, complete goblin variant of the walking corpse and soulless.
20091011 08:33:26< Espreon> Oh thank God.
20091011 08:33:34 * Espreon is happy that those are finally in.
20091011 08:37:39-!- zookeeper [n=l@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20091011 08:49:15-!- EdB [n=edb@56.117.88-79.rev.gaoland.net] has joined #wesnoth-dev
20091011 09:08:51< CIA-13> jetryl * r39238 /trunk/data/core/ (20 files in 2 dirs): Added new hammer frames for the dwarvish lord and steelclad.
20091011 09:09:02< CIA-13> mordante * r39239 /trunk/ (changelog po/CMakeLists.txt): 
20091011 09:09:02< CIA-13> Improve po-update to avoid nop updates.
20091011 09:09:02< CIA-13> The last change caused a lot of files be updated with the only change
20091011 09:09:02< CIA-13> the pot-update header. In the Wesnoth wiki there was a hint to use
20091011 09:09:03< CIA-13> msgattrib to cleanup the changes poedit makes. Using this on the file
20091011 09:09:05< CIA-13> seems to give the wanted output, proper files and no "dummy" updates.
20091011 09:09:07< CIA-13> Ivanovic please test this change.
20091011 09:13:08-!- silene [n=plouf@wesnoth/developer/silene] has quit ["Leaving."]
20091011 09:19:02-!- Zarel_ [n=Zarel@warzone2100/developer/Zarel] has quit ["This computer has gone to sleep"]
20091011 09:25:22-!- noy [n=Noy@d75-157-29-32.bchsia.telus.net] has joined #wesnoth-dev
20091011 09:36:04-!- EdB [n=edb@56.117.88-79.rev.gaoland.net] has quit [Remote closed the connection]
20091011 09:39:52-!- Blueblaze [n=nick@adsl-99-171-161-30.dsl.hstntx.sbcglobal.net] has quit [Remote closed the connection]
20091011 09:42:09-!- Netsplit pratchett.freenode.net <-> irc.freenode.net quits: shikadibot, Smar, kimrhh, dfranke_, rtfb, yann
20091011 09:42:09-!- Sirp [n=user@pool-71-164-166-178.dllstx.fios.verizon.net] has quit [Read error: 104 (Connection reset by peer)]
20091011 09:43:30-!- Smar [i=smar@smar.fi] has joined #wesnoth-dev
20091011 09:43:43-!- kimrhh [n=kimhh@static.51.211.46.78.clients.your-server.de] has joined #wesnoth-dev
20091011 09:44:11-!- Netsplit over, joins: shikadibot
20091011 09:45:34-!- noy [n=Noy@wesnoth/developer/noy] has quit []
20091011 09:46:17-!- noy [n=Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20091011 09:49:28-!- kimrhh [n=kimhh@static.51.211.46.78.clients.your-server.de] has quit [Killed by douglas.freenode.net (Nick collision)]
20091011 09:49:28-!- kimhh_ [n=kimhh@static.51.211.46.78.clients.your-server.de] has joined #wesnoth-dev
20091011 09:49:33-!- rtfb [n=read-the@78-56-0-35.static.zebra.lt] has joined #wesnoth-dev
20091011 09:49:33-!- yann [n=dwitch@nan92-1-81-57-214-146.fbx.proxad.net] has joined #wesnoth-dev
20091011 09:49:33-!- kimrhh [n=kimhh@exherbo/developer/kimrhh] has joined #wesnoth-dev
20091011 09:49:33-!- dfranke_ [i=dfranke@wesnoth/developer/dfranke] has joined #wesnoth-dev
20091011 09:49:40-!- kimrhh [n=kimhh@exherbo/developer/kimrhh] has quit [Connection reset by peer]
20091011 09:53:20< Espreon> mordante: Would fun translations be accepted?
20091011 09:55:16< mordante> Espreon, what do you consider a fun translation?
20091011 09:55:29< Espreon> An en@internets-bayeux translation.
20091011 09:56:41< Espreon> ... which is essentially a translation of the dialect of Internets English that is used on the Bayeux tapestry.
20091011 09:56:42< mordante> what kind of translation would that be?
20091011 09:59:32< mordante> I'm still not sure what kind of dialect that is and how well defined it is
20091011 10:03:33-!- wesbot changed the topic of #wesnoth-dev to: string/feature freeze active! | 102 bugs, 237 feature requests, 10 patches | logs: http://irclogs.wesnoth.org/ | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com | http://imagebin.org
20091011 10:11:38-!- EdB [n=edb@79.88.117.56] has joined #wesnoth-dev
20091011 10:16:32-!- boucman [n=rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20091011 10:25:02-!- loonybot [n=loonybot@79.139.139.50] has joined #wesnoth-dev
20091011 10:26:16-!- loonycyborg [n=sergey@79.139.139.50] has joined #wesnoth-dev
20091011 10:47:01-!- DDR [n=chatzill@66.183.125.196] has quit [Read error: 110 (Connection timed out)]
20091011 10:58:15-!- euschn [n=chatzill@85-127-115-137.dynamic.xdsl-line.inode.at] has joined #wesnoth-dev
20091011 10:58:20< euschn> hi
20091011 10:58:46< Espreon> Hello.
20091011 10:59:08< euschn> shadowmaster: yes I usually read the logs. I was quite busy recently, thats why I was rarely available on irc. hopefully this should get better by next week
20091011 10:59:48< mordante> hi euschn 
20091011 11:00:04< euschn> hi mordante 
20091011 11:12:07< CIA-13> caslav_ilic * r39240 /branches/resources/image-localization/tutorial/ (27 files): Images used in the tutorial on image localization.
20091011 11:18:54< Ivanovic> moin
20091011 11:20:42< Espreon> Ivanovic: Are fun translations accepted?
20091011 11:21:08< Ivanovic> define "fun translations"
20091011 11:21:40< mordante> hi Ivanovic 
20091011 11:21:40< Espreon> Things such as Internets English. In this case, the dialect that is used on the Bayeux tapestry.
20091011 11:23:17< Espreon> Ivanovic: ^
20091011 11:31:59< loonycyborg> How about LOLspeak? :P
20091011 11:32:22< Espreon> Maybe.
20091011 11:32:36< Espreon> I believe that would be the main form of Internets English.
20091011 11:32:40< loonycyborg> (As used in lolcat pictures)
20091011 11:32:55< Espreon> But, I am only interested in the Bayeux dialect.
20091011 11:33:00< Ivanovic> to be honest: you ain't serious, right?
20091011 11:33:09< Espreon> I am serious.
20091011 11:33:23< Espreon> loonycyborg: No, you're thinking of the lolcat dialect.
20091011 11:34:11< loonycyborg> I suspected that I've got that one wrong..
20091011 11:34:36< boucman> hey all
20091011 11:34:42< Espreon> Hello boucman.
20091011 11:34:43< boucman> mordante: any new clues ?
20091011 11:34:46< mordante> hi boucman 
20091011 11:35:07< mordante> boucman, no not really but also haven't looked further at it yet
20091011 11:35:11< boucman> ok
20091011 11:36:08< loonycyborg> Actually, no. I was right. It's lolspeak: http://en.wikipedia.org/wiki/Lolcat
20091011 11:36:29< ardesh> o rly?
20091011 11:36:43< Espreon> Yes, that is the lolcat dialect of Internets English.
20091011 11:37:04< CIA-13> ivanovic * r39241 /trunk/po/wesnoth-low/en@shaw.po: updated Shavian translation
20091011 11:40:48< Espreon> Ivanovic: What is the verdict?
20091011 11:42:44< Ivanovic> mordante: i have the impression that the update stuff now works nicely
20091011 11:43:16< mordante> Ivanovic, :-)
20091011 11:43:27< Ivanovic> that is: i removed the linebreaks by hand in some en_GB files and ran the po-update on them afterwards
20091011 11:43:36< Ivanovic> those linebreaks were removed and no other changes introduced
20091011 11:43:52< mordante> I did the same on latin and reordered the file as well
20091011 11:44:15< Ivanovic> i also saved those files via poedit (which compeltely garbles translations) and the pot-date was not updated afterwards
20091011 11:44:18< mordante> only one change a line break, but that line might have been too long before
20091011 11:44:28< Ivanovic> (sure, the author field and last change field changed, but thats it)
20091011 11:45:08< mordante> I used the trick to ungarble the poedit mess as listed in the wiki
20091011 11:45:36< Ivanovic> and i completely rely on pofix and po-update to solve the mess
20091011 11:45:38< Ivanovic> ;)
20091011 11:45:53< mordante> :-)
20091011 11:46:43< Espreon> Uh, Ivanovic?
20091011 11:47:06< Ivanovic> Espreon: to be honest, i don't think adding those "pure fun translations" is such a good idea
20091011 11:47:20< Espreon> Oh, I see... :(
20091011 11:47:37< Ivanovic> it is likely to end in another "yeah, was once fun but stopped it right after a week"
20091011 11:47:49< Ivanovic> there are already so many dead translations...
20091011 11:47:56< Espreon> Nah, not over here.
20091011 11:48:23< Ivanovic> uhm, ain't you also working on some other (not complete yet!) translation?
20091011 11:48:30< Espreon> No.
20091011 11:48:40< Espreon> Where did you get that idea?
20091011 11:48:52< Ivanovic> wishful thinking?
20091011 11:49:03< Ivanovic> still sleepy?
20091011 11:49:44< loonycyborg> Espreon: So you aren't working on Spanish translation?
20091011 11:49:55< Espreon> loonycyborg: I am only localizing the images for that.
20091011 11:50:17< Espreon> I am not fluent enough to maintain the Spanish translation.
20091011 11:50:29-!- stikonas [n=and@bcm-131-111-216-119.girton.cam.ac.uk] has joined #wesnoth-dev
20091011 11:50:45-!- kimhh_ is now known as kimrhh
20091011 11:51:10< Espreon> Ivanovic: If I get a decent amount done, then what?
20091011 11:51:53< Ivanovic> hmm, lets say "it has to be basically core complete"
20091011 11:52:25< loonycyborg> Ivanovic: Are unmaintained languages that don't have any upt-to-date strings garbage-collected?
20091011 11:52:34< Espreon> Oh, then I guess that I shouldn't have began with UtBS...
20091011 11:52:50< Espreon> *begun
20091011 11:52:59< Ivanovic> core consists of the following textdomains:
20091011 11:53:02< Ivanovic> wesnoth wesnoth-lib wesnoth-editor wesnoth-units wesnoth-multiplayer wesnoth-tutorial
20091011 11:53:26< Ivanovic> loonycyborg: no, translations don't go through any kind of garbage collection
20091011 11:53:52< loonycyborg> So if a language added it remains forever?
20091011 11:54:09-!- allefant [n=elias@allegro/developer/allefant] has joined #wesnoth-dev
20091011 11:54:10< Ivanovic> that is: for a possible 2.0 we will probably have a policy of "only translations that are core complete are in and only those parts of the translation that are 100% done are shipped"
20091011 11:54:16< Ivanovic> loonycyborg: basically yes
20091011 11:54:37< Ivanovic> since there was once done some work that is most likely recycable
20091011 11:55:21< Ivanovic> (what i just tried to outline was the 1.0 policy)
20091011 12:08:52< mordante> IMO we can GC translations that are added and have nothing done to them
20091011 12:09:06 * mordante looks at mf fur_IT
20091011 12:09:38-!- euschn [n=chatzill@wesnoth/developer/euschn] has quit [Read error: 113 (No route to host)]
20091011 12:10:04< mordante> but like pt has some work done and if a maintainer shows up s/he as a small head start
20091011 12:10:52< mordante> I'm not fond of the idea to drop not ready translations for 2.0, can be a regression for a lot of users
20091011 12:11:21< mordante> instead of having most translated in 1.x.y they are suddenly dropped to English only
20091011 12:15:30-!- noy [n=Noy@wesnoth/developer/noy] has quit []
20091011 12:19:44< zookeeper> fun translations seem very pointless, a huge amount of effort, and no one will actually use them, the fun factor is only in being able to say that there is such a translation
20091011 12:20:18< zookeeper> if someone actually makes one and completes it, then i don't see why not add it, but i fail to see the point of doing the work
20091011 12:20:40< Ivanovic> zookeeper: that is why i said that it should be at least core complete before inclusion
20091011 12:21:06< mordante> Ivanovic, when is the next release planned?
20091011 12:22:19< Espreon> zookeeper: I am doing it for the lulz.
20091011 12:22:39< Espreon> Besides, I often have periods of time in which I can do nothing but lulz-seeking.
20091011 12:23:17< Espreon> In addition, I am sure that very few people use the Latin translation, yet it exists.
20091011 12:23:53-!- SonIcco [n=SonIcco@pD9510BF9.dip0.t-ipconnect.de] has joined #wesnoth-dev
20091011 12:25:02< Ivanovic> i originally though about doing it today, but i won't
20091011 12:25:12< Ivanovic> maybe wednesday?
20091011 12:25:38< Ivanovic> mordante: i don't think dropping translations that are incomplete is too much of a regression
20091011 12:25:50< Ivanovic> think of a translation that was complete in eg 1.2 but unmaintained since then
20091011 12:26:04< Ivanovic> those are likely to be "one sentence in english, one translated" and so on
20091011 12:26:20< Ivanovic> this is not too much fun and in such a situation i as player would switch to english
20091011 12:27:10-!- Crab_ [i=crab@wesnoth/developer/crab] has joined #wesnoth-dev
20091011 12:28:06< mordante> I agree that half translated is no fun but I don't like to drop a 95% translation
20091011 12:28:34< mordante> good then I can get some of my new commits in today, they're a bit intrusive (again)
20091011 12:29:53< Ivanovic> :)
20091011 12:34:36-!- ilor [n=user@wesnoth/developer/ilor] has joined #wesnoth-dev
20091011 12:35:03< Espreon> Hello ilor.
20091011 12:35:17< ilor> hi Espreon
20091011 12:35:42< ilor> hi mordante
20091011 12:36:58< mordante> hi ilor 
20091011 12:37:32< mordante> ilor, I've been working on the new server with boost asio for a few weeks now
20091011 12:37:35-!- blarumyrran [n=minaise@81-20-159-197.levira.ee] has joined #wesnoth-dev
20091011 12:38:16< ilor> mordante: the addon server?
20091011 12:38:22< mordante> yup
20091011 12:38:27< ilor> hmm
20091011 12:39:21-!- Nayela [n=Nayela@cpc3-lich6-0-0-cust673.brhm.cable.ntl.com] has joined #wesnoth-dev
20091011 12:39:28< CIA-13> mordante * r39242 /trunk/src/gui/auxiliary/event/dispatcher.hpp: Allow to disconnect signals.
20091011 12:39:38< CIA-13> mordante * r39243 /trunk/src/gui/auxiliary/event/dispatcher.hpp: Remove assert so the compiler can properly warn.
20091011 12:40:18< mordante> but what kind of project are you exactly looking for?
20091011 12:41:04< ilor> mordante: a standalone project that can go through an overly formal specification and design phase
20091011 12:41:12< mordante> btw I mentioned here a few weeks ago that I was about to start on that new server
20091011 12:42:33< mordante> how standalone does it need to be? can it be working on another part of Wesnoth?
20091011 12:44:01< ilor> mordante: not sure how that would work out, this course is all about formal designs and stuff
20091011 12:44:09< freim> ilor: add features to wesnoth do it can be a platform for making games like Master of Magic :)
20091011 12:44:18< freim> s/do/so/
20091011 12:44:31< ilor> might be hard to have the project be deep inside a largely undocumented heap of Wesnoth ;)
20091011 12:46:14< mordante> large undocumented heap of Wesnoth, do we have that?
20091011 12:46:17< mordante> :-p
20091011 12:46:26< ilor> mordante: how far along are you with the addon server anyway?
20091011 12:46:56< mordante> I started to work on the network handling in boost and test with that part
20091011 12:47:22< mordante> the rest of the current server is rather small
20091011 12:47:33< freim> mordante: working just on the underlying stuff, or the interface also?
20091011 12:48:06< mordante> freim, the server only, but want to work on the GUI part as well
20091011 12:48:10< freim> what has struck me with regards to user content is that I see heaps of campaings both on the server and on the forum, but I have no idea which are any good
20091011 12:48:14< ilor> I see. the project for the course would likely be specifications for the first few months with an actual implementation somewhere next semester
20091011 12:48:20< freim> since there is no "top list", rating etc
20091011 12:48:42< freim> neither if they are complete
20091011 12:48:58< mordante> freim, still want to implement your gui ideas for it, but the new gui engine can't handle a button on a toggle button properly yet
20091011 12:49:23< mordante> freim, that what's my event rewriting is about, to allow events to be properly forwarded
20091011 12:49:36< freim> ok
20091011 12:52:01< mordante> ilor I see and other gsoc ideas are not interesting big enough?
20091011 12:52:30< ilor> well, AI stuff or WML optimization are too technical and not suitable for this course
20091011 12:53:29< ilor> ditto about WML validation I guess
20091011 12:54:07< Ivanovic> ilor: uhm, "validator for markup language" is something you can *perfectly* design
20091011 12:54:12< Ivanovic> so it should be possible
20091011 12:54:46< Ivanovic> that is: validator for a markup language using a preprocessor
20091011 12:54:48-!- Crab_ [i=crab@wesnoth/developer/crab] has quit ["Leaving."]
20091011 12:54:54< Ivanovic> (the other one would be boring, right? ;) )
20091011 12:55:23< ilor> Ivanovic: I'm afraid it might be too much of a technical fun and not enough of a user thing, but I can try
20091011 12:56:14< Ivanovic> if you mean with "user thing" "graphical interface": sure
20091011 12:56:52< Ivanovic> if you mean soemthing along the lines of "the one responsible for the course does not see much use in it": easy to comunicate that it does make sense to have a tool like this
20091011 12:57:36< mordante> ilor, are you around longer this afternoon?
20091011 12:57:59< ilor> mordante: yeah, some time early evening I'll be out for around 2 hours then available later
20091011 12:58:09< ilor> otherwise not going anywhere
20091011 12:58:35< freim> ilor: if you want something more geared towards UI, I would suggest a rewamp of the help section. Imo the current one has a rather poor layout and is a bit fiddly to use
20091011 12:58:50< freim> ilor: you can do a lot of user testing etc also there if that is something int.
20091011 12:58:55< mordante> ilor, good since I'm afk now will be back later this afternoon to discuss it further
20091011 12:59:02< ilor> mordante: okay
20091011 12:59:07< Ivanovic> hmm, different topic: since the date of next fosdem is already out: shall we start planning sometime soon?
20091011 12:59:20< ilor> freim: again, this would be knee deep in wesnoth code, probably a no-go
20091011 12:59:23< mordante> Ivanovic, sounds like a good plan
20091011 12:59:37< mordante> afk 
20091011 12:59:42< freim> ilor: I'm more thinking about the design and layout. Make something that looks good and is usable
20091011 12:59:59< ilor> Ivanovic: ouch. fosdem is smack in the middlem of my exams :/
20091011 13:00:05< Ivanovic> :(
20091011 13:00:22< Ivanovic> though you don't have exams on saturdays and sundays, right?
20091011 13:00:24< Ivanovic> ;)
20091011 13:00:48< ilor> well, no, but I might have exams on the following Monday
20091011 13:01:06< Ivanovic> hmm, not nice either
20091011 13:02:01< freim> ilor: do you have to implement your design as part of the task?
20091011 13:02:09< ilor> freim: yes
20091011 13:02:12< freim> hmm, ok
20091011 13:02:52< freim> I would still consider it
20091011 13:29:17< ilor> Ivanovic: I had a look at the officially okayed projects for this year's course and I guess the WML validator might actually work if given some appropriate spin
20091011 13:30:54< ilor> one of the okayed projects is a gdb-related tool to parse logs and simulate program flow
20091011 13:30:58< Ivanovic> it should be possible to do a heavy design approach for a tool like this (it is probably the only way that is *can* work since WML is a damn complex lang by now)
20091011 13:32:58< ilor> I'm somehwat worried there might be not enough UI in it, I mean it'd be a great tool if it just parsed a file or tree and reported errors\
20091011 13:33:45< ilor> it's more of a single-purpose tool than a full system and there might not be enough use cases and stuff to make the prof happy
20091011 13:33:58< Ivanovic> use cases:
20091011 13:34:08< Ivanovic> 1) run by an author to verify data
20091011 13:34:31< Ivanovic> 2) automatted server side running including rejection of the data if invalid
20091011 13:34:51< Ivanovic> 3) possibility of "semi smart" conversion of broken things
20091011 13:35:10< Ivanovic> (that is to allow automatic updates at syntax changes in the language)
20091011 13:35:57< ilor> so 3) is pretty much doing some of the work of wmllint
20091011 13:36:12< Soliton> recent thread on the topic: http://forum.wesnoth.org/viewtopic.php?f=10&t=27389
20091011 13:36:25< Ivanovic> ilor: yes, it is, but does your prof know that we have it?
20091011 13:36:27< Ivanovic> ;)
20091011 13:36:45< ilor> also 4) modified by a developer to accomodate new tags etc/
20091011 13:36:58< Ivanovic> jupp
20091011 13:37:10< Soliton> it'd presumably be a different approach then wmllint since it'd work on the expanded WML i imagine.
20091011 13:37:12< Ivanovic> testing if new tags will break validity for existing content, lovely
20091011 13:37:32< Ivanovic> Soliton: i'd call it extended wmllint because of this
20091011 13:38:03< Soliton> well, both approaches have their dis-/advantages.
20091011 13:38:44< Ivanovic> the problem for the extended version is when changes are required, in which files should they best go?
20091011 13:38:53< Ivanovic> this part is "a little" tricky
20091011 13:44:48-!- Bellerophon_true [n=chatzill@89.142.149.100] has joined #wesnoth-dev
20091011 13:46:51-!- ilor [n=user@wesnoth/developer/ilor] has quit [Read error: 60 (Operation timed out)]
20091011 13:47:04-!- ilor [n=user@wesnoth/developer/ilor] has joined #wesnoth-dev
20091011 14:51:50< mordante> ilor, does your project need UI as well?
20091011 14:52:21< ilor> mordante: it's probably best if it's a self contained system, so yeah with some sort of UI
20091011 14:52:38< loonycyborg> CUI or GUI? :P
20091011 14:52:52< ilor> most of the proposed projects are some webapps, bugtrackers or the like
20091011 14:53:02< mordante> Ivanovic, the main idea about validation is to parse whether or not WML is broken, not to fix it
20091011 14:53:24< mordante> also don't know whether it would be a good idea to mix them
20091011 14:54:01< ilor> loonycyborg: one project specifies "at least a commandline interface" but it's just the one
20091011 14:54:05< mordante> ilor, what kind of UI do you want to add for the addon server?
20091011 14:54:32< ilor> mordante: most likely a web ui for submitting addons and moderation
20091011 14:55:15< mordante> moderation, so we really want to get rid of inane stuff ;-)
20091011 14:56:22< mordante> the idea I'm working on for the addon server is mainly aimed at rewriting the current server to use boost asio instead of SDL net
20091011 14:56:29< ilor> I also thought that interfacing with the forum database for login info would be nice to avoid another authentication system
20091011 14:56:50< mordante> in that way we can test whether boost asio is up for the task
20091011 14:57:10< mordante> and gives me the option to see how best integrate it with the network progress widget
20091011 14:57:28< mordante> so I don't want to make a big project out of it
20091011 14:57:29< ilor> I see
20091011 14:58:01< ilor> that can be seen as somwhat orthogonal to what I'd want to do, which would be rewriting the whole thing with new logic for places etc
20091011 14:58:03< mordante> when would you start coding on your prokect
20091011 14:58:06< mordante> project
20091011 14:58:25< ilor> we're supposed to have a prototype somewhere around January I guess
20091011 14:58:50< mordante> of the code or the design?
20091011 14:59:26< ilor> the design should have gone through 2 or 3 iterations by that time, and a prototype app should be done
20091011 14:59:54< ilor> actual implementation is scheduled for the next semester (march+)
20091011 15:00:01< mordante> I would love to do more for the addon server but I have enough to do with the gui
20091011 15:00:28< mordante> but the network progress touches the gui, that's why I got interested to work on the addon server
20091011 15:00:44< ilor> I think that project I'll be doing can reuse the network layer you're working on without much problems
20091011 15:00:57< mordante> I think so too :-)
20091011 15:00:57< Ivanovic> possible next soc project: redo the addon server, here is my proposal: "{$detailed_design_docs}"
20091011 15:00:59< Ivanovic> ;)
20091011 15:01:32< ilor> Ivanovic: I'm supposed to have the final product by the end of May, so not really ;)
20091011 15:01:44< Ivanovic> ahh, okay
20091011 15:01:55< ilor> besides, the project is a four-people thing
20091011 15:02:01< mordante> and after acceptation directly send the code to google ;-)
20091011 15:02:05< ilor> ;)
20091011 15:03:03< ilor> mordante: I also guess the prototype can just as well be the logic only without the network interface for the game, so that won't depend on your work
20091011 15:03:48< mordante> ilor, yes I also see less overlap then
20091011 15:04:25< mordante> only then I've no need to polish the actual server handling code, which I haven't started yet
20091011 15:04:49< ilor> mordante: also I hope you'll have some good general ideas for the addon server for the requirements specification phase I'll be doing ;)
20091011 15:05:46< CIA-13> mordante * r39244 /trunk/ (6 files in 3 dirs): 
20091011 15:05:46< CIA-13> Add a new event distributor class.
20091011 15:05:46< CIA-13> This class is meant to replace the current event handler class. The replacement
20091011 15:05:46< CIA-13> will be done in a later commit.
20091011 15:06:50< mordante> I've some ideas here http://wiki.wesnoth.org/wiki/User:SkeletonCrew#Addon_server
20091011 15:08:08< mordante> also it would be nice if the addon server would store the content in a gzipped archive so it can be send directly instead of gzipping it every time it's send
20091011 15:08:56< ilor> I'll have a read
20091011 15:09:24< mordante> ok
20091011 15:09:51< mordante> luckily you started about this project before I started to work on the real server code :-)
20091011 15:11:11-!- ardesh [n=ardesh@port-92-206-60-157.dynamic.qsc.de] has quit [Read error: 110 (Connection timed out)]
20091011 15:11:23< ilor> mordante: :)
20091011 15:11:37-!- ardesh [n=ardesh@port-92-206-72-92.dynamic.qsc.de] has joined #wesnoth-dev
20091011 15:12:36< ilor> I'll be meeting the prof for the course tomorrow and will discuss the idea, official deadline for the project topic is thursday so if he doesn't like something I'll still have the time to convince him
20091011 15:14:06< mordante> ok good luck with it
20091011 15:20:47-!- Bellerophon_true [n=chatzill@89.142.149.100] has quit ["ChatZilla 0.9.85 [Firefox 3.5.3/20090824101458]"]
20091011 15:35:34-!- ardesh [n=ardesh@port-92-206-72-92.dynamic.qsc.de] has quit [Read error: 110 (Connection timed out)]
20091011 15:36:50< CIA-13> mordante * r39245 /trunk/src/gui/widgets/ (5 files): 
20091011 15:36:50< CIA-13> Let the new dispatcher do the event handling.
20091011 15:36:50< CIA-13> The old code remains but ifdef'ed out so can be restored if needed.
20091011 15:36:52-!- ardesh [n=ardesh@92.206.28.2] has joined #wesnoth-dev
20091011 15:55:56-!- Espreon [n=espreon@wesnoth/developer/espreon] has quit ["WRYYYYYYYYYYYYYYYYYYYY!"]
20091011 16:04:35< Arc> mordante: why not just switch to pango for font rendering there?
20091011 16:06:09< mordante> Arc, that's a solution, but it still takes time to do that
20091011 16:06:38< mordante> Arc, but it's also possible with SDL_TTF not sure what the fastest fix is
20091011 16:07:37< Arc> how about testing for a surrogate and carrying it over to the next loop to combine with the next character
20091011 16:08:13< Arc> i would like to get a proper screenshot of the shavian translation to solicit some help from some of the teens in the shavian groups
20091011 16:09:35< Arc> i'm not very familiar with c++ but i'm very familiar with cairo
20091011 16:09:53< Arc> if all you're using that class for is to render single lines i could easily replace it with cairo
20091011 16:10:35< Arc> i just need to know which function is being called from the outside to render and what needs to be outputted
20091011 16:13:14< Arc> actually - itor can be incremented within the loop right?
20091011 16:13:52< Arc> U+D800-U+DBFF U+DC00-U+DFFF are the two blocks to look for
20091011 16:14:29< Arc> so could just test to see if the character falls in one of those two ranges, if so, cat two ucs2 characters together and itor++
20091011 16:14:32< Arc> else, do it the normal way
20091011 16:14:42< Arc> heh sorry that is one big range
20091011 16:14:55< mordante> like I said this morning I'm rather busy with fixing the event handling
20091011 16:15:26< Arc> ok mind if i take a crack at it?
20091011 16:15:31< mordante> no not at all
20091011 16:15:43< mordante> I also don't remember who wrote that code
20091011 16:16:04< mordante> do you want to change it to cairo or fix it using SDL?
20091011 16:16:44< mordante> and somehow all gui related bugs are thrown in my lap, also to ones in the old gui I try to get rid of :-/
20091011 16:17:04< mordante> (which means I'm not familiar at all with that code and don't want to either)
20091011 16:18:13< Arc> i don't entirely understand whats going on with the TTF
20091011 16:18:32-!- ardesh [n=ardesh@92.206.28.2] has quit [Read error: 145 (Connection timed out)]
20091011 16:18:35< Arc> if you tell me what in this code is called to render text and what it needs to output, i will take a crack at migrating it to cairo
20091011 16:18:57-!- ardesh [n=ardesh@port-92-195-111-27.dynamic.qsc.de] has joined #wesnoth-dev
20091011 16:19:28-!- stikonas [n=and@wesnoth/translator/stikonas] has quit [Remote closed the connection]
20091011 16:19:39< loonycyborg> I wonder why only shavian is affected. Other languages with multi-byte characters work fine even with old GUI.
20091011 16:19:58< Arc> loonycyborg: all other multi-byte characters were on plane0 which only requires 16 bits
20091011 16:20:11< Arc> only plane1+ uses surrogate pairs
20091011 16:20:20< Arc> shavian is on plane 1
20091011 16:20:46< loonycyborg> Even chinese and Japanese are on plane 0?
20091011 16:20:52< Arc> yes.
20091011 16:21:03< Arc> plane1+ are for scripts with fewer than 1 million users
20091011 16:21:10< mordante> Arc, the sdl library we use for rendering text is called SDL_TTF
20091011 16:21:15< Arc> ie, native american languages, other tribal languages, etc
20091011 16:21:16-!- stikonas [n=and@bcm-131-111-216-119.girton.cam.ac.uk] has joined #wesnoth-dev
20091011 16:21:52< loonycyborg> Well.. Then it's either an SDL_ttf bug or wesnoth fails to fetch plane1 characters.
20091011 16:21:57< Arc> mordante: i'm aware, and i'm also aware that you're migrating to using pango with cairo backend, which means cairo is a dependency
20091011 16:22:23< Arc> we can eliminate SDL_TTF as a dependency and the concept of "codepoints" at the same time, likely speed things up and trim the code quite a bit
20091011 16:23:06< Arc> we use pango and cairo with PySoy so i've had about 1.5 years of experience working with it
20091011 16:24:08< mordante> loonycyborg, it's our bug we use the unicode routines instead of utf8
20091011 16:24:15< Arc> i can try to fix the sdl_ttf code if you'd prefer but i think a cairo rewrite of it would be easier and more useful
20091011 16:24:44< mordante> Arc, yes we can, I just wanted to know whether you wanted to fix it in SDL_TTF or in pango
20091011 16:25:31< Arc> i would prefer cairo since the result would be less wesnoth code and thus less place for bugs
20091011 16:25:57< Arc> i just need to know the API which this code is used from the outside with
20091011 16:26:10< loonycyborg> If you use cairo directly instead of pango, won't it screw arabic text?
20091011 16:27:34< Arc> *checking*
20091011 16:28:22-!- silene [n=plouf@wesnoth/developer/silene] has joined #wesnoth-dev
20091011 16:29:57-!- Sirp [n=user@wesnoth/developer/dave] has joined #wesnoth-dev
20091011 16:31:22-!- lizard_r [n=Miranda@wesnoth/umc-dev/developer/lizard] has joined #wesnoth-dev
20091011 16:33:04< Arc> you're right, we're going to need pango
20091011 16:35:14< loonycyborg> AFAICT you'll need api from text.hpp
20091011 16:36:44< Arc> im trying to discover the API thats actually used externally
20091011 16:36:55< Arc> i may be able to just migrate the external API calls to use text.cpp
20091011 16:39:13< silene> Arc: no, this is not sufficient; or at least this is not 1.8 material if you do it that way, as you will break all the things that rely on wml markup
20091011 16:39:14< loonycyborg> font.hpp seems to have text drawing functions that use sdl-ttf
20091011 16:40:48< mordante> loonycyborg, ???
20091011 16:41:39< silene> mordante: by the way, it is not surprising you don't remember who wrote this code, since it happened before you joined wesnoth ;-) (namely, the authors of this part of the code are yann, philippe, and i)
20091011 16:41:52< Arc> silene: wml markup is implemented in font.cpp?
20091011 16:42:16< loonycyborg> mordante: ?
20091011 16:42:22< Arc> silene: what would you suggest as a fix, just surrogate testing and pairing?
20091011 16:43:51< silene> Arc: yes, but you also have to use a better data structure than just an array containing all the characters from unicode as it is currently; for instance a patricia map for finding fonts would be both compact and efficient
20091011 16:44:59< Arc> i'm not familiar with patricia map
20091011 16:45:00< mordante> silene, I know it was written before I joined ;-) 
20091011 16:45:27< mordante> loonycyborg, in what way does font.hpp depend on SDL_TTF?
20091011 16:46:33< loonycyborg> It contains declarations of functions like draw_text_line()
20091011 16:46:58< Arc> so?
20091011 16:47:07< loonycyborg> Which seem to have fribidi conditionals in them.
20091011 16:47:25< loonycyborg> Therefore those are probably pre-pango functions.
20091011 16:48:10< mordante> ah I see
20091011 16:49:17< Arc> i'm sorry im not familiar with fribidi
20091011 16:53:28-!- ardesh [n=ardesh@port-92-195-111-27.dynamic.qsc.de] has quit [Read error: 110 (Connection timed out)]
20091011 16:54:19< Arc> it looks like subsets are ok
20091011 16:54:27< silene> Arc: patricia things for integers (which is the case here) are not too complicated, they are just (binary) trees labeled by integer intervals and children nodes have intervals included in their parent nodes; the only difficulty is in balancing them
20091011 16:57:21-!- EdB [n=edb@79.88.117.56] has quit [Remote closed the connection]
20091011 16:57:33-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has joined #wesnoth-dev
20091011 16:57:55< Arc> silene: im sorry that's really pushing what i can do in c++
20091011 17:00:34< Arc> would you be willing to try?
20091011 17:01:48< Arc> im barely understanding what the code in this file is doing
20091011 17:02:50< Arc> i dont understand why ucs2 is even used given that it triples the amount of storage with even latin text and sdl_ttf supports RenderUTF8
20091011 17:08:48< CIA-13> mordante * r39246 /trunk/src/gui/widgets/ (button.cpp button.hpp): 
20091011 17:08:48< CIA-13> Wire in the first public widget signal handler.
20091011 17:08:48< CIA-13> These signals should replace the callbacks, but this project will be
20091011 17:08:48< CIA-13> done post 1.8. But it shows the design idea.
20091011 17:08:52< CIA-13> mordante * r39247 /trunk/src/gui/dialogs/mp_connect.cpp: 
20091011 17:08:52< CIA-13> Convert a button callback to a signal handler.
20091011 17:08:54< CIA-13> Just a proof of concept change, the rest of the callbacks will be
20091011 17:08:56< CIA-13> changed post 1.8.
20091011 17:08:58< CIA-13> mordante * r39248 /trunk/src/gui/auxiliary/window_builder.cpp: 
20091011 17:09:00< CIA-13> Set the window type as id.
20091011 17:09:02< CIA-13> This helps debugging since the id is printed in the event debug
20091011 17:09:04< CIA-13> messages.
20091011 17:09:06< CIA-13> mordante * r39249 /trunk/src/gui/auxiliary/event/distributor.cpp: 
20091011 17:09:10< CIA-13> Disable the recursion locker.
20091011 17:09:12< CIA-13> This locker isn't needed for a top level dispatcher and it seems to
20091011 17:09:14< CIA-13> break (at least in this case). I however assume this commit fixes a
20091011 17:09:16< CIA-13> symptom not a cause...
20091011 17:09:18< CIA-13> mordante * r39250 /trunk/src/gui/dialogs/mp_connect.cpp: 
20091011 17:09:20< CIA-13> Force a redraw.
20091011 17:09:24< CIA-13> Just like the last commit it looks like it's hiding another bug.
20091011 17:09:35 * mordante neither and it's probably the cause of your problem
20091011 17:12:37-!- alink [n=alink@wesnoth/developer/alink] has joined #wesnoth-dev
20091011 17:13:24< mordante> hi alink 
20091011 17:13:31< alink> hello mordante
20091011 17:14:07< CIA-13> ivanovic * r39251 /trunk/po/wesnoth-dm/de.po: updated German translation
20091011 17:15:15< alink> mordante: you were right the implementation of xp_sort is really ugly, I will probably recode the fraction parsing, and maybe use it for HP and MP too
20091011 17:15:57< mordante> thanks alink 
20091011 17:16:07< alink> I quickly tried, and seems easy to write it in few lines by resuing our other string utility functions
20091011 17:16:54< mordante> alink, maybe you can also assign https://gna.org/patch/?1165 to yourself
20091011 17:18:13 * alink was not aware of this patch, looking
20091011 17:18:25< mordante> I thought so ;-)
20091011 17:19:16< mordante> Soliton, can you test whether the double click bug is now fixed in trunk?
20091011 17:19:22< Arc> this should take care of the current problem with codepoint mapping, right?
20091011 17:19:23< Arc> -	font_map.reserve(0x10000);
20091011 17:19:23< Arc> +	font_map.reserve(0x10480);
20091011 17:19:32< alink> but good, i was planning to write a level sorting breaking the equal cases by using xp difference, seems that patch does that
20091011 17:19:38< Arc> but the rendering is also wrong since the glyphs are getting doubled up
20091011 17:20:19< mordante> alink, no but I assume when you're done that patch will become won't do
20091011 17:20:59< mordante> and the patch seems abandoned 
20091011 17:22:18< alink> mordante: yes the patch itself, probably. But will keep the idea about level sort
20091011 17:22:45-!- SonIcco_ [n=SonIcco@pD9510B1E.dip0.t-ipconnect.de] has joined #wesnoth-dev
20091011 17:23:29< alink> in fact that patch uses code that i already deleted
20091011 17:24:21< mordante> :-)
20091011 17:25:01< alink> ilor: i saw that you were looking for a project for some webapp about wesnoth. Perhaps something for translation ?
20091011 17:25:48< ilor> alink: for now I'm counting on the addon server idea working, what do you have in mind?
20091011 17:25:53< alink> i know that there is http://www.wesnoth.org/gettext/ but maybe something else is needed/possible. I don't really follow translation stuff but just throwing the idea
20091011 17:26:10-!- grzywacz [n=grzywacz@wesnoth/developer/grzywacz] has joined #wesnoth-dev
20091011 17:26:41< alink> nothing precise, but for example I saw a open-source game with a simple webapp for translation
20091011 17:26:44< alink> http://www.mtp-target.org/mt/translation.php?lid=fr
20091011 17:27:07< alink> you can tie that to login of translator teams and stuff like that
20091011 17:27:19< alink> in case po tools are a barrier for new translators
20091011 17:27:52-!- Aethaeryn [n=Michael@wesnoth/umc-dev/developer/aethaeryn] has joined #Wesnoth-dev
20091011 17:28:06< Soliton> if you lower the barrier you most likely lower the quality.
20091011 17:28:55< alink> well you can lower the technical barrier, translators doens't need to know technical stuff
20091011 17:29:09< Arc> speak for yourself
20091011 17:29:11< Chusslove> Yes, they do need to know.
20091011 17:29:33< Chusslove> Just the amount of technical stuff required by PO format and surrounding tools.
20091011 17:30:00< Chusslove> To have a chance of making -- and *maintaining* -- a reasonable quality of translation.
20091011 17:30:28< alink> Chusslove: yes currently they do need to know, but that doesn't seem needed to be able to translate dialogues
20091011 17:31:27< Chusslove> Oh, it's needed alright. To translate a dialog, translator has to know to look into PO comments to see who is speaking (and before that was there, he had to look in WML file).
20091011 17:32:51< alink> ah yes indeed, need to know WML at least
20091011 17:32:52< mordante> ilor, Crab_ r39246 and r39247 show how the gui callbacks will look in the future
20091011 17:33:32< Arc> you could always use launchpad
20091011 17:33:36< alink> well the webapp may have a link displaying the WML file where it highlight the current line 
20091011 17:33:51< Arc> its free software now and contains a pretty good translations management tool
20091011 17:34:14< Arc> users can submit fixes for strings for review, for example
20091011 17:34:58< Chusslove> alink: The point is, when you add all that's needed into a web app, you'll come up with something having the same "barrier" as the current tools.
20091011 17:35:34< Chusslove> With only distinction that it will be slower and harder to use than a local tool :)
20091011 17:35:42< alink> Chusslove: maybe, i don't know I never worked with translation tools :-)
20091011 17:36:00< Chusslove> Ok, I'm just giving my estimate :)
20091011 17:36:09-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has quit []
20091011 17:36:34< mordante> personally I also think a local tool works better for translations
20091011 17:37:06-!- giusef [n=giusef@unaffiliated/giusef] has joined #wesnoth-dev
20091011 17:37:12< mordante> not that I did a lot of translating, only reading and correction the German one
20091011 17:39:21< Soliton> mordante: same thing.
20091011 17:40:18< Soliton> mordante: i wonder if some preference or so can influence the behaviour and explain why we see differences?
20091011 17:40:52-!- SonIcco [n=SonIcco@pD9510BF9.dip0.t-ipconnect.de] has quit [Read error: 101 (Network is unreachable)]
20091011 17:41:21< alink> ok but maybe as way to get new translators ? I don't know how translation team recruit, but a webapp-game presenting few untranslated sentences with a "can you translate this", that may trigger some interests in someone who never thought joining the translator team
20091011 17:41:56< Chusslove> Never happened, in my experience so far.
20091011 17:42:23< alink> ok, again, I have no idea how this currently works
20091011 17:44:26< alink> ilor: so, in fact, my suggestion would be simply "ask translators if they need something" :-)
20091011 17:45:20< alink> but apparently they seems happy with the current system
20091011 17:45:32< Chusslove> Oh, there are many things I'd like.
20091011 17:46:00-!- allefant [n=elias@allegro/developer/allefant] has quit [Connection timed out]
20091011 17:47:44< Chusslove> For example, if I've been editing a PO file in SVN, and hit, say, "posvn update" after they have been updated on server, I want to get true conflicts or no conflicts. That posvn is the tool I'd like someone to write.
20091011 17:48:09< Chusslove> (Conflicts on line-level are worthless in PO files.)
20091011 17:49:30< Chusslove> Of course, that someone would have to have good experience in working with PO files, so that he has an idea what a true conflict is and how to show it.
20091011 17:53:44< boucman> mordante: after some studying, I think my part is good, the pre_movement made visible a bug in the halo engine, probably linked to how haloes are removed. My guess is that they don't invalidate properly their old positions before being removed... trigering extra redraws don't seem to help
20091011 18:03:18< Arc> ok so the utf8_iterator is broken
20091011 18:03:44< mordante> Soliton, no idea which preference could do that, but I still have more chances pending so one of them might fix your issue
20091011 18:03:45< Arc> i would like to include http://utfcpp.sourceforge.net/ - it's small and mit licensed - is this permissible?
20091011 18:04:30< alink> at least, it seems that fiddling with the "show halo" option may trigger an assert
20091011 18:05:52< mordante> boucman, ok who'll take care of the bug? (it is a 1.8 blocker IMO)
20091011 18:06:33< mordante> Arc, what's broken?
20091011 18:06:37< alink> I can see the glitch now, I can take a look at it
20091011 18:07:05< alink> btw the assert cause by "halo option" : wesnoth: halo.cpp:330: void halo::unrender(std::set<map_location, std::less<map_location>, std::allocator<map_location> >): Assertion `invalidated_haloes.size() == 0' failed.
20091011 18:07:05< mordante> alink, the halo bug boucman was working on?
20091011 18:07:31< mordante> alink, IIRC I fixed that recently
20091011 18:07:57< alink> mordante: i'll update my yesterday trunk
20091011 18:08:14< mordante> alink, well then my fix didn't work did it a while ago
20091011 18:08:52< Arc> mordante: the iterator only seems to support 2 byte values
20091011 18:09:07< alink> mordante: to reproduce it: no halo, create a mage of light, reactivate halo. then some trigger the assert
20091011 18:09:18< Arc> instead of rewriting and maintaining it, it would seem better just to incorporate this library that does this and more
20091011 18:09:44< alink> sometime even just cliking the option does it (when it try to update the halo, I guess)
20091011 18:10:09< Arc> i think ive narrowed down the problem
20091011 18:10:29< Arc> but the iterator is part of it
20091011 18:10:34< mordante> alink, well feel free to fix it
20091011 18:11:23< Arc> we need the 32-bit codepoints from the iterator, not 16 bit chunks
20091011 18:11:30< mordante> the iterator seems to support more characters but it always returns a wchar_t type which is 32bit on Linux and 16bits on Windows
20091011 18:11:40< alink> mordante: ok I'll check if i can at least identify the source of the bug (and fix it if i see how)
20091011 18:12:54< Arc>         for(std::string::const_iterator c = current_substr.first+1;
20091011 18:12:54< Arc>                         c != current_substr.second; ++c) {
20091011 18:14:14< mordante> thanks alink
20091011 18:14:44< Arc> or is this an issue of me not understanding this code
20091011 18:14:50< loonycyborg> Arc: It looks like author had tried to submit utfcpp to boost judging by presense of boost/ dir in its repo.
20091011 18:15:07< loonycyborg> The question is why it wasn't included :P
20091011 18:18:10< mordante> Arc, it adds several chars to a wchar_t the first and second have nothing to do with the number of chars, they are the beginning and end of a range
20091011 18:18:50< mordante> off to hunt some food
20091011 18:19:37< Arc> well still wchar_t != uint
20091011 18:20:04< mordante> Arc, no so what does that mean?
20091011 18:20:18< Arc> well you said yourself, its 16 bits in windows
20091011 18:20:40< Arc> i am eliminating the ucs2 code, going pure utf8, and i need to be able to properly append the codepoint from the iterator to the chunk text
20091011 18:21:24< Arc> utfcpp has a function called utf8::append that does this
20091011 18:22:42< Arc> between that and utf8::next the utf8_iterator should no longer be needed
20091011 18:24:13< Arc> its my general philosophy to use a library whenever possible
20091011 18:30:19-!- Zarel [n=Zarel@warzone2100/developer/Zarel] has joined #wesnoth-dev
20091011 18:33:47< boucman> was afk, I'm back but not for long
20091011 18:34:16< boucman> mordante: yes it's a 1.8 blocker, and I have no idea how to handle this...
20091011 18:35:48< alink> boucman: i canl quickly try if  i see something (for the halo bug)
20091011 18:38:03-!- Netsplit pratchett.freenode.net <-> irc.freenode.net quits: shikadibot, Tigge, SonIcco_
20091011 18:38:08-!- SonIcco__ [n=SonIcco@pD9510B1E.dip0.t-ipconnect.de] has joined #wesnoth-dev
20091011 18:38:57< mordante> Arc, I agree with libraries, when they are good
20091011 18:39:21< boucman> alink: i have to leave for the evening, but mordante know how to test (basically move a MoL up and down)
20091011 18:39:34< mordante> but since you're on Linux returning wchar_t should work
20091011 18:39:40< alink> boucman: yes i see the glitch here too
20091011 18:39:43< boucman> just report your findings in bug 14405
20091011 18:39:52< boucman> see you all later
20091011 18:39:59< mordante> bye boucman 
20091011 18:40:09< alink> boucman: ok will do
20091011 18:40:23-!- Netsplit over, joins: shikadibot, Tigge
20091011 18:41:26< ilor> off for a bit, will be around late in the evening
20091011 18:41:36-!- ilor [n=user@wesnoth/developer/ilor] has quit []
20091011 18:41:37< mordante> bye ilor 
20091011 18:42:33< Arc> mordante: but wesnoth runs on windows
20091011 18:44:37< mordante> Arc, I know just using wchar_t should work for you
20091011 18:45:00< mordante> and obviously it needs to be fixed so it works properly on all systems
20091011 18:45:22-!- giusef [n=giusef@unaffiliated/giusef] has quit ["exit (-1);"]
20091011 18:45:25< Arc> ok then where is the utf8 append function
20091011 18:46:16< Arc> i'm sorry for being distrustful, i just see a lot of code in here that makes poor assumptions and i'm thus more trusting of a library that's been properly tested
20091011 18:46:43< Arc> i hope you can understand how frustrating it is when you don't know a language very well and trying to debug it
20091011 18:47:28< mordante> I understand, but code often tends to make assumptions that were good at the time
20091011 18:48:09< mordante> I'm only reluctant to change code which we know that has been working for a long time for library we don't know
20091011 18:48:19< mordante> especially since we're in a feature freeze
20091011 18:48:24-!- Netsplit pratchett.freenode.net <-> irc.freenode.net quits: shikadibot, Tigge
20091011 18:50:13-!- Netsplit over, joins: shikadibot, Tigge
20091011 18:52:16< Arc> i think we can reasonably test this in a few minutes.
20091011 18:52:53< Arc> i would rather use this library than write a function that tries to properly append a utf8 string with a wchar knowing it wont work right on windows anyway
20091011 18:53:11< loonycyborg> btw there's another boost unicode lib: http://blogloufoque.free.fr/unicode/doc/html/ it's this year's GSOC project.
20091011 18:55:02< Arc> i dont think boost has anything to do with it
20091011 18:58:18-!- EdB [n=edb@214.153.82-79.rev.gaoland.net] has joined #wesnoth-dev
20091011 18:58:30-!- Netsplit pratchett.freenode.net <-> irc.freenode.net quits: shikadibot, Tigge
20091011 18:58:36< loonycyborg> Well.. both of those libraries are in boost's vault, so it's related at least tangentially :P
20091011 18:58:39< Arc> but it seems moot since i dont know what is going on with c++ for iterators
20091011 18:58:56-!- Netsplit over, joins: shikadibot
20091011 19:00:55< Arc> im going to give up.  there's just too much of a curb here
20091011 19:01:11< loonycyborg> C++ iterators in general or only wesnoth's utf8_iterator? :P
20091011 19:01:26< Arc> c++ iterators in general
20091011 19:01:50< Arc> utils::utf8_iterator end = utils::utf8_iterator::end(utf8_text); << that makes no sense to me at all
20091011 19:01:57< mordante> what's wrong with them
20091011 19:02:06< Arc> for(i=0;i<end;i++) is what im used to
20091011 19:02:23< mordante> utils::utf8_iterator::end(utf8_text) is not a typical C++ iterator
20091011 19:02:26< Arc> but what the hell does that line do? its a string of pseudocode nonsense
20091011 19:02:55< loonycyborg> It constructs a utf8_iterator from utf8_text :P
20091011 19:03:06< Arc> but it already did it earlier!
20091011 19:03:09< mordante> create an iterator to the end of the utf8_text
20091011 19:03:13< Arc> utils::utf8_iterator ch(utf8_text);
20091011 19:03:23< Arc> ch should be the iterator
20091011 19:03:28< Sirp> Arc: well, imagine it was like this: utf8* end = utf8_end(utf8_text);  <-- what it would look like in C
20091011 19:03:33< silene> mordante: as a matter of fact, it is; begin and end are free functions too in c++
20091011 19:03:38< Arc> so i would expect for(ch; ch!=end; ++ch)
20091011 19:04:04< Sirp> Arc: think of iterators as being like pointers. ch is a pointer to the beginning, end is a pointer to the end
20091011 19:04:08< mordante> silene, not how you typically use them
20091011 19:04:41-!- Tigge [n=tigge@c-5ba372d5.015-28-73746f34.cust.bredbandsbolaget.se] has joined #wesnoth-dev
20091011 19:04:42< Sirp> Arc: in C you can iterate over an array like this:
20091011 19:04:47< silene> mordante: that's the only ones you can use when doing templates for instance, since you can't use the member functions on arrays or things like that
20091011 19:04:58< Arc> Sirp: so why not just use the normal end statement
20091011 19:05:24< Sirp> utf8* ch = utf8_text; utf8* end = utf8_end(utf8_text); for(; ch != end; ++ch) { ... }
20091011 19:05:31< Sirp> Arc: what "normal end statement"?
20091011 19:05:40< Arc> i<end
20091011 19:05:49< Arc> or in this case ch != end
20091011 19:05:50< Sirp> Arc: instead of != end you mean?
20091011 19:06:19< Sirp> Arc: sure, that's the normal way to test ending iteration, ch != end
20091011 19:06:35< mordante> silene, I know, but I meant that most of the time when you use iterators you use it on a container with a begin() and end() member
20091011 19:06:38< Arc> i am comparing to this:
20091011 19:06:40< Arc> std::string s = "example";
20091011 19:06:40< Arc> utf8::iterator i (s.begin(), s.begin(), s.end());
20091011 19:06:52< Sirp> ahhhh
20091011 19:06:55< Arc> why are strings functions?
20091011 19:07:15< mordante> type're not they're classes
20091011 19:07:32< mordante> objects*
20091011 19:07:40< Arc> *sigh* ok im going to try something
20091011 19:08:03-!- DDR [n=chatzill@66.183.125.196] has joined #wesnoth-dev
20091011 19:10:40< Arc> wait, why do you even need a for statement when you have iterators?
20091011 19:11:34< mordante> yes why not
20091011 19:11:45< Sirp> Arc: because iterators are like pointers.
20091011 19:11:52< Sirp> Arc: iterators are simply a generalization of pointers
20091011 19:12:26< loonycyborg> pointers can only point to memory locations while iterators can point to everything :P
20091011 19:12:40< Sirp> if you had two pointers to the beginning and ending of an array you would need a for loop to iterate over the array
20091011 19:12:51< Sirp> likewise if you had two iterators to the beginning and ending of a sequence you need a for loop
20091011 19:13:53< loonycyborg> An iterator is an object that can be dereferenced giving pointed to object and incremented giving iterator to next object.
20091011 19:14:51-!- EdB [n=edb@214.153.82-79.rev.gaoland.net] has quit [Remote closed the connection]
20091011 19:15:06< loonycyborg> So if you have begin and end iterators and want to loop over all objects between them then you'll need to make a loop that'll increment the begin iterator until it equals end iterator.
20091011 19:15:13< Arc> then  why does the constructor for this iterator take the end argument?
20091011 19:16:28< loonycyborg> You mean 'utf8_iterator(std::string::const_iterator const &begin, std::string::const_iterator const &end);'? Probably to operate on substrings.
20091011 19:16:30< Arc> in the above example would for (utf8::iterator i (s.begin(), s.begin(), s.end()); i != i.end; ++i) be correct?
20091011 19:22:45< loonycyborg> i (s.begin(), s.begin(), s.end() <- why do you pass s.begin() twice?
20091011 19:23:24< Arc> presumably for substrings
20091011 19:23:43-!- blarumyrran [n=minaise@81-20-159-197.levira.ee] has quit []
20091011 19:24:00< Arc> if interators are pointers, then why is this code dereferencing an array based on the size of the pointer?
20091011 19:25:39< loonycyborg> utf8_iterator is defined in serialization/string_utils.hpp and I see only one and two argument constructors there..
20091011 19:25:54< silene> Arc: size_t is not the size of pointer; just assume it is not there for the purpose of understanding the code
20091011 19:28:53< loonycyborg> utils::utf8_iterator::end(utf8_text) calls a static function that returns an iterator to the end of utf8 sequence
20091011 19:29:17< loonycyborg> While utils::utf8_iterator ch(utf8_text); creates an iterator to it's beginning.
20091011 19:30:07< loonycyborg> *its
20091011 19:34:33< loonycyborg> size_t(*ch) is a C-style cast. size_t is a standard type.
20091011 19:35:30< Arc> silene: its the size of whats pointed to, and i do need to understand what it does
20091011 19:36:09< silene> Arc: no, it is not
20091011 19:36:18< Arc> what does it do then?
20091011 19:36:34< silene> Arc: it does nothing, this is just a type cast
20091011 19:36:48< silene> that's why i said you could just ignore it for now
20091011 19:36:51< loonycyborg> Most likely it suppresses compiler warnings or something :P
20091011 19:37:29< mordante> I'm off bye
20091011 19:37:36< silene> loonycyborg: no, it is not that useless, it prevents segfaults due to possible signedness
20091011 19:38:06< Arc> im changing the type of ch from wchar to uint, so can i just unwrap the sizet()?
20091011 19:38:15-!- mordante [n=mordante@wesnoth/developer/mordante] has quit ["Leaving"]
20091011 19:44:38-!- Blueblaze [n=nick@adsl-99-171-161-30.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20091011 19:49:32< loonycyborg> Arc: Probably. size_t is most likely uint too :P
20091011 19:50:47< CIA-13> silene * r39252 /trunk/src/actions.cpp: Removed duplication of attack code. Ensured that battle contexts are always refreshed after WML events.
20091011 19:53:21< zookeeper> silene, hmm, is that going to change something about what sort of info can be manipulated or retrieved in mid-combat events?
20091011 19:54:05< silene> zookeeper: if it does change a thing, this is a bug
20091011 19:54:32< zookeeper> ah, ok
20091011 19:56:08-!- ardesh [n=ardesh@port-92-206-90-178.dynamic.qsc.de] has joined #wesnoth-dev
20091011 19:58:51-!- loktor [n=quassel@c83-254-63-195.bredband.comhem.se] has joined #wesnoth-dev
20091011 20:06:54-!- ardesh [n=ardesh@port-92-206-90-178.dynamic.qsc.de] has quit ["Quis custodiet ipsos custodes"]
20091011 20:09:12< Arc> what are "template arguments"
20091011 20:09:27< Arc> src/font.cpp:116: error: missing template arguments before ‘ch’
20091011 20:09:53< Arc> utf8::iterator ch(utf8_text.begin(), utf8_text.begin(), utf8_text.end());
20091011 20:10:41-!- Crab__ [n=Miranda@economical.offerer.volia.net] has joined #wesnoth-dev
20091011 20:13:35< zookeeper> btw, the loading times seem to have increased a lot. the "initializing teams" phase takes forever, and sometimes the transition from a story screen to the scenario takes forever too.
20091011 20:14:01< zookeeper> (no, not literally forever, so you can skip the funny remarks)
20091011 20:17:58< CIA-13> caslav_ilic * r39253 /trunk/ (2 files in 2 dirs): Updated localized images for German.
20091011 20:18:23< loonycyborg> Arc: Why do you believe that it should take three arguments? :/
20091011 20:19:05< CIA-13> caslav_ilic * r39254 /branches/resources/image-localization/de/ (wesnoth--overlay.png wesnoth.xcf): Updated source localized images for German.
20091011 20:19:27< Arc> loonycyborg: because it says you do
20091011 20:20:03< loonycyborg> Ah. You're using utf8::iterator, not utils::utf8_iterator
20091011 20:20:09 * loonycyborg was confused.
20091011 20:20:10< Arc> yes.
20091011 20:21:30< loonycyborg> It should be utf8::iterator<SomeType> ch(
20091011 20:22:25< Crab__> mordante: noted  (about r39246 and r39247 )
20091011 20:22:41< Arc> i tried <char*> and ..
20091011 20:23:13< Arc> what is the something when the source is const?
20091011 20:23:45< loonycyborg> I'm not sure what that type should be. I've never used that library.
20091011 20:24:43< Arc> http://pastebin.ca/1613125
20091011 20:24:58< Arc> how do you "template" for const?
20091011 20:26:37-!- MrTree [n=marc@ip-88-153-48-155.unitymediagroup.de] has joined #wesnoth-dev
20091011 20:31:46< loonycyborg> <const char*>? :P
20091011 20:32:27< Arc> tried it
20091011 20:36:30< loonycyborg> Anyway, isn't it in the library's documentation?
20091011 20:50:22< loonycyborg> Arc: Try utf8::iterator<std::string::const_iterator> ch(
20091011 20:51:09< loonycyborg> Or convert it to C string with .c_str() .
20091011 20:54:56< alink> zookeeper: can you post a --log-info=engine output so using timestamp, maybe we can see what took this time 
20091011 20:56:04< Arc> i tried both
20091011 20:58:41< zookeeper> alink, sure, probably within an hour or so..
20091011 20:59:09< alink> ok, thanks, I will be away at that time but I will look at it
20091011 21:01:23< Arc> damnit
20091011 21:01:29< Arc> see this is why i shouldnt work on the code.
20091011 21:01:41< Arc>                 utf8::iterator<char*> ch(utf8_text.c_str());
20091011 21:01:47< alink> zookeeper: if possible something that i can compare with mine. Like test-scenario or a campaign start (or :cl). But if you have a special pathological case, it could be useful too
20091011 21:01:47< Arc> that should work
20091011 21:05:56-!- noy [n=Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20091011 21:07:45-!- loktor [n=quassel@c83-254-63-195.bredband.comhem.se] has quit [Remote closed the connection]
20091011 21:27:05-!- mjs-de [n=mjs-de@vpw.wh.uni-dortmund.de] has joined #wesnoth-dev
20091011 21:34:35< boucman> back...
20091011 21:41:18< alink> wesbot: log r36601
20091011 21:41:19< wesbot> mordante * r36601 : Fixes teleporting leaving a unit halo.The probkem has been spotted by MJ999, who also provided this patch. I haven'tbeen able to reproduce the problem, but the invalidation should have been therebefore.
20091011 21:41:23< wesbot> URL: http://svn.gna.org/viewcvs/wesnoth?view=rev&rev=36601
20091011 21:44:44< alink> revert it seems to fix the halo bug, but not sure yet why. I suspect it just unhide another bug
20091011 21:45:27-!- MrTree_ [n=marc@ip-88-153-48-155.unitymediagroup.de] has joined #wesnoth-dev
20091011 21:46:39< alink> anyway, must go, bbl
20091011 21:47:21-!- MrTree [n=marc@ip-88-153-48-155.unitymediagroup.de] has quit [Read error: 60 (Operation timed out)]
20091011 21:50:52-!- loktor [n=quassel@c83-254-63-195.bredband.comhem.se] has joined #wesnoth-dev
20091011 21:51:18< zookeeper> hmh...
20091011 21:51:44< zookeeper> 20091011 22:47:50 info engine: inserting unit for side 1
20091011 21:52:04< zookeeper> and when starting TB, there's 10 of those (10 preplaced units), and each one takes 1 full second
20091011 21:52:15< zookeeper> at least that's what those timestamps say
20091011 21:52:37< zookeeper> so the last one is 22:47:59 meaning placing those 10 units took...10 seconds
20091011 21:56:41-!- loktor [n=quassel@c83-254-63-195.bredband.comhem.se] has quit [Remote closed the connection]
20091011 21:57:16-!- Aethaeryn [n=Michael@wesnoth/umc-dev/developer/aethaeryn] has quit ["meh"]
20091011 21:58:58< boucman> alink: that weird, that line makes sense...
20091011 21:59:06< boucman> I'll add it in a comment to the bug
20091011 22:03:33-!- wesbot changed the topic of #wesnoth-dev to: string/feature freeze active! | 101 bugs, 237 feature requests, 10 patches | logs: http://irclogs.wesnoth.org/ | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com | http://imagebin.org
20091011 22:04:55-!- Aethaeryn [n=Michael@wesnoth/umc-dev/developer/aethaeryn] has joined #Wesnoth-dev
20091011 22:05:35< zookeeper> not exactly a good test case for reproducing i guess, but this is what i get when loading an IftU beginning-of-scenario save: http://wesnoth.pastebin.com/m8bbd47f
20091011 22:06:07< zookeeper> there's two places where there's a long delay for no obvious reason
20091011 22:12:15< zookeeper> this is the log from starting TB: http://wesnoth.pastebin.com/m5ebf914a
20091011 22:14:41-!- SonIcco__ [n=SonIcco@pD9510B1E.dip0.t-ipconnect.de] has quit [Read error: 113 (No route to host)]
20091011 22:16:22-!- Aethaeryn [n=Michael@wesnoth/umc-dev/developer/aethaeryn] has quit ["meh"]
20091011 22:20:40-!- Espreon [n=espreon@wesnoth/developer/espreon] has joined #wesnoth-dev
20091011 22:22:30< silene> zookeeper: can you compile?
20091011 22:27:59< loonycyborg> silene: AFAIK he uses windows binary that is cross-compiled on wesnoth.org server.
20091011 22:30:08< silene> that's unfortunate
20091011 22:31:24< loonycyborg> You want him to test changes that you don't want to commit?
20091011 22:32:19< zookeeper> silene, nope
20091011 22:33:20< silene> loonycyborg: no, i want additional log lines; i want to know where the 15 seconds are spent when initializing a side
20091011 22:34:00< zookeeper> there isn't anything more verbose than --log-info=engine i could use?
20091011 22:34:11< silene> my guess would be commit 38401, but i won't revert it without any testing
20091011 22:34:39< silene> zookeeper: not in this part of the code
20091011 22:38:07< Soliton> well, it's no problem to revert a commit and rebuild the windows binary.
20091011 22:39:36< zookeeper> if someone gets a build up then i can of course test it (tomorrow), but i'd imagine that something like this would be reproducible...or maybe it's just windows-specific, i wouldn't know.
20091011 22:41:39< silene> zookeeper: can you post the savegame you used for "a real confrontation"?
20091011 22:43:41< zookeeper> silene, yeah, http://filebin.ca/vrwcfp/IftU-1-A_Real_Confrontation
20091011 22:49:14< silene> i can't reproduce it
20091011 22:51:09< silene> (that is, the scenario takes only 252ms to initialize here, and i doubt my laptop is so much more powerful than your computer)
20091011 22:52:18-!- boucman [n=rosen@wesnoth/developer/boucman] has quit ["Leaving."]
20091011 23:03:00-!- zookeeper [n=l@wesnoth/developer/zookeeper] has quit []
20091011 23:04:53< CIA-13> espreon * r39255 /branches/1.6/data/ (6 files in 6 dirs): Backported the improved maps from trunk.
20091011 23:05:41-!- MrTree_ [n=marc@ip-88-153-48-155.unitymediagroup.de] has quit [Read error: 110 (Connection timed out)]
20091011 23:06:19-!- MrTree_ [n=marc@ip-88-153-48-155.unitymediagroup.de] has joined #wesnoth-dev
20091011 23:20:54-!- Zarel [n=Zarel@warzone2100/developer/Zarel] has quit ["This computer has gone to sleep"]
20091011 23:27:52-!- Appleman1234 [n=Appleman@131.181.101.183] has joined #wesnoth-dev
20091011 23:37:42-!- Crab__ [n=Miranda@economical.offerer.volia.net] has quit ["Miranda IM! Smaller, Faster, Easier. http://miranda-im.org"]
20091011 23:39:57-!- MrTree_ [n=marc@ip-88-153-48-155.unitymediagroup.de] has quit [Client Quit]
20091011 23:56:59-!- silene [n=plouf@wesnoth/developer/silene] has quit ["Leaving."]
20091011 23:58:47-!- Doppp [n=Doppp@unaffiliated/doppp] has quit [Connection timed out]
--- Log closed Mon Oct 12 00:00:30 2009