--- Log opened Mon Mar 30 00:00:18 2009
20090330 00:05:15-!- gozaboru [i=akhilpai@markley-135168.reshall.umich.edu] has quit []
20090330 00:07:48-!- boucman [n=rosen@wesnoth/developer/boucman] has quit [Remote closed the connection]
20090330 00:11:56-!- loonycyborg [n=sergey@wesnoth/developer/loonycyborg] has quit ["Zzzzzzzzzzzzzzzzzzzzzzzzz"]
20090330 00:12:06< Ivanovic> time for me to head off to bed, n8
20090330 00:12:20< CIA-30> esr * r34267 /trunk/data/ (5 files in 5 dirs): 
20090330 00:12:20< CIA-30> Encapsulate some Pango markup in WHISPER and ASIDE macros so these conventions
20090330 00:12:20< CIA-30> can be tweaked without changing translatable strings.
20090330 00:12:30-!- loonybot [n=loonybot@wesnoth/bot/loonybot] has quit [Remote closed the connection]
20090330 00:13:10< Shadow_Master> esr: I'm not sure about making text smaller when characters are whispering
20090330 00:13:43< Shadow_Master> it doesn't convey it for me.
20090330 00:15:02< esr> Shadow_Master: Not even with the parens?
20090330 00:15:03< Shadow_Master> esr: nor as a translator, nor as a reader
20090330 00:15:19< Shadow_Master> esr: the parenthesis will give you problem with translations, for starters
20090330 00:15:45< Shadow_Master> some languages have special rules for that kind of punctuation signs.
20090330 00:15:50< esr> We already had that problem - parebs were already in use this way.
20090330 00:16:01< Shadow_Master> so, are we supposed to make it worse?
20090330 00:16:31< Shadow_Master> esr: your usage won't expose the parethesis to translators
20090330 00:16:48< esr> No, the macro is supposed to make it possible to change the convention is necessary.
20090330 00:17:11< Shadow_Master> and why does the text have to be small?
20090330 00:17:57< Shadow_Master> in a book, are whispers written in a smaller font? no.
20090330 00:18:20< esr> Because that's the closest thing to a convebtion for "whisper" there is.  It's fairly common in English, at least.
20090330 00:18:37-!- csaunders_ [n=csaunder@ip216-239-92-243.vif.net] has joined #wesnoth-dev
20090330 00:18:50< Shadow_Master> I tend to believe that this kind of changes should be discussed prior to implementation..
20090330 00:18:51< esr> I've seen it in books, actually.  Thoufgh it's probably more common in comics.
20090330 00:19:03< Sapient> italics, maybe?
20090330 00:19:25-!- csaunders [n=csaunder@ip216-239-74-177.vif.net] has quit [Read error: 110 (Connection timed out)]
20090330 00:19:53< esr> Sapient: I'm using italics in parens for asides.  That's *well* established; playwrights use it a lot.
20090330 00:19:59< fabi> I thought wispering in comics are done by displaying the letters with dots.
20090330 00:20:38< esr> fabi: Sometimes, yes.
20090330 00:20:40< Shadow_Master> I treat wesnoth like movie subtitles :-/
20090330 00:20:56< Shadow_Master> caption. whatever the english word for it si.
20090330 00:21:20< Velory> boucman:around ?
20090330 00:21:50< esr> Shadow_Master: That's a reasonable approach.  The form is something like that and something like a play.
20090330 00:21:56< Velory> i think i finished my healing imp.
20090330 00:22:24< Shadow_Master> esr:  yeah. I never see bigger or smaller fonts in movie text captions
20090330 00:22:51< dfranke> movie subtitles are alongside spoken dialog
20090330 00:23:07< dfranke> so it's unnecessary there to have any written convention for whispering
20090330 00:23:24< esr> Shadow_Master: No, I guess you won't.  Those captioning systems are much more limited tha, for example, the typography iused in printed dramatic works.
20090330 00:23:32< Shadow_Master> dfranke: indeed :p
20090330 00:25:09< Shadow_Master> esr: I'll go off the tangent of this discussion and switch to a different related issue. does wmlxgettext expose to translators the names of the macros that include a translatable string in their argument list? when will you rewrite wmlxgettext in your favorite language so we can have a permanent maintainer for it?
20090330 00:26:00< Shadow_Master> if wmlxgettext does that, I can know by looking at the gettext comment lines that a string is a WHISPER or ASIDE. But totherwise I can't force my translators to read the WML
20090330 00:26:45< esr> Shadow_Master: Hm...interesting point, and rewriting wmlgettexxt has been on my to-do list for a long time now.
20090330 00:27:21< esr> If there's an FR for it, please attach your requirement.  If there isn't, start one and assign it to me.
20090330 00:27:27-!- Rrenys [n=villu@81-20-159-197.levira.ee] has joined #wesnoth-dev
20090330 00:28:11-!- Rrenys [n=villu@81-20-159-197.levira.ee] has quit [Remote closed the connection]
20090330 00:28:52< Shadow_Master> yeah. wmlxgettext doesn't include macro info. *brrr*
20090330 00:28:52-!- molgrum [n=molgrum@c83-249-53-163.bredband.comhem.se] has quit ["Lämnar"]
20090330 00:29:12< Shadow_Master> and I have all my [messages] encapsulated by the {MSG_UNIT} macro in my campaign. :((
20090330 00:29:18-!- iyonius [n=iyo@a02-0829c.kn.vutbr.cz] has quit [Read error: 104 (Connection reset by peer)]
20090330 00:29:19< CIA-30> soliton * r34268 /branches/1.6/src/server/server.cpp: make the diff from the right player after a kick(ban)
20090330 00:30:23< CIA-30> soliton * r34269 /trunk/src/server/server.cpp: make the diff from the right player after a kick(ban)
20090330 00:30:41-!- andreluiz [i=andrelui@unaffiliated/andreluiz] has joined #wesnoth-dev
20090330 00:32:25-!- Sapient [n=sapien-x@wesnoth/developer/sapient] has left #wesnoth-dev []
20090330 00:36:20< Shadow_Master> esr: added
20090330 00:36:39< esr> Thanks.
20090330 00:37:01< esr> I haave not forgotten that I have anither wmllint FR queued from you.
20090330 00:38:39-!- andreluiz [i=andrelui@unaffiliated/andreluiz] has quit []
20090330 00:40:26< CIA-30> fendrin * r34270 /trunk/data/campaigns/An_Orcish_Incursion/maps/ (2_Assassins.map 4_Valley_of_trolls.map 7_Showdown.map): AOI: 2,4,7: Map border cleanups.
20090330 00:41:20-!- West_ [n=West@hd5e2754c.gavlegardarna.gavle.to] has quit ["b'bye"]
20090330 00:43:15-!- PK [n=pk@r74-192-44-206.vctrcmta01.vctatx.tl.dh.suddenlink.net] has quit ["Java user signed off"]
20090330 00:50:40-!- gozaboru [i=akhilpai@markley-135168.reshall.umich.edu] has joined #wesnoth-dev
20090330 00:55:48-!- busfahrer [n=busfahre@unixboard/user/busfahrer] has joined #wesnoth-dev
20090330 00:56:18-!- busfahrer` [n=busfahre@ulmg-5f70eee1.pool.einsundeins.de] has joined #wesnoth-dev
20090330 01:00:31-!- csaunders_ is now known as csaunders
20090330 01:01:11-!- csaunders [n=csaunder@ip216-239-92-243.vif.net] has quit ["i <3 pork (http://dev.ojnk.net)"]
20090330 01:01:23< CIA-30> dfranke * r34271 / (6 files in 2 dirs): Ignore healing due to levelup when computing damage statistics. Fixes bug #13277.
20090330 01:02:25< esr> Nice catch, dfranke - I just read that bug report.
20090330 01:02:33< dfranke> esr: thanks
20090330 01:03:46-!- csaunders [n=chris@ip216-239-92-243.vif.net] has joined #wesnoth-dev
20090330 01:04:21-!- busfahrer` [n=busfahre@ulmg-5f70eee1.pool.einsundeins.de] has quit ["leaving"]
20090330 01:05:31< Velory> gnite everyone
20090330 01:05:39-!- Velory [n=GProcess@81.213.68.152] has left #wesnoth-dev []
20090330 01:07:12-!- thespaceinvader [n=chatzill@wesnoth/artist/thespaceinvader] has quit ["night all"]
20090330 01:08:01-!- yann [n=dwitch@nan92-1-81-57-214-146.fbx.proxad.net] has quit [Remote closed the connection]
20090330 01:10:52-!- stikonas [n=stikonas@wesnoth/translator/stikonas] has quit [Read error: 110 (Connection timed out)]
20090330 01:14:37-!- busfahrer [n=busfahre@unixboard/user/busfahrer] has quit ["leaving"]
20090330 01:17:06< Polarina> Ivanovic: I send you updated translations.
20090330 01:32:15-!- ryochan7 [n=ryochan7@173.9.238.161] has joined #wesnoth-dev
20090330 01:33:29-!- Ambushsabre [n=ambushsa@c-75-67-216-121.hsd1.nh.comcast.net] has joined #wesnoth-dev
20090330 01:46:04-!- csaunders [n=chris@ip216-239-92-243.vif.net] has quit ["BitchX-1.1-final -- just do it."]
20090330 01:46:54-!- csaunders [n=chris@ip216-239-92-243.vif.net] has joined #wesnoth-dev
20090330 01:48:27-!- Elvish_Pillage2 [n=eli@66-189-40-122.dhcp.oxfr.ma.charter.com] has quit ["Hi! I'm a quit message virus vaccine. If you see a quit message virus, don't replace your quit message with it!"]
20090330 01:53:47< Polarina> I HATE WEBSITES THAT LIE! IT DOESN'T MATTER WHICH SITE!  See http://www.wesnoth.org/gettext/index.lang.php?lang=is&version=branch and click on "W3C XHTML 1.0 Valid" button on the bottom.
20090330 01:55:26< esr> dfranke: I have a question about FR#13085.  What do you want tilde files renamed to?
20090330 02:05:54< dfranke> esr: replacing the tildes with underscores would probably work.
20090330 02:06:34< dfranke> esr: it might be a non-issue though.  Someone suggested that the tilde files on the 1.4 server might all be emacs backup files that accidentally got uploaded.
20090330 02:06:55< dfranke> esr: and I haven't checked them myself.
20090330 02:09:17< Soliton> Polarina: check out the resource branch and make it stop!
20090330 02:09:50< Polarina> Soliton: Ok.
20090330 02:09:54< Polarina> Tomorrow, tired.
20090330 02:14:34-!- Polarina [n=polarina@wesnoth/translator/Polarina] has quit ["Leaving."]
20090330 02:14:35-!- alink [n=alink@wesnoth/developer/alink] has joined #wesnoth-dev
20090330 02:17:37< Soliton> Rhonda: please remove wesbot from #wesnoth-mp again. it seems we'd rather like it to remain unlogged. :-)
20090330 02:23:02< orian_> cu all
20090330 02:23:04-!- orian_ [n=orian@abwz192.neoplus.adsl.tpnet.pl] has left #wesnoth-dev []
20090330 02:38:27-!- stevenj [i=4bba7fd7@gateway/web/ajax/mibbit.com/x-1e20a96e6a60691c] has joined #wesnoth-dev
20090330 02:49:56-!- Sirp [n=me@wesnoth/developer/dave] has joined #wesnoth-dev
20090330 03:04:39-!- res|laptop [n=res@pdpc/supporter/student/res2k] has quit [Read error: 60 (Operation timed out)]
20090330 03:07:30-!- alink is now known as alink_
20090330 03:08:25-!- alink_ [n=alink@wesnoth/developer/alink] has quit []
20090330 03:09:15-!- alink_ [n=alink@78.129.18.215] has joined #wesnoth-dev
20090330 03:09:37-!- alink_ [n=alink@78.129.18.215] has left #wesnoth-dev ["Konversation terminated!"]
20090330 03:13:25-!- alink [n=alink@78.129.18.215] has joined #wesnoth-dev
20090330 03:24:03-!- _deleter [n=_deleter@137.112.136.76] has joined #wesnoth-dev
20090330 03:24:52< esr> dfranke: In that case I think I'll just have wmllint throw a warning.
20090330 03:25:23-!- Ambushsabre [n=ambushsa@c-75-67-216-121.hsd1.nh.comcast.net] has quit [Read error: 110 (Connection timed out)]
20090330 03:25:30< dfranke> esr: sounds fair
20090330 03:25:33-!- csaunders [n=chris@ip216-239-92-243.vif.net] has quit ["csaunders has no reason"]
20090330 03:36:36< CIA-30> esr * r34272 /trunk/data/tools/wmllint: Address bug #13095 by having wmllint throw a warning on filenames ending in tilde.
20090330 03:38:25< dfranke> esr: it's not just files ending in tilde, it's tilde anywhere in the filename.  In fact if they end in tilde we already ignore those for upload as of somewhere in 1.5 prior to when I reported the security issue.
20090330 03:38:50< esr> OK, I can fix that trivially.
20090330 03:40:49< CIA-30> esr * r34273 /trunk/data/tools/wmllint: For bug #13095, warn on files with a tilde *anywhere*.
20090330 03:43:00-!- happygrue [n=George@wesnoth/developer/wintermute] has quit ["http://xkcd.com/91/"]
20090330 03:45:04< dfranke> ok, that should do.
20090330 03:45:19 * dfranke blows a rasberry at Micros~1 for making us deal with this
20090330 03:50:02< Shadow_Master> haha :)
20090330 03:50:53< Shadow_Master> (at the 8.3 name, of course. I don't laugh at vulner~1)
20090330 03:51:46-!- noy [n=Noy@wesnoth/developer/noy] has quit ["GO, GET TO THE CHOPPAH!!!"]
20090330 03:52:00-!- Sapient [n=sapien-x@c-71-204-10-166.hsd1.ga.comcast.net] has joined #wesnoth-dev
20090330 03:52:20< Sapient> hi Voris
20090330 03:52:36< Sapient> voris: you have been making many changes to the wiki, right?
20090330 03:55:27< Sapient> hmm, nevermind I think I found what I was looking for
20090330 03:55:45< Sapient> someone has changed the defender_* events to fire with attacker as primary unit
20090330 03:56:09< Shadow_Master> :-/
20090330 03:56:34< Sapient> Shadow_Master: shouldn't defender be the primary unit in the defender_hits and misses events?
20090330 03:56:35< Shadow_Master> http://www.wesnoth.org/forum/memberlist.php?mode=viewprofile&u=105651 
20090330 03:57:30< Shadow_Master> Sapient: don't ask me that kind of questions unless you really want my head to explode and leave its sticky remains all over the place
20090330 03:57:42 * Sapient asks again ;)
20090330 03:57:53-!- gozaboru [i=akhilpai@markley-135168.reshall.umich.edu] has quit []
20090330 03:58:00-!- Crab_ [n=crab_@c.218.175.a530.sta.adsl.cyfra.net] has left #wesnoth-dev []
20090330 03:58:09< Sapient> turin: oh, btw... I beat Up From Slavery on Hard today
20090330 03:58:24< Sapient> that's my first campaign to actually play all the way through it
20090330 03:58:32< Shadow_Master> whatever you decide to change, please add a note ALL IN CAPS in the changelog so nobodyu changes it again. I'm sure those vents' semantis have changed many times already
20090330 03:58:44< Shadow_Master> SSH is dropping some characters
20090330 03:59:01< Soliton> a likely story.
20090330 03:59:02< Sapient> Shadow_Master: I'm just trying to figure out who changed it in the first place, and why
20090330 03:59:25< Shadow_Master> I hoped Soly was sleeping. There goes my excuse ;(
20090330 03:59:40< Soliton> Sapient: got a file and line number?
20090330 03:59:55< Sapient> actions.cpp... I'm running blame on it now
20090330 04:00:07< Blueblaze> Sapient: Oh I finally figured out what was wrong with my ifs with the have_units
20090330 04:00:19< Shadow_Master> even I could've done it accidentally. But I'm sure we've already talked about this before. ummm..
20090330 04:00:27 * Soliton pities poor svn blame users.
20090330 04:00:56< Sapient> suokko again
20090330 04:01:08 * Soliton saw that one coming.
20090330 04:02:39< Sapient> could also be alink or cjhopmap, we got several authors on those lines
20090330 04:02:50< Sapient> but I'll see
20090330 04:03:08< Sapient> wesbot: log 22908
20090330 04:03:11< wesbot> suokko * r22908 : * cleaned up delaying end_level_exception
20090330 04:03:14< wesbot> URL: http://svn.gna.org/viewcvs/wesnoth?view=rev&rev=22908
20090330 04:05:02< Sapient> looks like it predates that
20090330 04:06:04< Shadow_Master> I still think that we talked about this before. Someone forgot to fix?
20090330 04:06:19< Shadow_Master> s/think/recall/ I think  too much
20090330 04:07:49< Shadow_Master> though I may as well be recalling one of my old discussions with suokko. My memory is so broken that I'm recalling talking with you about bananas.
20090330 04:07:54< Soliton> afair something there changed back and forth over many commits.
20090330 04:10:13-!- JW1 [n=X@c-71-57-85-168.hsd1.il.comcast.net] has joined #wesnoth-dev
20090330 04:11:24< Sapient> it may be too much trouble to fix at this point
20090330 04:11:41< Sapient> and too much trouble to figure out who broke it, apparently
20090330 04:12:27< CIA-30> esr * r34274 /trunk/data/tools/wmllint: Add transformations requested in bug #13246.
20090330 04:16:22-!- Zen_Clark [n=user@99-136-80-191.lightspeed.rcsntx.sbcglobal.net] has quit ["Trying to leave before they find out who broke it."]
20090330 04:16:43< Shadow_Master> ZC is declaring himself guilty
20090330 04:17:57< Sapient> lol
20090330 04:24:41< Soliton> crimson_penguin: http://www.wesnoth.org/forum/viewtopic.php?p=348442#p348442
20090330 04:27:19 * crimson_penguin still hasn't made the wesnothd, and didn't realize it was callable from within the game
20090330 04:27:34 * crimson_penguin will reply in a bit, unless someone else has already
20090330 04:27:55< Shadow_Master> it's been so since 1.5.2 or so :O
20090330 04:38:20< CIA-30> esr * r34275 /trunk/data/campaigns/Northern_Rebirth/ (8 files in 2 dirs): 
20090330 04:38:20< CIA-30> Character name change in NR: Anita -> Eryssa
20090330 04:38:20< CIA-30> My chief playtester commented that for an princess of Wesnoth elves,
20090330 04:38:20< CIA-30> "Anita" sounds way too Earth-Human. I should have noticed this sooner.
20090330 04:39:33-!- Espreon [n=espreon@wesnoth/developer/espreon] has joined #wesnoth-dev
20090330 04:44:03< Shadow_Master> esr: good catch.
20090330 04:44:24< Shadow_Master> I've been meant to mention it since long ago. Can you rename Sister Theta too? <runs away>
20090330 04:44:33< esr> Shadow_Master: Congatulate my wife Cathy...
20090330 04:45:11< Shadow_Master> you don't know how many nasty jokes can be made on ST In spanish :p
20090330 04:45:19< esr> Well, at least "Theta" isn't a common name in English.  I should probably fix "Father Marcus", though
20090330 04:45:45< esr> What, does "Theta" mean something unhrlpful?
20090330 04:46:03< Shadow_Master> Theta sounds like Teta in Spanish, which means ... tit.
20090330 04:46:38< Sapient> is that how you translated it? :p
20090330 04:46:41< Shadow_Master> of course, it sounds and reads the same whe nused as the greek letter...
20090330 04:46:44< Shadow_Master> Sapient: no!
20090330 04:47:22< Shadow_Master> I left it untranslated, but I still don't like it, specially in a female's name
20090330 04:48:26< Shadow_Master> blahblah, "Hermana Teta" wtf ? I'd not really like to have to pronounce that in RL :p
20090330 04:48:48< esr> Hm.  I guess I could fix that.  "Does "Thera" have any unhrelpful connotations in Spanish?
20090330 04:49:01< Shadow_Master> esr: no.
20090330 04:49:13< esr> Good, I'll fix.
20090330 04:49:49< Shadow_Master> but since you asked for it, I'll refer to her as "Hermana Terabyte" when exchanging emails with translators :P
20090330 04:49:52< Sapient> Thera P Healer
20090330 04:52:17< Shadow_Master> sapy wins.
20090330 04:53:37 * Shadow_Master throws sugar packet at Sapient -------> #
20090330 04:53:49 * Sapient throws sugar packet at Shikadi ------------>#
20090330 04:54:26< Shadow_Master> don't cross the beams!
20090330 04:54:29< Sapient> or you could keep it as Theta and change the portrait, lol
20090330 04:54:37 * Sapient ducks for cover
20090330 04:56:03< Shadow_Master> fool, now we'll have to fight a gigantic wesbot!
20090330 04:57:35 * Shadow_Master runs away instead.
20090330 04:59:36< CIA-30> esr * r34276 /trunk/data/campaigns/Northern_Rebirth/ (16 files in 3 dirs): 
20090330 04:59:36< CIA-30> Two more character renames:
20090330 04:59:36< CIA-30> Marcus is Earth-human: Father Marcus -> Father Morvin
20090330 04:59:36< CIA-30> Spanish doesn't have a "th" sound; ShadowMaster tells me that
20090330 04:59:36< CIA-30> Sister Theta's name sounds a lot like "Sister Tit" in that
20090330 04:59:37< CIA-30> language. So Theta -> Thera.
20090330 05:05:53-!- gozaboru [i=akhilpai@markley-135168.reshall.umich.edu] has joined #wesnoth-dev
20090330 05:07:10< Sapient> cya later, have fun.
20090330 05:07:10-!- Sapient [n=sapien-x@wesnoth/developer/sapient] has left #wesnoth-dev []
20090330 05:10:56-!- Sirp [n=me@wesnoth/developer/dave] has quit ["leaving"]
20090330 05:16:53-!- Sirp [n=me@wesnoth/developer/dave] has joined #wesnoth-dev
20090330 05:21:06-!- alink [n=alink@78.129.18.215] has quit [Remote closed the connection]
20090330 05:24:03< CIA-30> esr * r34277 /trunk/data/ (campaigns/Northern_Rebirth/_main.cfg tools/wmllint): Address feature request at bug #13096: check for space in email names.
20090330 05:37:07-!- JW1 [n=X@c-71-57-85-168.hsd1.il.comcast.net] has quit []
20090330 05:45:07-!- noy [n=Noy@d75-157-52-251.bchsia.telus.net] has joined #wesnoth-dev
20090330 05:58:33-!- noy [n=Noy@wesnoth/developer/noy] has quit ["GO, GET TO THE CHOPPAH!!!"]
20090330 05:58:43-!- noy [n=Noy@d75-157-52-251.bchsia.telus.net] has joined #wesnoth-dev
20090330 06:04:23-!- gozaboru [i=akhilpai@markley-135168.reshall.umich.edu] has left #wesnoth-dev []
20090330 06:04:42-!- gozaboru [i=akhilpai@markley-135168.reshall.umich.edu] has joined #wesnoth-dev
20090330 06:31:14-!- nerwa [n=nerwa@59.78.26.75] has joined #wesnoth-dev
20090330 06:38:26-!- ABCD [n=ABCD@wikipedia/ABCD] has quit ["No Ping reply in 30 seconds."]
20090330 06:38:55-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090330 06:45:39-!- viggy_prabhu [n=vignesh@210.212.5.150] has joined #wesnoth-dev
20090330 06:48:26-!- gozaboru [i=akhilpai@markley-135168.reshall.umich.edu] has left #wesnoth-dev []
20090330 07:12:33-!- ABCD_ [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090330 07:12:38-!- ABCD [n=ABCD@wikipedia/ABCD] has quit ["No Ping reply in 30 seconds."]
20090330 07:16:33-!- res|laptop [n=res@erft-4d07fee2.pool.mediaWays.net] has joined #wesnoth-dev
20090330 07:25:56-!- _deleter [n=_deleter@137.112.136.76] has quit [Read error: 110 (Connection timed out)]
20090330 07:37:11-!- iyonius [n=iyo@a02-0829c.kn.vutbr.cz] has joined #wesnoth-dev
20090330 07:39:36-!- [Relic] [n=[Relic]@adsl-76-229-202-137.dsl.milwwi.sbcglobal.net] has quit ["Leaving"]
20090330 07:44:20-!- ryochan7 [n=ryochan7@173.9.238.161] has quit ["Leaving"]
20090330 07:46:21-!- nerwa [n=nerwa@59.78.26.75] has left #wesnoth-dev []
20090330 07:49:29-!- Sirp [n=me@wesnoth/developer/dave] has quit ["leaving"]
20090330 08:02:41-!- ABCD_ [n=ABCD@wikipedia/ABCD] has quit [Read error: 104 (Connection reset by peer)]
20090330 08:03:56-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090330 08:09:47-!- turin [n=turin@168.215.250.1] has quit [Read error: 110 (Connection timed out)]
20090330 08:10:06< Rhonda> Soliton: Done, and removed the logs from irclogs.w.o too
20090330 08:15:33-!- silene [n=plouf@ASte-Genev-Bois-152-1-97-76.w86-218.abo.wanadoo.fr] has joined #wesnoth-dev
20090330 08:23:07< dfranke> heh... bug #9496 is indeed still around.
20090330 08:36:22-!- Blueblaze [n=nick@c-98-199-143-139.hsd1.tx.comcast.net] has quit [Remote closed the connection]
20090330 08:53:39< CIA-30> dfranke * r34278 / (2 files in 2 dirs): Add an input_blocker during init_turn(), closing various race windows. Please observe a moment of silence for the long-awaited passing of bug #9496.
20090330 09:11:53< CIA-30> esr * r34279 /trunk/data/campaigns/The_Rise_Of_Wesnoth/scenarios/01_A_Summer_of_Storms.cfg: Give the Wesfolk raid a bit more backstory as suggested on the forums.
20090330 09:20:16-!- YogiHH [i=d4ca9d15@wesnoth/developer/yogihh] has joined #wesnoth-dev
20090330 09:20:44< YogiHH> hello everyone
20090330 09:22:37-!- Espreon [n=espreon@wesnoth/developer/espreon] has quit [Remote closed the connection]
20090330 09:26:20< CIA-30> esr * r34280 /trunk/data/campaigns/ (14 files in 14 dirs): 
20090330 09:26:20< CIA-30> Pry the difficulty ratings loose from the campaign summary descriptions.
20090330 09:26:20< CIA-30> This minimizes the translators' work when we re-grade campaigns.
20090330 09:26:48< esr> Hello YogiHH. Got a few minutes?
20090330 09:30:59 * dfranke is running short of bugs
20090330 09:32:26< Rhonda> topic says 52 bugs?
20090330 09:32:41< dfranke> wesbot: topic
20090330 09:32:55< Rhonda> .. 51, wow, what a difference. :)
20090330 09:33:42< dfranke> only 13 unassigned.
20090330 09:34:28< Smar> try to make it bugless once while ;)
20090330 09:38:09< Rhonda> dfranke: Assigning them doesn't make the bug go away magicly. :)
20090330 09:38:27 * Rhonda expects some assigned ones to be rotting away ...
20090330 09:39:23-!- euschn [n=chatzill@tmp9.logic.tuwien.ac.at] has joined #wesnoth-dev
20090330 09:39:31< euschn> good morning
20090330 09:48:40< YogiHH> esr: yep
20090330 09:51:03-!- silene [n=plouf@wesnoth/developer/silene] has quit ["Leaving."]
20090330 09:51:53< dfranke> is there a date planned for a 1.6.1 release?
20090330 09:52:01< esr> YogiHH: I've had a proposal on the table for nearlly a year to revise the savefile-handling interface so the things are threaded together.  Each would get a unique ID, and each would contain the user IDs of its ancestors.  With this, I could implement two-level browsing of your saves - one level of game threasds, each one expanding to a savefile sequence.
20090330 09:52:20< esr> Sound like a good idea sio far?
20090330 09:53:05< YogiHH> yes, although i would heavily recommend to wait with that until the savegame reorganization is done
20090330 09:53:11< Rhonda> dfranke: IIRWIIR
20090330 09:53:37< Rhonda> ... or is there something you want to have rushed in?
20090330 09:54:15< esr> Well, that's what I'm asking about.  I looked at doing this, but the code that calls save_game() is horrible.  Incomprehensible.  Is that likely to change at all soon?
20090330 09:54:22< dfranke> no, I'm just looking at the bug tracker and noting that there have been a sizable number of fixes since 1.6a.
20090330 09:54:29< esr> 26
20090330 09:54:56< YogiHH> esr: yes it is. I am working on that atm :-)
20090330 09:55:42< esr> Good.  That's what I needed to know.  Please keep the threading idea in mind as you refactor.
20090330 09:55:49< YogiHH> esr: What will you do when there are forks? That is you load the same save two times and play on from that?
20090330 09:56:48< esr> In the first rev, forking won't be possible; it would complicate the broweser considerably.  But I have some ideas about how to support it eventually.
20090330 09:57:01< YogiHH> i see
20090330 09:58:08< dfranke> esr: was that '26' in reply to me?
20090330 09:58:31< esr> There are tree stages to the plan (1) get the ids fgor threading in place, (2) support thread browsing without forking, (3) support forking.  There are significant challenges at each level.
20090330 09:58:37< esr> dfranke: Yes.
20090330 09:59:51< YogiHH> esr: (1) sounds pretty straightforward and easy to implement to me
20090330 10:00:57< esr> The challenge there is undestanding justy how and when save_game() grets called and what invariants are guaranteed when it's called.  I bounced off that wall, alas.
20090330 10:02:36< esr> I suspect you're the only person who actually understands that.
20090330 10:03:08< YogiHH> That will soon be history, hopefully. I plan to have a file savegame.cpp that is solely responsible for loading and saving. It will provide an interface like save_game, save_replay, save_autosave and that is it.
20090330 10:03:35-!- Espreon [n=espreon@wesnoth/developer/espreon] has joined #wesnoth-dev
20090330 10:03:39< YogiHH> esr: i suspect the same :-)
20090330 10:03:45< esr> Oh, *good*.
20090330 10:04:21< esr> Once you've got it factored like that, adding ID threading won't be hard.
20090330 10:04:32< YogiHH> yes, i think so, too
20090330 10:06:58-!- euschn [n=chatzill@tmp9.logic.tuwien.ac.at] has quit [Read error: 104 (Connection reset by peer)]
20090330 10:07:21-!- Dragonking [n=dk@wesnoth/developer/dragonking] has quit [Read error: 110 (Connection timed out)]
20090330 10:08:47-!- euschn [n=chatzill@tmp9.logic.tuwien.ac.at] has joined #wesnoth-dev
20090330 10:09:03< esr> YogiHH: OK, thanks, you've tpold me what I nneed to know for medium-term planning.
20090330 10:09:18< YogiHH> you are welcome
20090330 10:10:13< euschn> esr, YogiHH: sorry for interrupting, i missed part of your conversation due to disconnect - were you discussing something that affects avegame reorganization?
20090330 10:10:43< esr> euschn: We were discussing something I want to do when the reorg is finished.
20090330 10:10:58< euschn> i see
20090330 10:14:01< YogiHH> euschn: you can lookup the whole conversation on http://irclogs.wesnoth.org. I do that from time to time. I won't read the whole stuff but search for my name. So if you want to tell me something, you can as well drop it here. Some other users do that as well.
20090330 10:15:09< euschn> YogiHH: thanks
20090330 10:20:39< corn> dfranke: I will have the patch done for linux builds by the end of today. do you have any suggestions for another patch?
20090330 10:21:12< dfranke> corn: cool.  No, though, not offhand.
20090330 10:21:30< corn> windows is going to take longer since I don't have a build environment set up for it
20090330 10:22:20< corn> (yet)
20090330 10:32:50< dfranke> Good thing esr is planning to kill marked-up_text... I just thought of a really stupid trick that's possible in MP.
20090330 10:33:13< dfranke> hex labels allow marked-up text, and there's markup for setting text as green, red, and gray.
20090330 10:33:29< esr> Well, not kill it, but move it all to Pango.  Mordante's building the infrastructure.
20090330 10:33:49< dfranke> so if you were playing against a side with one of those colors...
20090330 10:33:59< esr> *snort*
20090330 10:34:16< dfranke> then you could set labels during their partners' turn, and make it look like they were instructions coming from them.
20090330 10:37:43< esr> Soliton: Is terrain2wki still relevant?  It rtefers to 1.2-style terrain codes.
20090330 10:38:34-!- busfahrer [n=busfahre@unixboard/user/busfahrer] has joined #wesnoth-dev
20090330 10:39:21< Soliton> esr: no idea, though i'm sure there is still a wiki page with terrain codes and their meaning so updating it might be nice.
20090330 10:41:14< esr> Soliton: It's undocumented and looks to me like a 1.2->1.4 transition aid that just never got removed.
20090330 10:47:34< esr> Actually---hm, I suppose it could still be useful after creating or importing new terrains.  The old_terrain stuff should be removed, though.
20090330 11:00:49-!- loonybot [n=loonybot@79.139.246.190] has joined #wesnoth-dev
20090330 11:01:25-!- loonycyborg [n=sergey@79.139.246.190] has joined #wesnoth-dev
20090330 11:05:09-!- zookeeper [n=l@dsl-tkubrasgw1-fe9ff800-215.dhcp.inet.fi] has joined #wesnoth-dev
20090330 11:07:31< CIA-30> esr * r34281 /trunk/data/tools/terrain2wiki.rb: 
20090330 11:07:31< CIA-30> Remove most of the machinery relevant to obsolete 1.2 terrains,
20090330 11:07:31< CIA-30> we'll clean up the rest on the next pass.
20090330 11:09:58-!- Dragonking [n=dk@wesnoth/developer/dragonking] has joined #wesnoth-dev
20090330 11:11:58< Ivanovic> moin
20090330 11:13:18< YogiHH> hi Ivanovic 
20090330 11:13:29< euschn> hi Ivanovic
20090330 11:14:20< CIA-30> esr * r34282 /trunk/data/tools/ (README terrain2wiki.rb wmlflip): Improved documentation.
20090330 11:15:55-!- fabi [n=fabi@wesnoth/developer/fendrin] has quit [Remote closed the connection]
20090330 11:22:29-!- noy [n=Noy@wesnoth/developer/noy] has quit []
20090330 11:23:42< CIA-30> ivanovic * r34283 /trunk/po/ (7 files in 7 dirs): updated Icelandic and Russian translation
20090330 11:23:51< CIA-30> ivanovic * r34284 /branches/1.6/po/ (7 files in 7 dirs): updated Icelandic and Russian translation
20090330 11:26:17-!- KnightR [n=anonymou@bb121-7-175-14.singnet.com.sg] has joined #wesnoth-dev
20090330 11:26:24< KnightR> hey folks
20090330 11:27:12< Ivanovic> hi
20090330 11:30:39< KnightR> are there any plans currently to implement a custom memory manager for Wesnoth?
20090330 11:31:28< KnightR> seeing as how the possibility of porting the game to smaller and memory constricted devices/platforms, i think it might be a good idea to look into the possibility of implementing a memory manager
20090330 11:32:51< KnightR> whoops i meant "there's a possibility" 
20090330 11:33:03< Ivanovic> hehe
20090330 11:33:18< Ivanovic> yeah, there is a basic implementation of a custom memory allocator already
20090330 11:33:35< Ivanovic> though this one is inactive by default since it leads to "some problems" on x86_64 systems
20090330 11:33:43< KnightR> i see
20090330 11:34:19< Ivanovic> but yes, in general a well done memory allocator would be a nice project, too
20090330 11:34:36< KnightR> great :) i was thinking of looking at that possibility as a GSoC project
20090330 11:34:46< loonycyborg> Moreover, that custom allocator doesn't save much memory..
20090330 11:34:47< KnightR> this could affect the WML implementation as well
20090330 11:35:10< KnightR> well it would help with fragmentation and memory budgeting
20090330 11:35:32< KnightR> and if implemented right, i believe it could provide some valuable memory allocation debugging information if needed
20090330 11:35:47< KnightR> rather than having to write tests each time we need to look into how memory is being allocated in the game
20090330 11:35:55< KnightR> or for specific areas of the game
20090330 11:37:18< YogiHH> KnightR: From what i have understood so far, optimization of using information in the code has a way greater potential for saving memory than a custom allocator
20090330 11:37:23< Ivanovic> KnightR: like i said, this is of course a possible project, too
20090330 11:37:36< Ivanovic> though the question is how much this will reduce our memory usage
20090330 11:38:00< YogiHH> KnightR: For example, if you load a campaign, all scenarios get read into memory, although only one of them is really needed at any certain moment.
20090330 11:38:06< Ivanovic> since IIRC our main problem are the data structures we use
20090330 11:38:44< Ivanovic> and yeah, some lazy loading could help, too
20090330 11:39:15< KnightR> ah okay..
20090330 11:39:35< KnightR> someone should really write up a technical doc on how WML works and is parsed :P
20090330 11:40:07< YogiHH> KnightR: It doesn't mean a memory manager won't help, just that there are other things to optimize first if memory saving is the primary goal
20090330 11:40:13< KnightR> i see
20090330 11:40:25< KnightR> i was looking at dave's post regarding std::map's overhead with regard to storing WML attributes
20090330 11:41:47< KnightR> i'm not entirely sure about the inner workings for map memory allocations but would writing a separate allocator for std::map help instead of writing hash table from scratch?
20090330 11:42:30< KnightR> also, the key values for the std::map are currently stored in std::strings correct? and from what i understand, those key values remain relatively static
20090330 11:42:51< KnightR> in that case, would it be better to simply use C-style strings instead of STL ones?
20090330 11:43:29< KnightR> seeing as how that could cut off a significant overhead for each attribute that is stored
20090330 11:46:28< KnightR> one of the issues i thought about when implementing a custom hash table instead of using std::map would be that since wesnoth is open source, it would be easier for developers to work and use an STL container rather than understand how a custom implementation works.
20090330 11:47:08< YogiHH> KnightR: That's probably something you should discuss with Sirp
20090330 11:47:45< KnightR> ah okay.. well he's not on now lol.. just thought i'd get some input from the rest if you guys have anything to add :)
20090330 11:48:00 * YogiHH found that low-level stuff always suspect. He is more familiar with higher levels of organization.
20090330 11:48:20< KnightR> heh alright :) i'll flesh it out more again with Sirp later
20090330 11:48:23< KnightR> thanks Yogi :)
20090330 11:48:29< KnightR> oh one more thing to add...
20090330 11:48:41 * YogiHH thinks that computers should do the low-level stuff ;-)
20090330 11:48:42< KnightR> or ask rather
20090330 11:49:03< KnightR> lol i wish theyd do most of the this nittygritty stuff for us
20090330 11:49:36< KnightR> i believe that STL containers are mostly bug-free and generally fairly efficient
20090330 11:49:59< YogiHH> Well, other languages do that, but there is a price to pay, of course
20090330 11:50:15< KnightR> as with everything else :)
20090330 11:50:19< KnightR> but implementing a custom map-like structure for the use of WML might introduce additional complications when it comes to the usage of our container class
20090330 11:51:32< KnightR> of course when done right, the benefits over using the STL "jack of all trades but master of none" would be that memory allocations would definitely be more efficient
20090330 11:52:12< KnightR> but I believe there to be other solutions to tackling the problem rather than throwing out std::map as a memory-hog
20090330 11:56:45< KnightR> oh and yes
20090330 11:56:52< KnightR> i found an interesting container class that we may be able to use :)_
20090330 11:56:54< KnightR> http://goog-sparsehash.sourceforge.net/
20090330 11:57:15< KnightR> apparently it can be optimised for space or for speed so i think i'll research a little more into that
20090330 11:57:21< KnightR> see if it might be a viable replacement
20090330 12:11:26-!- viggy_prabhu [n=vignesh@210.212.5.150] has quit [Remote closed the connection]
20090330 12:11:35-!- Espreon [n=espreon@wesnoth/developer/espreon] has quit [Remote closed the connection]
20090330 12:16:13-!- faktor4u [n=faktor@chello089077167144.chello.pl] has joined #wesnoth-dev
20090330 12:40:25-!- busfahrer is now known as lkwfahrer
20090330 12:49:23< CIA-30> esr * r34285 /trunk/data/tools/wmlxgettext: A start on a wmlgettexct replacement in Python.
20090330 12:50:26-!- andrey23 [n=Andrey@93.153.159.72] has joined #wesnoth-dev
20090330 12:53:50< CIA-30> esr * r34286 /trunk/ (src/Makefile.am utils/update_unit_tree): Address bug #13293.
20090330 12:57:11< CIA-30> esr * r34287 /trunk/src/multiplayer_connect.cpp: Address bug #13291: Compilation problem in multiplayer_connect.cpp.
20090330 13:01:50< CIA-30> esr * r34288 /trunk/Makefile.am: 
20090330 13:01:50< CIA-30> Address bug #13292: Installation problem caused by unknown find
20090330 13:01:50< CIA-30> arguments.
20090330 13:11:23< euschn> I implemented Easycoding/random village renaming to the following extent: villages now use lake, river, bridge and road names when close to these (e.g., $road|'s Rest), and just use special suffixes near forests, mountains, hills, as these do not have any names themselves (e.g., $name|mont)
20090330 13:11:57< euschn> is this considered a valid solution to the task or should i make mountains, hills, etc. have their own names too so i can use patterns like $mountain|mont instead of $name|mont?
20090330 13:13:14< euschn> additionally, I added a second random name $name2 to river,hill,grassland,regular,etc. villages
20090330 13:21:25-!- lkwfahrer is now known as busfahrer
20090330 13:25:56-!- squallwc [n=chatzill@118.100.22.81] has joined #wesnoth-dev
20090330 13:30:28< Soliton> i think it would be easiest to understand if you showed some screenshots, euschn.
20090330 13:30:43-!- ilor [n=user@wesnoth/developer/ilor] has joined #wesnoth-dev
20090330 13:31:17< euschn> Soliton: ok, now this is a little awkward for me: how do I make village names show in game?
20090330 13:31:38< euschn> i cant seem to find the option - i have done all my testing by commandline and debugging
20090330 13:32:02-!- Crab_ [n=crab_@c.218.175.a530.sta.adsl.cyfra.net] has joined #wesnoth-dev
20090330 13:32:09< Crab_> hi
20090330 13:32:19< euschn> hi
20090330 13:33:03-!- squallwc [n=chatzill@118.100.22.81] has quit ["ChatZilla 0.9.84 [Firefox 3.0.8/2009032609]"]
20090330 13:33:07-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090330 13:33:22< Ivanovic> euschn: in general you should get village names when creating a random map in multiplayer
20090330 13:33:28< YogiHH> hey Crab_
20090330 13:34:20< euschn> Ivanovic: yes I had that initially, but somehow it disappeared, before I even touched the code, and I cannot get it to work at the moment
20090330 13:34:59-!- busfahrer [n=busfahre@unixboard/user/busfahrer] has quit ["Changing server"]
20090330 13:35:13< euschn> it works neither for 1.6 stable nor trunk - maybe its some option in ~/.wesnoth?
20090330 13:35:20< Ivanovic> uhm, all i can say about this is: strange thing...
20090330 13:36:32< euschn> ok, so can anyone confirm that random village names show for them in the current revision of trunk?
20090330 13:39:59< Ivanovic> hmm, i don't get random names either
20090330 13:40:09< Ivanovic> no idea when this feature was deactivated though...
20090330 13:40:38< euschn> Ivanovic: ok, good to know I did not break anything there
20090330 13:41:25< Ivanovic> so a possible task for you: 1) submit a bugreport
20090330 13:41:40< Ivanovic> 2) check when it was removed and if it was intentional
20090330 13:41:53< Ivanovic> 3) provide a patch to readd it in a nice way...
20090330 13:41:55< Ivanovic> ;)
20090330 13:42:12< euschn> ok, I will look into that
20090330 13:42:32< euschn> can you give me a clue on where in the code I should start looking?
20090330 13:42:44< Ivanovic> to be honest: no idea
20090330 13:43:06< Ivanovic> maybe searching the commit mail archive for something like "removed village names" or something like this works
20090330 13:43:31< euschn> ok I will start with that
20090330 13:44:07< Ivanovic> though the easiest might be to download older versions and check when it stopped to work, basically bisecting
20090330 13:44:33-!- busfahrer [n=busfahre@unixboard/user/busfahrer] has joined #wesnoth-dev
20090330 13:44:37< euschn> mhm
20090330 13:45:15< Crab_> euschn, probably, it is easier to find the code that set village labels on generated maps in the version where it actually worked... IMO, fgrep -Rn '"label"' ./ of sorts may help..
20090330 13:46:09< euschn> Crab_: thanks for the tip
20090330 13:48:34< euschn> aha - they still show in the current ubuntu repository version that I ran initially, which is v1.4.5
20090330 13:51:28-!- faktor4u [n=faktor@chello089077167144.chello.pl] has quit ["..."]
20090330 13:51:42< Ivanovic> if your comp is fast at compiling: get the sources for some 1.5.x releases, compile them and check if it is still working there or not
20090330 13:52:13< Ivanovic> once you got the exact release where it "stopped working" you can search for the change breaking it (probably it was intentional, not sure though)
20090330 13:52:23< Ivanovic> afk for a while
20090330 13:57:06-!- Kalatina [n=tobypwri@5ac6446d.bb.sky.com] has joined #wesnoth-dev
20090330 13:58:41-!- Kalatina is now known as BobBobson
20090330 13:59:17-!- busfahrer [n=busfahre@unixboard/user/busfahrer] has quit ["leaving"]
20090330 13:59:32-!- busfahrer [n=busfahre@unixboard/user/busfahrer] has joined #wesnoth-dev
20090330 14:01:45< BobBobson> Hey all - I'm new here
20090330 14:02:10< BobBobson> Started playing Wesnoth a couple of days ago, and very grateful to all those who helped write it
20090330 14:03:09< BobBobson> I'm an out of work coder so, once I've seen more of the game firsthand, I'd be happy to contribute my skills to the project
20090330 14:03:10< zookeeper> what, even suokko? :>
20090330 14:03:32< BobBobson> Even everyone
20090330 14:03:58< zookeeper> there's always room for more contributors, that's for sure
20090330 14:04:21< BobBobson> Aint that always the way
20090330 14:05:13< BobBobson> Nothing gives you spoilers faster than code reading so I wouldn't be digging into the source just yet - but I thought I'd introduce myself early
20090330 14:05:16< zookeeper> any particular are you'd want to work on?
20090330 14:05:35< zookeeper> don't worry, reading the engine's code don't give you any spoilers ;)
20090330 14:05:53< BobBobson> I generally find bug fixing works best for getting to know a large project as a first job
20090330 14:07:01< BobBobson> I wouldn't want to go near anything that affects balance
20090330 14:07:19< zookeeper> don't worry, they wouldn't let you ;)
20090330 14:07:25< BobBobson> Good good
20090330 14:07:57< BobBobson> Who decides what feature requests aren't harmful?
20090330 14:08:02< zookeeper> the engine is really separated from the actual content like campaigns and units anyway
20090330 14:08:27< zookeeper> usually the devs in charge of that area of the game
20090330 14:09:27< YogiHH> BobBobson: There is hardly such a decision, basically. Once you get commit rights, you choose what you want to do. However, talking with others about it, especially if you are relatively new, is always helpful.
20090330 14:09:50< YogiHH> Most of the time, though, there is a broad consensus about what should be done and what not
20090330 14:10:42< BobBobson> Cool - you can't always trust people not to add blink text to menus or whatever so someone has to have some kind of control
20090330 14:10:58< zookeeper> you usually get commit access only after a couple of accepted patches
20090330 14:11:09< BobBobson> Righto
20090330 14:11:36< zookeeper> and if someone actually tried to slip something silly in, someone would probably notice anyway, since we have a lot of people developing and reading commit mails ;)
20090330 14:12:57< zookeeper> so people tend to do what they want unless it's something on which there'd be notable disagreement with
20090330 14:13:27< zookeeper> s/with/about
20090330 14:14:47-!- BenUrban [n=benurban@unaffiliated/benurban] has quit ["Power failu"]
20090330 14:18:25-!- VladimirSlavik [n=chatzill@wesnoth/translator/VladimirSlavik] has joined #wesnoth-dev
20090330 14:40:46< Rhonda> Soliton: Shadow_Master wants you to throw pizza into his face
20090330 14:41:42-!- stikonas [n=stikonas@ctv-79-132-187-171.vinita.lt] has joined #wesnoth-dev
20090330 14:42:09< Shadow_Master> ;(
20090330 14:42:39< YogiHH> Pizza is too valuable to throw it into someones face ;-)
20090330 14:42:45-!- jmunro[a] is now known as jmunro
20090330 14:43:10< Ivanovic> Soliton: be carefull to catch well, good pizza should not be wasted...
20090330 14:43:12< Ivanovic> ^^
20090330 14:43:57< CIA-30> ivanovic * r34289 /trunk/po/wesnoth-tutorial/ru.po: updated Russian translation
20090330 14:43:59< CIA-30> ivanovic * r34290 /branches/1.6/po/wesnoth-tutorial/ru.po: updated Russian translation
20090330 14:45:17 * YogiHH votes for having a separate svn repository for translations. Would shrink the log considerably ;-P
20090330 14:45:57< Ivanovic> hehehe
20090330 14:47:12< Shadow_Master> wesnoth-um-dev kinda used a separate svn repository for UMC tanslatiof at some point. It didn't work well
20090330 14:47:28< Shadow_Master> I'm half-asleep. I meant wesnoth-umc-dev and translations.
20090330 14:48:34< Ivanovic> :)
20090330 14:49:02-!- stikonas [n=stikonas@wesnoth/translator/stikonas] has quit [Remote closed the connection]
20090330 14:53:07 * Soliton doesn't have a pizza.
20090330 14:53:13< Shadow_Master> good!
20090330 14:53:49< Crab_> Attached a new version of my patch (which changes AI lifecycle management and implements AI hot redeployment and basic console history capabilities ) - https://gna.org/patch/index.php?1146  - For me, question of the day is: has I broken anything in team.cpp handling of AI parameters in the case that there is only 1 AI per side ? (my intent was to not change interface or semantics in that case)
20090330 14:54:16< Soliton> formula console or others as well?
20090330 14:54:58< Crab_> only formula console. and its very basic at the moment (view list of commands entered before, re-execute last command or any other command by number)
20090330 14:56:02< Shadow_Master> Mordante: when I upload my add-on the the server, it won't be bundled with the translations from wescamp-i18n automatically, even if it has translate=true in the .pbl?
20090330 14:57:07< Crab_> Soliton: If nothing is broken in team.cpp (I'll test a bit), I'll  want to commit it later today, to be able to make an 'AI arena' for testing various AIs.
20090330 14:58:32-!- BobBobson [n=tobypwri@5ac6446d.bb.sky.com] has quit []
20090330 15:04:51< Soliton> Crab_: that's some big patch. would be good to split that into several commits.
20090330 15:05:15< Crab_> Soliton: it would break the trunk if it is splitted.
20090330 15:05:30< Crab_> Soliton: most of the changes are "lookup code" 
20090330 15:06:11< Crab_> Soliton: there's only a small part that can be splitted off
20090330 15:06:32< Soliton> well, split that part off then. :-)
20090330 15:06:42< Crab_> ok
20090330 15:17:52-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit ["To err is human.. but to really foul things up requires a computer"]
20090330 15:22:18-!- Crab_ [n=crab_@c.218.175.a530.sta.adsl.cyfra.net] has quit ["Leaving."]
20090330 15:22:53-!- Crab_ [n=crab_@c.218.175.a530.sta.adsl.cyfra.net] has joined #wesnoth-dev
20090330 15:29:29< euschn> Ivanovic: I think I solved the village labeling problem: at mapgen_dialog.cpp:395, if labels is not NULL, it is replaced by an empty labels map
20090330 15:29:51< euschn> I guess its supposed to be the other way round, s/if(labels)/if(!labels)
20090330 15:30:06< euschn> if i change it that way, village names display again
20090330 15:30:22< euschn> does not look like intentional removal to me ;)
20090330 15:31:19< Ivanovic> this was commited by ilor in rev 25544
20090330 15:31:27< Ivanovic> so talk to him about it and if this was intentional
20090330 15:31:57< euschn> ok, will do
20090330 15:32:14< Ivanovic> wesbot: log r25544
20090330 15:32:15< wesbot> ilor * r25544 : Applied patch #1027 fixing possible label issues with repeated calls to the random map generator
20090330 15:32:18< wesbot> URL: http://svn.gna.org/viewcvs/wesnoth?view=rev&rev=25544
20090330 15:35:15-!- Ivan_Popov [n=killer@78-83-64-179.spectrumnet.bg] has joined #wesnoth-dev
20090330 15:36:08-!- Ivan_Popov [n=killer@78-83-64-179.spectrumnet.bg] has quit [Client Quit]
20090330 15:37:22< YogiHH> hmm, that might be related to the loop that keeps trying to generate a map if it failed before (up to 10 times)
20090330 15:38:00< euschn> yes I think so too
20090330 15:38:14-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090330 15:38:37< YogiHH> so that seems to reset the labels to start freshly again, if the map generation failed
20090330 15:39:04< euschn> mhm, in case the map generation somehow returned a nullpointer for labels
20090330 15:39:43< euschn> but if that is the intention, the 2nd condition should be if (!labels)
20090330 15:40:16< YogiHH> i think it is the other way round: If labels have already been generated and the it fails afterwards, you need to reset for the next try
20090330 15:43:23< YogiHH> euschn: You can test that. Map generation will fail, if you push the number of hills and hill size to the max.
20090330 15:43:56< euschn> ok will do
20090330 15:44:52-!- happygrue [n=George@wesnoth/developer/wintermute] has joined #wesnoth-dev
20090330 15:52:05-!- Reisiger_ [n=Reisiger@adsl-62-167-57-12.adslplus.ch] has joined #wesnoth-dev
20090330 15:52:17-!- Reisiger_ [n=Reisiger@adsl-62-167-57-12.adslplus.ch] has left #wesnoth-dev []
20090330 15:52:18< euschn> YogiHH: ok it works like this: If map generation succeeds, no labels are returned from the generation; if it fails with max hill(size), some labels are returned and added to the config in mapgen_dialog.cpp:429
20090330 15:52:40-!- Reisiger_ [n=Reisiger@adsl-62-167-57-12.adslplus.ch] has joined #wesnoth-dev
20090330 15:52:42< euschn> surely it should not add labels if the map generation fails?
20090330 15:54:29-!- Reisiger_ [n=Reisiger@adsl-62-167-57-12.adslplus.ch] has left #wesnoth-dev []
20090330 15:55:00< YogiHH> euschn: I am only recalling out of my head, i don't have the code in front of me. I thought that generate_map function cares for the labels as well, but i am not sure.
20090330 15:56:04-!- _deleter [n=_deleter@adamsrj-1.dhcp.rose-hulman.edu] has joined #wesnoth-dev
20090330 15:56:18< zookeeper> happygrue, btw, one thing i'd really like to fix in TSG now would be "vengeance" since it's such a dorky scenario :P
20090330 15:56:23< zookeeper> any ideas?
20090330 15:56:30< happygrue> heh
20090330 15:56:47< happygrue> there was just some feedback in umc-dev from someone who really liked it
20090330 15:56:51< happygrue> but I tend to agree with you
20090330 15:56:59< zookeeper> yeah, it seems that 1/10 will like it and the rest hate it
20090330 15:57:04< happygrue> it might be interesting for the very first time you see it, but there is no replay value
20090330 15:57:08< YogiHH> The method i told you to have the map generation fail does so, because it can't place the castles onto the map, which happens pretty much at the end of generate_map. So if some other stuff has already been done, we might have to reset that first.
20090330 15:57:36< YogiHH> but as i said, this is all out of my head
20090330 15:57:41< zookeeper> happygrue, now, i was considering that one option would be to have the elves be some kind of elvish cult of mebrin. that is, necromancers and all that, instead of just regular elfsies who think you're the bad guy.
20090330 15:57:46< happygrue> well, one nice thing about TSG is how the elves end up pissed off at you no matter what you do
20090330 15:57:52< ilor> euschn: try making it use labels_copy in the mapgen call, I think that was intended
20090330 15:57:53-!- Reisiger [n=Reisiger@adsl-62-167-57-12.adslplus.ch] has joined #wesnoth-dev
20090330 15:58:40< happygrue> as in, the overall storyline is similar no matter which branch you take, which is good continuity I think.  Though your idea is interesting
20090330 15:58:42< euschn> ilor: you mean in the call of default_generate_map?
20090330 15:58:46< ilor> euschn: yes
20090330 15:59:14< zookeeper> happygrue, i think it needs to change a bit since it really makes no sense at all as it is now, especially when ethiliel is with you.
20090330 15:59:38< happygrue> zookeeper: I think it could be made more interesting, and more sensical also
20090330 16:00:41< happygrue> one idea is to not have your elves leave, but have a civil war between the cult elves and ethiliel's elves?
20090330 16:01:21< happygrue> though that has some problems
20090330 16:02:03< zookeeper> well, anything goes as long as it provides a good alternative for that scenario. of course we can make small adjustments to the earlier scenarios too if it's needed.
20090330 16:02:53< happygrue> zookeeper: do you have any ideas in mind?
20090330 16:03:08< zookeeper> nothing beyond what i said already, i'm afraid
20090330 16:03:10-!- Reisiger [n=Reisiger@adsl-62-167-57-12.adslplus.ch] has quit ["Trillian (http://www.ceruleanstudios.com"]
20090330 16:03:13< happygrue> ok
20090330 16:03:24-!- Reisiger [n=Reisiger@adsl-62-167-57-12.adslplus.ch] has joined #wesnoth-dev
20090330 16:05:41 * happygrue is reminded that he needs to go back and finish the last scenario of the outlaw branch still
20090330 16:05:44< benetnash> Hi! Is it normal, that in 2p Basilisk Cave, enemy_units contains that black units turned to stone? It makes my simple custom AI crazy ;)
20090330 16:06:51< Reisiger> Evening everyone. I recently got my hands on MSVC (VC9) through my university and tried setting everything up to compile Wesnoth from the SVN. Yet I'm getting reports of missing boost and SDL? libraries. I do have boost installed on my PC (including precompiled libraries for MSVC 2008 but couldn't figure out how to change the settings to get it to work. Does anyone know what settings need to be changed?
20090330 16:06:58< YogiHH> benetnash: Well, that's not "normal", except for that map. In any way, the AI should be able to ignore these units
20090330 16:07:03< euschn> ilor: if i use labels_copy the labels show again at generation success, but at generation fail labels are still generated and added to the config
20090330 16:07:41< ilor> euschn: is this a problem?
20090330 16:07:51< Reisiger> Agh... back in 5... RL calls.
20090330 16:08:05< euschn> this could be fixed if the map_gen exception would be thrown before labels_copy is swapped back to labels
20090330 16:08:19< zookeeper> happygrue, of course if there was an actual big dangerous elvish cult then it'd make little sense why they wouldn't appear in the bandit branch too
20090330 16:08:24< euschn> ilor: actually, I dont know, probably not
20090330 16:09:18< benetnash> YogiHH: so should I report it as a bug? IMO there is no way to distinguish background units from real enemies from the level of FormulaAI scripts
20090330 16:09:34< happygrue> zookeeper: yes, I was having ideas of even some kind of "undead elf" follower, but I think that is too crazy perhaps
20090330 16:09:43< Ivanovic> Dragonking: i gave benchmarking wesnoth compile times another go, this time with a varying amount of threads: http://wesnoth.pastebin.com/m302deaf2
20090330 16:09:57< Ivanovic> looks like there is at least some impact when going to >2 threads
20090330 16:10:40< zookeeper> happygrue, why so? i already frankensteined a dark adept elf quite a long time ago :p
20090330 16:10:55< happygrue> I kind of like the defensive scenario, almost being overrun idea, but what about more of a time limit?  maybe ethiliel has to go obtain proof of the Merbin's corruption and return to show the other elves (who are normal and rational)
20090330 16:10:56< YogiHH> Reisiger: It's in the project properties, Configuran properties, Linker, second option ("Eingabe = input" in german) lists your libraries
20090330 16:11:02< Crab_> benetnash, you can try to filter them out
20090330 16:11:39< happygrue> zookeeper: well, it is a nice idea I think, but I think as you say, if there are lots of undead elves running around then the story needs to be updated in both branches
20090330 16:11:44< happygrue> though maybe that is fine
20090330 16:12:20< happygrue> zookeeper: so you already have art for an undead elf?
20090330 16:12:32< zookeeper> dark adepts aren't undead
20090330 16:12:39< happygrue> er yes, good point
20090330 16:12:42< happygrue> I mean corrupted
20090330 16:13:02< zookeeper> art is a non-issue, it's trivial to make elf versions of the dark adept line and that's about the only one we'd have, i guess
20090330 16:13:19< happygrue> well, a corrupt fighter would be nice too I think
20090330 16:13:22< happygrue> but not needed per say
20090330 16:13:52< zookeeper> fighters could just be regular fighters IMO, i don't see a reason for a new unit or especially new artwork
20090330 16:13:55< YogiHH> benetnash: WML qualifies those units with stoned="on". I can't tell you however how to deal with that in formulas.
20090330 16:14:12< zookeeper> they can always just be given slightly modified stats
20090330 16:15:04< happygrue> just making the AI player black and having the black TC with some stats would work nicely I guess.  But the story I guess is the thing
20090330 16:15:19< happygrue> so how do they come into the outlaw campaign?
20090330 16:15:20< Crab_> benetnash: 1) file a bug report about enemy_units  2) you can use a 'filter(enemy_units,self.side=2)'  [where 2 is the enemy side number, 1 based)  3) you can use that 'stoned' state they have to filter them
20090330 16:15:28< zookeeper> the branches are a bit strange anyway in that the other you kill the bad guy early and need to have some other bad guys for the end, whereas in the other branch you kill the bad guy in the end but lost the other bad guys somewhere ;)
20090330 16:15:33< happygrue> because they need to be defeated everywhere I think
20090330 16:15:43< benetnash> Crab_: I'll make bug report and use another map for tests ;)
20090330 16:15:45< happygrue> heh
20090330 16:17:07< happygrue> well, I guess corrupt elves just show up with merbin in the outlaw branch, and then his remaining followers show up in the elf branch
20090330 16:18:35< CIA-30> ivanovic * r34291 /trunk/po/wesnoth-trow/ru.po: updated Russian translation
20090330 16:18:36< Crab_> benetnash: (3) is:  filter(enemy_units,index_of('stoned',keys(self.states))<0)
20090330 16:18:39< CIA-30> ivanovic * r34292 /branches/1.6/po/wesnoth-trow/ru.po: updated Russian translation
20090330 16:20:50< Reisiger> Back; YogiHH, found it but there's no mention of the boost libraries, and the SDL ones are listed
20090330 16:22:05< YogiHH> hmm, boost is not in there? weird...
20090330 16:22:33< happygrue> zookeeper: or esr or anyone, do you know when the rogue mages and such from Liberty come into the timeline?  They *could* be decedents of outlaws teamed with these rogue elves that somehow survive after this is all over
20090330 16:22:43< Reisiger> Shouldn't the SVN trunk have a libs folder?
20090330 16:22:44< YogiHH> Reisiger: well, i have a slightly different project file, so ilor might be of greater help, then
20090330 16:22:52< benetnash> Crab_, YogiHH thanks for help
20090330 16:23:03< YogiHH> Reisiger: no, we don't hold libs in svn
20090330 16:23:22< YogiHH> benetnash: np
20090330 16:23:28< happygrue> I don't really know that much about liberty though, so perhaps it doesn't work
20090330 16:23:32< Reisiger> so all those libs listed are expected in the trunk folder?
20090330 16:24:48< YogiHH> Reisiger: you can specify a folder somewhere in the project settings
20090330 16:25:19< YogiHH> Reisiger: to tell the linker where to search for libraries
20090330 16:25:57< Crab_> Reisiger, it depends on the build system you are using. for example, scons allows you to set the directories where boost, lua, ..., are installed (that is for specifying them in the case it hasn't found them automatically)
20090330 16:27:00< YogiHH> Crab_: i think he is not using any build system, just the VC++ 2009 IDE
20090330 16:27:12< YogiHH> oops, 2008 :-)
20090330 16:27:16< Ivanovic> i think in the IDE you can somewhere specify include paths, too
20090330 16:27:17< Reisiger> Ok, so I'll get all the required libraries and correct the project file to point to the right places?
20090330 16:27:22< Crab_> YogiHH: then "VC++ 2008 IDE" IS his buildsystem )
20090330 16:27:26< Ivanovic> or you could use cmake to generate the project files
20090330 16:27:41< Ivanovic> in the cmake gui you can enter which folder should be used for which libs
20090330 16:27:48< KnightR> i wish there was a linux alternative to the VC++ debugger...
20090330 16:27:48< YogiHH> Reisiger: yup
20090330 16:28:22< Reisiger> M'kay... will do so, but now I got to disapear for 2 hours ;)
20090330 16:28:23< Reisiger> Thanks and until later
20090330 16:28:46< YogiHH> KnightR: Code::Blocks is probably next to it from what i saw, although not as comfortable
20090330 16:28:54-!- Reisiger [n=Reisiger@adsl-62-167-57-12.adslplus.ch] has quit ["Trillian (http://www.ceruleanstudios.com"]
20090330 16:29:19< KnightR> i've tried it but i don't think it's close to what the VS debugger can provide though...
20090330 16:30:55< YogiHH> KnightR: I agree, i used it for a while as well
20090330 16:31:19-!- _deleter [n=_deleter@adamsrj-1.dhcp.rose-hulman.edu] has quit [Nick collision from services.]
20090330 16:31:32-!- _deleter [n=_deleter@adamsrj-1.dhcp.rose-hulman.edu] has joined #wesnoth-dev
20090330 16:31:36< CIA-30> ivanovic * r34293 /trunk/ (8 files in 7 dirs): updated Czech translation
20090330 16:31:43< CIA-30> ivanovic * r34294 /branches/1.6/ (8 files in 7 dirs): updated Czech translation
20090330 16:31:54< VladimirSlavik> danke schön
20090330 16:32:01< Ivanovic> you're welcome
20090330 16:32:05< KnightR> and eclipse... well it just feels kinda clunky heh
20090330 16:32:14 * Ivanovic says ddd
20090330 16:32:16< Ivanovic> ^^^
20090330 16:32:25< Ivanovic> s/^^^/^^
20090330 16:33:20< KnightR> has anyone thought about uploading a set of development dependencies for wesnoth?
20090330 16:33:30< KnightR> like what Ogre does
20090330 16:33:34< KnightR> building boost is such a pain :(
20090330 16:33:59< Ivanovic> KnightR: uhm, most of us are on linux, there we got great tools called "package manager"
20090330 16:34:11< Ivanovic> they handle the "get me all the deps to build XYZ"-part
20090330 16:34:48< YogiHH> KnightR: I think for windows, that's a real good idea *waves to loonycyborg*
20090330 16:35:05< KnightR> i'm stuck on windows :( don't really have too much of a choice cos my school uses VS and windows
20090330 16:35:37< KnightR> would switch to linux in a heartbeat if i could do everything there though lol
20090330 16:35:37< euschn> this is what my current implementation of random village names looks like: http://imagebin.org/43728
20090330 16:35:42< KnightR> but no such world exists :(
20090330 16:36:13< Ivanovic> KnightR: you could use linux in general and run windows in a virtual machine
20090330 16:36:24< Ivanovic> unless you need 3d stuff in windows this works nicely
20090330 16:36:24< YogiHH> KnightR: I am more or less stuck on windows, because my work requires it and i don't have the time to learn another OS to the extend i can use it fluently
20090330 16:36:32< euschn> the prefixes like Great use a new variable called $name2, which can of course be used as a second sylabus
20090330 16:37:20< KnightR> I would but I'm doing game development on windows (with stupid DX)
20090330 16:37:31< Ivanovic> ugh, two variables makes things *really* tricky when it comes to translations
20090330 16:37:37< Ivanovic> s/two/three
20090330 16:38:02< euschn> well there are only 2 variables right now
20090330 16:38:03< Ivanovic> if there is always a "this is a prefix" and "this is a postfix" it works nicely for translations, too
20090330 16:38:20< Ivanovic> but with three vars it would get really ugly
20090330 16:38:47< KnightR> YogiHH: let's hope that Win7 improves our experience lol
20090330 16:39:18< KnightR> although I heard they're trying to greatly improve the VS experience with the next version
20090330 16:39:41< KnightR> like an intellisense update
20090330 16:39:49< euschn> Ivanovic: agreed, I just did that to see if it works properly: the 3rd variable doesnt actually exist, as the suffixes like ham, ton, etc. are hardcoded choices
20090330 16:39:50 * YogiHH have heard that one time too often to believe in everything
20090330 16:40:02< KnightR> haha
20090330 16:40:19< euschn> so I suppose its best to use the 2nd variable $name2 for those postfixes?
20090330 16:40:19< Ivanovic> euschn: okay, just wanted to make sure that you are aware of this possible problem
20090330 16:40:32< euschn> yes
20090330 16:40:34< zookeeper> happygrue, liberty takes place before TSG, i'm quite sure
20090330 16:40:47< KnightR> might I ask if there's a person in charge of the VS project files for Wesnoth?
20090330 16:41:11< Ivanovic> euschn: to be honest: no idea how things are built best
20090330 16:41:35 * YogiHH points to ilor
20090330 16:41:41< happygrue> zookeeper: I see.  Too bad, that seemed like a nice way of tying campaigns together
20090330 16:41:47< KnightR> ilor: hey ilor :D
20090330 16:42:00< KnightR> oh wait he's not here
20090330 16:42:07< KnightR> oh wait he is
20090330 16:42:09< KnightR> oops
20090330 16:42:18< Ivanovic> KnightR: to be honest, i think as coder under windows, when knowing linux, you probably feel crippled in several areas...
20090330 16:42:24< euschn> Ivanovic: well, that way I cannot see it being a big problem for translations
20090330 16:42:42< KnightR> I have coded a fair bit in linux and find that linux is a helluva lot more developer friendly compared to Windows...
20090330 16:42:45-!- elias [n=allefant@allegro/developer/allefant] has joined #wesnoth-dev
20090330 16:42:48< Ivanovic> euschn: i would need some "real world examples" to understand the stuff without diving into your code
20090330 16:42:50< Ivanovic> ;)
20090330 16:43:07 * YogiHH is glad he doesn't know linux, then
20090330 16:43:38< KnightR> does anyone here work in a professional game dev. company by the way?
20090330 16:43:40< ilor> KnightR: so, what's the problem?
20090330 16:43:43< Ivanovic> KnightR: this is a difference between "hey, the end users have to be able to handle it" and a coder saying "man, i want to do *work* with the box i work on..."
20090330 16:43:59< KnightR> Ivanovic: i know -exactly- what you mean...
20090330 16:44:27< Ivanovic> in the end it comes down to stuff like the terminal
20090330 16:44:42< KnightR> I'm kinda hoping that KDE for windows really takes off though... I could possibly replace crappy explorer :D
20090330 16:44:48< Ivanovic> most end users will not be able to really handle it, though if you know what you are doing it is a damn efficent tool
20090330 16:45:43< KnightR> I'm not too great with terminal but I do know that you miss it when you get back on Windows heh
20090330 16:46:04< KnightR> well there's always cygwin i guess... heh ;)
20090330 16:46:07< euschn> Ivanovic: currently, the name patterns are something like $name|ton, $name|ham etc. with $name being one of {Kin, Olf, Oul...} - I think the easycoding task suggests it would be better to have $name|$name2 with $name in {ton,ham,...}
20090330 16:46:37< Ivanovic> ah, okay
20090330 16:46:39< euschn> what my implementation currently does is e.g. $name2$name|ton with $name2 being Upper
20090330 16:46:51< euschn> but that was just to better see if it works
20090330 16:47:31< euschn> but yes, the task description is not completely precise, so I wanted to know what you devs prefer
20090330 16:47:50-!- elminse [n=elminsei@netblock-208-127-84-182.dslextreme.com] has quit [Read error: 110 (Connection timed out)]
20090330 16:49:38< Ivanovic> hmm, i think the $name2 stuff is to be left out since it makes the generator to difficult to handle from a creators/translators point of view
20090330 16:50:18< Ivanovic> and to be honest: "It would be nice to have a separate list of "first sylabus" and "last sylabus" for naming. That's not really needed in english, but some translations could use it " does not make too much sense either...
20090330 16:50:39< Ivanovic> that is: due to the var in the old string stuff is already covered for different orders
20090330 16:51:10< Ivanovic> since instead of "$name|ton" you can you "ton$name" to reverse order when needed in the translation
20090330 16:51:42< Ivanovic> which will get damn difficult with a generic "$name1|$name2" generic entry
20090330 16:51:47-!- _deleter [n=_deleter@adamsrj-1.dhcp.rose-hulman.edu] has quit [Read error: 110 (Connection timed out)]
20090330 16:52:01< Ivanovic> there you could only switch the two in general, but not have some at the front and some at the back
20090330 16:52:19< Ivanovic> (even valid for some german names IIRC)
20090330 16:52:38< euschn> I am not sure I understand
20090330 16:52:50< Ivanovic> sorry for talking to you about this so late in your process of providing the patch
20090330 16:52:54< euschn> it all depends on what you set as name2 doesnt it?
20090330 16:53:11< Ivanovic> euschn: in the translations you got basically this at the moment:
20090330 16:53:31< euschn> if you set $name2 as {ton,...} you can do exactly the same
20090330 16:54:06< Ivanovic> English: "$name|bury,$name|ham,$name|ton"
20090330 16:54:16< Ivanovic> German: "Sankt-$name,$name|heim,$name|hausen"
20090330 16:54:39< euschn> ok now I get it
20090330 16:54:47< Ivanovic> after the change there would be:
20090330 16:55:00< Ivanovic> list of suffixes, list of prefixes, matching term
20090330 16:55:27-!- anakayub [n=anakayub@203.82.91.34] has joined #wesnoth-dev
20090330 16:55:31< Ivanovic> with the matching term just being "$name1|$name2" and it would catch *all* terms, which would make the system more inflexible
20090330 16:56:31< YogiHH> i'm off, might be getting here again in the late evening
20090330 16:56:33-!- ikarius__ [n=ross@216.27.182.3] has quit []
20090330 16:56:34< Ivanovic> and like i already said: sorry for not spotting and bringing this up earlier
20090330 16:56:37< Ivanovic> cu YogiHH
20090330 16:56:37< YogiHH> see you
20090330 16:56:45< euschn> ok I am not sure I fully understand it yet
20090330 16:57:12< euschn> so if all suffixes go into $name and all prefixes into $name2, it should be fine?
20090330 16:57:24-!- YogiHH [i=d4ca9d15@wesnoth/developer/yogihh] has quit ["http://www.mibbit.com ajax IRC Client"]
20090330 16:57:27< Ivanovic> no, that is the problem
20090330 16:57:48< Ivanovic> in general this would *force* things to always be translated with a prefix and suffix
20090330 16:58:08< Ivanovic> which would in german eg not work with Sankt-SOMETHING
20090330 16:58:17< Ivanovic> since this can not be a suffix
20090330 16:58:20< euschn> ok
20090330 16:58:50< euschn> so how does the change you suggested fix that?
20090330 16:59:22< Ivanovic> basically what i suggest is: remove point two and three from this list
20090330 16:59:35< euschn> and just implement 1)?
20090330 16:59:39< Ivanovic> jupp
20090330 16:59:57< euschn> ok, I did that, but ahve a question about it
20090330 17:00:05< Ivanovic> just ask
20090330 17:00:52< euschn> how I coded it, villages can take the names of rivers, lakes, bridges and roads (as on the screenshot: Path of Oul-> Oul's Rest)
20090330 17:01:14< euschn> the task says tis should work for mountains, swamps, etc too, but the problem is those do not have any names themselves
20090330 17:01:14< Ivanovic> okay, this sounds great
20090330 17:01:41< Ivanovic> hmm, i think once upon the time there was some naming possible for mountains, not sure though
20090330 17:01:55< euschn> only the 4 above have names, for mountains you can do only extra suffixes like "$name|mont"
20090330 17:02:00< euschn> i see
20090330 17:02:10< Ivanovic> uhm, yeah, this *is* a name
20090330 17:02:58< euschn> yes it is, but the pattern is $name|mont, not $mountain|mont - as opposed to $road|'s Rest which is possible
20090330 17:03:04< Ivanovic> ingame it will then be something like "Oulmont"
20090330 17:03:32< euschn> yes, just that Oul is a random name, and not the name of the mountain, as in the road example
20090330 17:03:33< Ivanovic> ah, i see what you mean
20090330 17:03:59< euschn> in short: is the task considered solved like that, or do I need to add mountian naming too?
20090330 17:04:18< Ivanovic> sounds like a refactoring is in place to make the mountains explicit $mountain|suffix instead of $name|suffix
20090330 17:04:21< euschn> which I am not sure how to do :)
20090330 17:04:40< Ivanovic> this should in general be identical to eg the naming of roads
20090330 17:04:54< euschn> yes,
20090330 17:04:54< Ivanovic> that is something will be defined in a wml file in data/
20090330 17:05:17< euschn> yes, thats what I did for roads and bridges
20090330 17:05:36< euschn> earlier, it was just $lake,$river,$name, and I added $road,$bridge
20090330 17:06:01< loonycyborg> [18:33:16] <KnightR> has anyone thought about uploading a set of development dependencies for wesnoth?
20090330 17:06:01< euschn> that was easy because these get named in the generation process anyways
20090330 17:06:15< loonycyborg> I did, but boost poses some problems here..
20090330 17:06:39< euschn> adding $mountain is more complicated, as mountains do not get names when the height map is generated
20090330 17:08:22< Ivanovic> so part of the task is to *give* mountains names
20090330 17:08:49< loonycyborg> Unfortunately, because C++ ABI isn't standatrized like C's I'd had to provide boost binaries for several different compilers and compiler versions..
20090330 17:09:10< euschn> ok this was unclear to me
20090330 17:09:28< Ivanovic> euschn: and this can most likely be done in a manner comparable to eg rivers of lakes
20090330 17:09:37< Ivanovic> s/of/or
20090330 17:12:00< euschn> Ivanovic: yes, I think that is doable
20090330 17:12:23< Ivanovic> great
20090330 17:12:31< euschn> so basically: forget about $name2, but add support for $mountain, $hill etc, correct?
20090330 17:12:54< Ivanovic> yeah
20090330 17:13:18< euschn> I just want to know for sure this time what to do exactly before I spend another day on it
20090330 17:13:29< euschn> but I guess I should have asked earlier
20090330 17:14:57< euschn> I will do finish my proposal (YogiHHs questions) first though, and hope that I will manage to complete it before friday evening
20090330 17:15:11< euschn> the naming, that is
20090330 17:15:13< Ivanovic> and i should probably have had a closer look at all the easy coding tasks before, too (just to make sure that they are not more problematic than anything else when it comes to translations...)
20090330 17:15:33< euschn> mhm
20090330 17:17:14< euschn> I just want to show I am commited to this and I really want to do all to improve my chances of getting accepted, so I hope working on a problem in the wrong way like this did not make me look to bad now
20090330 17:17:26< euschn> s/to/too
20090330 17:17:34< Ivanovic> no, it is not negative
20090330 17:17:41< Ivanovic> you showed that you look into stuff, this is positive
20090330 17:17:44< euschn> thanks
20090330 17:17:48< Ivanovic> and you also found a bug
20090330 17:17:49< Ivanovic> ;)
20090330 17:17:56< euschn> hehe, with help
20090330 17:18:31< Ivanovic> this is not uncommon when exactly finding out what the bug is
20090330 17:18:37< euschn> mhm
20090330 17:18:39< Ivanovic> hey, we are somehow a team working on things
20090330 17:18:50< euschn> true
20090330 17:19:38< euschn> anyways, I gotta go, be back later, thanks for the lengthy discussion again
20090330 17:19:45-!- orian_ [n=opera@afje2.neoplus.adsl.tpnet.pl] has joined #wesnoth-dev
20090330 17:19:52-!- euschn [n=chatzill@tmp9.logic.tuwien.ac.at] has quit ["ChatZilla 0.9.84 [Firefox 3.0.7/2009030423]"]
20090330 17:19:53-!- orian_ [n=opera@afje2.neoplus.adsl.tpnet.pl] has left #wesnoth-dev []
20090330 17:24:20-!- turin [n=turin@168.215.250.1] has joined #wesnoth-dev
20090330 17:26:12-!- sthenuc [n=omegabul@fw1.ursinus.edu] has joined #wesnoth-dev
20090330 17:35:15-!- ikarius [n=ross@smtp.gridironsystems.com] has joined #wesnoth-dev
20090330 17:37:20-!- cjhopman [n=chris@wesnoth/developer/cjhopman] has quit ["Konversation terminated!"]
20090330 17:40:09-!- thespaceinvader [n=chatzill@wesnoth/artist/thespaceinvader] has joined #wesnoth-dev
20090330 17:40:41-!- Polarina [n=polarina@wesnoth/translator/Polarina] has joined #wesnoth-dev
20090330 17:43:39-!- _deleter [n=_deleter@137.112.104.103] has joined #wesnoth-dev
20090330 17:47:28< Ivanovic> Sapient: you should now be removed from the mentor list, asked leslie to do so and she just answered: [17:46:46] <lh> Ivanovic: done
20090330 17:47:39< Ivanovic> please ping me if it is not working as expected
20090330 17:49:59-!- sthenuc [n=omegabul@fw1.ursinus.edu] has left #wesnoth-dev []
20090330 17:55:01< KnightR> hey guys, I would like to clarify something about the way WML works
20090330 17:55:29< KnightR> do the contents of a WML document stored in memory grow throughout the duration of a game?
20090330 18:02:35-!- molgrum [n=molgrum@c83-249-53-163.bredband.comhem.se] has joined #wesnoth-dev
20090330 18:02:46-!- shadowmaster [i=929b360f@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20090330 18:10:04< Dragonking> Hello everyone
20090330 18:10:17< Crab_> hi Dragonking
20090330 18:11:28< Dragonking> Ivanovic: Interesting results. :)
20090330 18:11:38< Ivanovic> Dragonking: jupp
20090330 18:11:42-!- Velory [n=GProcess@88.254.129.247] has joined #wesnoth-dev
20090330 18:11:51< Velory> Hi everyone
20090330 18:11:57< Crab_> hi Velory
20090330 18:12:15< Dragonking> hi Velory 
20090330 18:12:17-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit [Read error: 104 (Connection reset by peer)]
20090330 18:12:26-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090330 18:16:16-!- busfahrer [n=busfahre@unixboard/user/busfahrer] has quit ["leaving"]
20090330 18:17:51< Crab_> Dragonking, can you tell me about the situations in which the default ai fallbacks to formula_ai ? http://wesnoth.pastebin.com/m722448b1
20090330 18:18:17< Crab_> Dragonking: basically, I want to know about 'what is desired to happen'
20090330 18:18:53< Dragonking> I don't think that default AI ever fallback to formula AI.
20090330 18:19:22< Crab_> but, in current trunk, it at least tries to )
20090330 18:19:53< Dragonking> It is no fallback.
20090330 18:20:26< Dragonking> It is only recruitment if recruitment=formula is specified.
20090330 18:22:29< Crab_> and the first part ? (in new_turn) 
20090330 18:23:47< Dragonking> It calls formula AI new turn, which invalidates move_maps and calls ai::new_turn IIRC
20090330 18:24:09< Dragonking> It is not like any side formula or something is executed AFAIK
20090330 18:26:28< Dragonking> BTW I have never used this recruitment=, just some time ago I checked if it is called, it was, so I left it in as sometihng that may be used by someone.
20090330 18:28:25-!- Turuk_ [n=Turuk@FUSE-WS1-208-102-202-188.fuse.net] has joined #wesnoth-dev
20090330 18:29:46-!- Velory [n=GProcess@88.254.129.247] has quit [Read error: 110 (Connection timed out)]
20090330 18:30:40< KnightR> I just posted a thread on the development forum regarding my ideas on WML memory optimisation and game memory management. please provide input if you can :) thanks!
20090330 18:30:44-!- noy [n=Noy@d75-157-52-251.bchsia.telus.net] has joined #wesnoth-dev
20090330 18:33:47< Crab_> Dragonking: I asked because I need to modify ai lookup code, and there is a potential infinite loop situation (ai::play_turn() calls formula_ai::play_turn, and formula_ai can fallback to ai::play_turn() again, thus creating an infinite loop.             This is avoided by using the master_ variable (only *master* ai is supposed to be allowed to fallback). This master_ variable is stored in ai_interface::info.master and is set in ai_int
20090330 18:34:04-!- oak_ [i=oakus@c02--112b.kn.vutbr.cz] has quit ["( www.nnscript.de :: NoNameScript 4.02 :: www.XLhost.de )"]
20090330 18:34:17< zookeeper> KnightR, no, the WML document is created at startup (starting the game, starting a campaign, loading a campaign save, starting MP, ...)
20090330 18:34:22-!- brahim [n=lippe-er@ensihp303.imag.fr] has joined #wesnoth-dev
20090330 18:34:29< brahim> hello
20090330 18:34:36< KnightR> ah i see...
20090330 18:34:38< zookeeper> i'm not sure how exactly the presumed lazy-loading of unit types works though, maybe it's an exception, i don't know.
20090330 18:34:39< KnightR> hi brahim
20090330 18:34:41< shadowmaster> someone said masteR?
20090330 18:34:41-!- flav [n=flav@mne69-3-82-225-23-140.fbx.proxad.net] has joined #wesnoth-dev
20090330 18:34:44< shadowmaster> *master
20090330 18:34:50< KnightR> zookeeper: so in that case, it doesn't really expand or shrink right?
20090330 18:35:00< Crab_> shadowmaster: it's a variable named master_ )
20090330 18:35:25< KnightR> it would mean that memory fragmentation might be less of an issue as the WML document doesn't do heavy allocs/deallocs
20090330 18:35:51< zookeeper> KnightR, that's right. or at least i'm not aware of any such thing...unless perhaps for that unit type lazy-loading of which i really know nothing about.
20090330 18:35:58-!- _deleter [n=_deleter@137.112.104.103] has quit [Nick collision from services.]
20090330 18:36:11-!- _deleter [n=_deleter@adamsrj-1.dhcp.rose-hulman.edu] has joined #wesnoth-dev
20090330 18:36:20-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit [Read error: 110 (Connection timed out)]
20090330 18:36:20< KnightR> i see... thanks for the info :)
20090330 18:36:26< Dragonking> Crab_: You meant infinite loop if recruitment formula has fallback command?
20090330 18:36:33< KnightR> someone *really* should document the technical part of WML lol
20090330 18:37:03-!- Polarina [n=polarina@wesnoth/translator/Polarina] has quit ["Leaving."]
20090330 18:37:20< flav> Hi, I'm here for the GSoC, and i'm quite interested but i need more informations... can i have some ?
20090330 18:37:29< KnightR> hi flav
20090330 18:37:33< KnightR> i think you have to be more specific :)
20090330 18:37:46< Crab_> Dragonking: yes. and it is controlled with an 'if (master_)' statements. so I'm trying to guess 'is it a sole purpose of that variable to avoid that?'
20090330 18:37:46< flav> ok i'll try :)
20090330 18:38:18< flav> sorry because i'm french and my english is not perfect
20090330 18:38:30< zookeeper> look at the channel topic..
20090330 18:39:04< flav> yes i'have a quick look at the wiki
20090330 18:39:15< Crab_> Dragonking: and note that " ai_manager::get_ai("formula_ai", info_)->new_turn();" in ai::play_turn()  is not protected by if (master_)  ))
20090330 18:39:37< Dragonking> Crab_: Actually I wonder if this do_recruitmant in formula AI really work.
20090330 18:39:39< KnightR> zookeeper: what other uses does WML server apart from save games, replays and campaigns?
20090330 18:39:47< KnightR> which parts of the game does WML drive?
20090330 18:39:50< Crab_> Dragonking: so, it allows for formula_ai (master) => ai(not master) => formula_ai(not master) in some situations 
20090330 18:40:07< shadowmaster> KnightR: themes, terrain graphics layout, terrain tile definitions, internal game config (data/game_config.cfg), ...
20090330 18:40:20< shadowmaster> networking protocol, add-on storage on the server,...
20090330 18:40:22< zookeeper> yep
20090330 18:40:38< KnightR> damn thats a lot of WML heh
20090330 18:40:42< shadowmaster> unit type definitions, race definitions, preferences file,...
20090330 18:40:51-!- Blueblaze [n=nick@c-98-199-143-139.hsd1.tx.comcast.net] has joined #wesnoth-dev
20090330 18:40:52< zookeeper> basically all the actual content there is
20090330 18:40:56< KnightR> i see
20090330 18:40:57< zookeeper> (and then some)
20090330 18:41:32< Crab_> Dragonking: for me, now, it doesn't matter if it works or not ) I want to commit my ai_manager patch and I want to keep the semantics the same in that commit, even if the current behavior is wrong/don't works. So, I'm trying to learn what it tries to do now )
20090330 18:41:51< KnightR> in that case, what would you think of the idea of caching some of the WML documents into disk?
20090330 18:42:12< Dragonking> Crab_: How would it return to formula AI from AI if it AI is not master?
20090330 18:42:15< KnightR> i doubt ALL the WML documents are used extensively throughout the duration of a game right?
20090330 18:42:18< Dragonking> Bye executing what?
20090330 18:42:29< Crab_> Dragonking: http://wesnoth.pastebin.com/m722448b1,  ai_manager::get_ai("formula_ai", info_)->new_turn();
20090330 18:42:36< shadowmaster> KnightR: some are, some aren't.
20090330 18:42:37< Crab_> it is not inside 'if'
20090330 18:42:50< zookeeper> KnightR, i'm not a coder, but hard to say. i presume you have some profiling done on how much memory the document uses?
20090330 18:42:56< Crab_> so, it ignores that 'master/not master' stuff.
20090330 18:43:05< KnightR> i'm basing it on Dave's findings at the moment
20090330 18:43:05< shadowmaster> some are deserialized into C++ classes with custom fields (e.g. unit class), others are used/unused as config objects all the time
20090330 18:43:11< shadowmaster> KnightR: ^
20090330 18:43:21< KnightR> zookeeper: ^
20090330 18:43:33< flav> but first i need to know more about the subscribtions and how it works with google ... do you always needs students for GSoC ?
20090330 18:43:51< Dragonking> Crab_: But what heppens if ai_manager::get_ai("formula_ai", info_)->new_turn(); is called?
20090330 18:43:55< KnightR> shadowmaster: are these unused objects stored in memory? or unloaded when they are not needed?
20090330 18:44:10< Dragonking> Crab_:  It only invalidates move map, no formula is executed/called.
20090330 18:44:22< Turuk_> flav: http://code.google.com/opensource/gsoc/2008/faqs.html
20090330 18:44:31< shadowmaster> I recall at least one class whose name I can't recall, that keeps a config object of its own floating around until its parent (that class) is destroyed
20090330 18:44:42< Crab_> Dragonking: I've managed to get it to infinite-loop once )
20090330 18:44:43< Turuk_> whoops old one
20090330 18:44:49< Turuk_> flav: for 2009   http://socghop.appspot.com/document/show/program/google/gsoc2009/faqs
20090330 18:44:51< Crab_> but maybe it's my luck )
20090330 18:45:11< flav> yes i have this link but somes questions get no answer
20090330 18:45:17< Dragonking> Crab_: Infinite loop with current code with no changes?
20090330 18:45:28< shadowmaster> I don't recall what the class was or what its lifetime is however. I think it had something to do with unit types, races or animations
20090330 18:45:59< KnightR> shadowmaster: ok ill dig deeper into the way WML is handled and see what I can find
20090330 18:46:05< Crab_> Dragonking: no, not with current code.
20090330 18:46:06< KnightR> dave knows a lot more about this stuff right?
20090330 18:46:36< shadowmaster> KnightR: he knows well the underpinnings of WML. However he'd be surprised if he knew of all the clever and clumsy ways we use it for
20090330 18:46:49< KnightR> shadowmaster: i think the idea of caching unused objects to disk might be useful depending on how much is unused during a game.
20090330 18:47:03< KnightR> haha ;)
20090330 18:47:08< Crab_> Dragonking: ok, thanks with clearing out about the new_turn
20090330 18:47:38< brahim> what means "validation schemes"?
20090330 18:47:49< Crab_> Dragonking: so, about the 'master' variable: had I guessed its purpose right (only allow master to fallback) ?
20090330 18:48:01< shadowmaster> brahim: are you familiar with DTDs?
20090330 18:48:14< brahim> no
20090330 18:48:39< shadowmaster> http://en.wikipedia.org/wiki/Document_Type_Definition
20090330 18:48:46< zookeeper> KnightR, if you want to lower the WML memory footprint, it might be a good idea to first find out which things are taking up how much memory...basically, the biggest piles of WML are the terrain graphics definitions (basically only [terrain_graphics] tags) and unit type definitions ([unit_type] tags). when playing campaigns, also [scenario]s and when playing MP [multiplayer]s and maybe [era]s.
20090330 18:48:47< flav> I'have a look at your list of Ideas and i would like to know if some students have already contacted you for some of the ideas ?
20090330 18:48:51< brahim> ok thx
20090330 18:48:57< zookeeper> that's where the biggest WML is in terms of tags
20090330 18:49:00< shadowmaster> brahim: it's something like that, but applied to WML
20090330 18:49:17< shadowmaster> so it may involve Increased Complexity(tm)
20090330 18:49:57< brahim> ok
20090330 18:50:02< shadowmaster> zookeeper is making me want to rush out of here to steal a power outlet to connect my lappy and examine the terrain builder.cpp code
20090330 18:50:15< esr> shadowmaster: I did your wmllint FR, have yo seen mine about conditionals in [part]?
20090330 18:50:22< shadowmaster> generally I avoid having a single glare at builder.cpp
20090330 18:50:26< Dragonking> Crab_: IK am not sure, but I'd guess the same.
20090330 18:50:45< shadowmaster> esr: haven't you received the email after I marked it "fixed"?
20090330 18:50:50< shadowmaster> days ago
20090330 18:50:59< Dragonking> Crab_: And if you manage to make it infinite loop I'd be happy to know way to reproduce it. :)
20090330 18:51:02< KnightR> zookeeper: i see.. do you happen to know how these tags are stored in WML?
20090330 18:51:16< KnightR> an example of how a tag is stored into the std::map that WML uses
20090330 18:51:28< shadowmaster> esr: ahhh, I see 
20090330 18:51:46< Crab_> Dragonking: ok. thanks. I'll review my changes to formula_ai.cpp and ai.cpp, taking this into consideration, to ensure it works like now.
20090330 18:51:48< shadowmaster> esr: I can do it, but first I want to refactor intro.cppp
20090330 18:52:01< shadowmaster> without that, it'll be a PITA to do it IMO
20090330 18:52:32< shadowmaster> IIRC.
20090330 18:52:36< esr> shadowmaster: That answer doesn't surprise me, as the right wayy to do it would involve a mutual recursion the code cannot support as-is.
20090330 18:53:20< esr> I could *almost* do it myself.  But you know that code better.
20090330 18:54:01< shadowmaster> I want to refactor a lot of things this cycle (just not the "suokko way")
20090330 18:54:21 * esr makes encouraging noises
20090330 18:54:52< Crab_> Dragonking: so, 'master' AIs should be those AIs which are created by the 'game', and 'not master' AIs should be those AIs which are created by other AIs ?
20090330 18:55:23< Dragonking> Crab_: I think It would be best to wait for Sirp to make sure 
20090330 18:55:36< Dragonking> There may be something more into it
20090330 18:55:46< Dragonking> But what you say makes sense. :)
20090330 18:55:52 * shadowmaster rushes to steal a power outlet
20090330 18:55:53< Crab_> master is not used anywhere else, except fallback code
20090330 18:56:06-!- shadowmaster [i=929b360f@wesnoth/developer/shadowmaster] has quit [Client Quit]
20090330 18:56:08< Dragonking> Ok, then it should be exactly like that I think.
20090330 18:56:13< Crab_> ok
20090330 18:57:56-!- Turuk_ [n=Turuk@FUSE-WS1-208-102-202-188.fuse.net] has quit [Read error: 104 (Connection reset by peer)]
20090330 18:58:11-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090330 18:58:36-!- orian_ [n=orian@afje2.neoplus.adsl.tpnet.pl] has joined #wesnoth-dev
20090330 19:02:31< esr> wesbot: topic
20090330 19:05:58< brahim> what is a comment based scheme definition?
20090330 19:07:07< brahim> i did not understand what comment based means?
20090330 19:08:53< Shadow_Master> I'd ask Sirp if he was here
20090330 19:09:17< brahim> ok thx for the advice
20090330 19:11:41< Crab_> btw, Dragonking, what do you think about ai parameter handling (all those - ai parameters, global ai parameters, ai memory, effective ai parameters, some fields in team_info which act as parameters, ... ) ? (I haven't touched these yet, just moved their location without changing the interface to get them)
20090330 19:13:33< Dragonking> Crab_: Can you be a bit more specific? What you would like to do with these? And I do tihnk that it is best to talk with Sirp about it. :)
20090330 19:16:25< CIA-30> loonycyborg * r34295 /trunk/projectfiles/ (4 files in 4 dirs): Committed McShark's projectfiles update.
20090330 19:21:34< Crab_> Dragonking: more specific: 1) should the fields in team_info which deal with ai parameters be removed (this will require small changes to default ai to get them from effective ai parameters instead) ?  2) should the ai memory be merged with ai parameters ? 
20090330 19:23:03< Dragonking> Crab_: I'm not really fammiliar with team_info, so I don't know th answer.
20090330 19:25:42< Crab_> Dragonking, basically, there's a mess there ) the same info is duplicated in a number of places
20090330 19:26:09< Dragonking> Crab_: If so, then you may try fixing it. :)
20090330 19:26:13< Crab_> and about the second question ?
20090330 19:27:12< Dragonking> That's prolly also a question for Sirp.
20090330 19:27:42< Dragonking> I meself did not get too deep into default AI code.
20090330 19:28:13< Dragonking> I planned it but well.. then we had some big mees with default AI and I lost interest for a while. :)
20090330 19:29:25< Crab_> ok. and then a 3rd question. I intend to make a specific folder where AI .cfg files (basically - contents of the side tag) may be dropped to make them available as AIs (in multiplayer, for example). What will be a good place/name for such a folder ? (I am thinking about ./data/ai/ais , but I am not very familiar with wesnoth data dir structure)
20090330 19:30:19< Crab_> (a folder where each file will correspond to ai which will appear in the dropdown list of AIs in user interface when selecting opponents)
20090330 19:30:55-!- boucman [n=rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20090330 19:31:13< Dragonking> Crab_:  I think you may make use of the data/ai 
20090330 19:31:41< Crab_> ok
20090330 19:32:43< boucman> hey all
20090330 19:32:49< Crab_> boucman: hi boucman
20090330 19:32:56< Dragonking> hi boucman 
20090330 19:35:24-!- mordante [n=mordante@wesnoth/developer/mordante] has joined #wesnoth-dev
20090330 19:35:38< Crab_> Dragoking, boucman, mordante: so, I'm ready to commit that big patch of mine (2.5 kloc's), mostly touches team.cpp and 4 new files. If you want to review *before* I commit, please say so. It adds very basic history capabilities, ai hot-redeployment, and changes ai lifecycle management.
20090330 19:35:42< mordante> hi
20090330 19:35:52< Crab_> s/Dragoking/Dragonking, sorry
20090330 19:35:57< boucman> ouch, that's a lot
20090330 19:36:13< boucman> Crab_: yeah I'd gladly do a (quick) review
20090330 19:36:13< mordante> Crab_, is it possible to commit it in smaller chunks?
20090330 19:36:28< Crab_> mordante: no. mostly - lookup code changes.
20090330 19:36:55< mordante> :-(
20090330 19:37:09< Crab_> boucman: then wait a sec, I'll post a latest version on patches. the main question is : has I broken anything that uses team.cpp ? (i intended to preserve team.cpp interface and semantics)
20090330 19:37:42< boucman> Crab_: that what I want to check
20090330 19:38:51-!- _deleter [n=_deleter@adamsrj-1.dhcp.rose-hulman.edu] has quit [Nick collision from services.]
20090330 19:39:04-!- _deleter [n=_deleter@adamsrj-1.dhcp.rose-hulman.edu] has joined #wesnoth-dev
20090330 19:39:31< Crab_> I've checked it myself, by rereading through, plus on test scenario, plus on 1st level of HttT, plus in multiplayer, and it looked good. basically, I have a stack of AIs for each AI, and I want it to behave the same when there's only one AI per team.
20090330 19:40:11-!- jmunro is now known as jmunro[a]
20090330 19:40:11< boucman> Crab_: I was a bit away of wesnoth recently, but I'll check all that
20090330 19:44:30< zookeeper> Crab_, does your patch include anything that a WML author should know about?
20090330 19:44:40< Crab_> zookeeper: no, nothing yet.
20090330 19:44:52< zookeeper> ok
20090330 19:45:46< Crab_> zookeeper: I intended to stick to the old config for the 1st commit of it, so no new changes yet (apart from the fact that one can write AI in standalone config file (basically - contents of [side] tag) and redeploy using it)
20090330 19:45:47-!- anakayub [n=anakayub@203.82.91.34] has quit ["Later"]
20090330 19:47:02-!- cyber_11 [n=muratcan@78.165.120.116] has joined #wesnoth-dev
20090330 19:50:42< CIA-30> crab * r34296 /trunk/data/multiplayer/factions/ (12 files): removed team_formula elements from multiplayer faction configuration files, for they use a broken fai script
20090330 19:52:58< mordante> Shadow_Master, what's the exact problem with wescamp? I haven't looked at it last week and it isn't working for 1.6 at the moment
20090330 19:52:59< Crab_> boucman, mordante, Dragonking, zookeeper, here is it -> https://gna.org/patch/index.php?1146   Do not forget to update to r34296 before trying. and use --log-debug=ai_manager,ai_configuration
20090330 19:53:22< boucman> ok, will review tonight
20090330 19:54:09< zookeeper> boucman, did you get my note about flicking hp bars?
20090330 19:54:43< boucman> no
20090330 19:54:53< boucman> zookeeper: in what circumstances ?
20090330 19:55:06< zookeeper> <zookeeper> boucman, when recruiting, the leader's hp bars flicked off. probably caused by the default recruiting anim.
20090330 19:55:15< mordante> KnightR, boost is easy for msvc http://www.boostpro.com/products/free
20090330 19:55:29< zookeeper> s/flicked/flicker
20090330 19:55:35< boucman> zookeeper: yeah, probably
20090330 19:55:49< boucman> I'll make the HP bar visible during recruit anim, this should solve the problem
20090330 19:56:02< Crab_> boucman: there is code there which is marked with @todo - there are next in my work plans, so if there's some problems with them, that's not critical (for example, the entire history manager is a 100-line hack which will be rewritten very soon). The main thing is is 'the patch should not break anything that worked before, and should open the road for further improvement'
20090330 19:57:19< boucman> Crab_: noted
20090330 19:57:36< Crab_> boucman: If it 'doesn't break anything that worked before', then it's good for committing. That's because I don't want to 'make it ideal and then commit a 10-kloc-patch' ))
20090330 19:58:14< boucman> Crab_: indeed, your patch is already pretty big
20090330 19:58:23< boucman> but that's ok, we're not in a feature freeze right now
20090330 19:58:31< Crab_> that's good )
20090330 19:58:33< boucman> try not to make it bigger, though ;)
20090330 19:58:45< Ivanovic> in fact now is the best time to make "huge changes" since we are at the start of a new dev line
20090330 19:59:05< Ivanovic> though eventually there is a limit in "sane size for patches"
20090330 19:59:07< Ivanovic> ;)
20090330 20:00:00< Ivanovic> Dragonking: please add your idea page on the SoC ideas page, too
20090330 20:00:07< boucman> Crab_: the main problem with big patches is that if they break something, it's impossible to bisect the problem
20090330 20:00:17< Ivanovic> http://www.wesnoth.org/wiki/SummerOfCodeIdeas#GSoC_Student_pages
20090330 20:00:19< Dragonking> Ivanovic: Yea I know no worries
20090330 20:00:23< boucman> that's why i'd rather have more "less tested" commits than one big commit
20090330 20:00:42< Ivanovic> Dragonking: we just want to start to have a closer look at proposals listed in the wiki so far tonight
20090330 20:00:51< Dragonking> Ivanovic: ok :)
20090330 20:01:03< Ivanovic> and i was somehow suprised by boucman saying "hey, dk submitted a proposal" and me not seeing it so far
20090330 20:01:05< Ivanovic> ;)
20090330 20:01:08< Dragonking> Still my proposal is inclomplete ;p
20090330 20:01:25< Crab_> Dragonking: but it is interesting )
20090330 20:01:26< Ivanovic> tststs, you perfectionist...
20090330 20:01:30< boucman> no worries, that's fine, we know you have a week to finalize
20090330 20:02:13< Ivanovic> our main focus today will be getting a good idea what students atm are up to and where we should provide some more guidance
20090330 20:02:33-!- Velory [n=GProcess@85.99.0.63] has joined #wesnoth-dev
20090330 20:02:38< Velory> Hi everyone
20090330 20:02:39< Crab_> boucman: yes. but, in that regard, my patch is not that big. it's more like 'many small changes which are needed at the same time because it's lookup code' plus 2*2 new files with a well-defined interface, plus an attempt to clean up half of the mess in team.cpp (because i had to extract 'get ai from config' part fromit to separate function)
20090330 20:02:42-!- oak_ [n=oakus@c02--112b.kn.vutbr.cz] has joined #wesnoth-dev
20090330 20:03:03< Crab_> Velory: hi Velory
20090330 20:03:08< Velory> Hi crab
20090330 20:03:21< Dragonking> Ivanovic: Added.
20090330 20:03:25< Ivanovic> thanks
20090330 20:03:40< boucman> Crab_: I have already seen your code, I was not too worried,
20090330 20:04:04< boucman> it's more about wesnoth policy, (wrt searching bugs from a MIA dev) and general good practices
20090330 20:04:41< Ivanovic> yeah, huge mass commits fixing several, completely different bugs at the same time are *lovely*
20090330 20:05:04< Ivanovic> led us to some swearing when trying to get the AI back to a sane level in the last months...
20090330 20:05:06< Crab_> boucman: yes. I was trying to say that it was hard to make it as 'many small commits'
20090330 20:05:39< Crab_> Ivanovic: yes, I've seen those IFDEF's and svn logs :)
20090330 20:05:56< Ivanovic> Crab_: a smaller chunk could eg be the reorg of team.cpp
20090330 20:06:14< Ivanovic> and yeah, you can state "this patch requires that patch ABC is applied first"
20090330 20:06:17< Ivanovic> ;)
20090330 20:06:24< Crab_> Ivanovic: ))
20090330 20:06:28-!- brahim [n=lippe-er@ensihp303.imag.fr] has quit []
20090330 20:06:51-!- grzywacz [n=grzywacz@wesnoth/developer/grzywacz] has joined #wesnoth-dev
20090330 20:06:52-!- Reisiger [n=Reisiger@adsl-62-167-57-12.adslplus.ch] has joined #wesnoth-dev
20090330 20:07:05< Velory> boucman: by the way i fixed my healing improvement and submitted again :) https://gna.org/patch/index.php?1145
20090330 20:07:23< boucman> ok, I might be short in time tonight, though
20090330 20:07:43-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit [Read error: 104 (Connection reset by peer)]
20090330 20:07:58-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090330 20:08:22-!- cjhopman [n=chris@wesnoth/developer/cjhopman] has joined #wesnoth-dev
20090330 20:08:48< Crab_> Ivanovic: probably, yes  (although it would require me to 'patch team.cpp to some temporary-but-working state in patch A,  then immediately refactor that  temporary-but-working state' away in patch B.)
20090330 20:09:34< Ivanovic> Crab_: lovely, isn't it?
20090330 20:09:36< Ivanovic> ;)
20090330 20:09:39< Crab_> yea )
20090330 20:10:01 * Ivanovic heard that foo like this was possible with git...
20090330 20:10:38-!- ettin_ [n=jorda@wesnoth/developer/ettin] has joined #wesnoth-dev
20090330 20:11:14-!- VladimirSlavik_ [n=chatzill@88.103.147.178] has joined #wesnoth-dev
20090330 20:11:29< Crab_> Ivanovic: yes, it would be easier with git (eg, I could have done a local branch before i started diverging away with this 'temporary-but-working state ', and continued my work from here, without need to copy files around)
20090330 20:12:02< boucman> Crab_: some devs have set up a git mirror
20090330 20:12:32< Crab_> boucman: where ? I've was only told about a git tarball of *public* repo
20090330 20:12:33-!- Dragonking [n=dk@wesnoth/developer/dragonking] has quit []
20090330 20:12:53< boucman> Crab_: I might have misunderstood, I don't use git myself
20090330 20:13:05-!- Dragonking [n=dk@wesnoth/developer/dragonking] has joined #wesnoth-dev
20090330 20:13:16< Crab_> and I'm not skilled with git enough to *relocate* it to use svn+ssh
20090330 20:13:42< boucman> wesbot: seen aziz009
20090330 20:13:43< wesbot> boucman: Sorry, I don't know of aziz009.
20090330 20:15:28-!- VladimirSlavik [n=chatzill@wesnoth/translator/VladimirSlavik] has quit [Nick collision from services.]
20090330 20:15:38-!- VladimirSlavik_ is now known as VladimirSlavik
20090330 20:17:40< Crab_> and now I need to do the following: create a scenario (based on 'Den of Onis', on Dragonking's suggesion), change it a bit to make 'prisons' for 3 leaders (human,ai,ai) outside the main map, and hack up a way (c++ or WML or a combo of them) to, on user-specified trigger, execute a specific console command (=specific part of c++ code) and delete almost all units of map and create some more units which will be defined according to that u
20090330 20:17:52< Crab_> creating/deleting units on map are WML stuff, right ?
20090330 20:18:08< Dragonking> Crab_: If you want to make it multi purpose, then maybe select something biger than 2p den of onis
20090330 20:18:27< Dragonking> If you want to have small area, den of onis can be nice choice
20090330 20:18:45< Crab_> Dragonking: why ? it must only be big enough for one purpose at a time.  it would be 'repopulated' according to purpose :)
20090330 20:19:01< Crab_> so, it would be almost empty (only 3 leaders).
20090330 20:19:37< Dragonking> So if you think small map will be ok, then Den of onis could be nice.
20090330 20:19:39-!- stikonas [n=stikonas@ctv-79-132-187-171.vinita.lt] has joined #wesnoth-dev
20090330 20:19:56< Crab_> yes. those leaders will be "in prison pockets" outside the main map.  and then, user will do *something* and the main map will be repopulated according to the situation the user has chosen.
20090330 20:20:31< Crab_> then, any AI is to be redeployed , and voila - end turn and see how this ai fares in that situation.
20090330 20:21:05-!- ettin [n=jorda@wesnoth/developer/ettin] has quit [Read error: 110 (Connection timed out)]
20090330 20:21:12-!- gozaboru [i=akhilpai@markley-135168.reshall.umich.edu] has joined #wesnoth-dev
20090330 20:21:47< boucman> Crab_, Dragonking how complicated to have fallback() have the default AI play one unit, instead of playin all units (i.e having to call fallback() multiple times)
20090330 20:22:11< Crab_> boucman: it can be done quite easily
20090330 20:22:55< boucman> that would be a nice change (trunk only, obviously) to be able to let the C++ ai play a couple of stuff and then have some things moved at the end, like healers
20090330 20:23:00< Crab_> boucman: albeit a 'general solution' would be better. to make it possible to 'fallback' a single unit, a group of units, a recruitment, a village grabbing, etc..
20090330 20:23:08-!- gozaboru [i=akhilpai@markley-135168.reshall.umich.edu] has left #wesnoth-dev []
20090330 20:23:21-!- gozaboru [i=akhilpai@markley-135168.reshall.umich.edu] has joined #wesnoth-dev
20090330 20:23:21-!- gozaboru [i=akhilpai@markley-135168.reshall.umich.edu] has left #wesnoth-dev []
20090330 20:23:33< boucman> Crab_: ok, let's wait to have a better idea of what we want instead of rushing
20090330 20:24:11< Crab_> boucman: I'll see how easy is to do it as an 'option'...
20090330 20:24:20< loonycyborg> Crab_: You'll need to edit .git/config to enable commits through git-svn.
20090330 20:25:05< Crab_> loonycyborg: is this enough ? I've tried exactly that, but it was not enough for me.
20090330 20:25:32< loonycyborg> Hm.. It was the only thing I did IIRC.
20090330 20:25:37< Crab_> loonycyborg: or maybe I have to do this *immediately* after unpacking the tarball ?
20090330 20:26:01< Crab_> loonycyborg: thanks, i'll try again, later.
20090330 20:26:33< loonycyborg> Though that was long ago. I could forget some details :P
20090330 20:28:51< CIA-30> boucman * r34297 /trunk/src/unit_display.cpp: display bars during recruiting anims to avoid flickering
20090330 20:30:26< mordante> Crab_, did you change it to  svn+ssh://crab@svn.gna.org/svn/wesnoth
20090330 20:34:03< Crab_> mordante, yes
20090330 20:34:45< mordante> that should work, but I also had a problem with a new repo a little while ago, but suddenly it started to work
20090330 20:34:56< mordante> and I've no clue why it started to work...
20090330 20:35:07-!- iyonius [n=iyo@a02-0829c.kn.vutbr.cz] has quit ["Leaving."]
20090330 20:35:10< Crab_> ok. i'll refetch the git tarball & retry
20090330 20:37:42-!- iyonius [n=iyo@a02-0829c.kn.vutbr.cz] has joined #wesnoth-dev
20090330 20:38:35-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit []
20090330 20:40:36-!- Blueblaze [n=nick@c-98-199-143-139.hsd1.tx.comcast.net] has quit [Remote closed the connection]
20090330 20:46:45< Crab_> boucman: regarding your idea of 'let C++ default_ai play a couple of stuff' - it can be quite cleanly implemented 'as is'. For C++ default AI already allows to remove specific units fromits decision making process (this is for 'passive_leader' to work - it has to remove leader's moves from possible moves). So, we can easily hack 'void ai::do_move()' a bit to allow C++ default_ai to remove all the units, except those which we 'give' to 
20090330 20:47:31< boucman> Crab_: in that case, it would be good to have a way for FAI to "reserve for later" a unit
20090330 20:47:54< boucman> (in particular to be sure that healers are played at the end of turn)
20090330 20:48:12< Dragonking> boucman: Like what we discussed during FOSDEM :)
20090330 20:48:24< boucman> something like that, yeah
20090330 20:49:04< Crab_> boucman: reserving 'healers for last' it is not always a best approach. for example, Mage of Light is both a healer and a support unit
20090330 20:49:33< boucman> Crab_: yeah, but we need a way to do that
20090330 20:49:40< boucman> that doesn't mean we will always do that
20090330 20:49:41< Crab_> boucman: but, in general, I think that 'candidate move evaluation' stuff should be pluggable to allow different turn sequences.
20090330 20:50:11< boucman> there are all sort of cases where an evaluation needs to say "I'm pretty certain that what I want to do is a good idea, but I'd rather do it later"
20090330 20:50:18< boucman> berserker comes to mind to
20090330 20:50:35< Crab_> yes, you are right
20090330 20:51:04-!- Elvish_Pillager [n=eli@66-189-40-122.dhcp.oxfr.ma.charter.com] has joined #wesnoth-dev
20090330 20:51:04< boucman> Crab_: yeah, but we don't want that a "good idea that must wait for end of turn" is bypassed by "an average idea that can be done right away" and we have no way to specify that right now
20090330 20:51:54< Crab_> boucman: and a good idea that must wait for end of turn may suddenly become a bad idea at the end of turn (thanks to the holy RNG) 
20090330 20:52:26< boucman> yeah, that too, which means we also need a way to de-reserve a unit 
20090330 20:52:47< Crab_> boucman: for example, a unit that we want to heal may be killed when it attacks, or the enemy might die  faster that expected)
20090330 20:53:02< boucman> (or, probably simpler to follow, reset all unit reservation on a "regular" basis, where regular needs to be defined smartly
20090330 20:55:15< boucman> a good idea for "regular" would be "whenevear a real move is done (i.e not a reservation)
20090330 20:55:51< Crab_> boucman: I was thinking about those 'smart reservations' when I talked about AI decision cycle. I was thinking about AI understanding logic like 'I decide that I want to heal unit B with my healer A, so I expect this unit B to be in location where my healer A can reach it.' So, when considering moves for B, AI would take this expectation into account.
20090330 20:57:00-!- Blueblaze [n=nick@c-98-199-143-139.hsd1.tx.comcast.net] has joined #wesnoth-dev
20090330 20:57:49-!- elias [n=allefant@allegro/developer/allefant] has quit [Read error: 60 (Operation timed out)]
20090330 20:58:06< Crab_> so, AI can decide that 1) "moving\attacking with unit B to where healer can reach it." this keeps in line with earlier expectation for healer, so healer A's move is still reserved.  OR 2) "it is better to drop this expectation and pick another move with both A and B"
20090330 21:02:09< Crab_> this can also be extended with attacks: for example: AI evaluates move for unit A and sees that after attacking target C, it can also attack C with another unit B and have a good change to kill it. So, AI will attack with 1st unit and it will expect the target to be damaged enough for the second unit B to have a good chance to finish it off. That is another expectation which will 'reserve' unit B. And, for example, If none of A's attac
20090330 21:03:01< Crab_> it is, basically, dynamic reservation of units that may be useful if 'things go as we plan it'
20090330 21:03:24< boucman> Crab_: wrt your patch, you probably want to replacee "? ai" in ai_interface.cpp with a _( "unknown Artificial Intellignece") or something like that
20090330 21:03:29< boucman> note the translated string
20090330 21:04:03< boucman> I don't think we need reservation in your second example, reanalyzing the new situation should be sufficiant
20090330 21:04:14< boucman> however I agree with the first example and the general idea
20090330 21:05:44< boucman> hmm apparently none of the AI names are translatable, is this string for debug purpose or might it be user visible ?
20090330 21:05:52< Crab_> boucman: "reanalyzing the new situation should be sufficiant" - yes, but it will be slower.
20090330 21:05:59< Crab_> boucman: debug purpose
20090330 21:06:02< boucman> ok
20090330 21:06:41< boucman> Crab_: one thing I deeply believe is that trying to have the AI think in the long term is bound to failure, better do a deep analysis of the current situation
20090330 21:07:19< boucman> we have might have simple cases where we know a unit is more usefull at end of turn, but trying to predict the future will result in an overdesigned and ineffective AI
20090330 21:08:09< Crab_>  boucman: "better do a deep analysis of the current situation". This is not 'long-term thinking', this is basically *caching* of the decisions made before. For example, each unit will have to consider : is there any healers that are waiting till the end of turn to heal me ? if the said healer has two separate targets to heal, will it choose me?"
20090330 21:08:44< Crab_> so, we just *cache* the healers decision.
20090330 21:08:55< boucman> in the case of AI, I think that decideing if a cached decision is still good might be very tricky
20090330 21:09:14< boucman> that's why I thought rereserving unit after every unit move is a good idea
20090330 21:09:39< Crab_> boucman: yes,I agree.
20090330 21:09:50< boucman> because it's impossible to have a simple logic to know if a decision is still valid except reapplying the deciding logic
20090330 21:10:02< Crab_> but note the difference between "checking if cached decision is good enough" and reevaluating from scratch.
20090330 21:10:11< boucman> not even mentionning that most FAI programmers won't be able to do complicated stuff, so we need to keep the paradigm simple
20090330 21:10:27< boucman> yeah, well in that case I think that reevaluating is better
20090330 21:11:30< Crab_> imo, it wil be dawn slow :) but, of course, its better to code and see with own eyes.
20090330 21:11:53< Crab_> for example, if we have a healer..
20090330 21:12:13< Crab_> should we consider possible healing moves after *each* of our moves/attacks ?
20090330 21:12:22< boucman> Crab_: if you don't do that you need to give a "invalidate me" formulat with every formula, and there are very few cases wher I can imagine a coder will do something else than providing you the same code as "evaluate me"
20090330 21:13:08-!- Polarina [n=polarina@wesnoth/translator/Polarina] has joined #wesnoth-dev
20090330 21:14:05< boucman> the other way would be to check what variables are accessed to take the decision, but since we do everything with filtering, I'm not convinced we can do that
20090330 21:14:06< Crab_> boucman, at least 'locality' would be better. eg, "if something happens 90 tiles away from me, don't bother to reevaluate"
20090330 21:14:06< boucman> Crab_: the "invalidate me" formula will also probably very hard to code/debug for most coders
20090330 21:14:07< Crab_> boucman: i intend to write a debugger for fai )
20090330 21:14:07< boucman> Crab_: the engine has to know wher "me" want to be, which is not that simple
20090330 21:14:08< molgrum> hello, when is the last date to make a student idea proposal?
20090330 21:14:08< Crab_> boucman: but , of course, you are right about keeping invalidation rules simple. and 'invalidate always' is very simple :)
20090330 21:14:08< boucman> of course you could base yourself on the params given to the registered move, but my gut feeling is that it would be very error prone
20090330 21:14:28< Crab_> boucman: and what do you think of 'locality' invalidation ?
20090330 21:14:32< boucman> Crab_: I believe it's the only feasible one, except on a few cases
20090330 21:14:42< boucman> and so a dangerous and useless optimisation
20090330 21:14:50< molgrum> i have picked an idea already and started reading some code, and i want to know how much time i have on me to get into all this
20090330 21:15:20< boucman> we might add it later if everything else work, but it will be easy to add afterward since all evaluate can be used as invalidate, but the emphasis is on "if everything else works"
20090330 21:15:32-!- Elvish_Pillager [n=eli@66-189-40-122.dhcp.oxfr.ma.charter.com] has quit [Read error: 110 (Connection timed out)]
20090330 21:15:49-!- Netsplit verne.freenode.net <-> irc.freenode.net quits: shikadibot, esr, CIA-30, AI0867_, mordante, AI0867
20090330 21:16:03< boucman> molgrum: basically follow the google planning, so you have to submit your proposal to google before end of week, 
20090330 21:16:11< boucman> a little bit more to polish it on the wiki
20090330 21:16:26-!- Netsplit over, joins: mordante, shikadibot, esr, CIA-30, AI0867_, AI0867
20090330 21:16:26< Crab_> we can say that 'default =  invalidate always' , and if there will be some formulas where we can improve speed by caching, do it only for them.
20090330 21:16:32-!- Elvish_Pillage2 [n=eli@66-189-40-122.dhcp.oxfr.ma.charter.com] has joined #wesnoth-dev
20090330 21:16:53< boucman> Crab_: we could but again, that's only an optimisation, don't do it before everything else works
20090330 21:17:04< boucman> remember, early optimisation is the mother of all evil
20090330 21:17:14< boucman> it's not just evil, it's the mother of all evil :)
20090330 21:18:38< Crab_> boucman: ok, let it be 'support for future optimization'. And I think that in some cases it will be needed anyway - for example, note the startup time of formula_ai which uses recruitment.fai (disabled in current code). it is around 5-10 seconds.
20090330 21:18:39< Velory> :d
20090330 21:18:47< boucman> Crab_: ai.cpp line 981 (or something like that) you have a commented out if(master_) play_turn
20090330 21:19:00< boucman> that you convert into the uncommented equivalent, you sure about that ?
20090330 21:19:25< Crab_> yes
20090330 21:19:40< boucman> ok, fair enough
20090330 21:19:54< boucman> I'm rereading the patch right now, so I have little context to check with :)
20090330 21:21:55-!- Elvish_Pillage2 is now known as Elvish_Pillager
20090330 21:21:55< Velory> Crab_: are you send you proposal or did anything from GSOC to wesnoth ?
20090330 21:22:07< Velory> Crab_: i just registered as user
20090330 21:22:27< Crab_> Velory: if you are about google, I've made them happy today - my proposal is registered with them.
20090330 21:22:34< boucman> Crab_: ok, this looks good from a non-reg point of view
20090330 21:23:15< Velory> Crab_: i want to do that too how can i i should create document or what ?
20090330 21:23:49< Crab_> boucman: regarding 'master_', if master_=true, then this AI is considered 'master' ai which interacts with the game. if 'master' is false, then this ai was created by some other AI and is considered auxillary ai. it was there mainly to prevent an infinite loop of fallbacks
20090330 21:23:50< Velory> and boucman: if u have a time can u read my proposal comments are really improtant for me i edited today http://www.wesnoth.org/wiki/SummerOfCodeProposal_Velory
20090330 21:24:23< Velory> because really short time left
20090330 21:25:30< Crab_> boucman: in my opinion, that line *should* be commented away (IMO, plain default_ai should not call formula_ai). but, I wanted to preserve bug-for-bug compatibility.
20090330 21:25:58< boucman> Crab_: yes, make sense
20090330 21:26:18< boucman> (i mean being bug-for-bug compat, we try to keep bisecting in mind)
20090330 21:27:38< Crab_> boucman: so, I made it work like it was originally intended (and so I've removed [team_formula]'s from multiplayer .cfg files, in order to not reintroduce https://gna.org/bugs/index.php?13218 which is still not fixed - it needs recruitment.fai changes to work)
20090330 21:28:42-!- _deleter [n=_deleter@adamsrj-1.dhcp.rose-hulman.edu] has quit [Read error: 60 (Operation timed out)]
20090330 21:29:19< Crab_> formula_ai was originally introduces to ai.cpp in order to delegate some AI decisions (eg, recruitment) to the formula_ai. but the dev's haven't managed to make formula_ai recruitment bug-free in time for 1.6
20090330 21:29:28-!- euschn [n=chatzill@85-127-119-2.dynamic.xdsl-line.inode.at] has joined #wesnoth-dev
20090330 21:29:31< Crab_> s/introduces/introduced
20090330 21:29:35< euschn> evening
20090330 21:29:38-!- cjhopman [n=chris@wesnoth/developer/cjhopman] has quit [Success]
20090330 21:29:44< Velory> evening euschn
20090330 21:29:50< boucman> hey euschn
20090330 21:29:58< boucman> i'm rereading your village patch as we speak
20090330 21:30:28< euschn> ah yes - unfortunately Ill have to change most of it again - had a discussion about the task with Ivanovic
20090330 21:30:44< Ivanovic> :)
20090330 21:31:03< boucman> euschn: ok, i'll still do a quick review, there are a couple of things you might want to know to avoid doing the same mistakes again
20090330 21:31:14< Crab_> Velory: 1) register to google 2) register as a SoC student 3) submit proposal where "Hi! I'm Velory, see the wiki *here* " is written
20090330 21:31:16< zookeeper> dfranke, it seems the http://gna.org/bugs/?13267 guy wasn't the only one who experiences it: http://www.wesnoth.org/forum/viewtopic.php?f=4&t=24737
20090330 21:31:29< euschn> boucman: of course, sounds good
20090330 21:33:02-!- Gauteamus [n=chatzill@c43C4BF51.dhcp.bluecom.no] has joined #wesnoth-dev
20090330 21:39:17< boucman> euschn: done
20090330 21:39:42< euschn> boucman: thanks, checking email
20090330 21:39:52 * Reisiger goes to get a life while boost compiles... o_O
20090330 21:40:02< boucman> euschn: I commented directly in the patch submission, might be faster to check there
20090330 21:40:23< euschn> ok
20090330 21:41:38< Crab_> boucman, " Crab_: ok, this looks good from a non-reg point of view" - does that mean that you've done with reading through that patch ? 
20090330 21:41:55< boucman> yes
20090330 21:42:01< Crab_> thanks
20090330 21:42:04< Velory> Crab_: thanks so Hi i'm velory see my wiki page about proposal *http://www.wesnoth.org/wiki/SummerOfCodeProposal_Velory* ... is enough. or should i copy paste all proposal to there and in addition can i change my proposal after i submit i cant right?
20090330 21:42:22< boucman> it was actually pretty simple to review, since it's a simple patch semantically
20090330 21:42:48< Velory> s/is enough./is enough?
20090330 21:43:00< boucman> Velory: we don't really care about what you submit to google, what's really important is your wiki page (we read both of them of course,)
20090330 21:43:05< Crab_> Velory: I've heard that this is enough.
20090330 21:43:19< Velory> boucman: okey
20090330 21:43:32< Ivanovic> what is important is that every SoC student submits *something* to google, too
20090330 21:43:44< Ivanovic> and of course: make it easy for us to see what your wiki page is
20090330 21:43:45< Ivanovic> ;)
20090330 21:43:57< Velory> Ivanovic: OK :)
20090330 21:44:24< boucman> actually, if you had a one line application "please see my wiki page at ..." that would be enough for us :P
20090330 21:44:32< Ivanovic> jupp
20090330 21:44:45< Velory> boucman : woow :) 
20090330 21:45:09< Velory> boucman: one man :) should comment about my wiki page :)
20090330 21:45:24< Ivanovic> yes, we need to have info about your project (including a timeline!) and stuff like this, but it does not matter much for us if it is in the wiki or at google
20090330 21:45:37< Velory> its really really bad or its bad or its not really bad :p
20090330 21:45:38< boucman> Velory: I'm rereading it right now, I will post comments on the talk page
20090330 21:45:48< Velory> boucman: Ok thanks :)
20090330 21:47:35-!- silene [n=plouf@wesnoth/developer/silene] has joined #wesnoth-dev
20090330 21:47:50< silene> hi
20090330 21:47:54< Crab_> hi, silene
20090330 21:52:45< Ivanovic> euschn: currently reading over your application
20090330 21:52:59-!- Reisiger [n=Reisiger@adsl-62-167-57-12.adslplus.ch] has quit [Read error: 104 (Connection reset by peer)]
20090330 21:52:59< Ivanovic> euschn: found a (small) mistake: "Austria is UTC+1"
20090330 21:53:09< euschn> it isnt?
20090330 21:53:13< Ivanovic> IIRC you got daylight saving time in austria as well as in germany
20090330 21:53:19< Ivanovic> so atm you are at utc+2
20090330 21:53:25< euschn> ah right
20090330 21:53:30< euschn> thanks
20090330 21:53:54< Ivanovic> (really not important though...)
20090330 21:54:19< euschn> the application is still largely work in progress, especially the questions YogiHH posted as first guideline
20090330 21:54:36< Ivanovic> :)
20090330 21:54:56< euschn> Ill get them done once i finish random village naming :)
20090330 21:55:50< boucman> Velory: have you submited your proposal to google yet ?
20090330 21:55:59< Velory> no
20090330 21:56:04< Ivanovic> and regaring your timeline: personally i have the feeling that the stuff you listed under "may 16th - june 15th" is mostly what you do to get familiar with the code
20090330 21:56:09< Velory> boucman:no
20090330 21:56:30< boucman> Velory: ok, tell me once you're done, I just remembered that we said we would post reviews on google page so other mentors can see them too
20090330 21:57:14< Velory> Ok then
20090330 21:57:16-!- KnightR [n=anonymou@bb121-7-175-14.singnet.com.sg] has quit []
20090330 21:57:18< Ivanovic> so yeah, i don't see a real difference between "april 20th - may 15th" and "may 16th - june 15th"
20090330 21:57:55< Ivanovic> beside this: your timeline is *by far* too optimistic in the area of implementation
20090330 21:57:59< euschn> boucman: thanks for the comment, I was not aware that my editor does not track whitespace diff, proofreading the patch will help
20090330 21:58:11< euschn> what changelog should I add myself to? trunk/changelog?
20090330 21:58:20< Ivanovic> you won't be able to implement the stuff in ~5 to 6 weeks simply because of the problems you will encounter
20090330 21:58:22< boucman> euschn: being the official patch monkey, I'm used to spot that type of typoes :)
20090330 21:58:30< euschn> :)
20090330 21:58:39< Ivanovic> since the current implementation is a mess, you will encounter some problems and ask ilor how much this can slow you down
20090330 21:59:41< Velory> boucman: now i submitted
20090330 21:59:58< boucman> ok, thx
20090330 22:00:08-!- Reisiger [n=Reisiger@adsl-62-167-57-12.adslplus.ch] has joined #wesnoth-dev
20090330 22:00:48< boucman> Velory: commented
20090330 22:00:59< Velory> boucman: i filled abstract part by the way
20090330 22:01:07< Velory> boucman: were my user page?
20090330 22:01:12< Velory> s/were/where
20090330 22:01:18< boucman> I commented on the google page
20090330 22:01:28< Reisiger> Evening everyone *goes back to getting all prerequisites set up for compiling wesnoth in msvc 2008*
20090330 22:02:51-!- _deleter [n=_deleter@adamsrj-1.dhcp.rose-hulman.edu] has joined #wesnoth-dev
20090330 22:03:20-!- ettin_ is now known as ettin
20090330 22:03:48< Velory> boucman: thanks for comment
20090330 22:04:17< boucman> np
20090330 22:07:49-!- cyber_11 [n=muratcan@78.165.120.116] has left #wesnoth-dev []
20090330 22:08:42< CIA-30> crab * r34298 /trunk/ (21 files in 2 dirs): 
20090330 22:08:42< CIA-30> added basic history and hot-redeployment capabilities to in-game console.
20090330 22:08:42< CIA-30> changed AI lifecycle management to use the new ai_manager implementation
20090330 22:08:42< CIA-30> changed AI configuration parsing to use the new ai_configuration implementation
20090330 22:09:52-!- cyber_11 [n=muratcan@78.165.120.116] has joined #wesnoth-dev
20090330 22:10:08< euschn> Ivanovic: the first two periods are quite the same because Im still a bit confused about the first milestone of savegames, "getting familiar with the code" and what google calls the "bonding period" before the official start
20090330 22:10:14< mordante> euschn, can you ping me when you finished your application, then I'll read it
20090330 22:10:35< euschn> I admit the timeline is quite vague at the moment, I will improve it
20090330 22:10:44< Ivanovic> euschn: basically you should be familiar with the code when the "coding time" starts
20090330 22:10:51-!- flav [n=flav@mne69-3-82-225-23-140.fbx.proxad.net] has left #wesnoth-dev ["Konversation terminated!"]
20090330 22:11:04< Ivanovic> so this should be the latest time when you start on planning things eg via uml
20090330 22:12:45< euschn> Ivanovic: ok, so something like april 20th-may 15th: get familiar, may 16th-may30th: finish uml, june 1st - end: coding is more reasonable?
20090330 22:13:14< euschn> by the way it is likely I will have more time in June anyways, I will mention it in the application
20090330 22:13:15< Ivanovic> personally i'd say: add two weeks at the end to evaluate if your stuff really works the way as planned
20090330 22:13:26< Ivanovic> this time for testing is also an additional buffer if required
20090330 22:13:49< euschn> ok, I can see this would be a good idea
20090330 22:14:20< Ivanovic> since the savegames are a fragile and problematic area thorogh testing will be required
20090330 22:14:57-!- Sparks_ [n=Sparkste@78-105-212-150.zone3.bethere.co.uk] has joined #wesnoth-dev
20090330 22:15:04< euschn> mordante: ok, I was planning to finish my easycoding task first, which means Ill finish my application later in the week - or should i prioritize the other way round?
20090330 22:15:18< Ivanovic> no, finish your coding
20090330 22:15:25< Sparks_> hello guys!
20090330 22:15:30< euschn> ok
20090330 22:15:34< Ivanovic> we just comment already a little so that you have some feedback
20090330 22:15:46< mordante> doesn't really matter, I just like to read the application after it's finished
20090330 22:15:53< euschn> thanks, its much appreciated
20090330 22:16:10< mordante> you're welcome
20090330 22:16:40< boucman> hey Sparks_
20090330 22:16:58< Sparks_> boucman: good week? :)
20090330 22:17:05< boucman> busy :)
20090330 22:17:10< Sparks_> same.. :)
20090330 22:17:24< boucman> reviewing student applications+patches takes all my wesnoth time currently :)
20090330 22:17:46< Polarina> boucman: Then use your rest time too. :)
20090330 22:17:57< boucman> :P
20090330 22:18:04< Sparks_> boucman: studen apps are kind of bursting at the moment compared to about a week ago :)
20090330 22:18:06< Velory> my midterms starting at saturday :D but im giving all time to wesnoth :)
20090330 22:18:06< Polarina> And sleep every other night.
20090330 22:18:16< Velory> first exam is operating systems really hard one :) 
20090330 22:18:30< Polarina> Velory: That's the fun part.
20090330 22:18:50< boucman> Sparks_: that was to be expected :)
20090330 22:18:55< Velory> Poarina: hehe :)
20090330 22:19:37< Sparks_> Boucman:  How do I ask people for some feedback on my GSoC proposal?  Do i just ask here?
20090330 22:19:38< Polarina> Velory: Just press P on the keyboard, then Tab to get my name correctly. :)
20090330 22:19:45< Polarina> Sparks_: Yes.
20090330 22:19:46< Velory> Polarina:  wow :D
20090330 22:20:00< Velory> Polarina:  well like terminal :)
20090330 22:20:02< noy> can we stop with the smilies?
20090330 22:20:03< boucman> Sparks_: we are reviewing all aplications submited to google right now, I don't think yours was...
20090330 22:20:10< Polarina> Velory: Type "iv" and Tab.
20090330 22:20:15< Sparks_> Polarina: thanks for the tip about tabs ... 
20090330 22:20:29< Polarina> Sparks_: My pleasure.
20090330 22:20:33< Sparks_> boucman:  umm.. is the dedaline not the 3rd of april?
20090330 22:20:40< noy> Anyway, has a bug report been submitted for the fact that the turn bell nolonger works?
20090330 22:20:42< Velory> Sparks_: yes
20090330 22:20:46< cyber_11> Sparks_ yeah
20090330 22:20:59< boucman> it is, but don't worry, we are just providing some comments to help improve proposals, not doing a final review
20090330 22:21:21< Sparks_> boucman:  so should I fill in the google stuff now to get some feedback?
20090330 22:21:29< boucman> and as stated above, we juge both the google proposal and wiki page, so you can improve your proposal based on the feedback we give
20090330 22:21:31< cyber_11> boucman so when it will be ended?
20090330 22:22:11< cyber_11> boucman: what's your deadline?
20090330 22:22:18-!- yann [n=dwitch@nan92-1-81-57-214-146.fbx.proxad.net] has joined #wesnoth-dev
20090330 22:22:49< mordante> google's deadline is the 3th of april
20090330 22:23:06< boucman> not sure when we need to review for google, but don't worry you will be told in time
20090330 22:23:07< mordante> so after that we really start to judge the proposals 
20090330 22:23:43< Sparks_> could some one post the link to the google stuff to fill for GSoC?  I can't seem to find it..
20090330 22:25:33< mordante> Crab_, I also had a look at your proposal but no comments, but didn't review the project itself in detail, I leave that for sirp
20090330 22:25:56< Crab_> mordante: ok. I've changed my priorities in my proposal after IRC discussions.
20090330 22:26:18< boucman> Crab_: I just did comment on your google page
20090330 22:26:46< mordante> Crab_, ok, just wanted to inform you, that your proposal is being looked at ;-)
20090330 22:26:49-!- YogiHH [n=chatzill@c174106.adsl.hansenet.de] has joined #wesnoth-dev
20090330 22:26:55< Crab_> boucman: reading...
20090330 22:27:11< cyber_11> i wanna ask questions about extending multiplayer server idea and savegame reorganization
20090330 22:27:28< mordante> for savegames best ask YogiHH 
20090330 22:27:39< mordante> what do you want to know about the MP server?
20090330 22:27:39< Ivanovic> and multiplayer server stuff: just ask
20090330 22:27:45< Crab_> boucman: thanks
20090330 22:27:53< Reisiger> Q: is gettext stil a prerequesite? It's mentionned on the wiki/CompilingWesnoth site but not in the INSTALL note from the WesnothSVN
20090330 22:27:55< mordante> in general just ask on irc
20090330 22:27:59< Ivanovic> that is the way to go anyway: ask your questions and wait (patiently) for a reply
20090330 22:28:03< cyber_11> i'm trying to be familiar with STL but i have no time for writing a patch to 3rd of april
20090330 22:28:10< Ivanovic> Reisiger: it is
20090330 22:28:10< cyber_11> my midterms are starting
20090330 22:28:21< cyber_11> do i have any chance for gsoc?
20090330 22:28:46< Reisiger> Ivanovic thanks
20090330 22:28:48< noy> cyber_11: not likely tbh
20090330 22:28:53< Ivanovic> cyber_11: we at least do want to see some code you wrote yourself so that we can get an idea if you have the required skills
20090330 22:29:01< mordante> having a patch in greatly increases your changes
20090330 22:29:22< Velory> boucman: about loop idea yes i learned deep using filter function in my healing approach but my idea on loop structure is like for example suppose that we have a list with 10 variables and we want to do/learn/control something with [2].element to [7].element i dont know if we can do that with filters ?
20090330 22:29:22< mordante> but it will also depend on your proposal
20090330 22:29:52< cyber_11> Ivanovic: for exapmle what should i do in shortest way to prove my self?
20090330 22:29:54< Ivanovic> it depends a) on your proposal and b) on the "impression" we have about you
20090330 22:30:16< mordante> cyber_11, writing a patch
20090330 22:30:28< Ivanovic> the impression includes (but is not limited to) "do we think you are able to finish the project?"
20090330 22:30:37< boucman> Velory: i'm not sure what you mean, could you give a more detailed example ? I guess we can, it's just that FAI is very different of the usual functional approch of most languages
20090330 22:30:42< boucman> more like prolog 
20090330 22:30:55< Ivanovic> beside this it is "do we communicate well" and other stuff like this
20090330 22:31:31< Velory> boucman: hmm let me think about that...
20090330 22:31:47< cyber_11> yeah everything about teamworking is OK for me and for you also
20090330 22:32:29< cyber_11> do you have any suggestion on the patch especially on savegame reorganization and extending multiplayer server?
20090330 22:33:08< YogiHH> cyber_11: there are 2 EasyCoding Tasks related to savegames
20090330 22:34:07< Soliton> Crab_: still trouble committing with git svn?
20090330 22:34:39< YogiHH> cyber_11: You can also check the bugtracker for bugfixes. You might want to double-check here, though, because we probably can better estimate if the bug is suited for a beginner.
20090330 22:34:53< Crab_> Soliton: not tested yet. I will test again tomorrow.
20090330 22:35:24< Soliton> Crab_: after fixing .git/config you need to rebase again with new commits to make it work, afair.
20090330 22:36:29< Crab_> Soliton, I've googled up a procedure to do that, but it hasn't worked for me. Maybe I
20090330 22:36:46< Crab_> Maybe, I've done something wrongly. will try again )
20090330 22:37:14< Crab_> I'll tell the results.
20090330 22:39:23< euschn> silene: thanks for your comment, I agree that it is too english centered, it will be changed
20090330 22:39:28< Velory> boucman: for example suppose we filtered a list that carrying all units that have special attack,first we sorted and took size,finally we want to control units between 2nd order to last (or some specific order).Can we do that with filters do we have a power for controlling/tracing list which choosen order like (2 to 7 in 10 elements list) 
20090330 22:41:24< Velory> hmm maybe we can filter 2 to 7 ones too
20090330 22:42:42< boucman> map([2,3,4,5,6,7],'i',my_list[i])
20090330 22:42:48< boucman> something like that
20090330 22:43:54< boucman> a range(2,7) function might be a good addition, but it's not a complicated one...
20090330 22:47:51< Velory> hmm i see 
20090330 22:48:17< Velory> range(2,7) is really looking simply and thats what i want to do :)
20090330 22:48:56< Velory> boucman:thanks :)
20090330 22:49:18< Velory> s/want to do/im looking for
20090330 22:49:19< Velory> :p
20090330 22:49:53< CIA-30> esr * r34299 /trunk/data/tools/wesnoth/wmliterator.py: Improved convenience methods for testing element types.
20090330 22:51:43< Crab_> I have a question about "campaigns/Legend_of_Wesmere/scenarios/16_The_Chief_must_die.cfg". http://wesnoth.pastebin.com/m231f5e90   it is using ai_algorithm=formula_ai for one of the sides. is this intentional  ?
20090330 22:52:01-!- zookeeper [n=l@dsl-tkubrasgw1-fe9ff800-215.dhcp.inet.fi] has quit []
20090330 22:52:13< boucman> Crab_: IIRC yes
20090330 22:52:26< boucman> that scenario is a bit special zookeeper would know :P
20090330 22:52:37-!- Blueblaze [n=nick@c-98-199-143-139.hsd1.tx.comcast.net] has quit [Remote closed the connection]
20090330 22:52:46< Crab_> why ? i'm curious. it is the *only* mainline scenario which uses it, and, still, only for one of the sides )
20090330 22:55:24< boucman> Crab_: I don't know but I'm pretty sure it's on purpose, since we had to change it when removing python
20090330 22:55:51< Crab_> ok )
20090330 22:56:21< Reisiger> Q: How do I integrate gettext into MSVC 2008 ? Looking at the gettext folders it seems I need to compile them, yet neither ./configure (doesn't get recognized) nor make work.
20090330 22:56:30< Soliton> it's a scenario where the heros are invisible and only wolf riders can smell them and hunt them down or such.
20090330 22:56:35-!- cjhopman [n=chris@wesnoth/developer/cjhopman] has joined #wesnoth-dev
20090330 22:57:46< mordante> esr, something went wrong in commit 34263, the line "*Left click or press spacebar to continue..." in the tutorial hasn't been converted
20090330 22:58:07< esr> mordante: Looking...
20090330 22:58:34< loonycyborg> Reisiger: AFAICT you only need to add location of libintl.h to include path and link against libintl.
20090330 22:58:49< loonycyborg> That is unless you want to compile .mo files..
20090330 22:59:20< boucman> res|laptop: reviewed your google application, please see comment there
20090330 22:59:33< loonycyborg> Reisiger: You'd better use binary distribution of gettext.
20090330 23:00:00< Reisiger> loonycyborg: that explains it o_O there is no libintl.h file anywhere.
20090330 23:00:20< esr> mordante: I see why.  Multiline string comntinuation - no " before the *.   That's a weird edge case; I'll fix it by hand.
20090330 23:00:22< boucman> esr: I don't know if you noticed, but alink submited a shell script that uses ImageMagick to count the number of "out of hex" pixels of unit images
20090330 23:00:35< Reisiger> M'kay... I'll go looking through their ftp server
20090330 23:00:41< boucman> for the sake of uniformity it might be nice to rewrite it in python, and is probably very simple
20090330 23:00:55< mordante> esr, I think it happens more often also in UMC
20090330 23:01:16< esr> boucman: Is that what it does? I saw it, and was annoyed because it wasn't documented.
20090330 23:01:33< Dragonking> boucman, Ivanovic I'll improve and complete my proposal the day after tomorrow
20090330 23:01:42< Ivanovic> Dragonking: great
20090330 23:01:43< boucman> esr: it was a quick hack we did in an emergency, it was not meant to be "clean"
20090330 23:01:49< boucman> but yes, that's what it does :)
20090330 23:02:03< Dragonking> Still I have 2 tests - one tomorrow and one the day after, so i won't be able to do anything tomorrow
20090330 23:02:09< boucman> Dragonking: sure thing, submit it when you want us to "pre-review" it
20090330 23:02:23< Dragonking> ok
20090330 23:02:31< Dragonking> Good night everyone
20090330 23:02:35< boucman> night
20090330 23:02:37< Crab_> night, Dragonking
20090330 23:02:39< Ivanovic> n8 Dragonking
20090330 23:02:56-!- Dragonking [n=dk@wesnoth/developer/dragonking] has quit []
20090330 23:03:03< Sparks_> boucman: hi again... please would you check whether I have submitted a application via Google?  I think I have just done that...
20090330 23:04:20< boucman> Sparks_: got it,
20090330 23:04:23-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090330 23:04:24< esr> mordante: Looks like that case might not be difficult to lift automatically.  Testing now....
20090330 23:04:35< boucman> I'll try to review tonight, but it's already late, so no promise
20090330 23:04:42< boucman> if I don't another mentor will do it
20090330 23:04:46< Sparks_> boucman: many thanks :)
20090330 23:05:42< Sparks_> boucman: could you please put any questions etc. in the 'comments' section?  And is this all I need to do to submit my application now?  Just editing the proposal and wiki?
20090330 23:06:03< Sparks_> boucman: I don't need to fill in any other stuff.... ? :)
20090330 23:06:04-!- happygrue [n=George@wesnoth/developer/wintermute] has quit ["http://xkcd.com/91/"]
20090330 23:06:33< mordante> esr, nice
20090330 23:06:45< boucman> nope, not for us at least, providing a real link would be better, but we can find your proposal the way it is
20090330 23:06:56< boucman> and yes, we comment on google page to have a better tracability
20090330 23:07:57< Sparks_> boucman: what do you mean about better tracebility?
20090330 23:08:33< boucman> we can find out easily what we said about your proposal, and read each other's comments
20090330 23:09:22< Sparks_> boucman: so where abouts ont he google pages can I see any queries/comments?
20090330 23:09:38< boucman> below your proposal
20090330 23:10:13< Sparks_> boucman: I just see a list of my propsals at the moment?  will it be in the proposal itself? :)
20090330 23:10:32< boucman> yes, most likely
20090330 23:10:43< boucman> though I don't know the student interface, I never was a GSoC student
20090330 23:11:08< Sparks_> nevermind..  I found it ... :)
20090330 23:12:20< Sparks_> Boucman: so just to make sure...  if I don't fill in anything else from now,I will be succesfully 'applied' come april 3rd.. :)
20090330 23:12:32< CIA-30> mordante * r34300 /trunk/src/ai_manager.cpp: Comment out some unused variables.
20090330 23:12:38< boucman> Sparks_: AFAIK yes
20090330 23:12:38< CIA-30> esr * r34301 /trunk/data/campaigns/tutorial/scenarios/1_Tutorial.cfg: Some old markup can't be lifted aotomatically.
20090330 23:12:41< CIA-30> mordante * r34302 /trunk/src/ai_manager.hpp: Fix a compiler warning.
20090330 23:12:46< CIA-30> mordante * r34303 /trunk/src/ai_configuration.cpp: Fix a compiler warning.
20090330 23:12:51< CIA-30> mordante * r34304 /trunk/src/ai_interface.hpp: Comment out an unused parameter.
20090330 23:13:14< mordante> Crab_, ^any reason why the function isn't pure virtual, I ask since the class is abstract
20090330 23:13:26< Sparks_> Boucman: thanks :)
20090330 23:13:31< boucman> np
20090330 23:14:59< Crab_> describe_self or evaluate ?
20090330 23:15:38< mordante> evaluate
20090330 23:16:22< crimson_penguin> Ivanovic: my uncle confirmed that the font I use does actually come with the OS
20090330 23:16:34< Ivanovic> crimson_penguin: good
20090330 23:16:59< Crab_> because not all AIs need to implement it - only those AIs who can be used as console AIs. for example, default_ai hasn't got any language - so it cannot evaluate anything. formula_ai, on the other hand - can evaluate.
20090330 23:17:19-!- VladimirSlavik [n=chatzill@wesnoth/translator/VladimirSlavik] has left #wesnoth-dev []
20090330 23:17:53< mordante> ok
20090330 23:18:08< YogiHH> anyone knows, if it is possible to define an event dynamically via WML (that is, that an event is added during playing the game)?
20090330 23:18:28< boucman> YogiHH: i think it's possible yes
20090330 23:18:36< silene> YogiHH: yes, it is
20090330 23:18:38< mordante> I'm also quite sure it's possible
20090330 23:18:39< boucman> I remember reading some documentation about it, at least
20090330 23:18:49< boucman> [event] with [event] IIRC
20090330 23:18:58< boucman>  /with/within
20090330 23:19:03< YogiHH> urgh, wrong answer ;-)
20090330 23:19:10< silene> and it is used a lot
20090330 23:19:32< YogiHH> WML is a really evil beast for a coder :/
20090330 23:19:32< Crab_> mordante: and, for example, if there will be a LUA AI, it will, most likely, have this evaluate ability )
20090330 23:20:11-!- Gauteamus [n=chatzill@c43C4BF51.dhcp.bluecom.no] has quit ["ChatZilla 0.9.84 [Firefox 3.0.7/2009021910]"]
20090330 23:20:17-!- grzywacz [n=grzywacz@wesnoth/developer/grzywacz] has quit [Remote closed the connection]
20090330 23:20:31< mordante> that's why I wondered why not pure virtual in the base and the error message in the subclasses that don't support it ;-)
20090330 23:21:02< euschn> YogiHH: so that means WML events in both snapshot and replay_start are not redundant?
20090330 23:21:18-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit [Read error: 60 (Operation timed out)]
20090330 23:21:19< boucman> Crab_: describe_self could be made virtual though,
20090330 23:21:20< Crab_> mordante: imo, it is better write such an error message once (it is the same message, after all), to make derived classes simpler )
20090330 23:21:22< YogiHH> euschn: seems so, yes
20090330 23:21:26< boucman> not having it is very suspect
20090330 23:21:28< Crab_> boucman: yes
20090330 23:22:55< mordante> Crab_, I use both methods, depending on the case
20090330 23:23:58-!- andrey22 [n=Andrey@82.196.90.17] has joined #wesnoth-dev
20090330 23:24:27< silene> euschn: yes, you can have more event in the snapshot (created events are added), and you can also have less events in it (consumed events are removed)
20090330 23:24:54< esr> wesbot: seen Sapient?
20090330 23:24:54< wesbot> esr: The person with the nick Sapient last spoke 18h 17m ago.  18h 17m ago person left:
20090330 23:27:23< Reisiger> Q: I linked both the lib and the include directories in MSVC, yet I'm getting the following error:
20090330 23:27:23< Reisiger> e:\projects_ext\wesnoth\trunk\src\text.hpp(23) : fatal error C1083: Cannot open include file: 'pango/pango.h': No such file or directory
20090330 23:27:40< Reisiger> Did I miss a place for a setting?
20090330 23:28:10< euschn> silene: thanks for the clarification
20090330 23:28:10-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090330 23:28:22< mordante> did you include the proper directory and do you have pango installed
20090330 23:28:38< mordante> it looks for pango.h in the subdirectory pango
20090330 23:28:54< Reisiger> got the minary for win32, altho the incluse has an odd folder structure: ...
20090330 23:29:04< Reisiger> E:\Projects_Ext\pango\pango-dev_1.24.0-1_win32\include\pango-1.0\pango
20090330 23:29:40< Reisiger> I used ..\..\..\..\pango\pango-dev_1.24.0-1_win32\include\pango-1.0\pango to have everything relative to the wesnoth folder
20090330 23:30:02< mordante> not sure whether relative paths are wise ;-)
20090330 23:30:19< mordante> but I think the patch should be ..\..\..\..\pango\pango-dev_1.24.0-1_win32\include\pango-1.0
20090330 23:30:23< mordante> path*
20090330 23:30:26< Reisiger> heh neither are absolute ones ^^
20090330 23:31:01< boucman> Sparks_: commented on your application
20090330 23:31:02< mordante> maybe I should have said, I think developing on Windows is not wise ;-)
20090330 23:31:06< boucman> time for bed now
20090330 23:31:12< boucman> mordante: :)
20090330 23:31:27< mordante> night boucman 
20090330 23:31:35< Reisiger> heh o_O
20090330 23:31:35< Reisiger> Try developping on Win7
20090330 23:31:42< Reisiger> Thanks that solved it :)
20090330 23:31:47< Sparks_> boucman: thanls , goo night :)
20090330 23:31:59< Crab_> good night boucman )
20090330 23:32:04< Sparks_> boucman: oops... thanks , good night :)
20090330 23:33:34< CIA-30> silene * r34305 /trunk/src/Makefile.am: Added missing files.
20090330 23:33:37< CIA-30> silene * r34306 /trunk/src/ai_manager.cpp: Removed some bloat.
20090330 23:34:18-!- andrey23 [n=Andrey@93.153.159.72] has quit [Read error: 110 (Connection timed out)]
20090330 23:34:36< mordante> Reisiger, you're welcome
20090330 23:34:45< mordante> I'm off night
20090330 23:35:34-!- mordante [n=mordante@wesnoth/developer/mordante] has quit ["Leaving"]
20090330 23:35:41< Polarina> <sarcasm> Remove only *some* bloat?!? What is wrong with you people? </sarcasm>
20090330 23:36:12-!- Espreon [n=espreon@wesnoth/developer/espreon] has joined #wesnoth-dev
20090330 23:38:18-!- Appleman1234 [n=Appleman@131.181.47.2] has joined #wesnoth-dev
20090330 23:39:12< Reisiger> Night everyone ;)
20090330 23:39:23< Polarina> Good night, Reisiger.
20090330 23:39:32-!- Reisiger [n=Reisiger@adsl-62-167-57-12.adslplus.ch] has quit ["Trillian (http://www.ceruleanstudios.com"]
20090330 23:40:02-!- noy [n=Noy@wesnoth/developer/noy] has quit []
20090330 23:42:47-!- boucman [n=rosen@wesnoth/developer/boucman] has quit [Remote closed the connection]
20090330 23:43:53-!- busfahrer [n=busfahre@unixboard/user/busfahrer] has joined #wesnoth-dev
20090330 23:46:20-!- silene [n=plouf@wesnoth/developer/silene] has quit ["Leaving."]
20090330 23:47:19-!- Sapien-X99 [n=sapien-x@c-71-204-10-166.hsd1.ga.comcast.net] has joined #wesnoth-dev
20090330 23:47:29< Sapien-X99> esr: thou hast summoned me?
20090330 23:47:35-!- Sapien-X99 is now known as Sapient
20090330 23:48:19< Espreon> Heh...
20090330 23:48:41-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has left #wesnoth-dev []
20090330 23:48:56< esr> Sapient: Yes, I'm haing a bit of trouble with wnliterator.  It doesn't seem to detect macro closes properly.  (I'm trying to rewrite wmlxgettext for Ivanovic.)
20090330 23:49:30< Sapient> macro closes are not considered elements
20090330 23:50:04< Sapient> maybe if you described what you are trying to do?
20090330 23:50:19< esr> Hm, then when iterating wity this code how do you recommend I recognize a multiline macro?
20090330 23:51:06 * Sapient checking
20090330 23:51:09< esr> Wghat I'm trying to do: report on all translatable strings -- including macro args -- with a representation of their contecxt.
20090330 23:52:27< Sapient> aha
20090330 23:54:33< Sapient> well you could always check item in itor.scopes for item.element.startswith("{")
20090330 23:54:58< Sapient> that will let you know if you are inside a macro scope
20090330 23:55:05-!- JW1 [n=X@c-71-57-85-168.hsd1.il.comcast.net] has joined #wesnoth-dev
20090330 23:55:48< esr> Now, my problem is getting a notification on *end* of scope.  Hold on while I pastebin...
20090330 23:56:26< esr> Here's an example file: http://rafb.net/p/rKQDBr18.html
20090330 23:57:23< esr> If you check out the prototype wmlxgettextt in data/tools and run it against this, you'll see the problem - I don't get an element or event when the multiline macro scope closes.
20090330 23:57:40-!- thespaceinvader [n=chatzill@wesnoth/artist/thespaceinvader] has quit ["ChatZilla 0.9.84 [Firefox 3.0.8/2009032609]"]
20090330 23:58:13< Sapient> it is possible that I should have made "}" an element
20090330 23:58:26< Sapient> just didn't think there would be any use for that
20090330 23:58:59< esr> This is the case where that's a problem.
20090330 23:59:11-!- Gauteamus [n=chatzill@c43C4BF51.dhcp.bluecom.no] has joined #wesnoth-dev
20090330 23:59:47< Sapient> actually the element is not '}', it is wmliterator.closeMacroType which is defined as 'end of macro'
--- Log closed Tue Mar 31 00:00:20 2009