--- Log opened Wed Jul 07 00:00:03 2010
20100707 00:05:11-!- billynux [~billy@wesnoth/developer/billynux] has quit [Remote host closed the connection]
20100707 00:06:12-!- billynux [~billy@wesnoth/developer/billynux] has joined #wesnoth-dev
20100707 00:07:32-!- elvish_sovereign [~Zoltan@pool-74-109-67-183.phlapa.east.verizon.net] has joined #wesnoth-dev
20100707 00:23:19< boucman> night all
20100707 00:23:22-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Quit: Leaving.]
20100707 00:32:59-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20100707 00:33:02-!- boucman [~rosen@wesnoth/developer/boucman] has left #wesnoth-dev []
20100707 00:37:49-!- eleazar [~eleazar@ppp-70-226-197-86.dsl.spfdil.ameritech.net] has joined #wesnoth-dev
20100707 00:58:15< alink> arff this tod_manager code is really bad :-/
20100707 01:01:16< billynux> I'm off
20100707 01:01:27< billynux> bye y'all
20100707 01:01:29-!- billynux [~billy@wesnoth/developer/billynux] has quit [Quit: Leaving]
20100707 01:03:21< CIA-87> eleazar * r43968 /trunk/data/core/ (10 files in 2 dirs): Got the dry-mountain transitions working pretty well now.
20100707 01:06:32-!- stikonas_ [~and@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20100707 01:23:10< alink> yes really really bad. For example connect::update_and_send_diff() construct a temporary tod_manager reading level data, because it knows that deep inside, tod_manager breaks the constness of level to edit current_tod. The difference in level are then send by network
20100707 01:23:28< alink> and of course zero documentation 
20100707 01:25:05< alink> and this is the kind of things painfull to test if you want to clean it
20100707 01:26:10< AI0867> 18:33 < eleazar> is "Alink" and "AI0867" the same guy? <-- no
20100707 01:26:52< alink> AI0867: I already told him 
20100707 01:26:56< alink> alink: ah ok
20100707 01:28:33< alink> AI0867: btw, stupid question but does 0867 is  a SF reference to some AI ?
20100707 01:29:04< AI0867> not yet ;)
20100707 01:29:10< alink> hehe
20100707 01:31:24< alink> Jenny I've got your number ?
20100707 01:32:52< alink> anyway, back to the ugliness 
20100707 01:45:53-!- jennyf [~jennyf@212-198-248-35.rev.numericable.fr] has joined #wesnoth-dev
20100707 01:47:07-!- jennyf [~jennyf@212-198-248-35.rev.numericable.fr] has quit [Client Quit]
20100707 01:50:39-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20100707 01:51:13< alink> wait, when you say "Jenny", a Jenny join ?
20100707 01:51:58< alink> wesbot: seen jennyf
20100707 01:51:58< wesbot> alink: The person with the nick jennyf 4m 51s ago they left with the message: Client Quit
20100707 02:02:32-!- King_Elendil [~King_Elen@75.143.235.104] has joined #wesnoth-dev
20100707 02:12:22-!- King_Elendil [~King_Elen@75.143.235.104] has quit [Read error: Connection reset by peer]
20100707 02:13:03-!- King_Elendil [~King_Elen@75.143.235.104] has joined #wesnoth-dev
20100707 02:13:18-!- King_Elendil [~King_Elen@75.143.235.104] has quit [Client Quit]
20100707 02:20:08< alink> pfff, it seems that you can define an alternate ToD cycle in time_area, but it doesn't take account of the starting one :-/
20100707 02:20:29 * alink is tired of this mess, gn
20100707 02:20:38-!- alink [~alink@wesnoth/developer/alink] has quit [Remote host closed the connection]
20100707 02:35:17-!- Espreon [~espreon@wesnoth/developer/espreon] has joined #wesnoth-dev
20100707 03:01:35-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20100707 03:18:27-!- Tycale_ [~Thibault@viki-network.net] has joined #wesnoth-dev
20100707 03:19:47-!- Tycale [~Thibault@viki-network.net] has quit [Quit: Disconnecting from stoned server.]
20100707 03:20:18-!- loonycyborg [~sergey@wesnoth/developer/loonycyborg] has quit [Quit: Zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz]
20100707 03:20:35-!- Tycale_ is now known as Tycale
20100707 03:20:52-!- loonybot [~loonybot@wesnoth/bot/loonybot] has quit [Remote host closed the connection]
20100707 03:21:07-!- Appleman1234 [~Appleman1@CPE-60-226-178-176.qld.bigpond.net.au] has quit [Ping timeout: 245 seconds]
20100707 03:24:48-!- Aethaeryn [~Michael@wesnoth/umc-dev/developer/aethaeryn] has joined #wesnoth-dev
20100707 03:30:21< CIA-87> espreon * r43969 /trunk/src/whiteboard/typedefs.hpp: Ran umcpropfix.
20100707 03:31:40-!- Appleman1234 [~Appleman1@CPE-60-226-178-176.qld.bigpond.net.au] has joined #wesnoth-dev
20100707 03:38:43-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20100707 03:40:09-!- Appleman1234 [~Appleman1@CPE-60-226-178-176.qld.bigpond.net.au] has quit [Ping timeout: 240 seconds]
20100707 03:42:12-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20100707 03:42:20-!- noy [~Noy@wesnoth/developer/noy] has quit [Client Quit]
20100707 03:57:36-!- Appleman1234 [~Appleman1@CPE-60-226-178-176.qld.bigpond.net.au] has joined #wesnoth-dev
20100707 03:58:12-!- Espreon [~espreon@wesnoth/developer/espreon] has quit [Quit: Lost terminal]
20100707 04:24:16-!- wesbot changed the topic of #wesnoth-dev to: 141 bugs, 282 feature requests, 15 patches | logs: http://irclogs.wesnoth.org | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com | http://imagebin.org
20100707 04:27:15-!- Gambit [~quassel@pa-67-234-73-7.dhcp.embarqhsd.net] has quit [Quit: Oh I almost forgot to ask. Did you hear of my enlistment?]
20100707 04:33:29-!- Elvish_Pillager [~eli@71-10-224-192.dhcp.oxfr.ma.charter.com] has quit [Ping timeout: 260 seconds]
20100707 05:22:06-!- ancestral [~ancestral@166.137.142.151] has joined #wesnoth-dev
20100707 05:37:58-!- elvish_sovereign [~Zoltan@pool-74-109-67-183.phlapa.east.verizon.net] has quit [Remote host closed the connection]
20100707 05:40:57-!- elvish_sovereign [~elvish_so@pool-74-109-67-183.phlapa.east.verizon.net] has joined #wesnoth-dev
20100707 05:42:27-!- espreon [~espreon@wesnoth/developer/espreon] has joined #wesnoth-dev
20100707 05:43:30-!- espreon is now known as Espreon
20100707 05:43:55-!- elvish_sovereign [~elvish_so@pool-74-109-67-183.phlapa.east.verizon.net] has quit [Client Quit]
20100707 05:50:06-!- ancestral_ [~ancestral@97-116-126-66.mpls.qwest.net] has joined #wesnoth-dev
20100707 05:50:14-!- ancestral [~ancestral@166.137.142.151] has quit [Quit: Colloquy for iPhone - http://colloquy.mobi]
20100707 05:50:15-!- ancestral_ is now known as ancestral
20100707 05:53:25-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20100707 06:08:26-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20100707 06:09:35-!- Mythological [Mythologic@77.29.7.251] has quit []
20100707 06:16:06-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20100707 06:22:20-!- Blueblaze [~Blueblaze@adsl-99-186-67-245.dsl.hstntx.sbcglobal.net] has quit [Read error: Connection reset by peer]
20100707 06:23:02-!- Blueblaze [~Blueblaze@adsl-99-186-67-245.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20100707 06:29:27-!- Blueblaze [~Blueblaze@adsl-99-186-67-245.dsl.hstntx.sbcglobal.net] has quit [Quit: Blueblaze]
20100707 06:32:48-!- Blueblaze [~Blueblaze@adsl-99-186-67-245.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20100707 06:47:57-!- lauanana [~lauanana@212-198-248-35.rev.numericable.fr] has joined #wesnoth-dev
20100707 06:49:27-!- lauanana [~lauanana@212-198-248-35.rev.numericable.fr] has quit [Client Quit]
20100707 07:10:34-!- [Relic] [~relic@99-58-54-211.lightspeed.milwwi.sbcglobal.net] has quit [Quit: Leaving]
20100707 07:18:37-!- fendrin [~fabi@wesnoth/developer/fendrin] has quit [Remote host closed the connection]
20100707 07:42:13< CIA-87> gabba * r43970 /trunk/data/scenario-test.cfg: Removed annoying error test case and dialog popup on debug mode from the test scenario. Some of use actually start wesnoth with './wesnoth --debug -t' ...
20100707 07:42:20< CIA-87> gabba * r43971 /trunk/src/whiteboard/ (side_actions.cpp side_actions.hpp): Whiteboard: reorganised methods in side_actions
20100707 07:42:41< CIA-87> gabba * r43972 /trunk/src/whiteboard/validate_visitor.cpp: Whiteboard: proper invalidation of hexes to update numbers correctly.
20100707 07:42:42< CIA-87> gabba * r43973 /trunk/src/ (8 files in 2 dirs): Whiteboard: Actions can be bumped up and down the queue with the 1 and 2 keys. Hotkeys are temporary, but let me know how it works on non-qwerty keyboards. Also, reworked the way highlighting is done.
20100707 07:42:44< CIA-87> gabba * r43974 /trunk/src/whiteboard/ (6 files): Whiteboard: various tweaks and moving code around.
20100707 07:42:48< CIA-87> gabba * r43975 /trunk/data/core/hotkeys.cfg: Whiteboard: changed keys to move actions up and down the queue to pgup and pgdown, since 1, 2 were taken but not listed in hotkeys.cfg
20100707 08:12:16-!- zookeeper [~l@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20100707 08:18:25-!- crimson_penguin [~ben@wesnoth/developer/crimsonpenguin] has quit [Quit: crimson_penguin]
20100707 08:21:12-!- shadowm_laptop [ignacio@wesnoth/developer/shadowmaster] has quit [Quit: good night]
20100707 08:30:22-!- stikonas [~and@wesnoth/translator/stikonas] has joined #wesnoth-dev
20100707 08:37:24-!- timotei [~Timotei@wesnoth/developer/timotei] has joined #wesnoth-dev
20100707 08:37:32< timotei> yey ! internet
20100707 08:37:43< Espreon> Yay...
20100707 08:37:58< timotei> stupid internet providers
20100707 08:38:27< timotei> and of course, I've got another 2 spam mails:(
20100707 08:40:00< Espreon> Even though I get some spam, it almost never touches my inbox.
20100707 08:41:35< timotei> zookeeper: what's your timezone?
20100707 08:44:16< timotei> hmm... the Unit Wizard is the most ... "advanced" wizard in terms of GUI things
20100707 08:45:55< zookeeper> timotei, gmt+2
20100707 08:46:17< timotei> no wonder everytime I'm online you're here and can answer my question
20100707 08:46:20< timotei> I'm in GMT+2 too
20100707 08:46:24< timotei> :)
20100707 08:49:13< zookeeper> right
20100707 08:51:41-!- Unnheulu [~ieuan@cpc5-pnth2-0-0-cust800.5-2.cable.virginmedia.com] has joined #wesnoth-dev
20100707 09:04:32-!- timotei21 [~Timotei@193.34.191.4] has joined #wesnoth-dev
20100707 09:05:33< CIA-87> gabba * r43976 /trunk/src/whiteboard/ (highlight_visitor.cpp highlight_visitor.hpp): Whiteboard: adjustment to highlighting
20100707 09:05:39< CIA-87> gabba * r43977 /trunk/src/whiteboard/ (action.hpp typedefs.hpp): Whiteboard: added a log domain for the highlight visitor.
20100707 09:05:46< CIA-87> gabba * r43978 /trunk/src/whiteboard/ (highlight_visitor.cpp highlight_visitor.hpp typedefs.hpp): Whiteboard: changed to weak pointers in the highlight visitor to avoid keeping the move alive after its deletion from the queue.
20100707 09:08:16-!- timotei [~Timotei@wesnoth/developer/timotei] has quit [Ping timeout: 276 seconds]
20100707 09:13:55-!- stikonas [~and@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20100707 09:14:47-!- Aethaeryn [~Michael@wesnoth/umc-dev/developer/aethaeryn] has quit [Quit: ...]
20100707 09:16:14-!- Unnheulu [~ieuan@cpc5-pnth2-0-0-cust800.5-2.cable.virginmedia.com] has quit [Quit: Ex-Chat]
20100707 09:21:34-!- ancestral [~ancestral@97-116-126-66.mpls.qwest.net] has quit [Quit: And that’s the end of THAT chapter.]
20100707 09:23:18-!- Appleman1234 [~Appleman1@CPE-60-226-178-176.qld.bigpond.net.au] has quit [Remote host closed the connection]
20100707 09:24:36-!- timotei21 is now known as timotei
20100707 09:24:54-!- timotei [~Timotei@193.34.191.4] has quit [Changing host]
20100707 09:24:54-!- timotei [~Timotei@wesnoth/developer/timotei] has joined #wesnoth-dev
20100707 09:37:03-!- Espreon [~espreon@wesnoth/developer/espreon] has quit [Ping timeout: 265 seconds]
20100707 09:47:13-!- Mythological [Mythologic@77.29.93.157] has joined #wesnoth-dev
20100707 10:22:06< timotei> zookeeper: [+units] shouldn't have before a [units] declared?
20100707 10:22:14< timotei> zookeeper: so we have something to append to
20100707 10:24:16-!- wesbot changed the topic of #wesnoth-dev to: 142 bugs, 282 feature requests, 15 patches | logs: http://irclogs.wesnoth.org | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com | http://imagebin.org
20100707 10:24:42< zookeeper> timotei, the core data has it
20100707 10:24:48< zookeeper> all add-ons just append to it
20100707 10:25:26< timotei> well, I've searched for "[units]" in the data/ directory and didn't find it
20100707 10:25:33< timotei> that's why I was curious
20100707 10:25:38< zookeeper> hmh
20100707 10:26:09< zookeeper> well, the one i was thinking would be in data/core/units.cfg, but it seems [+units] is used there as well. my guess would be that you'd get the same behaviour by changing that to [units] there.
20100707 10:26:40< zookeeper> probably they're interchangeable if no such tag has been encountered already
20100707 10:28:11-!- Valkier [~IceChat7@c-174-55-104-2.hsd1.pa.comcast.net] has quit [Quit: REALITY.SYS Corrupted: Re-boot universe? (Y/N/Q)]
20100707 10:31:03-!- mjs-de [~mjs-de@vpw.wh.uni-dortmund.de] has joined #wesnoth-dev
20100707 10:37:56< timotei> ok
20100707 10:38:01-!- euschn [~euschn@wesnoth/developer/euschn] has joined #wesnoth-dev
20100707 10:43:05-!- eleazar_ [~eleazar@ppp-70-226-197-86.dsl.spfdil.ameritech.net] has joined #wesnoth-dev
20100707 10:43:05-!- eleazar [~eleazar@ppp-70-226-197-86.dsl.spfdil.ameritech.net] has quit [Read error: Connection reset by peer]
20100707 10:43:06-!- eleazar_ is now known as eleazar
20100707 10:44:12-!- eleazar_ [~eleazar@ppp-70-226-197-86.dsl.spfdil.ameritech.net] has joined #wesnoth-dev
20100707 10:44:12-!- eleazar [~eleazar@ppp-70-226-197-86.dsl.spfdil.ameritech.net] has quit [Read error: Connection reset by peer]
20100707 10:44:12-!- eleazar_ is now known as eleazar
20100707 10:47:43-!- fendrin [~fabi@wesnoth/developer/fendrin] has joined #wesnoth-dev
20100707 11:02:21< CIA-87> gabba * r43979 /trunk/src/display.hpp: Moved arrow typedefs in display.hpp to private visibility.
20100707 11:10:12-!- gabba [~gabba@wesnoth/developer/gabba] has quit [Quit: Leaving.]
20100707 11:19:05-!- Blarumyrran [~Blarumyrr@unaffiliated/blarumyrran] has joined #wesnoth-dev
20100707 11:24:08-!- loonybot [~loonybot@ppp79-139-137-245.pppoe.spdop.ru] has joined #wesnoth-dev
20100707 11:24:08-!- loonybot [~loonybot@ppp79-139-137-245.pppoe.spdop.ru] has quit [Changing host]
20100707 11:24:08-!- loonybot [~loonybot@wesnoth/bot/loonybot] has joined #wesnoth-dev
20100707 11:25:02-!- loonycyborg [~sergey@wesnoth/developer/loonycyborg] has joined #wesnoth-dev
20100707 11:42:41-!- dtiger [~dtiger@dynamic-vpdn-91-149-132-81.telecom.by] has joined #wesnoth-dev
20100707 11:52:01-!- mjs-de [~mjs-de@vpw.wh.uni-dortmund.de] has quit [Quit: On the road again]
20100707 11:53:23-!- Blueblaze [~Blueblaze@adsl-99-186-67-245.dsl.hstntx.sbcglobal.net] has quit [Quit: Blueblaze]
20100707 12:06:48-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20100707 12:12:02-!- euschn [~euschn@wesnoth/developer/euschn] has quit [Quit: Leaving.]
20100707 12:13:49< timotei>  zookeeper in the [race], there are: name and female_name, male_name. is there any case where it needs all 3 to be written?
20100707 12:14:09< timotei> zookeeper: for e.g. when it's neither female/male - I'm wondering if that's possible or not
20100707 12:14:28< timotei> zookeeper: I'm thinking of disabling writing in female/male textboxex when you put a "name"
20100707 12:17:06< zookeeper> timotei, if the wiki doesn't specify, then i don't know
20100707 12:28:49< timotei> fendrin: hi
20100707 12:55:34< timotei> well ok, I'll go that way then
20100707 12:56:54< fendrin> timotei: hi
20100707 12:57:08< timotei> fendrin: just in time. ^
20100707 12:57:32< fendrin> timotei: Do you know that there is a midterm something in the near future?
20100707 12:57:41< timotei> yes.
20100707 12:57:48< fendrin> I will have to write some text about how you are doing.
20100707 12:58:03< timotei> yeah, and I'll have to complete a survey or something like that
20100707 12:58:11< timotei> I think it's between 12 and 16
20100707 12:59:36< timotei> do you need me to do anything else than what's in the schedule of the proposal?
20100707 12:59:46< timotei> or maybe extra/reorder of the things?
20100707 13:02:36< fendrin> timotei: Maybe you can mark what goals have been reached and which not. In a first step.
20100707 13:03:46< fendrin> I have time from now until Monday to look at your code in detail and test out things that are already working.
20100707 13:04:03< timotei> ok
20100707 13:04:20< fendrin> I would like you to mark what has been reached and is good enough before I do so.
20100707 13:06:32-!- Netsplit *.net <-> *.split quits: ABCD
20100707 13:09:35-!- ABCD [~abcd@gentoo/developer/abcd] has joined #wesnoth-dev
20100707 13:10:01< freim> is trunk usable again now or is it still dead slow?
20100707 13:13:13< timotei> fendrin: ok, marked in the wiki the things done/in-progress
20100707 13:13:33< timotei> will you be online today whole day? (till 21 PM)?
20100707 13:13:56< timotei> I have some quiestions about the frontend for the wmltools, but I would better ask you later as now I'm working on the unit wizard
20100707 13:14:02< zookeeper> freim, still dead slow, i think
20100707 13:15:33-!- Mythological [Mythologic@77.29.93.157] has quit []
20100707 13:23:10< freim> zookeeper: so, nothing is being done about it?
20100707 13:33:25-!- alink [~alink@wesnoth/developer/alink] has joined #wesnoth-dev
20100707 13:40:35< zookeeper> freim, i'm sure something is, but i don't really know what
20100707 13:40:44< zookeeper> (or when)
20100707 13:41:27< zookeeper> alink, what's the plan for getting rid of the cache generation slowness?
20100707 13:41:59< alink> zookeeper: AFAIK reduce or optimize macro usage
20100707 13:42:49< zookeeper> ok...any idea if anyone's gonna actually do that anytime soon?
20100707 13:43:05< alink> reduce is for boucman, and optimize need someone skilled enough in c++ optimization and wesnoth cache/WML stuff, so I see only silene for now
20100707 13:43:51< alink> maybe Sirp too if he likes an optimization challenge
20100707 13:47:08< alink> my hack was maybe temporary acceptable on dev-trunk (and revert it before release), but such temporary hack just hides the problem and nobody will fix it
20100707 13:50:37< alink> timotei: race "name" is a fallback when gender names are missing, so disable them when "name" is there seems to not reflect well what the code is doing
20100707 13:50:51< timotei> alink: thanks
20100707 13:51:23< alink> in fact, plural_name should fall-back to it too, but it doesn't for now
20100707 13:51:30< timotei> well...
20100707 13:51:35< timotei> maybe yes
20100707 13:52:22< alink> also missing id or plural_name use "name" in the error, which is a bit silly without controlling if it's there
20100707 13:52:55-!- Gambit [~quassel@pa-67-234-73-7.dhcp.embarqhsd.net] has joined #wesnoth-dev
20100707 13:53:01< alink> probably remains of code's evolution
20100707 14:06:16< timotei> hmm...
20100707 14:06:34< timotei> the global [trait] in [units] isn't a single one piece?
20100707 14:06:41-!- Ken_Oh [~briang@static-71-178-174-220.washdc.fios.verizon.net] has joined #wesnoth-dev
20100707 14:07:07< timotei> since: All races with the attribute ignore_global_traits=no will have this trait
20100707 14:07:12< timotei> zookeeper, alink: ^
20100707 14:10:10< alink> no idea, not even sure to understand the question, but i see "ignore_race_traits" too in unit_type
20100707 14:10:30< alink> (having the same effect)
20100707 14:11:35< timotei> I'm saying... one piece, then it's a single [trait] ... [/trait] thingy
20100707 14:11:38< timotei> not multiples
20100707 14:13:03< fendrin> timotei: I am away from 18 till ~23 
20100707 14:13:17< alink> I see each trait in individual [trait][/trait] in macros/traits.cfg
20100707 14:13:32< timotei> fendrin: it's just 14.13 at you now, right?
20100707 14:13:36< fendrin> timotei: My sister celebrates her birthday by watching that awful soccer match between Germany and Spain.
20100707 14:13:46< timotei> fendrin: there is no problem :D
20100707 14:13:52< timotei> I still have to work on the unit wizard
20100707 14:13:54< fendrin> timotei: Right, that is my local time.
20100707 14:14:00< timotei> it's like the most uber wizard done until now :))
20100707 14:14:10< timotei> it has many sub-tags
20100707 14:14:25< timotei> alink: yeah, that's another thing (I think)
20100707 14:14:25< fendrin> timotei: Take your time. It's better to have one thing done finished and right than 10 things half.
20100707 14:15:36< timotei> alink: but you're right
20100707 14:15:40< timotei> it's still multiple
20100707 14:15:46< timotei> then I didn't understand the wiki too good
20100707 14:15:52< timotei> let's undo things then
20100707 14:17:03< alink> timotei: I suggest reading c++ code instead of wiki. After that, if there is imprecision, improving the wiki could be very nice
20100707 14:17:22< timotei> well... yeah
20100707 14:18:10< timotei> :(( my undo is screwed up
20100707 14:18:16< alink> but, unfortunately trait code is very complex
20100707 14:18:57< timotei> alink: where's that code in the source files?
20100707 14:19:43< timotei> I was searching in src/serialization/preprocessor.cpp everytime
20100707 14:20:15< alink> grepping "trait" with quotes will show most use of the "trait" key
20100707 14:20:25< alink> s/use/uses
20100707 14:20:30< timotei> I don't know why my grep doesn't work as needed:|
20100707 14:20:50< timotei> tried to search in data/core for things (e.g. "[trait]", but incorrect result appeared 
20100707 14:21:43< alink> grepping "trait" with quotes will show most use of the "trait" key <- in c++ code
20100707 14:21:59< alink> timotei: maybe a regex thing?
20100707 14:22:06< timotei> oh damn it
20100707 14:22:09< timotei> you're right
20100707 14:22:21< timotei> haha, didn't figure it out:|
20100707 14:22:25< fendrin> timotei: I am not sure if it is a good idea to hardcode every wizard.
20100707 14:22:50< timotei> fendrin: well, some things I tried to generalize
20100707 14:23:13< timotei> fendrin: do you have an example of what shouldn't be?
20100707 14:23:21< timotei> fendrin: or you're talking in general?
20100707 14:23:35< fendrin> timotei: I was thinking about a class that generates wizards out of the schema.cfg. 
20100707 14:23:49< timotei> fendrin: there is extra/custom/different things that need to be bound between
20100707 14:24:37< fendrin> timotei: Yes, we can think about enhancing the schema.cfg to trigger that extra stuff. 
20100707 14:24:48< timotei> hmm
20100707 14:25:21< fendrin> Or you can derive the generic class (or use it in another way) to extend it with the extras.
20100707 14:26:09< timotei> yes, there is a generic wizard class right now, that I'm using
20100707 14:27:18< timotei> and the "hard"/not-so-enjoying thing about creating a class to generate the wizards is about positioning them in the ... wizard
20100707 14:27:24< timotei> the creation of the GUI
20100707 14:38:54-!- Johannes13 [~Johannes@unaffiliated/johannes13] has joined #wesnoth-dev
20100707 14:46:55< CIA-87> alink * r43980 /trunk/src/ (5 files): Remove useless indirections between tod_manager and play_controller
20100707 14:46:59< CIA-87> alink * r43981 /trunk/src/ (tod_manager.cpp tod_manager.hpp): Remove one-use-only function
20100707 14:47:08< CIA-87> alink * r43982 /trunk/src/tod_manager.cpp: 
20100707 14:47:08< CIA-87> small cleaning in tod_manager
20100707 14:47:08< CIA-87> - probably fix [store_time_of_day] for a much earlier turn
20100707 14:47:08< CIA-87> - remove useless validation (we add a dummy when empty)
20100707 14:47:14< CIA-87> alink * r43983 /trunk/src/ (tod_manager.cpp tod_manager.hpp): More robust set_time_of_day (always set a valid value)
20100707 14:47:19< CIA-87> alink * r43984 /trunk/src/tod_manager.cpp: Small optimization of time_of_day_at
20100707 14:47:23< CIA-87> alink * r43985 /trunk/src/race.cpp: 
20100707 14:47:23< CIA-87> In [race] use "name" as "plural_name" if missing
20100707 14:47:24< CIA-87> (as we do for gender names and as the wiki says)
20100707 14:51:43< timotei> fendrin: I'm going to think about generating the wizards from schema.cfg
20100707 14:52:16< timotei> we'll let out the campaign out of this method as it's the only one that's "different" than the others  - it creates a project
20100707 14:56:34-!- Elvish_Pillager [~eli@71-10-224-192.dhcp.oxfr.ma.charter.com] has joined #wesnoth-dev
20100707 14:56:44< Ivanovic> moin
20100707 14:56:51< timotei> hi Ivanovic 
20100707 14:58:04< timotei> oops, I'm the only one who got this email: 	
20100707 14:58:04< timotei> software@heise.de
20100707 14:58:04< timotei> 		
20100707 14:58:04< timotei> [Wesnoth-dev] Your software is now being hosted ?
20100707 14:58:54< timotei> don't have thunderbird right now to "print preview" it
20100707 15:03:18< loonycyborg> timotei: Everyone who is subscribed to the ML got it, including me.
20100707 15:03:27< timotei> oh good
20100707 15:03:48< timotei> it isn't spam right?
20100707 15:04:33< loonycyborg> Not really spam, but I'm not sure whether it belongs on the ML..
20100707 15:05:32< loonycyborg> It seems to be autogenerated, so it's at least related to spam :P
20100707 15:05:50< timotei> :D
20100707 15:06:18< timotei> fendrin: are you here?
20100707 15:06:31< alink> damn spammers, they send one each time we release, that's several mails by year !!!
20100707 15:06:35< timotei> god....
20100707 15:06:46< timotei> just "marked" as read some spam email :((
20100707 15:07:20< timotei> Oh mein Gott!
20100707 15:08:43< alink> ah 4 of them, and only one with the error "we recently started offering version 1.8.2 (wesnoth-1.8.3-win32.exe) for download"
20100707 15:09:30< timotei> :D
20100707 15:11:11-!- wesbot [~wesbot@wesnoth/bot/wesbot] has quit [Ping timeout: 260 seconds]
20100707 15:13:24-!- wesbot [~wesbot@wesnoth/bot/wesbot] has joined #wesnoth-dev
20100707 15:15:47-!- Netsplit *.net <-> *.split quits: wesbot, loonybot
20100707 15:17:05-!- Netsplit over, joins: wesbot
20100707 15:17:49< timotei> fendrin: hmm...It may be possible to create that wizards from schema.cfg, but they won't look too "customized"
20100707 15:17:57< timotei> fendrin: the question is, 
20100707 15:18:38< timotei> fendrin: is the list on the wiki: http://wiki.wesnoth.org/SoC_Ideas_Eclipse_Plugin#Wizards the complete list of wizards that need to be done?
20100707 15:18:53< timotei> fendrin: because, if yes, I will finish the unit wizard and we're done
20100707 15:19:18< timotei> fendrin: if we need more wizards, then I could take some hours to experiment on the wizards generator
20100707 15:24:50< alink> AI0867: around? about the #warning and #error in test scenario
20100707 15:25:25-!- eleazar_ [~eleazar@ppp-70-226-197-86.dsl.spfdil.ameritech.net] has joined #wesnoth-dev
20100707 15:25:25-!- eleazar [~eleazar@ppp-70-226-197-86.dsl.spfdil.ameritech.net] has quit [Read error: Connection reset by peer]
20100707 15:25:26-!- eleazar_ is now known as eleazar
20100707 15:29:05< alink> AI0867: ah sorry, nevermind, gabba already fixed it
20100707 15:30:39< timotei> alink: I don't think that was "to fix":P
20100707 15:30:49< timotei> alink: he tested if it was working in the scenario-test
20100707 15:31:10< alink> yeah, probably, but he "fixed" the problem
20100707 15:31:16< timotei> :)
20100707 15:39:47-!- DesertPanther [~Khalid@unaffiliated/desertpanther] has joined #wesnoth-dev
20100707 15:55:27-!- loonycyborg [~sergey@wesnoth/developer/loonycyborg] has quit [Quit: I ATE'NT DEAD]
20100707 15:56:47-!- loonycyborg [~sergey@wesnoth/developer/loonycyborg] has joined #wesnoth-dev
20100707 15:58:47-!- dtiger [~dtiger@dynamic-vpdn-91-149-132-81.telecom.by] has quit [Remote host closed the connection]
20100707 16:05:21-!- e_s-iOS [~esios@pool-74-109-67-183.phlapa.east.verizon.net] has joined #wesnoth-dev
20100707 16:08:50-!- e_s-iOS [~esios@pool-74-109-67-183.phlapa.east.verizon.net] has quit [Remote host closed the connection]
20100707 16:16:54< timotei> fendrin: ping
20100707 16:22:40-!- Unnheulu [~ieuan@cpc5-pnth2-0-0-cust800.5-2.cable.virginmedia.com] has joined #wesnoth-dev
20100707 16:28:49-!- timotei21 [~Timotei@193.34.191.4] has joined #wesnoth-dev
20100707 16:31:32-!- timotei [~Timotei@wesnoth/developer/timotei] has quit [Ping timeout: 245 seconds]
20100707 16:31:33-!- timotei_ [~Timotei@193.34.191.4] has joined #wesnoth-dev
20100707 16:35:08-!- timotei21 [~Timotei@193.34.191.4] has quit [Ping timeout: 260 seconds]
20100707 16:40:17-!- timotei_ is now known as timotei
20100707 16:40:29-!- timotei [~Timotei@193.34.191.4] has quit [Changing host]
20100707 16:40:29-!- timotei [~Timotei@wesnoth/developer/timotei] has joined #wesnoth-dev
20100707 16:41:26-!- timotei21 [~Timotei@193.34.191.4] has joined #wesnoth-dev
20100707 16:41:27-!- timotei_ [~Timotei@193.34.191.4] has joined #wesnoth-dev
20100707 16:41:55-!- timotei_ [~Timotei@193.34.191.4] has quit [Client Quit]
20100707 16:42:03-!- timotei21 [~Timotei@193.34.191.4] has quit [Client Quit]
20100707 16:46:33-!- stikonas [~and@wesnoth/translator/stikonas] has joined #wesnoth-dev
20100707 16:56:14-!- stikonas [~and@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20100707 16:58:50-!- stikonas [~and@wesnoth/translator/stikonas] has joined #wesnoth-dev
20100707 17:22:07-!- crimson_penguin [~ben@wesnoth/developer/crimsonpenguin] has joined #wesnoth-dev
20100707 17:50:06-!- mjs-de [~mjs-de@vpw.wh.uni-dortmund.de] has joined #wesnoth-dev
20100707 18:04:34< CIA-87> alink * r43986 /trunk/src/ (game.cpp log.cpp log.hpp): 
20100707 18:04:34< CIA-87> --logdomains became a bit unusable to find the one you want.
20100707 18:04:34< CIA-87> so now display one domain by line and allow to specify a filter
20100707 18:04:34< CIA-87> example: "--logdomains gui" will shows all gui/.. domains
20100707 18:04:36< CIA-87> alink * r43987 /trunk/src/serialization/preprocessor.cpp: 
20100707 18:04:36< CIA-87> Better version of my temporary hack to speed cache rebuild (still almost 2x faster):
20100707 18:04:36< CIA-87> now only show the location of the error without the inclusion chain
20100707 18:04:36< CIA-87> That is "error at file.cfg:14" without "included from file1.cfg:10 included from file2.cfg:20..."
20100707 18:04:36< CIA-87> This seems to be an acceptable loss, especially if temporary.
20100707 18:04:53< alink> silene: you may want to review this ^
20100707 18:05:21< timotei> alink: you using git right?:D
20100707 18:05:30< alink> timotei: git-svn
20100707 18:05:34< timotei> yeah
20100707 18:05:44< alink> zookeeper: ^that seems ok for you
20100707 18:05:57< timotei> you're with "git-svn" and not "git svn" ?
20100707 18:06:22< alink> zookeeper: ? as a temporary hack, about the loss of inclusion chain for debugging macro
20100707 18:06:41< alink> timotei: yes I started with git-svn
20100707 18:06:53< timotei> what distro are you using?
20100707 18:07:01< alink> timotei: didn't look 'git svn' yet
20100707 18:07:09< alink> or is it the same thing ?
20100707 18:07:14< timotei> yes
20100707 18:07:16< alink> timotei: debian
20100707 18:07:32< timotei> on sabayon I had git svn instead of git-svn (or gentoo)
20100707 18:07:41< timotei> on windows too
20100707 18:08:29< alink> I knew that git command = git-command, but I was not sure about git(-)svn 
20100707 18:08:53< timotei> :-??
20100707 18:09:01< timotei> git-svn not found :D
20100707 18:09:19< alink> both are equal here, and the doc seems to say that too
20100707 18:09:45< timotei> weird then
20100707 18:10:53< alink> http://www.kernel.org/pub/software/scm/git/docs/git-svn.html
20100707 18:11:08< timotei> http://pr09studio.deviantart.com/art/My-glassy-box-170386554?q=boost%3Apopular+in%3Acustomization%2Fscreenshots+max_age%3A8h&qo=5 omg this looks so cool :O
20100707 18:11:23< alink> but my local man page don't say that
20100707 18:11:43< timotei> alink: you're right, but when tried "git-svn" on sabayon I got error, and on windows too..
20100707 18:11:51< timotei> maybe it dependes on how/what you install :)
20100707 18:12:39< alink> I never understand this love of glassy windows
20100707 18:12:57< alink> it's not real window, you know ;-p
20100707 18:13:44< timotei> well... I like it :D
20100707 18:13:53< timotei> of course too much transparency hurts
20100707 18:13:56< timotei> but it looks cute
20100707 18:14:38< alink> it's like these movie where a math guy writes on glass, as if it was usefull for him. The only point is probably to be able to see his face when he writes
20100707 18:15:08< alink> but I never tried, maybe it works ok
20100707 18:15:25< timotei> this is nice too: http://tuutis.deviantart.com/art/My-Simple-2010-Desktop-170388767?q=boost%3Apopular+in%3Acustomization%2Fscreenshots+max_age%3A8h&qo=52
20100707 18:15:34< timotei> :D
20100707 18:16:41< alink> I am not a fan of visual desktop stuff. I am more focused on keyboard/mouse shortcuts
20100707 18:17:15< timotei> that's true, I don't see my desktop too much, too
20100707 18:17:28< timotei> but when the PC starts... :D
20100707 18:17:54< timotei> btw, where are you from?
20100707 18:17:55< alink> timotei: linux box start less often ;-p
20100707 18:18:01< alink> belgium
20100707 18:18:22< alink> speaking of that, I really need to go see what's behind these transparent windows
20100707 18:18:23< alink> bbl
20100707 18:18:25-!- happygrue [~George@wesnoth/developer/wintermute] has joined #wesnoth-dev
20100707 18:18:32< timotei> what?
20100707 18:18:36< timotei> ok bye
20100707 18:21:23< fendrin> timotei: pong
20100707 18:21:47< timotei> you didn't go?:D
20100707 18:22:20-!- happygrue_ [~George@wesnoth/developer/wintermute] has joined #wesnoth-dev
20100707 18:25:19-!- happygrue [~George@wesnoth/developer/wintermute] has quit [Ping timeout: 265 seconds]
20100707 18:27:11-!- happygrue_ [~George@wesnoth/developer/wintermute] has quit [Read error: Connection reset by peer]
20100707 18:27:40-!- happygrue [~George@wesnoth/developer/wintermute] has joined #wesnoth-dev
20100707 18:31:16< zookeeper> alink, eh, it might be ok...i think the inclusion chain isn't very useful usually
20100707 18:31:44< zookeeper> if it's supposed to be temporary then i don't mind at all
20100707 18:57:22-!- elvish_sovereign [~elvish_so@pool-74-109-67-183.phlapa.east.verizon.net] has joined #wesnoth-dev
20100707 19:02:13-!- Blueblaze [~Blueblaze@adsl-99-186-67-245.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20100707 19:11:28-!- Unnheulu [~ieuan@cpc5-pnth2-0-0-cust800.5-2.cable.virginmedia.com] has quit [Quit: Ex-Chat]
20100707 19:13:24-!- Elvish_Pillager [~eli@71-10-224-192.dhcp.oxfr.ma.charter.com] has quit [Quit: Hi! I'm a quit message virus vaccine. If you see a quit message virus, don't replace your quit message with it!]
20100707 19:17:41-!- EdB [~edb@tss37-1-89-84-18-220.dsl.club-internet.fr] has joined #wesnoth-dev
20100707 19:17:48< CIA-87> eleazar * r43988 /trunk/data/core/terrain-graphics.cfg: enabled lava-to chasm transitions for non-chasm lava.
20100707 19:25:10-!- Blueblaze [~Blueblaze@adsl-99-186-67-245.dsl.hstntx.sbcglobal.net] has quit [Quit: Blueblaze]
20100707 19:28:38-!- billynux [~billy@wesnoth/developer/billynux] has joined #wesnoth-dev
20100707 19:28:40-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20100707 19:28:46-!- Blueblaze [~Blueblaze@adsl-99-186-67-245.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20100707 19:30:47-!- gabba [~gabba@wesnoth/developer/gabba] has joined #wesnoth-dev
20100707 19:41:47< alink> zookeeper: yeah, temporary because it doesn't fix the root of the problem (too much macros) and we are not back at 1.8 speed yet
20100707 19:43:18< alink> btw, to anyone who was caring about my previous patch. In fact, the bottleneck was the inclusion chain, not just one linenumber and one textdomain
20100707 19:43:48< alink> I suppose that increasing the nesting level of macro make this worse
20100707 19:44:36-!- grzywacz [~grzywacz@wesnoth/developer/grzywacz] has joined #wesnoth-dev
20100707 19:45:13< alink> so, in fact, perhaps you could use that many macros but instead reduce the nesting
20100707 19:53:24< timotei> alink: but why would the "too much macros" slow down that?
20100707 19:53:33< timotei> weren't many in 1.8 too
20100707 19:53:34< timotei> ?
20100707 19:54:22< alink> timotei: because it seems that we use several times more macros in trunk compared to 1.8
20100707 19:54:38< alink> and cache building is mainly about that
20100707 19:55:04< timotei> several times?:O
20100707 19:55:28< timotei> I wonder, why nobody saw the partial growing times, but only at once
20100707 19:55:40< timotei> you know, you could try adding this feature , maybe temporary
20100707 19:55:47< timotei> to count down the time taken for the game to load
20100707 19:55:50< timotei> or something like that
20100707 19:56:04< timotei> let's say for: ./wesnoth --perfmon
20100707 19:56:17< alink> the hack was committed r43987
20100707 19:56:39< alink> but I have another idea that I want to test, maybe to replace that
20100707 19:56:59< timotei> yeah. but that timer could help you/other better test the new loading
20100707 19:57:11< timotei> since maybe we are not aproximating good enough the time taken :D
20100707 19:57:23< timotei> disturbed by environment, other tasks, etc
20100707 19:58:18< alink> well i win 17 sec here on 40, I think it's clear enough on my box :-)
20100707 19:58:57< timotei> oh
20100707 19:59:14< alink> but I am considering using the --log-info=config flag or something, so trunk user could still get the inclusion chain when needed
20100707 19:59:24< timotei> yeah
20100707 20:05:51< eleazar> alink: compared to my compiled version from a week ago, it seems about twice as fast.
20100707 20:06:09-!- [Relic] [~relic@99-58-54-211.lightspeed.milwwi.sbcglobal.net] has joined #wesnoth-dev
20100707 20:06:18< eleazar> takes about a minute to start and get to the editor now.
20100707 20:06:36< timotei> 1 minute?:))
20100707 20:06:37< alink> eleazar: you mean r43987 is also ~2x faster on your box, good 
20100707 20:07:05< alink> (only cache rebuild)
20100707 20:07:18< [Relic]> Hello :)
20100707 20:08:14-!- e_s-iOS [~esios@pool-74-109-67-183.phlapa.east.verizon.net] has joined #wesnoth-dev
20100707 20:08:15< eleazar> yeah it was aproximately 2 min
20100707 20:09:14< alink> yeah really very long, and still too long. We really need to improve that, it's bad for developement
20100707 20:09:30< alink> less for users, since they rebuild cache much less often
20100707 20:11:25< alink> eleazar: btw I checked ToD overlay do have a cost, maybe ~10% fps drop on full screen redraw
20100707 20:11:53< eleazar> unfortunately i need to rebuild 90% of the time i start up wesnoth.  and yes it does make development drag, when i'm spending more time loading wesnoth than making pixels
20100707 20:12:00< alink> mainly because blitting half-transparent pixels is more expensive, esp. on a whole hex and on all of them
20100707 20:13:12< alink> terrain (and their transition) have less alpha-blending, it's mostly fully transparent (nothing to blit)  and opaque pixels
20100707 20:18:05< crimson_penguin> Wesnoth 2: OpenGL, and an overhaul to the way WML is loaded to be much more incremental?
20100707 20:18:46 * crimson_penguin starts rumors
20100707 20:18:58< alink> crimson_penguin: split the cache would be easier and more usefull
20100707 20:19:10< crimson_penguin> split how?
20100707 20:19:17< alink> for example a cache for units and one for terrains
20100707 20:19:22< crimson_penguin> oh, I get it
20100707 20:19:43< crimson_penguin> but.... isn't the basic "load all content on startup" a bad idea anyway?
20100707 20:20:06< crimson_penguin> all wml that is
20100707 20:20:09< alink> well we always need terrain and units stuff
20100707 20:20:31< crimson_penguin> but not UMC stuff
20100707 20:20:31< alink> what is bad is  rebuild the whole cache each time we touch a file
20100707 20:21:24< alink> yes but terrain is the big guy (for macro use), units has a lot of data to parse, but that's less cache related
20100707 20:21:59< crimson_penguin> hmm
20100707 20:23:36< alink> we already have some lazy loading code for units
20100707 20:24:13< alink> but terrain rules used by builder is a too complex problem for this kind of trick
20100707 20:26:20-!- dtiger [~dtiger@dynamic-vpdn-91-149-190-2.telecom.by] has joined #wesnoth-dev
20100707 20:33:27< crimson_penguin> well then, split the cache! :P
20100707 20:35:16< alink> Certainly not for my level, and not easy. WML writers will also need to watch which macro they use and where they are defined
20100707 20:37:59-!- stikonas_ [~and@ctv-79-132-162-160.vinita.lt] has joined #wesnoth-dev
20100707 20:37:59-!- stikonas_ [~and@ctv-79-132-162-160.vinita.lt] has quit [Changing host]
20100707 20:37:59-!- stikonas_ [~and@wesnoth/translator/stikonas] has joined #wesnoth-dev
20100707 20:38:50< timotei> I don't get it. what's the main difference between [unit] and [units]?
20100707 20:38:59< alink> s
20100707 20:39:06< alink> was easy ;p
20100707 20:39:18< timotei> [unit] defines only *one* unit(type)?
20100707 20:39:45-!- stikonas [~and@wesnoth/translator/stikonas] has quit [Ping timeout: 240 seconds]
20100707 20:39:48< alink> more seriously the first create unit, and the second define all the unit_types, I think
20100707 20:40:07< timotei> zookeeper: ^
20100707 20:41:01< timotei> zookeeper: I hope you're not irritated for asking you so much of the WML things :D
20100707 20:41:46< alink> timotei: I suspect that the wiki is a bit jealous ;-p
20100707 20:41:53< timotei> :))
20100707 20:42:19-!- Aethaeryn [~Michael@wesnoth/umc-dev/developer/aethaeryn] has joined #wesnoth-dev
20100707 20:43:11-!- stikonas_ is now known as stikonas
20100707 20:44:41< zookeeper> timotei, [unit] spawns a single unit, [units] defines all units, races, movetypes etc, [unit_type] defines a unit type
20100707 20:44:46-!- Blitzmerker [~Blitzmerk@p3EE0950B.dip0.t-ipconnect.de] has joined #wesnoth-dev
20100707 20:50:26< Blitzmerker> hi, is there a way to seperate the value of a variable? for example the value is "unit1,unit2" and I want the values unit1 in one variable and unit2 in an other.
20100707 20:51:04< alink> silene: alternatively, we could reduce the garbage in the inclusion chain, esp. for terrain macro
20100707 20:51:07< alink> core/terrain-graphics/builder.cfg 11 core/_main.cfg 5 _main.cfg 62 core/terrain-graphics/builder.cfg 11 core/_main.cfg 5 _main.cfg 66 core/terrain-graphics/internal-border.cfg 11 core/_main.cfg 5 _main.cfg 106 core/terrain-graphics/internal-complex.cfg 11 core/_main.cfg 5 _main.cfg 181 core/terrain-graphics/internal-complex.cfg 11 core/_main.cfg 5 _main.cfg 126 core/terrain-graphics/transitions.cfg 11 core/_main.cfg 5 _main.cfg 599 core/terrain-
20100707 20:51:09< alink> graphics/transitions.cfg 11 core/_main.cfg 5 _main.cfg 550 core/terrain-graphics.cfg 60 _main.cfg
20100707 20:51:28< alink> ^seems quite redundant and not very usefull
20100707 20:52:09< alink> add a "#line 4" in front
20100707 20:52:27< alink> no wonder that it's some work to parse
20100707 20:52:58< alink> maybe we should use shorter filenames ;)
20100707 20:53:14-!- e_s-iOS [~esios@pool-74-109-67-183.phlapa.east.verizon.net] has quit [Quit: Colloquy for iOS - client quit]
20100707 20:53:44< timotei> alink: maybe 1 digit/letter file names?:D
20100707 20:54:21< zookeeper> Blitzmerker, use [set_variables] [split]
20100707 20:55:28< Blitzmerker> zookeeper: thx
20100707 20:56:02-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20100707 20:56:49< alink> boucman: that probabilities list in terrain macros should indeed be done at c++ level
20100707 20:58:47< alink> which remind me that we should also benchmark builder speed
20100707 21:05:51< alink> mmh, need to be carefull, if should be done by c++ only it's very fast to generate, since it will be called more often (each map build instead of each cache build)
20100707 21:06:05< alink> s/if should /it should
20100707 21:06:38< timotei> this is weird, what happens if no "attacks" in the [unit_type] are supplied?
20100707 21:07:06< alink> the unit has no attack ?
20100707 21:07:25< timotei> attacks 	 the number of times that this unit can attack each turn. 
20100707 21:07:38< alink> default is 1
20100707 21:07:42< timotei> oh
20100707 21:07:42< timotei> ok
20100707 21:07:48< alink> max_attacks_ = cfg["attacks"].to_int(1);
20100707 21:08:02< timotei> right, I'll open the .cpp file :-)
20100707 21:08:21< timotei> indeed this is the most complex thing
20100707 21:08:23< timotei> the units
20100707 21:08:29< alink> unit_types.cpp:808 for similar stuff
20100707 21:08:36< timotei> thanks
20100707 21:26:43-!- e_s-iOS [~esios@pool-74-109-67-183.phlapa.east.verizon.net] has joined #wesnoth-dev
20100707 21:29:03-!- EdB [~edb@tss37-1-89-84-18-220.dsl.club-internet.fr] has quit [Remote host closed the connection]
20100707 21:37:12-!- e_s-iOS [~esios@pool-74-109-67-183.phlapa.east.verizon.net] has quit [Ping timeout: 265 seconds]
20100707 21:37:55-!- e_s-iOS [~esios@pool-74-109-67-183.phlapa.east.verizon.net] has joined #wesnoth-dev
20100707 21:40:26< timotei> hmm... based on this: http://wiki.wesnoth.org/UnitTypeWML 
20100707 21:41:06< timotei> there is the "movetype". but in schema.cfg / some examples (data/core/units/bats/Bat_Blood.cfg) it's actually "movement_type"
20100707 21:41:11< timotei> should I correct that in wiki?
20100707 21:42:49< zookeeper> err
20100707 21:42:57< alink> watch out that there is a key "movetype"  and a tag [movement_type]
20100707 21:43:09< zookeeper> indeed
20100707 21:43:44< timotei> hmm
20100707 21:44:08< zookeeper> there's [units] [movetype] and then there's [unit_type] movement_type=
20100707 21:44:37< zookeeper> two different things, the latter just refers to the former's name
20100707 21:44:47< alink> ah yes I had it reversed, sry
20100707 21:47:24< timotei> well... yes.
20100707 21:47:47< timotei> but on the wiki it says the movetype instead movement_type, no?
20100707 21:48:20< timotei> since there's no "movetype" key in tag [unit_types]
20100707 21:49:16< zookeeper> you mean the broken "See movetype" link on the UnitTypeWML page?
20100707 21:49:28< zookeeper> if yes, then sure, that should be fixed
20100707 21:49:53< timotei> no. I mean there should be written "movement_type" instead of movetype ?
20100707 21:50:24< timotei> I mean, just like you said
20100707 21:50:38< timotei> to reference to that movetype you write: movement_type= (id from a [movetype])
20100707 21:51:38< alink> yes, indeed the key in movetype http://wiki.wesnoth.org/UnitTypeWML seems wrong
20100707 21:51:53< alink> *the key "movetype" in
20100707 21:52:30< alink> mmh I wonder why we don't use the same word
20100707 21:52:53< timotei> don't ask me :-)
20100707 21:53:20< alink> there is a "race" key and a [race] tag already working like that
20100707 21:57:09< timotei> alink: will you fix that on the wiki?
20100707 21:57:16< timotei> Or I should fix it?
20100707 21:58:01< alink> timotei: go ahead, you spotted it :)
20100707 21:58:04< timotei> ok
20100707 21:59:09< alink> maybe also fix the wiki link movetype to point http://wiki.wesnoth.org/UnitsWML
20100707 21:59:17< timotei> yeah
20100707 22:00:37< timotei> ok, done
20100707 22:00:47< alink> timotei: thanks
20100707 22:00:50< timotei> np
20100707 22:01:05< timotei> I can't wait till we generate the wiki from C++ code :D
20100707 22:01:19< timotei> then, nothing will be "weird" or wrong
20100707 22:01:48< timotei> pff, it's already time to go and I can't belive I haven't finished this damn unit wizard :))
20100707 22:05:31< billynux> timotei, I'm pretty stuck myself too :(
20100707 22:05:51< timotei> billynux: hey, remember the messages on the mailling list?:)
20100707 22:06:06< timotei> billynux: when other guys/girls feeld puzzled or things like this?:D
20100707 22:06:11< billynux> GSoC ML?
20100707 22:06:13< billynux> yeah
20100707 22:06:14< timotei> yeah
20100707 22:06:31< timotei> you remember the answer: they picked us because we were good enough to do this ;)
20100707 22:06:41< timotei> so, we'll end this somehow:D
20100707 22:06:50< billynux> Well, its natural... I'm gonna have nightmares about the current implementation of the network module :)
20100707 22:06:55< timotei> :))
20100707 22:07:22< billynux> timotei, I know I will, I couldn't stand not doing so. However, I'm pretty worried about revisions and timeline
20100707 22:07:23< timotei> well, I'm a bit under schedule :-S... and that's badder than yours
20100707 22:07:47-!- Upthorn [~ogmar@adsl-75-26-188-107.dsl.scrm01.sbcglobal.net] has quit [Ping timeout: 245 seconds]
20100707 22:08:09< billynux> I am too, even though I've added a lot of code, and the ana library is working quite good
20100707 22:08:24< billynux> The game still hasn't suffered any (functional)  changes :(
20100707 22:08:53< timotei> well, till the end I don't think it has to
20100707 22:09:06< timotei> the end of a work brings the happiness :D
20100707 22:09:18< timotei> ok, I think I'm going now to read something :D
20100707 22:09:25< timotei> good night alink zookeeper billynux 
20100707 22:09:27< timotei> goodluck
20100707 22:09:34< alink> timotei: gn
20100707 22:09:34< billynux> maybe so... but I would surely like to be testing a first complete implementation ATM
20100707 22:09:40< billynux> good night timotei, take care
20100707 22:09:45< timotei> billynux: u too
20100707 22:09:54< billynux> may bugs and obscure code stay clear of you :)
20100707 22:09:57-!- timotei [~Timotei@wesnoth/developer/timotei] has quit [Quit: Leaving]
20100707 22:10:00< CIA-87> alink * r43989 /trunk/src/ (4 files): Disable ToD color shift for [time_area] (where a different ToD is used)
20100707 22:10:23< alink> eleazar: ^done but, as expected, looks weird with terrain transitions :-/
20100707 22:11:00< eleazar> i guess i have to compile again.
20100707 22:11:05< eleazar> ;)
20100707 22:11:10< eleazar> thanks, i'll check it out
20100707 22:11:40< alink> btw the underlying code is in bad shape, it will take me sometimes to even have a prototype :-/
20100707 22:11:59< alink> for color-shift area I mean
20100707 22:12:06< eleazar> understand
20100707 22:12:48< alink> the good news is that it motivates me to clean this ugliness :)
20100707 22:13:33< alink> bbl
20100707 22:15:07-!- Blueblaze_ [~Blueblaze@adsl-99-186-66-240.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20100707 22:17:34-!- Blueblaze [~Blueblaze@adsl-99-186-67-245.dsl.hstntx.sbcglobal.net] has quit [Ping timeout: 276 seconds]
20100707 22:17:35-!- Blueblaze_ is now known as Blueblaze
20100707 22:18:32-!- Aethaeryn is now known as `\
20100707 22:19:27-!- `\ is now known as Aethaeryn
20100707 22:20:29-!- [Relic] [~relic@99-58-54-211.lightspeed.milwwi.sbcglobal.net] has quit [Quit: Leaving]
20100707 22:22:18-!- dtiger [~dtiger@dynamic-vpdn-91-149-190-2.telecom.by] has quit [Remote host closed the connection]
20100707 22:24:19-!- Crab_ [~Crab_@wesnoth/developer/crab] has joined #wesnoth-dev
20100707 22:24:29-!- e_s-iOS [~esios@pool-74-109-67-183.phlapa.east.verizon.net] has quit [Quit: Colloquy for iOS - client quit]
20100707 22:26:30-!- Ken_Oh [~briang@static-71-178-174-220.washdc.fios.verizon.net] has quit [Read error: Connection reset by peer]
20100707 22:30:38-!- grzywacz [~grzywacz@wesnoth/developer/grzywacz] has quit [Read error: Connection reset by peer]
20100707 22:30:42-!- grzywa [~grzywacz@chello089078180180.chello.pl] has joined #wesnoth-dev
20100707 22:32:08-!- ettin [~jorda@wesnoth/developer/ettin] has quit [Ping timeout: 240 seconds]
20100707 22:33:28-!- ettin [~jorda@wesnoth/developer/ettin] has joined #wesnoth-dev
20100707 22:35:14-!- Aethaeryn is now known as {`B|u3-b|^z3_}
20100707 22:42:02-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20100707 22:43:23-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20100707 22:43:29< boucman> hey all
20100707 22:43:37< Crab_> hi, boucman
20100707 22:43:53-!- {`B|u3-b|^z3_} is now known as Aethaeryn
20100707 22:44:24< eleazar> boucman: hi
20100707 22:45:31< boucman> alink: around ?
20100707 22:47:09< alink> boucman: yep
20100707 22:47:37< CIA-87> eleazar * r43990 /trunk/ (8 files in 3 dirs): added default transitions for castles to lava chasms.
20100707 22:47:59< boucman> alink: do you know the preprocessor error handling well ?
20100707 22:48:41< alink> boucman: nope, it's a bit mysterious, I just explored a little for my today's speed hack
20100707 22:48:54< boucman> mkay...
20100707 22:49:29< alink> which may even break something there, or at least make it log less info
20100707 22:49:51< boucman> because it seems generating error messages is what is taking most time, and most of the messages won't be used, so I was wondering if it would be possible to only generate them on errors, i.e when destacking the macro stack after a parsing errors
20100707 22:50:05< boucman> but I might not understand all the ins and outs of the macro system
20100707 22:51:42< alink> I am not sure that "destacking the macro stack after a parsing errors" is easily doable. And if doable, you probably need to store info for this, which risk to cause the same overhead problem
20100707 22:52:20< alink> but i don't really know what i am talking about
20100707 22:52:29< boucman> well no, the call locations are implicitely stored in the call stack, it's the building of the string that is long, but I guess Soliton is the one to ask...
20100707 22:53:44< boucman> btw, hello gabba :)
20100707 22:53:46< alink> AFAIK the bottleneck is the parsing of that string
20100707 22:53:55< gabba> hello boucman :)
20100707 22:54:12-!- shadowm_laptop [ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20100707 22:54:16< billynux> boucman, hi. I'm stuck after receiving the first message from the official server. Somehow the code is not sending what it's supposed to. Any help?
20100707 22:54:34< boucman> alink: I call much more macros than before, which explains the slowness, but my gut feeling is that error message creation shouldn't use 65% of the macro parsing time
20100707 22:54:35< billynux> I know... not a lot of info there
20100707 22:54:42< Crab_> billynux: have you read norbert's doc ?
20100707 22:54:49< boucman> not much indeed :P
20100707 22:55:06< billynux> Crab_, no, where is it?
20100707 22:55:21< boucman> billynux: did you try netcat to spy between the official server and official client to check you send the right content ?
20100707 22:55:38< billynux> status is: I connect to the official server, I get my ID, and receive the first 33 bytes message.
20100707 22:55:52< billynux> boucman, I'm using wireshark
20100707 22:56:18< billynux> And, yes... up to the part it gets stuck everything goes the same way as in the SDL implementation
20100707 22:56:24< boucman> billynux: ok, that works too
20100707 22:56:32-!- Blueblaze [~Blueblaze@adsl-99-186-66-240.dsl.hstntx.sbcglobal.net] has quit [Ping timeout: 245 seconds]
20100707 22:56:44< billynux> I construct the config using read_gz directly
20100707 22:56:55< billynux> which leaves me a confg that is both valid and non empty
20100707 22:57:06< boucman> you send the same thing, and the real client doesn't send any sort of ack at that point ?
20100707 22:57:06< billynux> (but I can't really tell if the info loaded fine)
20100707 22:57:30< billynux> the "same thing" ?
20100707 22:57:42< boucman> the same content...
20100707 22:57:49< billynux> yes
20100707 22:57:54< boucman> you could use a fix file with raw data to check... 
20100707 22:57:56< boucman> hmm
20100707 22:58:03< billynux> The problem is: in the SDLnet implementation after I receive this 33 byte msg, the game sends something back
20100707 22:58:08< gabba> boucman: I'm leaving in 5 min unfortunately. But current plans are to either get attack planning and execution in -- more infrastructure work is probably required to get there.
20100707 22:58:15< boucman> I don't remember any ACK mechanism in the protocol...
20100707 22:58:16< billynux> but in my code there is no such call (to send anythin)
20100707 22:58:36< boucman> gabba: ok, sounds good, I have little time to test tonight anyway
20100707 22:58:40< gabba> ^s/are to either get/are to get
20100707 22:58:45< gabba> ok
20100707 22:58:56< billynux> boucman, It's this: 1- Send 0, 2- Receive 32bit ID in network byte order, 3- Receive a normal message
20100707 22:59:11< gabba> see you all
20100707 22:59:13-!- gabba [~gabba@wesnoth/developer/gabba] has left #wesnoth-dev []
20100707 22:59:23< billynux> a normal message is preceded by a 32bit int (also in network byte order) indicating how many bytes are coming
20100707 22:59:32< boucman> ok, so the server seems to be waiting for a reply from you, and you don't know what that response is, isn't it ?
20100707 22:59:33< billynux> up to this point, all is good
20100707 22:59:43< boucman> wouldn't it be the encoded password or something ?
20100707 22:59:45< Crab_> billynux: http://wiki.wesnoth.org/MultiplayerServerWML#The_login_procedure and http://forum.wesnoth.org/viewtopic.php?f=15&t=29840&p=436824 for the C code
20100707 23:00:36< billynux> Thanks Crab_ 
20100707 23:00:51< Crab_> billynux: maybe you'll be able to reverse engineer the login procedure details from his C code.
20100707 23:00:55< billynux> boucman, yes, but in the SDLnet version I send something after this
20100707 23:01:05< billynux> in my code I'm not being called to send anything
20100707 23:01:29< billynux> Crab_, I'll look at it, I was looking for some literature on this. How did I miss it?
20100707 23:01:30< boucman> so SDLNET is ->login <-33bytes ->something <-more data
20100707 23:01:44< Crab_> billynux: he's not using SDL, he's using low-level sockets api, so 'what is going on' is simpler.
20100707 23:01:44< boucman> and your version is ->login <-33bytes blocked
20100707 23:01:52< boucman> if I understand correctly
20100707 23:02:05< billynux> boucman, yes
20100707 23:02:30< billynux> actually those 33 bytes is 37 (4 bytes first with the number 33)
20100707 23:02:36< boucman> hmm, so why don't you send the "something" because you don't know what it is ? because nothing in the 33bytes told you to ?
20100707 23:02:37< billynux> but otherwise, yes
20100707 23:02:57-!- stikonas_ [~and@wesnoth/translator/stikonas] has joined #wesnoth-dev
20100707 23:03:02< billynux> boucman, it's a good point
20100707 23:03:17< billynux> boucman, two answers: 1- I misread the config
20100707 23:03:38< billynux> (so the code that was to interpret it didn't make sense of it and didn't decide to send anything)
20100707 23:03:58< boucman> cool, I'm happy I helped :P
20100707 23:03:59< billynux> or, 2- In the network implementation I am supposed to send something after reading this first message
20100707 23:04:06-!- stikonas [~and@wesnoth/translator/stikonas] has quit [Ping timeout: 240 seconds]
20100707 23:04:27< billynux> (and since mine is a new network implementation I didn't contemplate this)
20100707 23:05:27< alink> boucman: I don't see "error message creation" as the problem. It's more one part of the code passing location info to another via WML directives like #line, which need to be parsed back into c++ data
20100707 23:05:44< billynux> boucman,  looking at Crab_'s links: version: The client's version string. ( <- maybe I need to send my version info)
20100707 23:06:46< boucman> alink: I don't have the code opened, I'm working from memory, but IIRC, the string built at the critical place was the one that ends on stdout in case of errors... though the parsing of line might be the costly part and not the string concatenation
20100707 23:07:30< billynux> wesbot, seen norbert
20100707 23:07:30< wesbot> billynux: Sorry, I don't know of norbert.
20100707 23:07:33< billynux> wesbot, seen norbert_
20100707 23:07:34< wesbot> billynux: The person with the nick norbert_ last spoke 1d 2h ago.  1d 2h ago they left with the message: Quit: Leaving
20100707 23:08:03< Soliton> billynux: http://wiki.wesnoth.org/MultiplayerServerWML  did you look there already?
20100707 23:08:39< billynux> Soliton, no, it's the first time I'm hearing of these links
20100707 23:08:51< alink> boucman: yes the bottleneck is skip_comment() which read all the #stuff. But my last change shows that most of what parsed there is generated by us, not by real WML data
20100707 23:08:54< billynux> Soliton, Crab_ : and they are just what I was looking for
20100707 23:09:05< boucman> mkay... 
20100707 23:09:48< alink> boucman: we generate huge "#line ..." for bookkeeping purposes
20100707 23:10:52< alink> boucman: and terrain macro is also a weird case for that, not sure why. I already posted an example earlier today
20100707 23:12:28< boucman> alink: don't have much time now, we'll try to look at it, the three of us with sapien whenever we meed
20100707 23:13:57< alink> boucman: ok
20100707 23:14:38< alink> what i mean in my last sentence is that there is some redundancy there, which seems even worse with terrain macro
20100707 23:15:20< alink> example http://wesnoth.pastebin.com/1GeVutKi
20100707 23:16:34< alink> (note that trunk now stopped doing that)
20100707 23:18:06< billynux> So, narrowing down my problem. I'm not sending my version information. Question: The request to send this info comes from the current SDLnet code or from some operator above that will issue a network::send_data( ... )
20100707 23:18:42< billynux> or... anyone know where this request to send the info comes from?  (file)
20100707 23:21:40< Crab_> multiplayer.cpp:184 ?
20100707 23:22:14< Crab_> there's if(data.child("version")) { ... cfg["version"] = game_config::version; ... network::send_data(res, 0, true); } in there, if that's what you're looking for
20100707 23:22:31< Crab_> so, sending the version is triggered by the server
20100707 23:23:12< alink> does that means that we have version stats for 1.8.x servers ?
20100707 23:23:24< alink> Soliton: do we have those^
20100707 23:23:25< Crab_> ... as the server can send us several different things, like a redirect, or a login/password/salt request, or a whisper (that's later)
20100707 23:23:51< Soliton> alink: i can produce version stats from the logs, yes.
20100707 23:24:17< billynux> Crab_, I'm interested in where the client sends its version info (looking at mp.cpp)
20100707 23:24:27< Crab_> multiplayer.cpp:184
20100707 23:25:32< Crab_> in case you have different line numbers, it's http://wesnoth.pastebin.com/Uaeju9my
20100707 23:25:33< billynux> Crab_, good, so, Apparently in my version the code doesn't reach line 186 -> network::send_data(res, 0, true);
20100707 23:25:43< alink> Soliton: ok, cool, I was curious how many were too lazy to update 1.8.x
20100707 23:26:03< alink> because of some forum thread, I think
20100707 23:26:08< Crab_> billynux: does it reach if(data.child("version")) { ?
20100707 23:26:18< Soliton> just the current session: http://yourpaste.net/5744/
20100707 23:26:45< alink> Soliton: thanks
20100707 23:26:48< Crab_> billynux: have you looked at the code of the bot ?
20100707 23:27:06< billynux> Crab_, not yet
20100707 23:27:10< alink> oh there is still RC and beta :-(
20100707 23:27:34< billynux> Crab_, let me check if it reaches that line
20100707 23:27:39< Crab_> billynux: take a look, then, v. 0.4 from that forum topic, last message.
20100707 23:27:58< alink> Soliton: are we supposed to allow betaN to join official server ?
20100707 23:28:04< Blitzmerker> How can I get the Image of a unit in WML?
20100707 23:28:23< boucman> alink: some of them, yes...
20100707 23:29:09< Crab_> billynux: you'll see that the version is sent only after the server has asked for it.
20100707 23:29:11< alink> at least, i am glad that the majority use 1.8.2
20100707 23:29:20< Soliton> alink: good question, only since 1.7.14 actually.
20100707 23:29:23< Crab_> billynux: in your wireshark dumps, do you see the server asking you for [version] ?
20100707 23:29:31< Soliton> i need to make the settings a bit stricter.
20100707 23:29:58< Soliton> (they got in because of *1.8* being allowed.)
20100707 23:30:02< billynux> Crab_, Its binary (so I don't really know). But it's the same message the SDLnet implementation gets
20100707 23:30:42< Crab_> billynux: well, can't you feed it to gunzip filter ?
20100707 23:30:51 * alink will check changelog for beta compatibility changes
20100707 23:31:11< billynux> Crab_, And I interpret these 33 bytes with read_gz, is this ok? (after this the cfg tests non empty)
20100707 23:31:39< billynux> Crab_, feed it to the filter how? (sorry)
20100707 23:31:39< Crab_> billynux: you can easily dump them to stderr, and see if they are garbage or not :)
20100707 23:31:59< billynux> std::cerr << cfg? :)
20100707 23:32:14< Crab_> billynux: raw received bytes after your code unzips them
20100707 23:32:39< Crab_> if wireshark allows you to save the raw data or copy them to clipboard, you can write a small c/c++/perl/python/whatever program to decompress them for you
20100707 23:33:08< billynux> yes, wireshark gives me the C array if I want to (it's compressed)
20100707 23:33:51< billynux> but ATM I assume the data is good. My question is: Is the call to read_gz fine? (to fill the config&)
20100707 23:34:13< alink> many bug bugfixes in 1.8beta..,  it's sad that some don't upgrade
20100707 23:34:24< alink> s/bug/big
20100707 23:34:32< Crab_> billynux: yes, it should be
20100707 23:34:47< Crab_> (norbert's bot does gzread(), from a different lib, I suppose)
20100707 23:35:18< billynux> ok, read_gz takes a config& as argument (the call was buried deep :) )
20100707 23:35:45< boucman> billynux: why don't you just call the debug() method on your received config to see if it's ok ?
20100707 23:36:18< billynux> Crab_, no, it's not reaching the aforementioned line: if (data.child...
20100707 23:36:22< billynux> boucman, I'll try that
20100707 23:37:33-!- Johannes13 [~Johannes@unaffiliated/johannes13] has quit [Ping timeout: 265 seconds]
20100707 23:37:35< Crab_> also, you receive only a small thing.. so, if it gives you an invalid config, you can just trace the parser's work..
20100707 23:38:29< Crab_> also, for a quick hack, you can just make a copy of read_gz which dumps unzipped stream contents to stderr
20100707 23:39:21< zookeeper> Blitzmerker, units have an .image variable
20100707 23:39:40< billynux> Crab_, well... I tried std::cout << cfg; and cfg.debug(); ang got [version] [/version]
20100707 23:40:07< boucman> billynux: so that's a request from the server to send version info...
20100707 23:40:18< boucman> do you have the code to handle it ?
20100707 23:40:26< billynux> boucman, good, so. so far so good
20100707 23:40:57-!- DesertPanther [~Khalid@unaffiliated/desertpanther] has quit [Ping timeout: 240 seconds]
20100707 23:41:06< billynux> the problem is thus reduced to: Why if I returned this cfg with the version request does my code not send this info back?
20100707 23:41:29< billynux> multiplayer.cpp:181 (or whereabouts) is not reached
20100707 23:41:34< Crab_> billynux: the control flow is different, then
20100707 23:41:44< billynux> Crab_, it appears so
20100707 23:41:47< Crab_> billynux: move up the stack until you find out what is actually called
20100707 23:42:17< billynux> Crab_, meaning: who called receive_data ?
20100707 23:42:39< Crab_> billynux: you do a synced call to receive_data from which you got that [version][version] ?
20100707 23:42:57< billynux> Crab_, yes
20100707 23:43:11< billynux> this is to implement the current network API (which is synced)
20100707 23:43:30< Crab_> billynux: ok, then you need to make a real or mental backtrace of the original control flow
20100707 23:43:33< billynux> to do this I use custom handlers and regular async operations (not the happiest of hacks)
20100707 23:43:44< Crab_> how that 'send version'  is called in old code ? 
20100707 23:44:05< billynux> Crab_, ok... it won't be fun :P
20100707 23:44:09< Crab_> it is in server_type open_connection
20100707 23:44:23< Crab_> can you paste your backtrace somewhere ?
20100707 23:45:00< billynux> Crab_, the backtrace using ana or SDL? and at which breakpoint?
20100707 23:45:05< Crab_> ana
20100707 23:45:12< Crab_> just after your cfg.debug call
20100707 23:45:37< boucman> billynux: that's the kind of cases wher using gdb gains a lot of time... put a breakpoint where you put the debug() and roll from there
20100707 23:46:09< Crab_> boucman: I think that it's better to see the backtrace and think about what is supposed to happen, first :)
20100707 23:46:27< Crab_> boucman: and then, yes, verify the theory via gdb :)
20100707 23:47:02< boucman> Crab_: except it's not the bt we're interested in, we know where the code comes from, we're trying to know where it goes (IIUC)
20100707 23:47:06< Blitzmerker> zookeeper: thx, found out myself I just did not find it on http://wiki.wesnoth.org/InternalActionsWMLUnitTags
20100707 23:47:13< billynux> Crab_, boucman : so... the backtrace leading up to after the call to cfg.debug() ?
20100707 23:47:16< Crab_> sdl backtrace looks like game_controller::play_multiplayer -> mp::start_client -> mp::open_connection -> send version stuff
20100707 23:47:37< billynux> this will give me a sense of how I got there (instead of what happens next)
20100707 23:47:51< Crab_> billynux: just a backtrace when we call cfg.debug(), a backtrace at the point of time when we *think* that we have the correct [version] request from server
20100707 23:48:20< Crab_> billynux: yes, we need to see which point we have in common with sdl code
20100707 23:48:30< billynux> http://wesnoth.pastebin.com/9V9FF8UW
20100707 23:49:15< Crab_> ok, looks pretty good
20100707 23:49:51< Crab_> so, in #3, we are in the same mp::open_connection function which is supposed send the version back to server
20100707 23:50:00< billynux> aha
20100707 23:50:33< Crab_> note the mp::check_response(data_res, data); part in open_connection
20100707 23:50:59< Crab_> after that, it's that 'data' which is checked for 'version' child
20100707 23:51:17< billynux> and how is it that it doesn't reach the check to see if there is a version request? (line ~181)
20100707 23:52:14< billynux> is there line-stepping in gdb? ( like a break in the next line of the file in this frame)
20100707 23:52:28< Crab_> let's see the value of data after the control flow returns to open_connection function
20100707 23:52:30< billynux> (that omits code executed elsewhere)
20100707 23:53:06< Crab_> yes, use 'next' instead of step or just set a breakpoint on the needed line.
20100707 23:53:29< Crab_> for example, we can set a bp on  if (!data_res) return ABORT_SERVER;
20100707 23:53:45< Crab_> at that point, we should have the data filled up
20100707 23:53:57< billynux> Aha! it returns 0!
20100707 23:54:18< billynux> I'm guessing this is wrong, it should inform the connection number it read from, right?
20100707 23:54:45< billynux> indeed, let me fix that
20100707 23:54:47-!- Talkative [97c48e37@gateway/web/freenode/ip.151.196.142.55] has joined #wesnoth-dev
20100707 23:55:02< Crab_> so, after you fix it to return a non-zero value on success, it should go further
20100707 23:56:06-!- Talkative [97c48e37@gateway/web/freenode/ip.151.196.142.55] has quit [Client Quit]
20100707 23:56:16< billynux> it should...
20100707 23:58:23< Crab_> boucman: I was more interested in seeing the backtrace because I wanted to find out the closest place from that bt to the place in open_connection where that version info is sent, without having to look at all intermediate code.
20100707 23:58:38< boucman> ok
20100707 23:58:50< boucman> well, bed time for me, see you all tomorow
20100707 23:58:55< Crab_> boucman: and, since we were lucky that the bt included the open_connection function... :)
20100707 23:58:58< Crab_> boucman: bye :)
20100707 23:59:03-!- boucman [~rosen@wesnoth/developer/boucman] has left #wesnoth-dev []
--- Log closed Thu Jul 08 00:00:13 2010