--- Log opened Sat Aug 01 00:00:09 2009
20090801 00:00:19< zookeeper> esr, you mean some of the directional animations?
20090801 00:00:31< zookeeper> i'm supposed to be taking a look at them sometime soon..
20090801 00:03:15< esr> zookeeper: I had to outright remove a movement animation because it referenced several images that no longer exist.  And there are several flying images that probably should have been used there but aren't used anywhere yet.
20090801 00:03:28< CIA-62> soliton * r37365 /trunk/src/server/ (server.cpp server.hpp): allow faking the issuer for all commands and mark responses specifically in that case
20090801 00:06:56< shadowmaster> boucman: re bug 13967
20090801 00:07:09< shadowmaster> boucman: the solution is to replace terrain= with terrain_type= ?
20090801 00:07:20< AI0867> shadowmaster: yes
20090801 00:07:24< boucman> shadowmaster: yes and no
20090801 00:07:36< boucman> terrain was correct in 1.6 but renamed in 1.7
20090801 00:07:46< shadowmaster> yes, I know.
20090801 00:07:50< boucman> however, wmllint was supposed to fix it and didn't
20090801 00:08:04< boucman> so we are correcting this on wmllint side, then run it and commit the fixez
20090801 00:08:08< boucman> fixes
20090801 00:08:17< shadowmaster> I thought only [terrain] was renamed (the node, not the attribute); that's a surprise for me.
20090801 00:08:33< shadowmaster> esr: was [terrain] terrain= in events WML renamed too?
20090801 00:08:43< boucman> shadowmaster: well, terrain= was at the same time in that case, not sure if it was done on purpose
20090801 00:08:47< boucman> esr is looking into it
20090801 00:09:02< esr> shadowmaster: I don't think so.
20090801 00:09:11< AI0867> esr: the descriptions you just patched were recently written in the writer's forum. You may want to review them *before* they get comitted.
20090801 00:09:14-!- wesbot changed the topic of #wesnoth-dev to: 66 bugs, 239 feature requests, 11 patches | logs: http://irclogs.wesnoth.org/ | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com | http://imagebin.org
20090801 00:09:43< shadowmaster> boucman: re 14039, what about [label]?
20090801 00:10:31< esr> AI0867: Whoever committed them should have run wmllint to check them before committing - there were Britishisms like "armour" in there.
20090801 00:10:41< shadowmaster> ugh, out of battery, going off
20090801 00:10:47< esr> I'll have to post to this effect on the forum.
20090801 00:10:52< boucman> shadowmaster: i removed floating labels, but I think normal labels set by WML should stay
20090801 00:11:05< AI0867> Turuk: ^^
20090801 00:12:25-!- yann [n=dwitch@nan92-1-81-57-214-146.fbx.proxad.net] has joined #wesnoth-dev
20090801 00:16:40-!- Alesis-Novik [n=alesis@78-61-40-192.static.zebra.lt] has quit [Read error: 104 (Connection reset by peer)]
20090801 00:17:15-!- Alesis-Novik [n=alesis@78-61-40-192.static.zebra.lt] has joined #wesnoth-dev
20090801 00:17:15< Crab_> is there a specific reason why bool terrain_filter::match(const map_location& loc); is not marked as const ?
20090801 00:18:12< Sapient> Crab_: probably because it updates the internal cache which is not marked mutable?
20090801 00:18:43< Sapient> you could make it const if you prefer mutable cache members
20090801 00:19:50< Sapient> also, IIRC it needs a method to clear the cache; I simply haven't added it yet
20090801 00:20:53< boucman> Crab_: add a const and see where the compiler yells ;)
20090801 00:20:59< boucman> but IIRC there is a reason
20090801 00:21:17< Crab_> Sapient, boucman: ok, I'll do that. I'm also thinking about adding a 'static const terrain_filter match_no_locations', which will match no locations (to facilitate returning a const ref from a template factory method)
20090801 00:21:53< Sapient> fine by me
20090801 00:22:54-!- ardesh [n=ardesh@port-92-206-108-65.dynamic.qsc.de] has quit ["Quis custodiet ipsos custodes"]
20090801 00:25:16< CIA-62> esr * r37366 /trunk/data/ (4 files in 4 dirs): 
20090801 00:25:16< CIA-62> Address bug #13967: Animation glitches with terrain filters.
20090801 00:25:16< CIA-62> This needed both fixes in mainline and a bit of wmllint logic to ensure
20090801 00:25:16< CIA-62> that incoming UMC gets converted.
20090801 00:25:22-!- Blueblaze [n=nick@c-98-199-143-139.hsd1.tx.comcast.net] has quit [Remote closed the connection]
20090801 00:25:45< esr> boucman: OK, fix in place and ready for test.
20090801 00:27:33< Crab_> Sapient: also, nowadays, pointer to unit_map::units is available as 'resources::units' does this mean that unit_map argument of SLF's constructor can be dropped (since the SLF can fetch the unit_map itself) ? or there are situations where SLF is called with hypothetical unit map ?
20090801 00:28:01< Crab_> s/unit_map::units/unit_map units
20090801 00:28:10< Sapient> Crab_: I was asking myself the same question a few days ago
20090801 00:28:37< Sapient> I feel you would know better than anyone and only the AI would need such trickery
20090801 00:28:48-!- zookeeper [n=l@wesnoth/developer/zookeeper] has left #wesnoth-dev []
20090801 00:29:00-!- zookeeper [n=l@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20090801 00:29:44< Sapient> silene suggested that moving unit_map into resources was a bad idea, for AI purposes it might be convenient to have hypothetical unit_maps
20090801 00:29:59< Crab_> Sapient: well, 1-2 hours ago we talked about AI usage of SLF, and we found out that ' it's ok to take into account "unavailable info"'
20090801 00:30:22< boucman> esr: testing
20090801 00:30:39-!- thespaceinvader [n=chatzill@wesnoth/artist/thespaceinvader] has quit ["night all"]
20090801 00:30:45< Crab_> Sapient: but, otoh, is for avoided_locations sort of thing... for other usages, there indeed might be needed to use hypothetical unit map.
20090801 00:30:54< melinath> esr: the dwarf scout line descriptions are not part of an ongoing project. They were a one-time thing (as far as I know) done by tsi.
20090801 00:30:58< Sapient> in my personal opinion, I'd like to see it removed from the constructor call if possible
20090801 00:31:12-!- zookeeper [n=l@wesnoth/developer/zookeeper] has quit []
20090801 00:31:28< Sapient> people who need weird C++ AIs can revert that change if it becomes an issue for them
20090801 00:32:37< Crab_> Sapient: it can be replaced with a set_unit_map function which will clean the cache and set a new unit map
20090801 00:32:53-!- YogiHH_ [n=chatzill@c204084.adsl.hansenet.de] has joined #wesnoth-dev
20090801 00:33:13< Sapient> yeah, let's leave such a monstrosity for when it's actually needed
20090801 00:33:15< Crab_> Sapient: so, it's possible to remove it from constructor call if (apart from potential AI usage) there are no other reasons to keep it
20090801 00:34:42< Sapient> good, the less things being passed to the constructor, the better. (and the less nightmares I have about MFC)
20090801 00:35:14< Crab_> what's MFC ? merge-from-current ? or 'microsoft foundation classes' ?
20090801 00:35:37< Sapient> microsoft foundation classes -- horrid stuff
20090801 00:35:38< Crab_> or 'monstrously f*ing code' ?
20090801 00:35:48< Crab_> ok
20090801 00:39:20< boucman> esr: ok, it's fixed
20090801 00:39:28< boucman> night all
20090801 00:39:30< esr> Excellent.
20090801 00:39:37< YogiHH_> night boucman
20090801 00:39:39-!- boucman [n=rosen@wesnoth/developer/boucman] has quit ["Leaving."]
20090801 00:40:23< Sapient> yeah, imagine constructor calls so long they wrap around your screen three or four times with multiple series of obscure constants being bitwise combined.
20090801 00:40:25< YogiHH_> Ivanovic, still awake?
20090801 00:40:36< Ivanovic> somehow
20090801 00:42:06< YogiHH_> Ivanovic: I am going onto vacation tomorrow for one week. I discussed everything important with euschn yesterday and i will write him a mail pointing to you for all organizational stuff.
20090801 00:42:42< YogiHH_> Ivanovic: Just so you know
20090801 00:42:58< esr> Since it's being mentioned, my availability will be limited from this coming Sunday through the following one.
20090801 00:43:25< esr> (Second half of vacation. World Boardgaming Championships.)
20090801 00:43:34< Ivanovic> YogiHH_: okay
20090801 00:43:57< esr> I'll have WifFi but not be at my computer often.
20090801 00:44:47-!- YogiHH [n=chatzill@c211151.adsl.hansenet.de] has quit [Read error: 110 (Connection timed out)]
20090801 00:44:48< esr> "yeah, imagine constructor calls so long they wrap around your screen three or four times with multiple series of obscure constants being bitwise combined." - so, if this C++ we're talking about, how is this in the leasty unusual or noteworthy? :-)
20090801 00:46:05< Sapient> well, I guess you could call it a prime example of what not to do
20090801 00:46:47< esr> But C++ looks like that to me all the time...
20090801 00:47:07< Sapient> just because the language allows you to design classes like that, doesn't mean you *should*
20090801 00:47:41< esr> I' am in nop way disagreeing with that statement.
20090801 00:48:47< Sapient> like just because python allows you to encode your entire program in whitespace on a website, fetch it decode and eval() it, doesn't mean you *should*
20090801 00:49:01< YogiHH_> Good night everyone, see you again on August 10th, probably
20090801 00:50:35< ilor> see you YogiHH_ :)
20090801 00:51:17-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit [Read error: 104 (Connection reset by peer)]
20090801 00:51:32-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090801 00:56:12< Sirp> esr: the obscure constants being bitwise combined is much more common in C than in C++ in my experience.
20090801 00:57:50< esr> True.  I was focusing on "imagine constructor calls so long they wrap around your screen three or four times"
20090801 00:58:31-!- YogiHH_ [n=chatzill@c204084.adsl.hansenet.de] has left #wesnoth-dev []
20090801 01:08:03-!- blarumyrran [n=minaise@81-20-159-197.levira.ee] has quit []
20090801 01:09:37< Sapient> in #python a few days ago they were talking about a guy who won a programming competition to write the shortest program that achieved the task
20090801 01:10:14< Sapient> apparently the contest organizers decided whitespace didn't count, so he encoded his whole program in a string of whitespace
20090801 01:10:31< Sapient> then he just had three lines to decode and eval() it
20090801 01:11:22< Sapient> ... I guess they had to modify the rules after that ;)   cya
20090801 01:11:29-!- Sapient [n=patrickp@wesnoth/developer/sapient] has left #wesnoth-dev []
20090801 01:12:49-!- Polarina [n=polarina@wesnoth/translator/Polarina] has quit [Connection reset by peer]
20090801 01:13:23-!- Polarina [n=polarina@wesnoth/translator/Polarina] has joined #wesnoth-dev
20090801 01:17:35-!- deekay [n=dk@wesnoth/developer/dragonking] has quit []
20090801 01:17:44< CIA-62> soliton * r37367 /branches/1.6/src/server/server.cpp: allow faking the issuer for all commands and mark responses specifically in that case
20090801 01:19:40-!- grzywacz [n=grzywacz@wesnoth/developer/grzywacz] has quit [Remote closed the connection]
20090801 01:20:10< Turuk> esr: is there a reason you are altering the committed descriptions beyond spelling changes for wmllint?
20090801 01:22:48-!- elias [n=allefant@allegro/developer/allefant] has quit ["Leaving"]
20090801 01:22:58-!- happygrue [n=George@wesnoth/developer/wintermute] has quit [Read error: 104 (Connection reset by peer)]
20090801 01:24:08-!- Chusslove [n=Chusslov@brsg-d9bef4ab.pool.mediaWays.net] has quit [Read error: 110 (Connection timed out)]
20090801 01:24:37< CIA-62> ilor * r37368 /trunk/src/server/ (room.cpp room.hpp room_manager.cpp): Add room logged parameter, settable by room query, fix room logging so chat actually ends up in the log by default
20090801 01:25:26< CIA-62> ilor * r37369 /trunk/src/server/room_manager.cpp: Make the lobby persistent and logged by default
20090801 01:25:32-!- loonycyborg [n=sergey@wesnoth/developer/loonycyborg] has quit ["Zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz"]
20090801 01:25:41< CIA-62> ilor * r37370 /trunk/src/server/room_manager.hpp: add some doxy comments to room_manager.hpp
20090801 01:25:58< CIA-62> ilor * r37371 /trunk/src/server/ (room_manager.cpp room_manager.hpp): add dirty flagto room manager
20090801 01:26:08-!- loonybot [n=loonybot@wesnoth/bot/loonybot] has quit [Remote closed the connection]
20090801 01:26:13< CIA-62> ilor * r37372 /trunk/src/server/ (room_manager.cpp server.cpp): save room data periodically in the server
20090801 01:26:30< CIA-62> ilor * r37373 /trunk/src/server/room_manager.cpp: send room topic with room joins and allow setting topics to admins
20090801 01:26:46< CIA-62> ilor * r37374 /trunk/src/gui/dialogs/lobby_main.cpp: show room topic on room join in new lobby
20090801 01:27:04< CIA-62> ilor * r37375 /trunk/src/ (3 files in 2 dirs): support for passworded games in new lobby
20090801 01:28:30< CIA-62> soliton * r37376 /trunk/src/upload_log.cpp: fixed a warning
20090801 01:28:40< CIA-62> soliton * r37377 /trunk/src/server/ (server.cpp server.hpp): changed the format of faked nicks a bit
20090801 01:29:24-!- Chusslove [n=Chusslov@brsg-d9befbfb.pool.mediaWays.net] has joined #wesnoth-dev
20090801 01:34:13-!- Crab_ [n=Crab_@wesnoth/developer/crab] has quit ["Leaving."]
20090801 01:35:33-!- happygrue [n=George@wesnoth/developer/wintermute] has joined #wesnoth-dev
20090801 01:41:19< corn> just found a bug in defaldor's memoirs
20090801 01:42:24 * shadowmaster holds his breath
20090801 01:44:28< shadowmaster> corn: do you have some time to share it with us? :)
20090801 01:44:37< corn> death of lionel in the Night in the Swamp scenario leads to defeat, but it isn't listed as a defeat condition in the scenario objectives
20090801 01:44:39< melinath> indeed. Out with it, man!
20090801 01:44:49< corn> played on easy in 1.7.2
20090801 01:46:31< corn> I'll file a bug report now
20090801 01:47:59-!- Aethaeryn [n=Michael@wesnoth/umc-dev/developer/aethaeryn] has joined #Wesnoth-dev
20090801 01:48:03< shadowmaster> corn: no need
20090801 01:48:12< shadowmaster> I'm fixing it now (and for scenario 8 "Ur Thorodor" too)
20090801 01:48:17< corn> ok
20090801 01:48:18< Aethaeryn> Damn, power went out.
20090801 01:48:36< corn> is it OK to modify a maintainer's campaign?
20090801 01:48:41< corn> to fix bugs like these
20090801 01:49:35< Soliton> you're allowed to fix obvious bugs anywhere.
20090801 01:49:57< corn> ok
20090801 01:50:17-!- giusef [n=giusef@unaffiliated/giusef] has joined #wesnoth-dev
20090801 01:50:52< CIA-62> shadowmaster * r37378 /trunk/data/campaigns/Delfadors_Memoirs/scenarios/ (07_night_in_the_swamp.cfg 08_ur_thorodor.cfg): 
20090801 01:50:52< CIA-62> List Lionel's death as a defeat condition in [objectives] for scenarios
20090801 01:50:52< CIA-62> 7 and 8 (reported by corn).
20090801 01:53:06< shadowmaster> esr: maybe you want to convert faux-formatting cases of storytext in places such as /data/campaigns/Delfadors_Memoirs/scenarios/16_dark_sky_over_weldyn.cfg and /data/campaigns/Legend_of_Wesmere/scenarios/12_Revelations.cfg so they use Pango markup instead
20090801 01:53:26< shadowmaster> (no idea if there are more cases such as these ones, but a grep should reveal them if there are)
20090801 01:53:33< corn> also, a minor gui irritation that I noticed
20090801 01:54:06< corn> when selecting a campaign, the position of the image at the bottom of the campaign description is dependent on the description length
20090801 01:54:32< shadowmaster> that'd be for mordante I guess
20090801 01:54:35< corn> ok
20090801 01:58:52< shadowmaster> Boost.Asio was introduced in 1.35, right?
20090801 02:01:41< shadowmaster> huh. Well, looks like it's not in Debian stable, not even amongst the 1.35 packages.
20090801 02:06:00-!- ilor [n=user@wesnoth/developer/ilor] has quit []
20090801 02:12:40-!- giusef [n=giusef@unaffiliated/giusef] has quit [Remote closed the connection]
20090801 02:31:11-!- ABCD [n=ABCD@gentoo/contributor/abcd] has quit [Read error: 104 (Connection reset by peer)]
20090801 02:31:20-!- ABCD [n=ABCD@gentoo/contributor/abcd] has joined #wesnoth-dev
20090801 02:44:29-!- mordocai [n=mordocai@66.119.9.243] has joined #wesnoth-dev
20090801 02:52:53-!- mordocai [n=mordocai@66.119.9.243] has quit ["Leaving"]
20090801 03:14:44-!- Elvish_Pillager [n=eli@71-10-231-36.dhcp.oxfr.ma.charter.com] has joined #wesnoth-dev
20090801 03:17:04-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit [Connection reset by peer]
20090801 03:17:19-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090801 03:23:31-!- BenUrban [n=benurban@c-68-50-54-86.hsd1.md.comcast.net] has joined #wesnoth-dev
20090801 03:30:13-!- Elvish_Pillage2 [n=eli@71-10-231-36.dhcp.oxfr.ma.charter.com] has quit [Read error: 110 (Connection timed out)]
20090801 03:43:18-!- Blueblaze [n=nick@c-98-199-143-139.hsd1.tx.comcast.net] has joined #wesnoth-dev
20090801 03:51:39-!- Aethaeryn [n=Michael@wesnoth/umc-dev/developer/aethaeryn] has quit ["meh"]
20090801 04:01:38-!- Ivanovic_ [n=ivanovic@dtmd-4db228e7.pool.einsundeins.de] has joined #wesnoth-dev
20090801 04:06:29-!- Espreon [n=espreon@wesnoth/developer/espreon] has quit ["WRYYYYYYYYYYYYYYYYYYYY!"]
20090801 04:06:45-!- Doppp [n=Doppp@unaffiliated/doppp] has quit [Read error: 104 (Connection reset by peer)]
20090801 04:06:48< melinath> esr: did you change "They spend this time watching for invaders, and combating the bandits and thieves who might spring up inside dwarvish territory." to "They spend this time watching for invaders, and fighting bandits and thieves who incur dwarvish territory."
20090801 04:07:04< melinath> esr: if so, why?
20090801 04:09:07< melinath> esr: (This is from the dwarf pathfinder desc, trunk)
20090801 04:09:20< corn> found another bug in memoirs of delfaldor
20090801 04:09:44< corn> in the wasteland scenario, the Death of Delfador objective is repeated when he encounters the ally leader
20090801 04:14:42-!- Doppp [n=Doppp@c-67-171-96-240.hsd1.pa.comcast.net] has joined #wesnoth-dev
20090801 04:17:15-!- ABCD_ [n=ABCD@gentoo/contributor/abcd] has joined #wesnoth-dev
20090801 04:18:34-!- Ivanovic [n=ivanovic@wesnoth/developer/ivanovic] has quit [Read error: 110 (Connection timed out)]
20090801 04:19:36-!- Ivanovic_ is now known as Ivanovic
20090801 04:22:14-!- ABCD [n=ABCD@gentoo/contributor/abcd] has quit [Read error: 104 (Connection reset by peer)]
20090801 04:22:54< corn> same duplication occurs in the terror at the ford of parthyn scenario
20090801 04:59:50-!- Ivanovic_ [n=ivanovic@dtmd-4db2a995.pool.einsundeins.de] has joined #wesnoth-dev
20090801 05:16:09-!- Ivanovic [n=ivanovic@dtmd-4db228e7.pool.einsundeins.de] has quit [Read error: 110 (Connection timed out)]
20090801 05:17:58-!- Ivanovic_ is now known as Ivanovic
20090801 05:25:59-!- Blueblaze [n=nick@c-98-199-143-139.hsd1.tx.comcast.net] has quit [Remote closed the connection]
20090801 05:47:36-!- Blueblaze [n=nick@c-98-199-143-139.hsd1.tx.comcast.net] has joined #wesnoth-dev
20090801 05:56:56-!- Chusslove [n=Chusslov@brsg-d9befbfb.pool.mediaWays.net] has quit [Read error: 110 (Connection timed out)]
20090801 06:02:06-!- Chusslove [n=Chusslov@brsg-d9bee4a1.pool.mediaWays.net] has joined #wesnoth-dev
20090801 06:09:14-!- wesbot changed the topic of #wesnoth-dev to: 65 bugs, 239 feature requests, 11 patches | logs: http://irclogs.wesnoth.org/ | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com | http://imagebin.org
20090801 06:23:49-!- Elvish_Pillager [n=eli@71-10-231-36.dhcp.oxfr.ma.charter.com] has quit [Read error: 110 (Connection timed out)]
20090801 06:25:15-!- melinath [n=melinath@pool-71-162-41-64.altnpa.east.verizon.net] has quit [Read error: 113 (No route to host)]
20090801 07:09:08-!- silene [n=plouf@AMarseille-251-1-73-132.w83-205.abo.wanadoo.fr] has joined #wesnoth-dev
20090801 07:47:04< corn> shadowmaster: the same objective duplication bug appears in the wasteland scenario, the "Death of Delfador" objective is repeated when he encounters the ally leader
20090801 08:47:59-!- Sirp [n=user@wesnoth/developer/dave] has quit ["leaving"]
20090801 09:10:17-!- lizard_r [n=Miranda@wesnoth/umc-dev/developer/lizard] has joined #wesnoth-dev
20090801 09:11:13-!- busfahrer [n=busfahre@unixboard/user/busfahrer] has joined #wesnoth-dev
20090801 09:18:50-!- silene [n=plouf@wesnoth/developer/silene] has quit ["Leaving."]
20090801 09:20:15-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit [Read error: 60 (Operation timed out)]
20090801 09:21:35-!- zookeeper [n=l@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20090801 09:27:06-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090801 09:28:25-!- silene [n=plouf@wesnoth/developer/silene] has joined #wesnoth-dev
20090801 09:49:47-!- boucman [n=rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20090801 09:50:01< boucman> morning all
20090801 10:07:47-!- elias [n=allefant@allegro/developer/allefant] has joined #wesnoth-dev
20090801 10:17:47-!- Noyga [n=lame-z@wesnoth/developer/noyga] has joined #wesnoth-dev
20090801 10:18:26-!- olik_ [n=olik@85-220-68-185.dsl.dynamic.simnet.is] has joined #wesnoth-dev
20090801 10:19:09-!- ABCD_ is now known as ABCD
20090801 10:23:13-!- busfahrer [n=busfahre@unixboard/user/busfahrer] has quit ["leaving"]
20090801 10:31:42< Ivanovic> moin
20090801 10:32:22< boucman> moin moin
20090801 10:37:44-!- Turuk_ [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090801 10:41:01-!- olik [n=olik@85-220-78-118.dsl.dynamic.simnet.is] has quit [Read error: 113 (No route to host)]
20090801 10:42:16-!- EdB [n=edb@143.117.88-79.rev.gaoland.net] has joined #wesnoth-dev
20090801 10:43:59-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit [Read error: 110 (Connection timed out)]
20090801 10:57:56-!- thespaceinvader [n=chatzill@wesnoth/artist/thespaceinvader] has joined #wesnoth-dev
20090801 11:01:53-!- loonybot [n=loonybot@79.139.138.234] has joined #wesnoth-dev
20090801 11:03:54-!- loonycyborg [n=sergey@79.139.138.234] has joined #wesnoth-dev
20090801 11:08:26< loonycyborg> shadowmaster: Boost.Asio is header only, therefore it's in boost-dev package.
20090801 11:09:16-!- ilor [n=user@wesnoth/developer/ilor] has joined #wesnoth-dev
20090801 11:17:56-!- Crab_ [n=Crab_@wesnoth/developer/crab] has joined #wesnoth-dev
20090801 11:18:02< Crab_> hi
20090801 11:18:04< boucman> morning Crab_
20090801 11:32:38-!- Dragonking [n=dk@wesnoth/developer/dragonking] has joined #wesnoth-dev
20090801 11:33:10-!- Dragonking is now known as deekay
20090801 11:35:08< boucman> hey deekay
20090801 11:36:23< deekay> hi
20090801 11:36:55< boucman> how is it going with FAI ?
20090801 11:37:15< deekay> Slowly ahead :)
20090801 11:38:26< deekay> Had some unexpected things come out this week, so my work was a bit stalled
20090801 11:45:48-!- Blueblaze [n=nick@c-98-199-143-139.hsd1.tx.comcast.net] has quit [Remote closed the connection]
20090801 11:46:10-!- Blueblaze [n=nick@c-98-199-143-139.hsd1.tx.comcast.net] has joined #wesnoth-dev
20090801 12:01:00-!- Blueblaze [n=nick@c-98-199-143-139.hsd1.tx.comcast.net] has quit [Remote closed the connection]
20090801 12:02:56< boucman> zookeeper: around ?
20090801 12:04:42< zookeeper> boucman, not very
20090801 12:04:51< boucman> ok, will talk later then
20090801 12:05:07< zookeeper> well, i can reply later..
20090801 12:06:30< boucman> I want to add take off/landing anim, and I was not sure what way to do it...
20090801 12:06:46< boucman> a specific anim played before first move/after last move
20090801 12:07:05< boucman> an extra param in the mvt anim telling how far in the anim we are
20090801 12:07:08< boucman> something else...
20090801 12:12:24< zookeeper> well, as i've said sometime long ago, IMO the best way would be to be able to specify loop points for an anim. but it's probably more complicated than just adding new anim types.
20090801 12:14:00< boucman> ???
20090801 12:14:09< boucman> i don't see how that is relevant...
20090801 12:14:51-!- loktor [n=quassel@AAubervilliers-151-1-62-124.w83-112.abo.wanadoo.fr] has joined #wesnoth-dev
20090801 12:15:23< zookeeper> any solution which allows you to just specify the takeoff, flying and landing animation in the same animation block would be just fine
20090801 12:17:13< boucman> the problem is that TO anim is only played on the first anim of the move
20090801 12:17:32< boucman> not on subsequent anim (remember, if we cross a different terrain, we might change the animation)
20090801 12:18:20< ilor> Soliton: could you increase the connections_allowed server config param from the default 5 to say 10 or 20? would make testing some stuff easier
20090801 12:22:13-!- Elvish_Pillager [n=eli@71-10-231-36.dhcp.oxfr.ma.charter.com] has joined #wesnoth-dev
20090801 12:24:11< zookeeper> boucman, right
20090801 12:24:31< zookeeper> well, it's not a big deal if they have to be separate animation blocks
20090801 12:26:55< Soliton> ilor: for trunk?
20090801 12:27:03< ilor> Soliton yes
20090801 12:28:11< Soliton> ilor: done.
20090801 12:28:17< ilor> thanka
20090801 12:28:23< ilor> *thanks ;)
20090801 12:37:24-!- ilor [n=user@wesnoth/developer/ilor] has quit []
20090801 12:47:04-!- melinath [n=melinath@pool-71-162-41-64.altnpa.east.verizon.net] has joined #wesnoth-dev
20090801 12:48:02< CIA-62> boucman * r37379 /trunk/src/unit_animation.hpp: rename an (unused) function with a name more consistant with conventions
20090801 13:15:21-!- loktor [n=quassel@AAubervilliers-151-1-62-124.w83-112.abo.wanadoo.fr] has quit [Remote closed the connection]
20090801 13:17:48-!- EdB [n=edb@143.117.88-79.rev.gaoland.net] has quit [Remote closed the connection]
20090801 13:27:38< CIA-62> soliton * r37380 /trunk/src/server/server.cpp: various logging tweaks for admin commands
20090801 13:33:04< boucman> Crab_: you didn't commit your stuff yet, did you ?
20090801 13:33:17< boucman> or is gna mail completely borked ?
20090801 13:33:34< Crab_> boucman: no, I still have one segfault in terrain_filter to fix
20090801 13:33:40< boucman> ok
20090801 13:33:50< boucman> that's a good reason to not commit :P
20090801 13:37:36-!- Baufo [n=quassel@91-115-185-138.adsl.highway.telekom.at] has joined #wesnoth-dev
20090801 13:54:57< Ivanovic> silene: http://www.wesnoth.org/forum/viewtopic.php?f=4&t=26476&start=0&st=0&sk=t&sd=a
20090801 13:55:27< Ivanovic> uhm, for whom is this one? is it for you, boucman? http://www.wesnoth.org/forum/viewtopic.php?f=4&t=26477&start=0&st=0&sk=t&sd=a
20090801 13:55:38< boucman> checking
20090801 13:55:43< Crab_> Ivanovic: this is fixed in trunk
20090801 13:56:04< Ivanovic> Crab_: which of those two?
20090801 13:56:09< Crab_> Ivanovic: (about unit advances no longer exists)
20090801 13:56:16< boucman> or at least we were not able to reproduce with trunk, melinath will probably report here after upgrading to trunk
20090801 13:56:32< boucman> about haloes
20090801 13:56:34< Ivanovic> ah, okay
20090801 13:57:12< Crab_> wesbot: bug 14014
20090801 13:57:13< wesbot> Bug #14014  Assigned to: Yurii Chernyi <crab>  Status: Fixed  Priority: 5 - Normal
20090801 13:57:16< wesbot> Summary: Advancing a max-level unit with :unit advances=N broken since r36979
20090801 13:57:19< wesbot> Original submission: A change in r36979 breaks the debug-mode :unit advances=N
20090801 13:57:22< wesbot>  command when the target unit is already at maximum level.
20090801 13:57:25< wesbot> URL: https://gna.org/bugs/?14014
20090801 13:57:26< Ivanovic> this one sounds interesting, too
20090801 13:57:35< Ivanovic> related to recalling and undoing
20090801 13:57:37< Ivanovic> http://www.wesnoth.org/forum/viewtopic.php?f=4&t=26479&start=0&st=0&sk=t&sd=a
20090801 14:41:24< CIA-62> soliton * r37381 /trunk/src/server/server.cpp: added some missing new lines in log output
20090801 14:56:37< CIA-62> soliton * r37382 /branches/1.6/src/server/server.cpp: various logging tweaks for admin commands
20090801 15:07:50< melinath> Crab_: The bug I reported is different than bug 14014...
20090801 15:08:21< Crab_> melinath: yet the engine error messages is the same.
20090801 15:08:29< melinath> Crab_: I had it with advance on any non-core unit.
20090801 15:08:55< melinath> Crab_: okay. So it's also fixed in trunk?
20090801 15:09:06< Crab_> melinath: yes, It should be fixed.
20090801 15:09:12< melinath> kk. thanks.
20090801 15:09:39-!- MJ999 [n=msflower@80.154.28.81] has quit [Remote closed the connection]
20090801 15:10:34< Crab_> melinath: the procedure that replaced a unit by another (which happens during any unit advancement) was buggy for a while.
20090801 15:10:49-!- MJ999 [n=msflower@80.154.28.81] has joined #wesnoth-dev
20090801 15:25:21< melinath> Crab_: I was just checking because I don't have any problems advancing core units indefinitely, but I can't advance a lvl 1 UMC unit to a level 2.
20090801 15:25:58< melinath> checking my statement...
20090801 15:26:34< Crab_> melinath: if that problem resurfaces in latest svn, then tell me about it and I'll recheck.
20090801 15:27:08< melinath> melinath: nvm, seems to also happen with core units over max level. Probably the same thing, then.
20090801 15:27:30< melinath> *Crab_, that is. And I'll let you know.
20090801 15:27:50< melinath> Though I'm only running off the repos, not trunk.
20090801 15:31:15< melinath> Crab_: Embarrassing. I retract my bug report. Was the same as 14014 after all.
20090801 15:31:51< Crab_> melinath: np 
20090801 15:38:48-!- Baufo [n=quassel@wesnoth/developer/baufo] has quit [Remote closed the connection]
20090801 16:00:02-!- thespaceinvader [n=chatzill@wesnoth/artist/thespaceinvader] has quit ["ChatZilla 0.9.85 [Firefox 3.0.12/2009070611]"]
20090801 16:00:20-!- thespaceinvader [n=chatzill@wesnoth/artist/thespaceinvader] has joined #wesnoth-dev
20090801 16:10:41-!- Aethaeryn [n=Michael@wesnoth/umc-dev/developer/aethaeryn] has joined #Wesnoth-dev
20090801 16:17:30< CIA-62> crab * r37383 /trunk/ (69 files in 15 dirs): new AI configuration syntax. Note: boost 1.35 required for compilation
20090801 16:31:08< CIA-62> loonycyborg * r37384 /trunk/src/ai/configuration.cpp: Fixed a compile warning.
20090801 16:33:30< CIA-62> loonycyborg * r37385 /trunk/SConstruct: Bump boost version requirement in SConstruct.
20090801 16:37:21< CIA-62> crab * r37386 /trunk/CMakeLists.txt: cmake: changed required boost version to 1.35
20090801 16:37:26< CIA-62> crab * r37387 /trunk/configure.ac: automake: changed required boost version to 1.35
20090801 16:37:32< CIA-62> crab * r37388 /trunk/INSTALL: INSTALL documentation: changed required boost version to 1.35
20090801 16:46:28-!- silene [n=plouf@wesnoth/developer/silene] has quit ["Leaving."]
20090801 16:49:12< Aethaeryn> What does boost do?
20090801 16:49:15-!- silene [n=plouf@wesnoth/developer/silene] has joined #wesnoth-dev
20090801 16:50:02-!- Kenpachi [n=chatzill@CPE-58-170-85-76.sa.bigpond.net.au] has quit [Read error: 104 (Connection reset by peer)]
20090801 16:50:18< loonycyborg> Aethaeryn: It's like wesnoth-umc-dev, only for C++ libraries :P
20090801 16:51:46< boucman> it provides everything C++ should provide but doesn't :P
20090801 16:53:38< Aethaeryn> loonycyborg: so wesnoth-umc-dev provides everything Wesnoth should provide but doesn't? 
20090801 16:57:15< loonycyborg> wesnoth-umc-dev provides many quality campaigns, some of which are eventually mainlined.
20090801 16:57:51-!- Sirp [n=user@wesnoth/developer/dave] has joined #wesnoth-dev
20090801 16:57:54< loonycyborg> And some boost libraries will be added to upcoming new version of C++ standard..
20090801 16:58:12< silene> no
20090801 16:58:16< Sirp> Hello.
20090801 17:01:33< loonycyborg> At keast their interfaces :P
20090801 17:01:37< loonycyborg> *least
20090801 17:03:48< Sirp> the new standard will specify a standard library which includes many components that have their design influenced by Boost components, to varying degrees.
20090801 17:04:28< Sirp> I don't think any Boost components are being lifted and placed in the standard library verbatim.
20090801 17:07:06< boucman> esr: around ?
20090801 17:07:42< esr> Yes.
20090801 17:08:14< CIA-62> silene * r37389 /trunk/ (RELEASE_NOTES configure.ac src/foreach.hpp): Taken new Boost requirement into account.
20090801 17:08:14< boucman> I would like to rename an attribute like we did yesterday, would you have 5' to do the wmllint side of it ?
20090801 17:08:27< esr> Sure.  What attribute?
20090801 17:08:46< boucman> swing= in animation filters, I would like to rename it to value2
20090801 17:08:53< boucman> hmm
20090801 17:09:01< boucman> maybe second_value would be better
20090801 17:09:05< silene> "value2"? come on...
20090801 17:09:10< esr> I think so.
20090801 17:09:21 * boucman is open to other names...
20090801 17:09:30< silene> what does it do?
20090801 17:09:35< esr> What doesit actually mean?
20090801 17:09:59< boucman> currently it allows to filter on the swing number, but it only makes sense for animations that are related to attacks
20090801 17:10:19< boucman> I want to have it more generic, so it will be an "animation dependent" integer
20090801 17:10:30< boucman> the exact meaning for each anim type documented in the wiki
20090801 17:11:21< esr> Wouldn't it be better to have separate attributes for each type with more informative names?
20090801 17:11:26< boucman> I already have a value= parameter that does that, but I need a second one
20090801 17:11:36< esr> Even if they map to the same thing internally...
20090801 17:11:50< boucman> esr: it brings other problems
20090801 17:12:08< boucman> imagine I have an animation which is used both as a leading anim and a recruiting anim
20090801 17:12:35< boucman> totally different, so would need to specify the parameter twice ? what if they have conflicting value ? (assuming I save them at the same place)
20090801 17:13:35< silene> but didn't you just say they have different meaning? then the meaning for one will break the other animation, so they should have different names so that they don't conflict
20090801 17:14:27< boucman> true 
20090801 17:14:38< boucman> ok, let's just go for second_value
20090801 17:15:09< boucman> or rather value_second to be consistant with other stuff in anim WML
20090801 17:15:11< esr> This is insidewhat tag?
20090801 17:16:08< boucman> [animation] and \[.*_anim\]
20090801 17:16:44< esr> is the swing= attribute used anywhere else?
20090801 17:16:52< boucman> not that I know of
20090801 17:17:42-!- loktor [n=quassel@AAubervilliers-151-1-62-124.w83-112.abo.wanadoo.fr] has joined #wesnoth-dev
20090801 17:17:46< esr> Sounds pretty easy.  OK, commit your namechange; I'll do the wmllint logic.
20090801 17:19:39< boucman> gimme a minute
20090801 17:20:48-!- Crab_ [n=Crab_@wesnoth/developer/crab] has quit ["Leaving."]
20090801 17:22:20< boucman> esr: done, i'm updating wiki right now
20090801 17:22:21< CIA-62> boucman * r37390 /trunk/src/ (unit_animation.cpp unit_animation.hpp): renamed swing to value_second (preparing for take-off and landing animation)
20090801 17:22:38< esr> OK, working on it...
20090801 17:27:25< esr> Change made to wmllint.  Waiting for compile top finish so I can apply it.
20090801 17:29:30< melinath> esr: did you change "They spend this time watching for invaders, and combating the bandits and thieves who might spring up inside dwarvish territory." to "They spend this time watching for invaders, and fighting bandits and thieves who incur dwarvish territory."
20090801 17:29:40< melinath> esr: this is the dwarvish pathfinder desc.
20090801 17:31:41< esr> Because pplants "spring up" in English but bandits and theives don't - it's a bad use of the idiom.  I should fix the use of "incur" there, too.
20090801 17:31:57< thespaceinvader> encroach upon would fit
20090801 17:32:15< esr> Yes, that would work well.
20090801 17:33:29< esr> "spring up" carries with it an image of a bandit popping out of the ground like an ear of corn in an Iowa summer.  Not what we want.
20090801 17:34:00< melinath> If TSI's ok with changing it, fine with me.
20090801 17:34:12< melinath> Not that I exactly have much say :-p
20090801 17:34:22< melinath> but incur definitely doesn't work.
20090801 17:34:50< esr> You're right.  I made that change too quickly - distracted buy something else.
20090801 17:36:44-!- Elvish_Pillage2 [n=eli@71-10-231-36.dhcp.oxfr.ma.charter.com] has joined #wesnoth-dev
20090801 17:37:28< thespaceinvader> mm, encroach upon sounds fine to me
20090801 17:46:33-!- silene [n=plouf@wesnoth/developer/silene] has quit ["Leaving."]
20090801 17:47:48-!- ilor_ [n=user@wesnoth/developer/ilor] has joined #wesnoth-dev
20090801 17:47:54< CIA-62> esr * r37391 /trunk/data/ (3 files in 2 dirs): Attrubute change swing= ->second_value= requested by boucman.
20090801 17:48:21< boucman> thx esr
20090801 17:50:16< CIA-62> esr * r37392 /trunk/data/core/units/dwarves/Pathfinder.cfg: Fix usage error: "incur" -> "encroach upon".
20090801 17:52:32-!- Elvish_Pillager [n=eli@71-10-231-36.dhcp.oxfr.ma.charter.com] has quit [Read error: 110 (Connection timed out)]
20090801 17:58:43-!- silene [n=plouf@wesnoth/developer/silene] has joined #wesnoth-dev
20090801 18:07:55-!- allefant [n=allefant@allegro/developer/allefant] has joined #wesnoth-dev
20090801 18:09:14-!- wesbot changed the topic of #wesnoth-dev to: 65 bugs, 239 feature requests, 12 patches | logs: http://irclogs.wesnoth.org/ | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com | http://imagebin.org
20090801 18:09:42-!- melinath [n=melinath@pool-71-162-41-64.altnpa.east.verizon.net] has quit [Read error: 113 (No route to host)]
20090801 18:24:26-!- elias [n=allefant@allegro/developer/allefant] has quit [Connection timed out]
20090801 18:25:16-!- Crab_ [i=crab@wesnoth/developer/crab] has joined #wesnoth-dev
20090801 18:28:45-!- ilor_ [n=user@wesnoth/developer/ilor] has quit []
20090801 18:29:44-!- grzywacz [n=grzywacz@wesnoth/developer/grzywacz] has joined #wesnoth-dev
20090801 18:40:20< CIA-62> boucman * r37393 /trunk/src/unit_display.cpp: small detail I forgot in last commit
20090801 18:51:20-!- noy [n=Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20090801 19:04:19-!- giusef [n=giusef@unaffiliated/giusef] has joined #wesnoth-dev
20090801 19:25:52< Crab_> ilor: around ?
20090801 19:32:11-!- ilor [n=user@wesnoth/developer/ilor] has joined #wesnoth-dev
20090801 19:56:43-!- ilor_ [n=user@wesnoth/developer/ilor] has joined #wesnoth-dev
20090801 19:56:57< Crab_> boucman: do you have some time now to take a look at one piece of code ?
20090801 19:57:14< boucman> sure
20090801 19:57:18< Crab_> it's battle_context::battle_context
20090801 19:57:30< Crab_> src/actions.cpp:380
20090801 19:58:03< Crab_> I am puzzled by '} else if (defender_weapon == -1) {' line.
20090801 19:58:40< Crab_> why there's  'else' there,  why choose defender weapon only if attacker weapon is not chosen yet ?
20090801 20:01:30< boucman> my understanding would be that the code assumes that both weapons can't be invalid at the same time
20090801 20:01:44< Crab_> ok
20090801 20:01:56< boucman> not sure where that assumption comes from, though
20090801 20:02:29< Crab_> boucman: then, see this http://wesnoth.pastebin.com/m64134e89
20090801 20:02:41< Crab_> this is from src/ai/actions.cpp
20090801 20:03:28< Crab_> the question is:  what should be done if we were unable to select attacker's weapon ?
20090801 20:03:44< Crab_> there are two choices:
20090801 20:03:45< Crab_> 1) disallow the attack
20090801 20:04:33< Crab_> 2) do the attack as follows: remove 1 attack from unit and set movement to 0, then return
20090801 20:04:49< Crab_> (2) comes from attack::attack in src/actions.cpp
20090801 20:05:56< Crab_> (there's "// no attack weapon => stop here and don't attack"  part there)
20090801 20:06:45< Crab_> (1) is more similar to the reaction to other failed pre-ai-attack checks (e.g. attacking friendly unit, attacking petrified unit, etc)
20090801 20:06:49< boucman> i'm a bit confused at this point
20090801 20:07:16< Crab_> boucman: there are a set of circumstances in which we're unable to select attacker weapon.
20090801 20:07:35< Crab_> for example, if fai orders a unit without any attacks to attack.
20090801 20:07:57< Crab_> the question is: what to do in this case?
20090801 20:08:48< boucman> ok, I get the context now
20090801 20:08:56< boucman> hmm
20090801 20:09:04< Crab_> choice 1) is: disallow the attack, exactly as if ai ordered a unit to attack a statue, or as if ai ordered a unit to attack a friendly unit
20090801 20:09:21-!- ancestral [n=ancestra@166.205.5.15] has joined #wesnoth-dev
20090801 20:09:22< boucman> probably go for 1) I guess
20090801 20:10:29-!- ancestral [n=ancestra@166.205.5.15] has quit [Client Quit]
20090801 20:10:33< Crab_> choice 2) go on with the attack (add it to replay, get mp seed, etc), and let attack::attack do  " a_.get_unit().set_attacks(a_.get_unit().attacks_left()-1);   a_.get_unit().set_movement(-1);"
20090801 20:10:44< Crab_> (2) was the old behavior
20090801 20:11:43< Crab_> and it worked. then it was (recently) broken (during addition of a srng, code slightly changed)
20090801 20:11:58-!- ilor [n=user@wesnoth/developer/ilor] has quit [Read error: 110 (Connection timed out)]
20090801 20:12:09< boucman> i think go for 1... it's an AI bug, more or less to give such an order
20090801 20:12:20< Crab_> ok, thanks
20090801 20:13:29-!- ilor_ [n=user@wesnoth/developer/ilor] has quit []
20090801 20:14:45< Crab_> boucman: also note that ' code assumes that both weapons can't be invalid at the same time' is not a true assumption. there's a call to battle_context's constructor with both attacker's weapon and defender's weapon set to -1.
20090801 20:15:46< boucman> Crab_: my guess is that this is a bug...
20090801 20:15:59< Crab_> basically, the AI wants to 'attack using the best weapon of an attacker [taking our aggression into account], with defender choosing whatever he likes). so both weapons are not known.
20090801 20:16:20< Crab_> ok
20090801 20:17:00< boucman> most likely, the AI case has not been taken into account, since we always have an attacker weapon when the player choose an attack
20090801 20:18:12< Crab_> ok. btw, I now intend to switch ai_default to new attack/movement/recruit routines (and fully remove old ones)
20090801 20:19:24< boucman> Crab_: and i'm sure you'll enjoy doing that ;)
20090801 20:20:10< Crab_> boucman: of course. removing unneeded (and somewhat buggy) code is good :) and new bugs are always more interesting :) 
20090801 20:20:19< Crab_> it shouldn't take long
20090801 20:22:28< Crab_> then I'll turn my attention to refactoring fai, and introducing better debugging capabilities into it
20090801 20:24:33< Crab_> boucman: what do you think is a best way to interface with a 'fai debugger' ? (I'm thinking about a 'spawn a listening socket on localhost if wesnoth is launched with a specific parameter' approach)
20090801 20:25:34< boucman> hmm
20090801 20:26:00< boucman> I would do it completely within wesnoth, using a type of command line like the current formula command line
20090801 20:26:17< boucman> you might even want to do it with a specific FAI command syntax
20090801 20:27:12< Crab_> boucman: that will mean, for example, that, if user used a command-line fai command, it must not 'block' the wesnoth ui.
20090801 20:27:37< boucman> hmm
20090801 20:28:05< boucman> ok, let's think a bit what we want from an AI debugger
20090801 20:28:29< boucman> unlike a real debugger, we want it "read only" i.e no seting of variables
20090801 20:28:40< CIA-62> oron * r37394 /trunk/ (5 files in 3 dirs): 
20090801 20:28:40< CIA-62> revised fribidi support:
20090801 20:28:40< CIA-62> * Updated to support both fribidi2 (>= 0.19.0) and old fribidi (>= 0.10.9)
20090801 20:28:40< CIA-62> * The cmake part was done by Mark with some fixes add by me.
20090801 20:28:40< CIA-62> * I tested on Fedora-11, Mark tested on Debian stable.
20090801 20:28:52< boucman> the basic thing it should do is answer the question "why did the AI behave in the way it did"
20090801 20:29:06< boucman> i.e it's mainly a helper for AI behaviour
20090801 20:29:09< boucman> are you fine with that ?
20090801 20:29:27< Crab_> boucman: no, it should also ease the development of formulas
20090801 20:29:50< boucman> hmm
20090801 20:30:04< Crab_> boucman:  for that, we want the ability to: 1) inspect fai data structures  2) set (conditional) breakpoints on fai execution
20090801 20:30:21< Crab_> to allow to see the 'state' of fai during a specific moment of the execution.
20090801 20:30:36< Crab_> now, various debug_print's are used for this purpose
20090801 20:31:35< Crab_> also, we would need to allow stepping through formulas.
20090801 20:32:02< boucman> breakpoints are tricky... 
20090801 20:32:34< Crab_> boucman: why ?
20090801 20:33:02< boucman> i was still thinking of a command line interface, and making the interface non-blocking, while still preventing user from doing silly stuff
20090801 20:36:11< Crab_> silly stuff = any fai commands, any moves/attacks
20090801 20:36:34< boucman> yes, that sort of stuff
20090801 20:38:42-!- ilor [n=user@wesnoth/developer/ilor] has joined #wesnoth-dev
20090801 20:38:55< boucman> Crab_: what do we try to debug, the AI engine, or the formula engine, or both as a common problem
20090801 20:39:12< Crab_> we want to debug a formula engine
20090801 20:39:38< Crab_> and, formula engine has access to the various ai parts
20090801 20:40:19< boucman> i don't know the intricates of FAI, but
20090801 20:40:49< boucman> how about a special formula "start_debug(<formula>,<condition>)
20090801 20:41:18< boucman> when start_debug is evaluated, a window pops up, in which you can go step by step and closes once start_debug is evaluated
20090801 20:41:26< Crab_> boucman: and how would ai turn be debugged without changing the fai formula scripts ?
20090801 20:41:27< boucman> might be doable with the new gui
20090801 20:41:54< boucman> Crab_: if you are debugging an AI, you have access to the formula scripts
20090801 20:42:28< Crab_> boucman: yes, there is access to them. but is there a reason to change them ?
20090801 20:42:51< Crab_> I agree about 'having a special window for debug'
20090801 20:43:02< boucman> well, that's an easy and generic way to place breakpoint for the most common case, which is the case of someone actually writing an AI
20090801 20:43:43< boucman> specify a breakpoint anywhere in a formula would be extremely complicated to do
20090801 20:44:45< Crab_> boucman: well, standard debuggers handle this case by allowing to set a BP on a 'line of source'
20090801 20:45:21< boucman> yes, but we don't have that info once the fai is parsed
20090801 20:45:59< Crab_> boucman: but, what if we 'store' it when the fai is parsed ? as  'debug info'.
20090801 20:46:35< boucman> Crab_: i'm trying to balance complexity with developement cost ans SoC time
20090801 20:46:50< boucman> parsing that info is of course possible, but...
20090801 20:47:07< boucman> do you think you would manage to add debugging info to formulas in the remaining time ?
20090801 20:47:55< Crab_> boucman: yes.  I've looked into it in spring, and it's possible.
20090801 20:48:03< boucman> ok
20090801 20:50:13< boucman> Crab_: today is aug 1, sugested pencil down is aug 10... I think the first thing you should do is update your timeline, because you won't do what was originally planned
20090801 20:50:42< boucman> that's not a problem, but I need a bit more visibility on what's comming next
20090801 20:51:10< Crab_> boucman: then, 1) I'll put up a wiki page section describing my proposed plans for fai debugger  2) I'll update my wiki page with progress/plans/timeline
20090801 20:51:31< boucman> Crab_: ok, do that, then we'll discuss some more
20090801 20:51:38< Crab_> ok
20090801 20:54:12-!- giusef [n=giusef@unaffiliated/giusef] has quit ["exit (-1);"]
20090801 21:09:41-!- BenUrban is now known as BenUrban_
20090801 21:11:18-!- BenUrban [n=benurban@c-68-50-54-86.hsd1.md.comcast.net] has joined #wesnoth-dev
20090801 21:13:19-!- BenUrban_ [n=benurban@unaffiliated/benurban] has quit [Nick collision from services.]
20090801 21:26:50-!- Blueblaze [n=nick@c-98-199-143-139.hsd1.tx.comcast.net] has joined #wesnoth-dev
20090801 21:32:06-!- blarumyrran [n=minaise@81-20-159-197.levira.ee] has joined #wesnoth-dev
20090801 21:34:47< CIA-62> thespaceinvader * r37395 /trunk/data/core/images/units/drakes/ (glider-fly-4.png glider-fly-5.png): Fix broken TC on two Drake Glider frames
20090801 21:35:57< thespaceinvader> wesbot seen jetrel
20090801 21:35:57< wesbot> thespaceinvader: The person with the nick Jetrel last spoke 22h 59m ago.  22h 44m ago person left:
20090801 21:40:09< Crab_> boucman: 1) http://wesnoth.org/wiki/FormulaAIDebugger , with plan to do a working version in 2 weeks.
20090801 21:40:26 * boucman clicks
20090801 21:41:28< boucman> Crab_: the problem of telnet to localhost is that it's something windows users don't know about
20090801 21:42:25< Crab_> boucman: but aren't we targeting AI/UMC developers, which know about things like svn / puTTy  ?
20090801 21:43:07< Crab_> plus, it can be a simple (for example, python) frontend
20090801 21:44:05< Crab_> but, in-game debug window is good, too
20090801 21:44:16< boucman> Crab_: umc devs don't know svn/tty that's for sure
20090801 21:44:23< Crab_> ok )
20090801 21:44:33-!- lizard_r [n=Miranda@wesnoth/umc-dev/developer/lizard] has quit ["Saurian Augur - I'll heal you by 4 hp if you leave next to me"]
20090801 21:44:38< Crab_> so, you are in favor of in-game debug window ?
20090801 21:45:36< Crab_> (it can be done by doing something like   'while (ai::debugger::stop_flag_raised()) { raise_user_interact(); }' from within fai code
20090801 21:46:14< boucman> Crab_: do it via telnet (or even stdin stdout) for a first go, 
20090801 21:46:19< boucman> and add the debug window later
20090801 21:48:07< Crab_> boucman: ok
20090801 21:48:16< Crab_> boucman: and what about the featurelist ?
20090801 21:49:17< boucman> it sounds good, but making it in time will be quite a challange
20090801 21:49:32< Crab_> I want to make a specific object which will 'intercept' fai execution at specific points
20090801 21:49:49< Crab_> then, it would be able to collect info about a 'fai call stack', sort of
20090801 21:50:43< boucman> ok
20090801 21:50:58< Crab_> then, it would be possible to check 'if current state of fai matches one of specified 'breakpoint \ logpoint ' expressions ', and react accordingly
20090801 21:51:32< Crab_> by making fai debugger out-of-band (via a socket), it would be possible, for example, to use the wesnoth window for something else during debugging
20090801 21:52:05< boucman> let's do it in the simplest and fastest way for the moment
20090801 21:52:12< Crab_> ok
20090801 21:52:20< boucman> once we see how it's used in practice, we will have more insight
20090801 21:52:38< boucman> I think in game window is most likely the way to go, but i'd need to see how it's used first
20090801 21:52:58< Crab_> ok
20090801 21:53:15-!- Turuk_ [n=Turuk@wesnoth/forumsith/turuk] has quit [Connection reset by peer]
20090801 21:53:30-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090801 22:21:40< shadowmaster> corn: okay, I don't know about objective duplication issues, that sounds like an engine problem
20090801 22:22:26< shadowmaster> btw, I'm afk atm; I bought a replacement HDD for my desktop and I'll have to install a new system on it, and afterwards, fix the house's ground connection
20090801 22:23:16-!- Amu [n=smar@a88-113-60-192.elisa-laajakaista.fi] has quit [Read error: 60 (Operation timed out)]
20090801 22:24:28-!- Amu [n=smar@a88-113-60-192.elisa-laajakaista.fi] has joined #wesnoth-dev
20090801 22:26:12-!- ABCD_ [n=ABCD@gentoo/contributor/abcd] has joined #wesnoth-dev
20090801 22:26:16-!- ABCD [n=ABCD@gentoo/contributor/abcd] has quit [Read error: 104 (Connection reset by peer)]
20090801 22:31:09-!- Elvish_Pillager [n=eli@71-10-231-36.dhcp.oxfr.ma.charter.com] has joined #wesnoth-dev
20090801 22:31:54-!- Elvish_Pillage2 [n=eli@71-10-231-36.dhcp.oxfr.ma.charter.com] has quit [Read error: 60 (Operation timed out)]
20090801 22:32:34-!- noy [n=Noy@wesnoth/developer/noy] has quit []
20090801 22:35:23< Crab_> boucman: do you (by chance), know the purpose of the hack on src/ai/default/ai.cpp:460 ? ( itor->second.set_movement(itor->second.total_movement()-1); ) ?
20090801 22:36:27-!- loktor [n=quassel@AAubervilliers-151-1-62-124.w83-112.abo.wanadoo.fr] has quit [Remote closed the connection]
20090801 22:38:22< boucman> Crab_: I can't understand how there could be a unit at "from" at that point
20090801 22:38:45< Crab_> boucman:  from = readwrite_context_proxy::move_unit(..)
20090801 22:39:07< Crab_> boucman: so, 'from' = 'new location of a leader, (on keep)'
20090801 22:39:10< boucman> oh, yes
20090801 22:39:34< Crab_> boucman: then the AI violates the rules of the game, and I'm wondering "why?"
20090801 22:40:48< boucman> no idea
20090801 22:40:50< Crab_> note that if the comment ' and we're going to move there immediately' is wrong - there's recruitment phase, and 'there' can be occupied by newly recruited unit.
20090801 22:40:52< boucman> still looking into it
20090801 22:41:13< boucman> I am not sure the original writer thought of that
20090801 22:42:04< boucman> I understand why the code "works" but I'm not sure why the code restores the MP in the middle
20090801 22:42:50< Crab_> well, note that readwrite_context_proxy::move_unit does '  u->second.set_movement(0);'
20090801 22:43:08< Crab_> but it is not clear why move_unit_partial is not used instead
20090801 22:43:28< CIA-62> silene * r37396 /trunk/src/scripting/lua.cpp: Ensured WML tables are not merged.
20090801 22:43:39< boucman> did move_unit_partial exist at the time it was written ?
20090801 22:43:53< boucman> at least you explain why it restores MP
20090801 22:47:33< Crab_> move_unit_partial is r5800
20090801 22:48:53-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit [Read error: 110 (Connection timed out)]
20090801 22:49:18< Crab_> and that hack is r1971
20090801 22:49:50-!- silene1 [n=plouf@AMarseille-251-1-70-114.w83-205.abo.wanadoo.fr] has joined #wesnoth-dev
20090801 22:49:53< Crab_> boucman: so, you're right 
20090801 22:50:05< boucman> :)
20090801 22:50:29< Crab_> thanks.
20090801 22:50:32-!- busfahrer [n=busfahre@unixboard/user/busfahrer] has joined #wesnoth-dev
20090801 22:50:33 * boucman 's eveyday job is to whack old code and understand it... you develop a state of mind...
20090801 22:51:19 * Crab_ is looking to kill set_movement(x) calls from within the ai. set_movement(x) is something that human is not allowed to do, and it can cause various subtle bugs.
20090801 22:52:49< boucman> you have my blessing
20090801 22:52:52< boucman> :P
20090801 22:53:25-!- noy [n=Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20090801 22:53:50-!- silene [n=plouf@wesnoth/developer/silene] has quit [Nick collision from services.]
20090801 22:53:55-!- silene1 is now known as silene
20090801 22:56:28< Crab_> boucman: thanks :)
20090801 23:01:38-!- Espreon [n=espreon@wesnoth/developer/espreon] has joined #wesnoth-dev
20090801 23:07:57< boucman> night all
20090801 23:08:15-!- boucman [n=rosen@wesnoth/developer/boucman] has quit ["Leaving."]
20090801 23:09:45-!- silene [n=plouf@wesnoth/developer/silene] has quit ["Leaving."]
20090801 23:11:16-!- busfahrer [n=busfahre@unixboard/user/busfahrer] has quit ["leaving"]
20090801 23:19:40< CIA-62> crab * r37397 /trunk/src/ai/ (actions.cpp actions.hpp): added some more pre-attack-checks to ai::attack_result_ptr
20090801 23:26:03-!- Espreon [n=espreon@wesnoth/developer/espreon] has quit [Remote closed the connection]
20090801 23:26:50-!- Espreon [n=espreon@wesnoth/developer/espreon] has joined #wesnoth-dev
20090801 23:27:38-!- Espreon [n=espreon@wesnoth/developer/espreon] has quit [Client Quit]
20090801 23:32:55-!- Espreon [n=espreon@wesnoth/developer/espreon] has joined #wesnoth-dev
20090801 23:48:59-!- BenUrban is now known as BenUrban_
20090801 23:53:14-!- BenUrban_ [n=benurban@unaffiliated/benurban] has quit ["Power failu"]
--- Log closed Sun Aug 02 00:00:17 2009