--- Log opened Wed May 13 00:00:16 2009
20090513 00:05:12-!- stikonas [n=quassel@wesnoth/translator/stikonas] has quit [Read error: 110 (Connection timed out)]
20090513 00:09:13-!- wesbot changed the topic of #wesnoth-dev to: 59 bugs, 239 feature requests, 9 patches | logs: http://irclogs.wesnoth.org/ | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com | http://imagebin.org
20090513 00:11:55-!- ABCD_ [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090513 00:12:31-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Read error: 104 (Connection reset by peer)]
20090513 00:13:18-!- ABCD_ is now known as ABCD
20090513 00:15:33-!- giusef [n=giusef@unaffiliated/giusef] has quit [Client Quit]
20090513 00:24:13-!- stikonas_ [n=quassel@ctv-217-147-36-179.vinita.lt] has quit [Remote closed the connection]
20090513 00:31:54-!- ABCD [n=ABCD@wikipedia/ABCD] has quit ["No Ping reply in 90 seconds."]
20090513 00:31:56-!- loonycyborg [n=sergey@wesnoth/developer/loonycyborg] has quit ["Zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz"]
20090513 00:32:26-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090513 00:32:36-!- loonybot [n=loonybot@wesnoth/bot/loonybot] has quit [Remote closed the connection]
20090513 00:42:35-!- Andrai [n=Andrai@150.165.63.86] has quit [Read error: 104 (Connection reset by peer)]
20090513 00:43:33-!- mjs-de [n=mjs-de@vpw.wh.uni-dortmund.de] has quit [Read error: 113 (No route to host)]
20090513 00:49:32< Ivanovic> time for my bed, n8
20090513 00:57:40-!- BenUrban [n=benurban@c-68-55-19-224.hsd1.md.comcast.net] has joined #wesnoth-dev
20090513 01:06:23-!- faryshta [n=faryshta@189.181.156.134] has joined #wesnoth-dev
20090513 01:18:55-!- BenUrban_ [n=benurban@c-68-55-19-224.hsd1.md.comcast.net] has joined #wesnoth-dev
20090513 01:21:29-!- Turuk_ [n=Turuk@wesnoth/forumsith/turuk] has quit [Read error: 104 (Connection reset by peer)]
20090513 01:24:02-!- YogiHH [n=chatzill@d001120.adsl.hansenet.de] has left #wesnoth-dev []
20090513 01:25:39-!- BenUrban [n=benurban@unaffiliated/benurban] has quit [Nick collision from services.]
20090513 01:25:42-!- BenUrban_ is now known as BenUrban
20090513 01:33:50-!- PK [n=pk@r74-192-44-206.vctrcmta01.vctatx.tl.dh.suddenlink.net] has joined #wesnoth-dev
20090513 01:34:32< PK> I'm guessing none of the devs here would know where I can find out how to define a Sound list for wesnoth.
20090513 01:37:16-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090513 01:52:06-!- Elvish_Pillage2 [n=eli@71-10-228-131.dhcp.oxfr.ma.charter.com] has quit ["Hi! I'm a quit message virus vaccine. If you see a quit message virus, don't replace your quit message with it!"]
20090513 01:55:29-!- ancestral [n=masterpr@97-116-120-23.mpls.qwest.net] has joined #wesnoth-dev
20090513 01:56:22-!- thespaceinvader [n=chatzill@wesnoth/artist/thespaceinvader] has quit ["ChatZilla 0.9.84 [Firefox 3.0.10/2009042316]"]
20090513 02:03:55-!- Polarina [n=polarina@wesnoth/translator/Polarina] has quit ["Leaving."]
20090513 02:05:17-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Read error: 104 (Connection reset by peer)]
20090513 02:05:38-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090513 02:08:14-!- ikarius [n=ross@smtp.gridironsystems.com] has quit []
20090513 02:10:43< ancestral> Don't know who oversees the download page, but it says Current Version and Previous Version for OS X both at 1.6
20090513 02:11:51< ancestral> (I'm assuming it's 1.6.2 and 1.61?)
20090513 02:13:20< Turuk> Fixed.
20090513 02:37:46< CIA-30> alink * r35615 /trunk/src/game_display.cpp: 
20090513 02:37:46< CIA-30> Optimize terrain rendering.
20090513 02:37:46< CIA-30> An old option allow to use Time-of-Day terrain mask, also affected by local ToD (like cavern).
20090513 02:37:46< CIA-30> But the used ToD function uselessly calculated lawful bonus given by adjacent units having
20090513 02:37:46< CIA-30> the illuminates ability, which was not cheap and done for each hex. Fixed now, no visible change.
20090513 02:39:59< alink> ideally we should divide ToD structure in two parts: game-rule and visual rule, so we could easily ask only what we need, but a lot of work for little gain
20090513 02:40:21< alink> btw does someone still use ToD terrain mask? 
20090513 02:40:56< alink> I don't mind let it live when it will have no cost but I still have one thing to clean for that
20090513 02:41:01-!- ABCD_ [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090513 02:41:53< alink> Using different ToD masks in cavern seems a good idea and is more flexible than the normal global ToD coloring but I think it will always look silly
20090513 02:42:15-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Read error: 104 (Connection reset by peer)]
20090513 02:42:23-!- ABCD_ is now known as ABCD
20090513 02:42:28 * alink will upload a screenshot to show what he mean
20090513 02:43:42< alink> http://img228.imageshack.us/img228/1230/todcavern.png
20090513 02:44:09< alink> ok it's not fair to use the fog image for that, but you get the idea
20090513 02:45:00< Crab_> alink, well, that old 'time_of_day' function is intended mainly for combat modifier purposes...
20090513 02:45:01< alink> whatever image you uses, these hard edges between hexagons will always look bad and never mix well with the pseudo-3D of our layer system
20090513 02:46:18< alink> Crab_: yes but we used it for rendering too, I spotted it in one of your graph (took half of draw_hex() IIRC )
20090513 02:46:44< Crab_> alink: *used* is a good word here :)
20090513 02:47:57< alink> If we use the global ToD mask in cavern too (like we do for normal ToD coloring) that could be more optimized
20090513 02:48:19< alink> and I think it look ugly so...
20090513 02:49:31< alink> So if no one can show a ToD mask working well with the limit separating cavern and outside, I will simplify that
20090513 02:52:50< alink> I guess I can quickly try to gimp something better and see how it look
20090513 02:54:00-!- Espreon [n=espreon@wesnoth/developer/espreon] has quit ["Tengo que ir... Yeahzorz..."]
20090513 02:56:26< ABCD> while reading through recent diffs, I noticed that there seems to be a strange mixture of tabs and spaces used in the various .cpp files... I think that either one or the other should be used, for consistency
20090513 02:56:55< Turuk> Is there a reason that the Sergeant, in the unit description, shows Advances to: Lieutenant, but that the Lieutenant does not have Advances from: Sergeant, only Advances to: General?
20090513 02:57:06< alink> ABCD: i think we all agree, but some IDE seems not ;)
20090513 02:57:39< alink> Turuk: i check 
20090513 02:58:09< Turuk> Thanks alink.  Nothing wrong with the .cfg, but it just does not display properly in the help due to some deeper matter I am sure
20090513 02:58:50< alink> ok with a transparent black ToD mask it loos less ugly http://img216.imageshack.us/img216/8422/todcavern2.png   but still not nice IMHO
20090513 02:59:51< Crab_> alink: imo, only the borders look bad...
20090513 03:01:36< Crab_> alink: btw, you can try to get 'average' light level from several nearby hexes - thus it will 'darken' gradually, and look nicer
20090513 03:01:39< alink> It's true that on the gray cavern tile, the limit is less visible, not sure that it work well other tiles though
20090513 03:02:54< alink> Crab_: me no that means adding again expensive code checking adjacent hexes etc, but I think it's doable in wml and can even affect game bonus
20090513 03:03:58< alink> like adding a third ToD zone at the limit
20090513 03:04:58< Crab_> alink: there's no reason to do that each turn, or take illumunate (=moving light source) into account. Can't it be done as "calculate-once-per-all-times-of-day-then-cache" ?
20090513 03:05:29< alink> yes we can do that, but this remove power from WML
20090513 03:05:36< Crab_> it will help "borders", too.
20090513 03:06:12-!- Blueblaze [n=nick@c-98-199-143-139.hsd1.tx.comcast.net] has joined #wesnoth-dev
20090513 03:06:40< alink> for example i can use an image having special detail in it (fog, snow), so the gradual progression will look less nice
20090513 03:07:09< alink> plus some hex may be between 3 different ToD zone, so the gradual thing continue to more complex
20090513 03:07:23< alink> *be more
20090513 03:08:01< Crab_> alink: nothing too complex, imo - just 'average value biased to favor current hex if it contains special effect'
20090513 03:08:31< alink> I suppope that if we really want to go in this direction, we should use the same transition system has we use for fog, so it will be fully customisable
20090513 03:10:12< alink> and allow some nice effect for casting shadow or light
20090513 03:10:46< Crab_> well, since the maps in wesnoth are almost static, we can pre-calculate sets of shadows (for each tod) (they won''t be changed often).
20090513 03:13:21< alink> another problem is that it is render on a high layer, a bit hard to make it work well in all pseudo-3D cases
20090513 03:14:26< alink> btw such "pre-calculate sets of shadows" could be standard terrain transitions
20090513 03:14:54< Crab_> sorry, I'm not familiar with "standard terrain transitions" :)
20090513 03:15:44< alink> look in data/core/images/terrain, there is little piece of tile used as "transition" between terrain
20090513 03:16:21< Crab_> alink: thanks, got it.
20090513 03:16:32< alink> especially fog-*.png because which transition is used is decided at render-time
20090513 03:17:05< alink> normal terrain transition are decided at map creation by more complex terrain WML rules
20090513 03:18:54< alink> recycling the fog transition system by use a shadows instead could be easy and maybe look good, just need to replace the fogged test on location by has-special-ToD test
20090513 03:19:09< alink> shadow or light, not sure what works better
20090513 03:23:04-!- faryshta [n=faryshta@189.181.156.134] has quit ["Saliendo"]
20090513 03:23:22< alink> Turuk in a campaign specifically ? (the sergeant thing)
20090513 03:23:29< alink> and also in debug  mode ?
20090513 03:23:35< alink> (all seems fine here)
20090513 03:23:46< Turuk> Hmm? I noticed it when playing in a campaign earlier
20090513 03:23:54< alink> ah and 1.6 or trunk ?
20090513 03:24:09< Turuk> 1.6.2
20090513 03:24:20< alink> ok i test with 1.6
20090513 03:24:34< Turuk> Lieutenant just did not display Advances from: Sergeant.
20090513 03:27:19< Turuk> Alink: Never mind, I tested it again with debug and local myself and it works fine
20090513 03:27:29< Turuk> It must just have been a random thing.
20090513 03:28:17< alink> ok maybe you can verify which related units are in your encountered_units key in your preference file
20090513 03:28:44< Turuk> Though now I cannot get General to display it comes from Lieutenant :P
20090513 03:29:12< alink> but not encountered unit are supposed to show link with a (?)
20090513 03:30:09< Turuk> Right, and this does not show that at all
20090513 03:30:30< Turuk> example: http://imagebin.org/48844
20090513 03:31:10< alink> weird, I check the 1.6 code
20090513 03:32:34< Turuk> Yes, I double checked my encountered_units and all 4 of that unit line are in there, so I am not sure why General does not show that it advances from Lietuenant
20090513 03:32:40-!- Sapient [n=sapien-x@c-71-204-10-166.hsd1.ga.comcast.net] has joined #wesnoth-dev
20090513 03:32:48< alink> hi Sapient
20090513 03:32:53< Sapient> hi alink
20090513 03:33:11< Sapient> I read the irc discussion; it sounds like your changes to the advances_from code
20090513 03:33:27< alink> Turuk: the only possibility that i see is missing unit or someone adding hide_help=yes in that unit
20090513 03:33:40< alink> Sapient: changed in trunk and it's an 1.6 bug
20090513 03:33:46< Turuk> Well, I'm testing it now in just a local multiplayer game to test it out
20090513 03:34:02< Turuk> And so the only code I would be using is the regular game
20090513 03:34:08< Sapient> oh.
20090513 03:34:26< Turuk> And with it in my encountered_units, I think it might have to do with the unit somehow not reading what it advances from
20090513 03:35:08< alink> just to be sure, in debug mode too ?
20090513 03:35:31< Turuk> Yup
20090513 03:36:39< alink> at least in trunk advances_from is simplified a lot, it has possibly fixed that one ;-)
20090513 03:37:00< Sapient> there's a reason it was complex...
20090513 03:38:07< Sapient> some bug fix involving shamans and UtBS IIRC
20090513 03:39:11< Turuk> I can try it with trunk, though I feel that I remember noticing this before
20090513 03:39:20< Turuk> (I know feelings count for little without facts) :P
20090513 03:39:39< alink> Sapient: not really, from what i saw, it tried to absolutely generate and cache advance_from at loading time (when all unit are not yet know) but we only needed that info when displaying an help page
20090513 03:39:55< Turuk> I also noticed the propensity now to display something like Unit A, Unit A (?)
20090513 03:40:16< Turuk> Instead of just showing it once
20090513 03:40:56< alink> Turuk: I think it always did that, but trunk has a new [hide_help] feature that may help to fix that
20090513 03:41:17< Turuk> So it just hides the inconsistencies?
20090513 03:41:27< Sapient> alink, since you removed future_advancefroms, I suspect the old bug might be reintroduced
20090513 03:42:15< alink> Turuk: no it allow WML author to hide such thing, but no one use it yet (only me and changelog know about it)
20090513 03:42:27< Turuk> Ah, I see
20090513 03:42:37< alink> Sapient: checking this future_advancefrom thing
20090513 03:44:10< alink> but i admit that i don't see why just checking all unit having advance_to pointing to a specific  unit will not work
20090513 03:44:11< Sapient> wesbot: bug 11813
20090513 03:44:11< wesbot> Bug #11813  Assigned to: Patrick P <sapient>  Status: Fixed  Priority: 6
20090513 03:44:14< wesbot> Summary: Gender-specific unit advancements cause segmentation fault
20090513 03:44:17< wesbot> Original submission: The attached WML code for a unit causes segmentation faul
20090513 03:44:20< wesbot> t on Wesnoth 1.5.0+svn when it is used in a custom campaign. The issue triggers 
20090513 03:44:23< wesbot> URL: https://gna.org/bugs/?11813
20090513 03:44:25< wesbot> Attached file (1st): https://gna.org/bugs/download.php?file_id=4419
20090513 03:44:39< Sapient> that was the bug
20090513 03:45:39< alink> well add_advancefrom is dead, so this bt is maybe not valid anymore
20090513 03:46:29< Turuk> Cavarlyman does it too
20090513 03:47:00< Turuk> http://imagebin.org/48846
20090513 03:47:05< Turuk> I'm just combing through them now
20090513 03:47:12< alink> Sapient: and i see that lazy_loading tried to work on advance_form before building all unit types, this is what i changed
20090513 03:47:20< Turuk> Longbowman....
20090513 03:47:40< alink> Turuk: really puzzling
20090513 03:47:44< Turuk> All of the HI line
20090513 03:48:14< alink> Turuk: also happening after just starting wesnoth and clicking on help ?
20090513 03:48:18< Turuk> At this point, I'm going to assume that this is fairly prevalent behavior
20090513 03:48:23< Turuk> alink: let me go look
20090513 03:49:58< Turuk> Odd ok
20090513 03:50:04< Turuk> If you access it from Help
20090513 03:50:05< Turuk> It works fine
20090513 03:50:15< Turuk> But when accessing it from Unit Description, it does not
20090513 03:50:45< alink> then it is probably causes by a campaign or a mutliplayer era
20090513 03:51:16< alink> i suggest removing (or just move) your add-on and check again
20090513 03:51:43< alink> or maybe just try with a mainline era 
20090513 03:52:00< Turuk> Done
20090513 03:52:12< Turuk> Fresh wesnoth, with only default everything :)
20090513 03:52:25< Turuk> Works fine when viewed from help, messed up when viewed from Unit Description
20090513 03:52:50< alink> with what campaign, so i can try that too
20090513 03:55:30< alink> Sapient : seems safe to me now, http://wesnoth.pastebin.com/d27749130  at least it's much simpler to debug
20090513 03:55:52< Sapient> alink, yes after looking at your code I see it has no chance to crash
20090513 03:56:01< Turuk> Tale of Two Brothers
20090513 03:56:17< alink> worst case scenario bad advance_to cause bad advance_from
20090513 03:56:31< alink> wesbot log r35514 
20090513 03:56:32< wesbot> alink * r35514 : Generate and cache advance_from info needed a lot of tricky work during lazy-loading.
20090513 03:56:35< Turuk> In particular, just use the horseman that you are given at campaign/scenario start
20090513 03:56:36< wesbot> URL: http://svn.gna.org/viewcvs/wesnoth?view=rev&rev=35514
20090513 03:56:43< Turuk> Shows up fine in Help, not in User Description
20090513 03:56:58< alink> sapient : previous code was horrible IMO 
20090513 03:57:03< alink> Turuk: checking
20090513 03:57:03< Sapient> agreed
20090513 03:57:37< Turuk> Hope that helps alink, I have to vacate my current premises as they are closing.  Good luck.  If you do figure it out, leave me a note in the logs please.
20090513 03:57:42-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit ["The man who smiles when things go wrong has thought of someone to blame it on."]
20090513 03:57:45< Sapient> alink: it was ugly for a purpose at least; lazy-loading offers a lot of speed boosts
20090513 03:58:11< Sapient> your change works well since the Help is the only place calling advances_from
20090513 03:58:21< Sapient> otherwise it might be a problem
20090513 03:59:20< alink> yes but caching advance_from was useless, so instead of fixing this caching to make it work with lazy_loading was a bad idea
20090513 03:59:56< alink> but indeed if someone start to need it at a time-critical place, some caching will be needed.
20090513 04:00:17< alink> but I can't imagine such use
20090513 04:01:10< alink> and i fear that fixing all the bug caused by lazy-loading hurted its initial speed boost
20090513 04:01:31< Sapient> possibly
20090513 04:02:00< alink> plus the code still need some cleaning to be easily maintenable
20090513 04:03:08< alink> Also, i am not sure that a hierarchic division was the optimal solution
20090513 04:03:43< alink> each time you need some info from the last loading level you need to build the previous ones
20090513 04:04:22< alink> so any function trying to access this info on a big number of units, fully build them
20090513 04:04:53< Sapient> I don't know enough about the speed problem cases... I think that was YogiHH's idea. though
20090513 04:05:25< alink> plus there is no 'auto-loading when not build enough' which is a source of potential bug
20090513 04:07:22< Sapient> anyways, good night
20090513 04:07:30< alink> Sapient: gn
20090513 04:07:32-!- Sapient [n=sapien-x@wesnoth/developer/sapient] has quit []
20090513 04:10:29-!- Ivanovic_ [n=ivanovic@dtmd-4db26f7b.pool.einsundeins.de] has joined #wesnoth-dev
20090513 04:13:17< alink> to be honest, it's because i recently found a "fully build all units" in AI code (to evaluate recruiting)
20090513 04:16:04< alink> which first killed the memory gain at first AI turn, and second was slow each times because of costly search calls (but changing the code flow helped that)
20090513 04:18:15< alink> we still such do such search (needed by lazy-loading) but at least only once by unit_type
20090513 04:26:18-!- Ivanovic [n=ivanovic@wesnoth/developer/ivanovic] has quit [Read error: 113 (No route to host)]
20090513 04:26:27-!- Ivanovic_ is now known as Ivanovic
20090513 04:32:06< alink> Turuk: I can reproduce your problem, seems fixed in trunk, so i suspect that lazy loading code recently cleaned there
20090513 04:33:10< alink> grrr i flushed that code to avoid dealing with its unneeded complexity and now i need to check its ghost in 1.6 :-/
20090513 04:36:23< alink> otoh that means that i fixed a bug without even knowing about it :-)
20090513 04:37:22< happygrue> alink: you are just _that_ good.  ;)
20090513 04:38:01< alink> let's just say that i try to see the good side of things :)
20090513 04:38:18< happygrue> hehe
20090513 05:00:28-!- ardesh_ [n=ardesh@port-92-206-240-86.dynamic.qsc.de] has joined #wesnoth-dev
20090513 05:17:48-!- ardesh [n=ardesh@port-92-206-127-115.dynamic.qsc.de] has quit [Read error: 110 (Connection timed out)]
20090513 05:25:59< alink> Turuk: found it
20090513 05:26:06-!- Crab_ [n=Crab_@wesnoth/developer/crab] has quit ["Leaving."]
20090513 05:26:42< alink> but sadly it's me who caused this little regression in 1.6.2
20090513 05:29:44< alink> i continue to think it's because of this bad code, but i noticed the ugly line and didn't changed it because it was 1.6.x
20090513 05:30:19< alink> I mean the ugly line causing unintuitive behavior
20090513 05:30:53< alink> hopefully the bug is minor, it only causes some missing advance_from, no possible crash
20090513 05:31:44< alink> and the only unit having these missing advance_from are the units created when starting a game
20090513 05:32:49< alink> ah no correction I mean units having advances from pointing to these units
20090513 05:34:57< alink> so for example in MP only advancements of the leader will have missing advance_from
20090513 05:35:26< alink> mmh a bit more annoying for campaign :(
20090513 05:37:32< alink> well you still have access to the info via unit_description on a lower unit or by clicking on the page in the left panel, but no advance_from links
20090513 05:40:03< alink> one good thing is that starting a campaign or a savegame where you begin with only your leader and maybe one other unit (common case i think) will only cause the bug for these 2 advancement trees
20090513 05:46:42< CIA-30> alink * r35616 /branches/1.6/src/unit_types.cpp: 
20090513 05:46:42< CIA-30> revert one line of r35513 causing some advance_from in help
20090513 05:46:42< CIA-30> for the advancement tree of the units present on map when starting
20090513 05:46:57< alink> fixed but killed the AI optimization
20090513 05:47:27< alink> but I think i can get it back
20090513 05:56:12< alink> damned lazy-loading, when a unit is 'fully build", in fact it is not, you need to half-build all the other to make this true
20090513 06:07:34-!- noy [n=Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20090513 06:09:13-!- wesbot changed the topic of #wesnoth-dev to: 58 bugs, 241 feature requests, 9 patches | logs: http://irclogs.wesnoth.org/ | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com | http://imagebin.org
20090513 06:10:00< alink> pfff i don't see a way to get the speed back, unless backporting the simplification done on trunk
20090513 06:10:34< alink> but even if safe, i feel less confident touching this in 2
20090513 06:10:45< alink> s/2/1.6
20090513 06:11:09< alink> and more decided than never to clean this mess in 1.7
20090513 06:13:00< alink> btw i quote myself :
20090513 06:13:12< alink> 04:54 <alink> I will backport it to 1.6 tomorrow, there is only a tiny change of the code flow, seems valid but I prefer check that with fresh eyes
20090513 06:13:34< alink> guess they were not fresh enough :-/
20090513 06:13:53< alink> again sorry for this 1.6.2 bug, guys
20090513 06:13:55< alink> bye
20090513 06:13:58-!- alink [n=alink@78.129.18.215] has quit [Remote closed the connection]
20090513 06:30:29-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has quit []
20090513 07:15:57-!- esr [n=chatzill@wesnoth/developer/esr] has quit [Read error: 104 (Connection reset by peer)]
20090513 07:17:02-!- BenUrban_ [n=benurban@unaffiliated/benurban] has joined #wesnoth-dev
20090513 07:19:38< sparr> 1.6 seems a lot less cpu intensive than 1.4
20090513 07:26:10-!- esr [n=chatzill@static-71-162-243-5.phlapa.fios.verizon.net] has joined #wesnoth-dev
20090513 07:30:35-!- mjs-de [n=mjs-de@vpw.wh.uni-dortmund.de] has joined #wesnoth-dev
20090513 07:34:07-!- BenUrban [n=benurban@unaffiliated/benurban] has quit [Read error: 110 (Connection timed out)]
20090513 07:37:31-!- BenUrban_ is now known as BenUrban
20090513 07:42:14-!- esr [n=chatzill@static-71-162-243-5.phlapa.fios.verizon.net] has quit [Read error: 104 (Connection reset by peer)]
20090513 07:46:23-!- YogiHH [n=chatzill@c169204.adsl.hansenet.de] has joined #wesnoth-dev
20090513 07:55:49-!- YogiHH [n=chatzill@c169204.adsl.hansenet.de] has left #wesnoth-dev []
20090513 08:02:00-!- AndreLuiz [n=deko__@unaffiliated/andreluiz] has quit ["we can be heroes, just for one day"]
20090513 08:06:23-!- Sirp [n=me@wesnoth/developer/dave] has quit ["leaving"]
20090513 08:28:50-!- ancestral [n=masterpr@97-116-120-23.mpls.qwest.net] has quit ["And that's the end of THAT chapter"]
20090513 09:15:02-!- euschn [n=chatzill@wesnoth/developer/euschn] has joined #wesnoth-dev
20090513 09:15:04< euschn> hi
20090513 09:26:35-!- Blueblaze [n=nick@c-98-199-143-139.hsd1.tx.comcast.net] has quit [Remote closed the connection]
20090513 09:34:07-!- busfahrer [n=busfahre@unixboard/user/busfahrer] has joined #wesnoth-dev
20090513 09:50:10-!- noy [n=Noy@wesnoth/developer/noy] has quit []
20090513 09:54:18-!- noy [n=Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20090513 09:55:16-!- noy [n=Noy@wesnoth/developer/noy] has quit [Client Quit]
20090513 10:05:59-!- ilor_ [n=user@wesnoth/developer/ilor] has quit []
20090513 10:16:11< euschn> happygrue: did you manage to find some broken saves of MP campaigns? (with respect to your post a few days ago in http://www.wesnoth.org/forum/viewtopic.php?f=15&t=25346&start=0&st=0&sk=t&sd=a )
20090513 10:36:41-!- busfahrer [n=busfahre@unixboard/user/busfahrer] has quit ["leaving"]
20090513 10:40:43-!- loonybot [n=loonybot@79.139.246.175] has joined #wesnoth-dev
20090513 10:41:46-!- loonycyborg [n=sergey@79.139.246.175] has joined #wesnoth-dev
20090513 10:43:27< Ivanovic> moin
20090513 10:53:48< Ivanovic> loonycyborg: http://www.wesnoth.org/forum/viewtopic.php?f=4&t=25443&start=0&st=0&sk=t&sd=a
20090513 10:58:00< loonycyborg> Ivanovic: I'm getting exactly the same under wine. Works nicely on winxp native though.
20090513 10:58:21< Ivanovic> interesting
20090513 10:59:58-!- isaac_ is now known as isaac
20090513 11:07:37-!- zookeeper [n=l@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20090513 11:17:19-!- euschn [n=chatzill@wesnoth/developer/euschn] has quit ["ChatZilla 0.9.84 [Firefox 3.0.7/2009030423]"]
20090513 11:47:33-!- AndreLuiz [n=deko__@unaffiliated/andreluiz] has joined #wesnoth-dev
20090513 12:14:17-!- alink [n=alink@wesnoth/developer/alink] has joined #wesnoth-dev
20090513 12:43:47< Ivanovic> hmm, where is fendrin when you need him?
20090513 12:52:25< CIA-30> ivanovic * r35617 /trunk/po/ (wesnoth-editor/zh_TW.po wesnoth-manpages/zh_TW.po): updated Chinese (Traditional) translation
20090513 12:52:28< CIA-30> ivanovic * r35618 /branches/1.6/po/ (wesnoth-editor/zh_TW.po wesnoth-manpages/zh_TW.po): updated Chinese (Traditional) translation
20090513 13:07:29-!- YogiHH [i=c3f5f398@wesnoth/developer/yogihh] has joined #wesnoth-dev
20090513 13:11:28-!- BenUrban_ [n=benurban@c-68-55-19-224.hsd1.md.comcast.net] has joined #wesnoth-dev
20090513 13:28:58-!- BenUrban [n=benurban@unaffiliated/benurban] has quit [Read error: 110 (Connection timed out)]
20090513 13:38:19-!- EdB [n=edb@103.12.95-79.rev.gaoland.net] has joined #wesnoth-dev
20090513 14:07:00-!- alink [n=alink@wesnoth/developer/alink] has quit [Remote closed the connection]
20090513 14:07:38< CIA-30> loonycyborg * r35619 /trunk/src/tests/test_lexical_cast.cpp: Made use of BOOST_CHECK_EXCEPTION in test_lexical_cast.
20090513 14:15:54-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090513 14:17:47-!- BenUrban_ is now known as BenUrban
20090513 14:19:03< Turuk> alink: Read the logs, thanks for tracking it down.  It's nice to hear that it should only happen for the units present at start, and not for others.  Hopefully you have got it figured out for the most part.
20090513 14:25:18-!- PK [n=pk@r74-192-44-206.vctrcmta01.vctatx.tl.dh.suddenlink.net] has quit [Read error: 104 (Connection reset by peer)]
20090513 14:44:08-!- EdB [n=edb@103.12.95-79.rev.gaoland.net] has quit [Remote closed the connection]
20090513 14:47:28-!- thespaceinvader [n=chatzill@wesnoth/artist/thespaceinvader] has joined #wesnoth-dev
20090513 14:48:15-!- jmunro[a] [i=james@217.112.83.12] has joined #wesnoth-dev
20090513 14:49:21-!- jmunro[a] is now known as jmunro
20090513 15:09:18-!- BenUrban [n=benurban@unaffiliated/benurban] has quit ["Power failu"]
20090513 15:22:30-!- Gnutoo [n=gnutoo@host109-10-dynamic.54-79-r.retail.telecomitalia.it] has joined #wesnoth-dev
20090513 15:45:50-!- euschn [n=chatzill@wesnoth/developer/euschn] has joined #wesnoth-dev
20090513 15:47:01-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit []
20090513 15:47:13< YogiHH> hi euschn
20090513 15:47:33< euschn> hi
20090513 15:59:33-!- Turuk_ [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090513 15:59:46-!- Polarina [n=polarina@wesnoth/translator/Polarina] has joined #wesnoth-dev
20090513 16:05:16-!- crimson_penguin [n=ben@64.201.60.216] has joined #wesnoth-dev
20090513 16:06:19-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090513 16:21:12-!- Turuk_ [n=Turuk@wesnoth/forumsith/turuk] has quit [Read error: 110 (Connection timed out)]
20090513 16:37:26< happygrue> euschn: I will PM you some saves shortly, thanks for the reminder
20090513 16:37:51< happygrue> shortly => later today ;)
20090513 16:44:27< Turuk> happygrue->leaves you for babies
20090513 16:44:39< Turuk> :P
20090513 16:44:48< euschn> happygrue: thanks a lot
20090513 16:50:54-!- AndreLuiz [n=deko__@unaffiliated/andreluiz] has quit ["we can be heroes.... just for one day!"]
20090513 17:01:08-!- euschn [n=chatzill@wesnoth/developer/euschn] has quit [Remote closed the connection]
20090513 17:03:55-!- allefant [n=allefant@allegro/developer/allefant] has joined #wesnoth-dev
20090513 17:14:57-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit []
20090513 17:17:36-!- Polarina [n=polarina@wesnoth/translator/Polarina] has quit [Remote closed the connection]
20090513 17:22:24-!- noy [n=Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20090513 17:22:28-!- West_ [n=West@hd5e2754c.gavlegardarna.gavle.to] has joined #wesnoth-dev
20090513 17:23:45-!- Polarina [n=polarina@wesnoth/translator/Polarina] has joined #wesnoth-dev
20090513 17:24:25-!- busfahrer [n=busfahre@unixboard/user/busfahrer] has joined #wesnoth-dev
20090513 17:34:02-!- EdB [n=edb@171.12.95-79.rev.gaoland.net] has joined #wesnoth-dev
20090513 17:38:53< West_> um, I'm getting a "Can't create directory" error when trying to commit a song. anyone know what I might be doing wrong?
20090513 17:40:23< allefant> did you create a new directory?
20090513 17:41:28< West_> allefant, a new directory where?
20090513 17:41:42< West_> it's been a while since I committed anything, so I don't remember the exact process
20090513 17:41:58< West_> oh, now I see what you mean
20090513 17:42:00< West_> no I didn't
20090513 17:42:20< West_> I put the song in the proper folder and edited the changelogs. that's it.
20090513 17:42:56< allefant> not sure then... should just be "svn add <file>" and "svn ci <file>" (with commandline svn)
20090513 17:43:34< West_> hm, I'm using TortoiseSVN
20090513 17:43:53< allefant> probably right-click the file and "add", then right-lick and "commit"
20090513 17:44:00< allefant> (never used it though)
20090513 17:45:24< West_> hmmm
20090513 17:45:28< West_> the message says
20090513 17:45:30< West_> Can't create directory '/var/svn/wesnoth/db/transactions/35619-1.txn':
20090513 17:46:13< West_> I'll try and update TortoiseSVN, there seems to be a new version out
20090513 17:46:23< allefant> that sounds like a problem on the svn server
20090513 17:47:02< West_> possibly, I have no idea what it means
20090513 17:48:02< allefant> i guess it either means that nobody can commit right now... or that you need to do a complete new checkout
20090513 17:48:33< West_> but I did an update just 30 minutes ago, and that worked fine as far as I could tell
20090513 17:48:57< West_> need to reboot, brb
20090513 17:49:02-!- West_ [n=West@hd5e2754c.gavlegardarna.gavle.to] has quit [Read error: 104 (Connection reset by peer)]
20090513 17:51:09-!- West [n=West@hd5e2754c.gavlegardarna.gavle.to] has joined #wesnoth-dev
20090513 17:51:56< West> nope, still doesn't work. :|
20090513 17:56:07< zookeeper> i sometimes get strange errors like that
20090513 17:56:30< zookeeper> usually they've just pretty much went away themselves at some point
20090513 17:57:38< zookeeper> but you can always just upload the file elsewhere and ask someone to commit it for you
20090513 17:57:38< West> zookeeper, ah ok. I'll just try again later then. just wanted to make sure I wasn't doing something wrong.
20090513 17:57:57< zookeeper> well, i'm not saying you aren't, just that you necessarily aren't ;)
20090513 17:58:05< West> true. I'll do that if it doesn't work later today
20090513 17:58:48< zookeeper> sure. btw, you're committing to trunk, not 1.6, right?
20090513 17:59:38< West> I'm comitting to trunk, yes
20090513 17:59:56< zookeeper> ok, was just checking
20090513 18:00:07< West> at least the changelog says Version 1.7.0-svn so... :)
20090513 18:00:36-!- ikarius [n=ross@68.166.34.14] has joined #wesnoth-dev
20090513 18:01:57< CIA-30> elias * r35620 /trunk/data/tools/wesnoth/wmlparser.py: Fixed bug #13516 which I had introduced with the previous commit to the Python parser.
20090513 18:02:56-!- allefant [n=allefant@allegro/developer/allefant] has quit ["Leaving"]
20090513 18:05:30-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090513 18:09:13-!- wesbot changed the topic of #wesnoth-dev to: 59 bugs, 241 feature requests, 9 patches | logs: http://irclogs.wesnoth.org/ | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com | http://imagebin.org
20090513 18:21:21-!- ikarius_ [n=ross@smtp.gridironsystems.com] has joined #wesnoth-dev
20090513 18:23:50-!- ikarius_ [n=ross@smtp.gridironsystems.com] has quit [Read error: 113 (No route to host)]
20090513 18:24:16-!- ikarius_ [n=ross@smtp.gridironsystems.com] has joined #wesnoth-dev
20090513 18:24:34-!- Polarina1 [n=polarina@85-220-26-48.dsl.dynamic.simnet.is] has joined #wesnoth-dev
20090513 18:25:02-!- Polarina [n=polarina@wesnoth/translator/Polarina] has quit [Nick collision from services.]
20090513 18:25:04-!- Polarina1 is now known as Polarina
20090513 18:25:06-!- Polarina is now known as Polarina1
20090513 18:25:31-!- Polarina1 is now known as Polarina
20090513 18:27:18-!- ikarius__ [n=ross@smtp.gridironsystems.com] has joined #wesnoth-dev
20090513 18:30:12-!- Polarina [n=polarina@wesnoth/translator/Polarina] has quit [Client Quit]
20090513 18:30:48-!- ikarius [n=ross@68.166.34.14] has quit [Read error: 110 (Connection timed out)]
20090513 18:34:21-!- Polarina [n=polarina@wesnoth/translator/Polarina] has joined #wesnoth-dev
20090513 18:36:45-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has quit [No route to host]
20090513 18:37:34-!- giusef [n=giusef@unaffiliated/giusef] has joined #wesnoth-dev
20090513 18:37:53-!- West [n=West@hd5e2754c.gavlegardarna.gavle.to] has quit ["b'bye"]
20090513 18:39:12-!- Sirp [n=me@wesnoth/developer/dave] has joined #wesnoth-dev
20090513 18:42:38-!- noy [n=Noy@wesnoth/developer/noy] has quit []
20090513 18:42:45-!- ikarius_ [n=ross@smtp.gridironsystems.com] has quit [Read error: 110 (Connection timed out)]
20090513 18:42:45-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit []
20090513 18:43:09-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090513 18:45:04-!- ikarius [n=ross@smtp.gridironsystems.com] has joined #wesnoth-dev
20090513 18:46:15-!- Polarina [n=polarina@wesnoth/translator/Polarina] has quit [Client Quit]
20090513 18:47:16-!- Crab_ [n=Crab_@wesnoth/developer/crab] has joined #wesnoth-dev
20090513 18:47:25< Crab_> hi
20090513 18:48:05< Ivanovic> hi Crab_
20090513 18:49:17< Crab_> Ivanovic: is 1.7 still scheduled for this Sunday ? (I'm just looking at the topic and wondering) 
20090513 18:49:19-!- crimson_penguin [n=ben@64.201.60.216] has joined #wesnoth-dev
20090513 18:49:33-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit []
20090513 18:49:39< Ivanovic> jupp, it is
20090513 18:49:54-!- Ivanovic changed the topic of #wesnoth-dev to: 1.7.0 scheduled for Sunday | 59 bugs, 241 feature requests, 9 patches | logs: http://irclogs.wesnoth.org/ | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com | http://imagebin.org
20090513 18:49:56-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090513 18:50:01< Ivanovic> better?
20090513 18:50:07< Crab_> yes :)
20090513 19:01:58-!- ikarius__ [n=ross@smtp.gridironsystems.com] has quit [Read error: 113 (No route to host)]
20090513 19:03:59-!- noy [n=Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20090513 19:11:06-!- Sirp [n=me@wesnoth/developer/dave] has quit ["leaving"]
20090513 19:12:07-!- mordante [n=mordante@wesnoth/developer/mordante] has joined #wesnoth-dev
20090513 19:12:23< mordante> servus
20090513 19:15:24-!- boucman [n=rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20090513 19:15:35< boucman> hey all
20090513 19:15:41< Crab_> hi boucman
20090513 19:18:28< boucman> so; what's up ?
20090513 19:18:34< boucman> did you test with silene's change ?
20090513 19:18:38< mordante> hi boucman 
20090513 19:19:34< Crab_> boucman: no, haven't tested yet.
20090513 19:20:12-!- Gnutoo [n=gnutoo@host109-10-dynamic.54-79-r.retail.telecomitalia.it] has left #wesnoth-dev ["Leaving"]
20090513 19:20:14< boucman> ok, good...
20090513 19:20:31< boucman> I want to test other optimisations, ant that particular optimisation would "hide" the other gains
20090513 19:20:46< boucman> (and I still intend to implement the same optimisation in a different way later)
20090513 19:21:15< boucman> I will probably (re)revert that commit, but I want to discuss it with silene first, obviously
20090513 19:21:42-!- YogiHH [i=c3f5f398@wesnoth/developer/yogihh] has left #wesnoth-dev []
20090513 19:40:22-!- maxy [n=maxy@80-219-0-199.dclient.hispeed.ch] has joined #wesnoth-dev
20090513 19:41:00-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Read error: 113 (No route to host)]
20090513 19:41:39-!- Polarina [n=polarina@wesnoth/translator/Polarina] has joined #wesnoth-dev
20090513 19:50:49-!- jmunro is now known as jmunro[a]
20090513 19:52:01< mordante> boucman, did you see r35500? IIRC you inlined unit::refresh and measured a speedup
20090513 19:52:09< zookeeper> i'm committing the song west requested now
20090513 19:52:37< boucman> mordante: yes, I noticed, but had no time to look at it, there were other problems in display code that creeped in during my week of
20090513 19:53:09< mordante> I read that in the logs also was away last week
20090513 19:53:21< zookeeper> urgh, at ~5.5kb/s -.-
20090513 19:55:19< CIA-30> zookeeper * r35621 /trunk/data/core/music/casualties_of_war.ogg: Committed Tyler Johnson's new track as requested by West.
20090513 19:55:45< zookeeper> uh, what...tortoisesvn said it was uploading at that pace and that it had transferred <1mb by the time it finished
20090513 19:55:48 * zookeeper is puzzled
20090513 19:56:08< boucman> zookeeper: what do you think of taking the habit of temporarly stickifying threads that contain commit approved art ?
20090513 19:56:30< zookeeper> boucman, well, i don't mind if people do that
20090513 19:57:11< zookeeper> sounds like a good idea when there's a delay to be expected, so it doesn't get forgotten
20090513 19:57:46< CIA-30> mordante * r35622 /trunk/ (4 files in 3 dirs): The tmessage dialog can now show up to 4 buttons.
20090513 19:58:17< boucman> k, I'll interview a bit around, and if nobody strongly disagree, I'll post something on the moderator forum 
20090513 19:59:42< mordante> boucman, I also don't mind to sticky them to avoid them getting forgotten
20090513 20:06:54-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit []
20090513 20:15:25< boucman> Crab_: do you have some time to rerun your test ?
20090513 20:15:49< Crab_> later (~30minutes) today, yes.
20090513 20:16:06< boucman> ok
20090513 20:16:17< Crab_> test latest trunk ?
20090513 20:17:05< boucman> yes, I need a point of comparison for my future commits
20090513 20:23:45-!- stikonas [n=quassel@ctv-217-147-36-179.vinita.lt] has joined #wesnoth-dev
20090513 20:38:05< mordante> Crab_, when you test again can you also look how much time is spend in display::draw_wrap ?
20090513 20:38:22< happygrue> omaha
20090513 20:38:32< happygrue> we are going as far as st. louis tonight
20090513 20:38:39< Crab_> mordante: yes, I'll post my profiling results
20090513 20:38:48< happygrue> is what I whould have said, in some other channel...
20090513 20:39:00< mordante> thanks
20090513 20:39:00< Crab_> boucman: ok, compiling r35622
20090513 20:43:59< boucman> mordante: ftp://ftp.terraninfo.net/wesnoth/ai_2009may11_after.jpg
20090513 20:44:06< boucman> from yesterday
20090513 20:45:53< Ivanovic> anyone any idea what this could be? http://www.wesnoth.org/forum/viewtopic.php?p=356805#p356805
20090513 20:46:12< Ivanovic> a case of "remove your addons and retest"? (aka broken/too man addons?)
20090513 20:46:19< Ivanovic> since the cpu is a fast one
20090513 20:47:43-!- busfahrer [n=busfahre@unixboard/user/busfahrer] has quit ["leaving"]
20090513 20:49:02< mordante> you can try to ask that, I wonder why my machine isn't crawling ;-)
20090513 20:49:23-!- grzywacz [n=grzywacz@wesnoth/developer/grzywacz] has joined #wesnoth-dev
20090513 20:49:48-!- busfahrer [n=busfahre@unixboard/user/busfahrer] has joined #wesnoth-dev
20090513 21:00:24-!- YogiHH [n=chatzill@d157127.adsl.hansenet.de] has joined #wesnoth-dev
20090513 21:00:51-!- euschn [n=chatzill@wesnoth/developer/euschn] has joined #wesnoth-dev
20090513 21:00:53< euschn> hi
20090513 21:01:00< happygrue> euschn: bad news
20090513 21:01:04< happygrue> and hello :)
20090513 21:01:10< YogiHH> hi euschn
20090513 21:01:23< euschn> happygrue: all your saves are gone for ever?
20090513 21:01:30< happygrue> I tried all the saves that I thought were bad, but it turns out that they were not  what I was hoping they were
20090513 21:01:37< euschn> i see
20090513 21:01:56< happygrue> well, I think I still have a lot of saves on the hard drive of my old computer, but I just didn't back them up
20090513 21:02:09< happygrue> and I will be out of town this week
20090513 21:02:14< happygrue> starting in about an hour
20090513 21:02:36< euschn> happygrue: oh, sounds like a lot of work for you to dig throguh those saves aswell
20090513 21:02:50< boucman> Crab_: do you have some time to quickly report to me how it is going on the AI front ?
20090513 21:02:52< happygrue> so I put in a request from one other person I think has some corrupted saves, and maybe next week I can take a look
20090513 21:03:06< boucman> that perf thingy had me almost forget about it ;)
20090513 21:03:21< happygrue> euschn: I want an excuse to fire up the old hard drive anyway, but I won't get to it 'soon'
20090513 21:03:44< happygrue> anyway, maybe I will get you a save at some point, but best not count on it ;)
20090513 21:03:50< happygrue> good luck anyway!
20090513 21:03:56< euschn> happygrue: no hurry in any case, thanks for lookiogn into it
20090513 21:03:57< Crab_> boucman: i think I'll include composite_ai in 1.7
20090513 21:04:05< happygrue> euschn: sure
20090513 21:04:20< boucman> when is it due ?
20090513 21:04:27< Crab_> boucman: this sunday
20090513 21:04:33< boucman> ok, cool
20090513 21:05:22< boucman> Crab_: if WML syntax is not backward compatible, you should warn esr so he can see if it's doable with wmllint
20090513 21:05:43< Crab_> boucman: no, nothing backward incompatible...
20090513 21:05:50< boucman> ok
20090513 21:06:08< Crab_> boucman: all is the same when ai_algorithm!=composite_ai
20090513 21:06:28< boucman> oh, you mean : "no backward compatibility issues"
20090513 21:06:31< boucman> even better :)
20090513 21:06:41< boucman> I read it as "it is not backward compatible"
20090513 21:06:50< Crab_> :)
20090513 21:09:42< boucman> Crab_: btw, it seems you background test machine is not playing any games anymore...
20090513 21:09:59< boucman> I would be interesting in seeing the average time per turn of latest svn...
20090513 21:10:56-!- EdB [n=edb@171.12.95-79.rev.gaoland.net] has quit [Remote closed the connection]
20090513 21:10:59< Crab_> boucman: yes. I thought that there was nothing so interesting to test at this point. but, you're right, I'll schedule a test now.
20090513 21:12:42< boucman> thx
20090513 21:26:25< boucman> Crab_: do you hae a new png for me ? :)
20090513 21:26:42< Crab_> boucman: test in progress...
20090513 21:26:47< boucman> ok,
20090513 21:27:56-!- happygrue [n=George@wesnoth/developer/wintermute] has quit [Read error: 104 (Connection reset by peer)]
20090513 21:31:04-!- Yermandu [n=daniel@201.28.104.101] has joined #wesnoth-dev
20090513 21:32:34-!- yann [n=dwitch@nan92-1-81-57-214-146.fbx.proxad.net] has joined #wesnoth-dev
20090513 21:34:49< Crab_> boucman: 9 minutes 40 seconds. ouch :)
20090513 21:34:56< boucman> ok
20090513 21:36:19< boucman> Crab_: please test with http://wesnoth.pastebin.com/m22658b2f
20090513 21:36:38< boucman> it should apply on current trunk, it reverts silene's change but does other changes
20090513 21:37:00< boucman> there are some more optim I can do after that, but they are more complicted and I would like to know if they are worth it
20090513 21:42:19< Soliton> euschn: http://www.wesnoth.org/replays/1.6/MP_campaign_-_World_Conquest_2_Turn_15_(6448).gz first save i tried and it's broken. though it looks like a WML error.
20090513 21:42:54< Crab_> boucman: ftp://ftp.terraninfo.net/wesnoth/ai_2009may13_r35622.jpg
20090513 21:43:05< Crab_> will try with the patch now...
20090513 21:43:11< boucman> k, thx a lot
20090513 21:43:11< Soliton> hmm, it might also be a problem of how it got saves. maybe there is some escaping needed.
20090513 21:43:47< euschn> Soliton: thanks a lot, I will try it out
20090513 21:45:05< Soliton> euschn: note that it is a replay saved by the server though which uses its own method so it might be a fault with that.
20090513 21:45:22< euschn> Soliton: i see
20090513 21:46:08< Soliton> i have no idea where the error you get when loading is even coming from though.
20090513 21:48:07-!- ardesh_ [n=ardesh@port-92-206-240-86.dynamic.qsc.de] has quit ["Quis custodiet ipsos custodes"]
20090513 21:56:43-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090513 21:57:13-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit [Success]
20090513 21:57:30-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090513 22:00:00< zookeeper> boucman, the wolf riders sometimes keep their heads up even when not in water in trunk. i also remember that we already talked about this ages ago, can't remember which one of us was supposed to take a look at it.
20090513 22:00:16< zookeeper> (the anim WML used is in the wolf rider .cfg)
20090513 22:00:40< boucman> zookeeper: ill take a quick look at the wml, if you're around to help me with the details
20090513 22:01:18< zookeeper> sure
20090513 22:01:31-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit [Client Quit]
20090513 22:01:43< zookeeper> (i haven't tested whether an alternative syntax would work)
20090513 22:02:02< boucman> zookeeper: terrain=Wo,Ww,Ww^Vm,Chs,Chw,Ss <== is that correct (I'm not good at terrain WML)
20090513 22:02:18< zookeeper> yes, that's a list of watery terrains
20090513 22:02:34< boucman> even Chs Ss ?
20090513 22:03:28< Crab_> boucman: can you, when you'll have time, look at unit::movement_cost and the profiling graph ?  note the overhead of (utils::string_bool(get_state("slowed"))  line.
20090513 22:03:51< Soliton> note that if you select the rider repeatedly it randomly has the head up or down regardless of terrain.
20090513 22:03:59< boucman> Crab_: yes, I just noticed that for the animation side, and removed the useless call to get_state("poisonned")
20090513 22:04:23< boucman> not sure what can be done for the movement_cost side of it, though
20090513 22:04:40< zookeeper> boucman, yes
20090513 22:04:44< boucman> Soliton: that is an interesting clue, thx
20090513 22:05:14< Crab_> boucman: pull that call up or cache it's result somehow ? that 1 line costs us 10%.
20090513 22:05:30< boucman> hmm
20090513 22:05:55< Crab_> note that even making "poisoned" a static std::string will give a small +
20090513 22:06:15< Crab_> s/poisoned/slowed
20090513 22:06:22< boucman> Crab_: not after my latest optim, all call to poisoned in the invalidate line have been removed
20090513 22:06:44< boucman> and they shouldn't be called unless needed in the redraw time, where it becomes over-optimization
20090513 22:06:47< Crab_> ( I'm about movement_cost )
20090513 22:07:04-!- stikonas [n=quassel@wesnoth/translator/stikonas] has quit [Read error: 104 (Connection reset by peer)]
20090513 22:09:05< boucman> Crab_: there are lot of small optims all around, and I don't want to get confused, so let's focus on animations for the moment
20090513 22:09:29< boucman> it's not trivial to add a cache for that sort of stuff, so let's do one thing at a time
20090513 22:09:39< boucman> (it would be a huge gain though, so let's not forget it :)
20090513 22:10:08< boucman> Crab_: if you have a todo list somewhere, you can add it there, if not, you can open a bug and assign to you/me
20090513 22:10:26-!- Turuk_ [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090513 22:10:47< Crab_> imo, it's better assigned to silene. since patching it up requires patching a*, and he touched it recently
20090513 22:11:06< boucman> no it doesn't...
20090513 22:11:14< boucman> or I misunderstood your idea
20090513 22:11:24< Crab_> why ? the root of the problem is with cost_calculator being not stateful enough.
20090513 22:11:52< boucman> oh, I though you only want to cache the result of get_state("slowed")
20090513 22:12:02< boucman> which would have been a change internal to the unit
20090513 22:12:03< Crab_> so, it calculates same things over and over when they don't change (for example,if while consider 10000 moves for 1 unit, his poisoning state doesn't change)
20090513 22:12:16< Crab_> s/if while/while
20090513 22:12:48-!- thespaceinvader_ [n=chatzill@91.108.133.37] has joined #wesnoth-dev
20090513 22:13:14< Crab_> " cache the result of get_state("slowed")" is a possible solution, but there may be others, better ones. so, it's better to start with examining A* and then moving down.
20090513 22:13:22-!- thespaceinvader [n=chatzill@wesnoth/artist/thespaceinvader] has quit [Nick collision from services.]
20090513 22:13:26-!- thespaceinvader_ is now known as thespaceinvader
20090513 22:13:41< boucman> the results of movement_cost_internal are already cached
20090513 22:14:14< boucman> the poisened state of a unit can change in game, but intercepting that particular change in set_state and adding a unit-wide bool would be trivial
20090513 22:14:45< boucman> else you have to cache at a* level the mvt cost of all terrains, which is almost like doubling the cache that is already in the unit, not worth it
20090513 22:14:54< YogiHH> boucman: Have you ever played "A New Land"?
20090513 22:15:03< boucman> YogiHH: not recently, why ?
20090513 22:15:44-!- Turuk_ [n=Turuk@wesnoth/forumsith/turuk] has quit []
20090513 22:15:45< YogiHH> boucman: i played it yesterday and i noticed, that "quick replay" doesn't catch some stuff
20090513 22:16:00< boucman> what do youmean by "catch" ?
20090513 22:16:08-!- Turuk_ [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090513 22:16:09< YogiHH> for example the floating labels from harvesting farms are not suppressed
20090513 22:16:21< boucman> hmm
20090513 22:16:23< YogiHH> as well as a lot of sounds
20090513 22:16:26< boucman> interesting
20090513 22:16:40< boucman> these are not cause by the anim engine, but the WML engine
20090513 22:16:49< boucman> so they went through the net when I checked all that
20090513 22:16:56< boucman> could you open a bug and assign to me ?
20090513 22:16:57< YogiHH> ah, that might be the reason then
20090513 22:17:06< YogiHH> yes
20090513 22:17:10< boucman> thx a lot
20090513 22:21:32-!- ardesh [n=ardesh@port-92-206-240-86.dynamic.qsc.de] has joined #wesnoth-dev
20090513 22:21:56-!- fendrin [n=fabi@g226143198.adsl.alicedsl.de] has joined #wesnoth-dev
20090513 22:22:11-!- stikonas [n=quassel@ctv-217-147-36-179.vinita.lt] has joined #wesnoth-dev
20090513 22:23:15< YogiHH> boucman: https://gna.org/bugs/index.php?13519
20090513 22:23:37< boucman> perfect, thx
20090513 22:26:49< YogiHH> night everyone
20090513 22:26:55< boucman> night
20090513 22:27:21-!- YogiHH [n=chatzill@d157127.adsl.hansenet.de] has left #wesnoth-dev []
20090513 22:31:26-!- ABCD [n=ABCD@pool-173-71-204-106.clppva.fios.verizon.net] has joined #wesnoth-dev
20090513 22:34:12-!- Blueblaze [n=nick@c-98-199-143-139.hsd1.tx.comcast.net] has joined #wesnoth-dev
20090513 22:37:15-!- Espreon [n=espreon@wesnoth/developer/espreon] has joined #wesnoth-dev
20090513 22:41:01< Crab_> boucman: 11.5 minutes with your patch
20090513 22:41:27< boucman> ok, thx
20090513 22:41:48< boucman> so stil a significant diff but not as big as it used to be
20090513 22:42:44-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090513 22:45:26< Crab_> yes. ftp://ftp.terraninfo.net/wesnoth/ai_2009may13_r35622_boucman.jpg
20090513 22:49:26< boucman> hmm
20090513 22:49:43< boucman> in 1.6 mouse seems totally frozen during AI turn, could anybody check ?
20090513 22:51:16-!- maxy [n=maxy@80-219-0-199.dclient.hispeed.ch] has quit []
20090513 22:52:55-!- Turuk_ [n=Turuk@wesnoth/forumsith/turuk] has quit [Read error: 110 (Connection timed out)]
20090513 23:05:39< boucman> Crab_: do you have time to try with this one ? http://wesnoth.pastebin.com/m6086d2ef
20090513 23:05:52< Yermandu> Every time that i "try" add on, the extra package "Survival Extreme" and other with about 9 MB, i recieve the msg "network::error thrown during transaction with add-on server; "Client disconnected" and "Remote host disconnected." What happens? How could i get the Survival Pack?
20090513 23:06:41< Crab_> boucman: apply to trunk or to your previous patch ?
20090513 23:06:48< boucman> to trunk
20090513 23:06:58< boucman> (it includes my previous patch)
20090513 23:07:20< boucman> there are still a few possible optim on the anim side, but they get more and more dangerous
20090513 23:08:31< Crab_> compiling..
20090513 23:10:39< boucman> it should be a bit faster than the previous one (got rid of some get_state calls and a useless call to the very long merge_parameters)
20090513 23:10:54< Crab_> we'll see :)
20090513 23:11:46< Soliton> Yermandu: try via the web interface: http://addons.wesnoth.org
20090513 23:11:52< boucman> the good news is that an important part of these optims affect all units, not just the offscreen ones
20090513 23:15:45< euschn> good night
20090513 23:15:47-!- euschn [n=chatzill@wesnoth/developer/euschn] has quit ["ChatZilla 0.9.84 [Firefox 3.0.7/2009030423]"]
20090513 23:15:51< boucman> night
20090513 23:15:56< Crab_> night
20090513 23:16:32-!- yann [n=dwitch@nan92-1-81-57-214-146.fbx.proxad.net] has quit [Remote closed the connection]
20090513 23:20:47-!- giusef [n=giusef@unaffiliated/giusef] has quit ["exit (-1);"]
20090513 23:22:53-!- noy [n=Noy@wesnoth/developer/noy] has quit []
20090513 23:29:25-!- stikonas [n=quassel@wesnoth/translator/stikonas] has quit [Read error: 110 (Connection timed out)]
20090513 23:29:34-!- mjs-de [n=mjs-de@vpw.wh.uni-dortmund.de] has quit ["On the road again"]
20090513 23:32:11< Yermandu> Soliton: Where i put the files? 
20090513 23:32:25< Crab_> boucman: 10 minutes 15 seconds
20090513 23:33:36< boucman> so the difference is ~55''
20090513 23:34:02< Crab_> more if " an important part of these optims affect all units, not just the offscreen ones"
20090513 23:34:41< boucman> well, the thing is at that point, I consider the anim engine to be "fast enough" it's not worth breaking features to improve
20090513 23:34:45< Crab_> ok :)
20090513 23:34:54< boucman> and there are still some improvements if a new need arise
20090513 23:35:07< Crab_> we can chop 1 minute from movement_cost, too ...
20090513 23:35:11< boucman> but as you pointed out, caching the poisoned stat will probably be the best way to gain
20090513 23:35:21< Crab_> (slowed)
20090513 23:35:27< boucman> I will commit as soon as I'm able to discuss it with silene
20090513 23:35:31< boucman> yes, sorry, slowed
20090513 23:35:41< Crab_> ok
20090513 23:35:59< boucman> did you do the png ?
20090513 23:36:35< Crab_> will rerun and do
20090513 23:36:43< boucman> k
20090513 23:39:23-!- Elvish_Pillager [n=eli@71-10-228-131.dhcp.oxfr.ma.charter.com] has joined #wesnoth-dev
20090513 23:40:39< Soliton> Yermandu: under data/campaigns/ in wesnoth's userdata dir.
20090513 23:44:52-!- Espreon [n=espreon@wesnoth/developer/espreon] has quit ["Tengo que ir... Yeahzorz..."]
20090513 23:50:35< mordante> Rhonda, I'm somewhat confused with bug 528587, I build trunk with G++-4.4 (in squeeze) today without a problem...
20090513 23:51:31-!- busfahrer [n=busfahre@unixboard/user/busfahrer] has quit ["leaving"]
20090513 23:52:55-!- alink [n=alink@wesnoth/developer/alink] has joined #wesnoth-dev
20090513 23:52:58< alink> hi
20090513 23:53:04< mordante> hi alink 
20090513 23:53:17< alink> good evening mordante
20090513 23:54:22< Soliton> squeeze != sid; do they have the same version?
20090513 23:54:38< Crab_> boucman: rerun, 10 minutes 12 seconds, ftp://ftp.terraninfo.net/wesnoth/ai_2009may13_r35622_boucman_2.jpg
20090513 23:54:39< mordante> yes unless you use hurd
20090513 23:55:15< mordante> it might be a 64 bit issue, but the code looks sane
20090513 23:56:36< Soliton> there is no 4.4 in sid or squeeze afaict.
20090513 23:56:55< mordante> http://packages.debian.org/search?keywords=gcc-4.4
20090513 23:57:28< Soliton> ah, a seperate package.
20090513 23:57:34< boucman> there are still 6% in image_mods that is probably gainable, and then we can do the "skip offscreen standing" thingy
20090513 23:57:51< boucman> but that's enough for today 
20090513 23:57:51< Soliton> that's odd then.
20090513 23:57:55< boucman> night all
20090513 23:57:58< alink> Crab_: About the pathfinding cost function I have actually a plan to cache repeated queries on the same location, but anything optimizing the first query will still be good. And indeed anything not location-specific can also use a global caching
20090513 23:58:01< mordante> night boucman 
20090513 23:58:03< alink> gn boucman
20090513 23:58:08< Soliton> nice work boucman.
20090513 23:59:49< Crab_> alink: 'cache repeated queries on the same location' won't help with that particular case, since it's because 'repeated queries using the same unit share some info (=is the unit slowed) which is expensive to get'
--- Log closed Thu May 14 00:00:25 2009