--- Log opened Fri Oct 09 00:00:57 2009
20091009 00:01:57-!- stikonas [n=and@wesnoth/translator/stikonas] has quit [Remote closed the connection]
20091009 00:05:14-!- stikonas [n=and@bcm-131-111-216-119.girton.cam.ac.uk] has joined #wesnoth-dev
20091009 00:10:48-!- stikonas [n=and@wesnoth/translator/stikonas] has quit [Remote closed the connection]
20091009 00:12:38-!- stikonas [n=and@bcm-131-111-216-119.girton.cam.ac.uk] has joined #wesnoth-dev
20091009 00:13:58-!- boucman [n=rosen@wesnoth/developer/boucman] has quit ["Leaving."]
20091009 00:14:14-!- thespaceinvader [n=chatzill@wesnoth/artist/thespaceinvader] has quit ["night all"]
20091009 00:21:40-!- giusef [n=giusef@unaffiliated/giusef] has quit ["exit (-1);"]
20091009 00:38:39-!- ancestral [n=ancestra@166.205.4.54] has joined #wesnoth-dev
20091009 00:42:50-!- Appleman1234 [n=Appleman@131.181.102.166] has quit [Read error: 60 (Operation timed out)]
20091009 00:43:29-!- ilor_ [n=user@wesnoth/developer/ilor] has joined #wesnoth-dev
20091009 00:52:52-!- loonycyborg [n=sergey@wesnoth/developer/loonycyborg] has quit ["Zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz"]
20091009 00:53:31-!- loonybot [n=loonybot@wesnoth/bot/loonybot] has quit [Remote closed the connection]
20091009 00:57:53-!- ancestral [n=ancestra@166.205.4.54] has quit ["Get Colloquy for iPhone! http://mobile.colloquy.info/"]
20091009 00:58:16-!- ancestral [n=ancestra@97-116-105-75.mpls.qwest.net] has joined #wesnoth-dev
20091009 01:00:21-!- ilor [n=user@wesnoth/developer/ilor] has quit [Read error: 110 (Connection timed out)]
20091009 01:01:42-!- stikonas [n=and@wesnoth/translator/stikonas] has quit [Remote closed the connection]
20091009 01:03:41-!- Espreon [n=espreon@wesnoth/developer/espreon] has quit ["WRYYYYYYYYYYYYYYYYYYYY!"]
20091009 01:05:38-!- Nayela [n=Nayela@cpc3-lich6-0-0-cust673.brhm.cable.ntl.com] has quit [Remote closed the connection]
20091009 01:37:09< shadowmaster> ilor_: thanks a lot for that "export selected coordinates to system clipboard" feature :)
20091009 01:45:16< CIA-62> alink * r39177 /trunk/src/ (6 files): 
20091009 01:45:16< CIA-62> Simplify one unit constructor by removing most of its messy bool parameters.
20091009 01:45:16< CIA-62> Also make :create spawn (randomly) male and female.
20091009 01:45:16< CIA-62> I noticed an error causing plague to only create male. I keep that behavior
20091009 01:45:16< CIA-62> for the moment but make it more clear.
20091009 01:46:44< alink> is there a reason why we don't give traits and name to debug-created units. It seems to be to prevent OOS because of the random calls but what's the point? Illegal creation of unit will cause OOS anyway. Am I missing something ?
20091009 01:47:05< alink> plus it prevent to easily test/debug trait and name generation
20091009 01:47:46< Crab_> alink: I think that there's no particular reason.
20091009 01:48:02< shadowmaster> I guess it's just because nobody has felt the need to give traits and name to debug-generated units
20091009 01:48:22< shadowmaster> I had planned to introduce that via the newer GUI2-based dialog, though. I only got to the name generation option before the string freeze started
20091009 01:48:38< shadowmaster> (use --new-widgets -d to see it in action)
20091009 01:48:41< alink> I feel the need now (to simplify and unify the code)
20091009 01:49:31< alink> shadowmaster: yes I saw that. but why make it optional and not always generate name?
20091009 01:50:04< shadowmaster> good question :P It wasn't there, I assumed there was a reason?
20091009 01:50:45< shadowmaster> I personally like to have an easy way to distinguish debug-created units from regular units
20091009 01:50:56-!- noy [n=Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20091009 01:51:02< Sirp> hi noy
20091009 01:51:05< alink> see in code or on map ?
20091009 01:51:19< shadowmaster> on the map, of course.
20091009 01:51:40< noy> hey
20091009 01:51:43< shadowmaster> any block of code invoking any unit constructor is too horrible to distinguish anything in it
20091009 01:52:14< alink> shadowmaster: agreed that's why i am simplifying that
20091009 01:52:30< shadowmaster> that's mainline because there are no named parameters in C++ (yes, I know there's a boost library that works around that)
20091009 01:52:46< shadowmaster> so I have to memorize a crapload of parameters and their possible default values to distinguish every case
20091009 01:53:00< shadowmaster> s/mainline/mainly/
20091009 01:54:03< CIA-62> ivanovic * r39178 /trunk/po/wesnoth-lib/en@shaw.po: updated Shavian translation
20091009 01:54:21< alink> i already removed 3 bool
20091009 01:54:59< shadowmaster> all -9 svn
20091009 01:55:03< shadowmaster> br.
20091009 01:55:27< alink> ?
20091009 01:55:40< shadowmaster> killing a svn client that's been stuck for a whole day
20091009 01:56:10< shadowmaster> I thought that destroying the terminal in which it was running would be enough...
20091009 02:11:40< CIA-62> crab * r39179 /trunk/ (10 files in 7 dirs): new candidate action: testing_ai_default::default_move_to_targets_phase - copy of default ai movements and targeting phase, as a candidate action
20091009 02:12:44< Crab_> alink: ^ it's pretty easy to fork this one into 'testing edition of movement and targeting phase' and check the result of any modifications on ai dumbness (in ai-vs-ai matchups)
20091009 02:13:27 * shadowmaster hands Ivanovic a soft pillow
20091009 02:15:23< alink> Crab_: and how such fork is done?
20091009 02:15:50< alink> ah i see it's already a copy
20091009 02:16:08< shikadibot> fork, fork, fork!
20091009 02:16:17< Crab_> shikadibot: yes, yes, yes
20091009 02:16:42< Crab_> alink: it's already a copy, so it allows to test 'default ai with old targeting vs RCA ai with this targeting' 
20091009 02:17:11< alink> Crab_: so any experimental change must be done in the testing version. And how do i play against this testing version ?
20091009 02:18:03-!- Arc [n=arc@pysoy/developer/ArcRiley] has joined #wesnoth-dev
20091009 02:18:08< Crab_> alink: it's included in RCA-based ai ( dev/testing_ai_default.cfg )
20091009 02:18:19< Crab_> alink: it's available in debug mode (start with -d )
20091009 02:18:38< Crab_> it's called 'RCA AI' there
20091009 02:18:42< Arc> any devs around? have a font issue trying to resolve
20091009 02:18:50< Crab_> alink: but, we can copy that file once more, into something like src/ai/testing/ca_testing_move_to_targets.?pp
20091009 02:19:12< Crab_> and rename it inside and reference it in src/ai/registry.cpp
20091009 02:19:35< Crab_> this way, we'll be able to test 'RCA ai with old targeting vs RCA ai with new targeting'
20091009 02:19:51< shadowmaster> Arc: basically, anyone in /who wesnoth/developer/*  is a developer ;)
20091009 02:19:55< shadowmaster> how can we help you?
20091009 02:20:16< Crab_> alink: I'll make such a fork now, if you're interested in tweaking that code
20091009 02:20:52< alink> well for the moment i mainly want to see with my own eye the difference in behavior, and indeed later check if it's make the AI better 
20091009 02:20:53< Crab_> alink: also, I can use my testing script to batch test those two ai's. If I fork once more, then "any" differences will be attributed solely to your changes.
20091009 02:21:28< Arc> shadowmaster: well that 7k font I uploaded, it's working in some places and not others
20091009 02:21:46-!- meowreka [n=Administ@c-69-242-154-116.hsd1.mo.comcast.net] has joined #wesnoth-dev
20091009 02:22:04< Arc> the terrain type shown in the far upper right is fine.  the yes/no dialog to skip a turn and the terrain shown in hover-over in the editor shows boxes
20091009 02:22:14< Arc> trying to discover what i may have done wrong, or whats barfing
20091009 02:22:26< meowreka> ew barf
20091009 02:22:37< Crab_> alink: so, I'll commit the forked code in 20-30 min.
20091009 02:23:05< shadowmaster> um.
20091009 02:23:13< shadowmaster> Arc: does it work in the campaign menu?
20091009 02:23:31< alink> Crab_: maybe i wait that i have the time to implement and experiment locally my AI change and I get back to you about this
20091009 02:23:41< alink> * maybe wait
20091009 02:24:20< Crab_> alink: well, it's fairly easy to do, only some copypasting and recompiling. it's better for it to be ready - it's not much work (most of that time is actually recompiling)
20091009 02:25:15< Arc> campaign menu?
20091009 02:25:24< alink> Crab_: yeah but if I see that the change make the AI crazy, no need to compare it to anything ;)
20091009 02:25:30< shadowmaster> Arc: yes, title screen -> Campaigns
20091009 02:25:35< Arc> http://www.wesnoth.org/gettext/index.lang.php?lang=en@shaw&version=trunk
20091009 02:25:47< Arc> only just started translation, most things are still in latin alphabet
20091009 02:25:56< Crab_> alink: well, that targeting phase is a mess. if your change won't work, than another should be done anyway :)
20091009 02:26:16< CIA-62> ivanovic * r39180 /trunk/ (52 files in 2 dirs): automatically add the new font (Andagii) into the font list when pofix.py is run
20091009 02:26:29< shadowmaster> um.
20091009 02:26:49< alink> Crab_: I will try to do it this week-end
20091009 02:26:57< Crab_> ok
20091009 02:28:33-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has joined #wesnoth-dev
20091009 02:28:38< shadowmaster> Arc: difficult to check exactly what the problem could be unless the files are already uploaded to trunk?
20091009 02:29:22< shadowmaster> also, my system doesn't like that locale :P
20091009 02:30:34< Crab_> alink: ok. the change is already done, waiting for compile to finish...
20091009 02:31:02-!- chains [n=Rylar@netblock-72-25-91-59.dslextreme.com] has quit [Remote closed the connection]
20091009 02:31:26< shadowmaster> Arc: also, you have made sure that all the .po files you are working with are using UTF-8?
20091009 02:32:45< alink> Crab_: that was indeed fast
20091009 02:33:33< Crab_> alink: that stuff is fairly modular and self-contained - copy a file, find-replace class name, add an entry to registry, fix buildsystems, copy&fix ai config.
20091009 02:35:27< Crab_> allowing such quick forks of 'only one component of ai' was one of major design goal behind making ai modular.
20091009 02:35:58< Crab_> as most of default ai is 'untouchable except for clear bugfixes'...
20091009 02:37:56< alink> yeah the untouchable aspect is annoying. And the last time it was touched, it was messy (1.6RC problems)
20091009 02:43:38-!- ilor [n=user@wesnoth/developer/ilor] has joined #wesnoth-dev
20091009 02:45:01-!- ShikadiLord [n=ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20091009 02:45:38-!- ShikadiLord [n=ignacio@wesnoth/developer/shadowmaster] has quit [Client Quit]
20091009 02:52:01-!- ilor_ [n=user@wesnoth/developer/ilor] has quit [Read error: 145 (Connection timed out)]
20091009 02:53:17< CIA-62> crab * r39181 /trunk/ (10 files in 7 dirs): forked candidate action: testing_ai_default::default_move_to_targets_phase into testing_ai_default::testing_move_to_targets_phase - set up everything for experiments on ai movement and targeting phase
20091009 02:53:38< Crab_> alink: ^ your playground is src/ai/testing/ca_testing_move_to_targets.?pp , it's available with -d as " RCA AI (experimental targeting) "
20091009 02:54:36< alink> Crab_: ok thanks
20091009 02:54:41< Crab_> alink: you can easily make the ai which does 'only' this phase (just throw away all unnecessary candidate actions from a copy of ai config)
20091009 02:55:26< Arc> shadowmaster: yes they are UTF-8.  you will need to add a syste locale for en_US@shaw UTF-8
20091009 02:55:43< Arc> en_US.UTF-8@shaw
20091009 02:55:52< Arc> or any other en_*.UTF-8@shaw
20091009 02:55:59< Arc> i can share the locale files if you would like
20091009 02:56:05< alink> Crab_: allowing easily such tweak seems usefull/interesting, thanks again :)
20091009 02:56:22< Crab_> alink: also, if you'll need it, I can either (1) batch test your committed/patched changes or  (2) modify the ai batch testing script to work without a db - so, you'll be able to launch it and (after some hours of testing) see the 'win %'.
20091009 02:57:05< Crab_> infrastructure for (1) is already in-place (that's the script that can feed results to aitest.wesnoth.org )
20091009 02:57:32< shadowmaster> (locale-gen) cannot open locale definition file `en_US@shaw': No such file or directory
20091009 02:57:48< shadowmaster> I think I'll just rebuild wesnoth with dummy locales enabled >.<
20091009 02:59:33< alink> yeah ok but as I said i want first test it with my eyes on custom map having targets at various distance and see which unit/targets the AI pick.
20091009 03:01:05< Arc> shadowmaster: doesnt work.
20091009 03:01:07< Crab_> alink: ok. I suggest throwing away testing_ai_default::get_villages_phase (at the beginning of testing) , and playing with units pre-available on map.
20091009 03:01:32< alink> (to check if the old broken code really do what is it supposed to do. Because, if not, at least comments must changes, regardless of AI fight perf)
20091009 03:01:54< Crab_> note that targeting phase has *lowest* priority, so it runs only if other phases have nothing to do
20091009 03:02:09< Arc> shadowmaster: or i can email it to you
20091009 03:02:13< Crab_> (it's explicit in ai config)
20091009 03:02:19< shadowmaster> Arc: well, it's because I haven't accepted the file :P
20091009 03:02:26< Arc> ah
20091009 03:02:36< Arc> just plug it into /usr/share/i18n/locales/en_US@shaw 
20091009 03:03:44< alink> Crab_: yeah i already noticed that low priority, that's a small problem. but OTOH I think some bad AI behavior were reported about this kind of cases (as the AI don't try enough things)
20091009 03:04:12< Crab_> alink: RCA-based ai have less such problems.
20091009 03:04:57< Crab_> alink: since if, after movement phase runs, new things become possible for other phases (candidate actions), they *will* be executed.
20091009 03:06:08< Crab_> the problem with targeting phase is that, for efficiecy, it needs to 'run multiple units per 1 go till no units remain or something unexpected happens'
20091009 03:06:09< alink> yes I remember that you explained that, much better for these cases indeed
20091009 03:06:15< Arc> its really just a dummy locale for now, i havent translated anything in it.  gets glib to let you through
20091009 03:07:13< shadowmaster> right, and what entry is it in the language dialog?
20091009 03:07:40< shadowmaster> (btw, does anyone have some eggs and a pan there? I could take advantage of my 84°C CPU)
20091009 03:09:48< Crab_> alink: note about testing_move_to_targets_phase::execute return value: it must return true ONLY if gamestate was changed (i.e. move was made).
20091009 03:09:51< alink> Crab_: about cpu-perf, my current position is to optimize current AI, and parallely to that, try do see what changes makes it smarter. and if a good one is found, then check what impact it has and how optimize it.
20091009 03:10:22< Crab_> alink: move&targeting phase is a bit specific - since we do only 'clean moves' there
20091009 03:11:05< Crab_> alink: so, in the future, we can actually avoid recalculating srcdst maps if nothing interesting happens during the move.
20091009 03:11:29< Arc> shadowmaster: it should be at the bottom.  it is not in a roman alphabet
20091009 03:11:33< Arc> http://www.omniglot.com/writing/shavian.htm
20091009 03:11:38< Crab_> alink: but, if you recode it to 'do 1 move at a time' - that's ok
20091009 03:12:06< alink> Crab_: good because for the moment, srcdst etc seems a bigger bottleneck than A*
20091009 03:12:19< shadowmaster> Arc: right, looksl ike I do need a build with dummy locales.
20091009 03:12:27 * shadowmaster smells fried human skin.
20091009 03:12:34< Arc> it doesnt let you select?
20091009 03:12:44< shadowmaster> nope, even after rerunning locale-gen
20091009 03:12:51< Crab_> alink: there's a bad place with those maps, which I'll optimize, it's in retreat stage.
20091009 03:13:03< shadowmaster> it'll be cooked in 2 minutes
20091009 03:13:34< Crab_> alink: enemy_fullmove_srcdst and enemy_fullmove_dstsrc are not cached there
20091009 03:13:43< alink> Crab_: and don't forget that terrain filter thing ;)
20091009 03:13:51< Crab_> yes, I remember
20091009 03:15:48< shadowmaster> I think I also needed to compile te .po.
20091009 03:15:52< shadowmaster> :P
20091009 03:16:12< shadowmaster> wait. why am I making a debug build again? :/
20091009 03:17:15-!- mjs-de [n=mjs-de@vpw.wh.uni-dortmund.de] has quit ["On the road again"]
20091009 03:17:39< Crab_> shadowmaster: because you want to run it under gdb in case you'll catch some interesting segfault :) ?
20091009 03:17:57< shadowmaster> apparently. I was sure I had run it earlier with build=release though.
20091009 03:21:32-!- noy_ [n=Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20091009 03:21:56< shadowmaster> okay, now it works (but only after throwing everything to swap memory) P
20091009 03:23:30< shadowmaster> Arc: with the "dialog to skip a turn", do you mean the "you have not started your turn yet" warning when I lclick on End Turn and I have not moved any units?
20091009 03:26:25-!- mjs-de [n=mjs-de@vpw.wh.uni-dortmund.de] has joined #wesnoth-dev
20091009 03:26:36-!- noy [n=Noy@wesnoth/developer/noy] has quit [Connection timed out]
20091009 03:26:37-!- noy_ is now known as noy
20091009 03:28:29< alink> does a very good AI should bet on probability to have 'quick' trait when deciding what to recruit to reach targets ?
20091009 03:28:42< Crab_> alink: why not ?
20091009 03:28:56< shadowmaster> hey, Chinese (traditional) wesnoth looks nicce :P
20091009 03:29:37< Crab_> it's, however, DKs area of expertise - he played with recruitment (via formula_ai) for a long time...
20091009 03:29:40< Arc> shadowmaster: yes, that is an example
20091009 03:30:02< Crab_> alink: also note the following 'feature' in phase-to-candidate_action conversion of targeting phase: since that targeting phase might do moves in several places, I've cheated a bit and done it in a 'always return >0 in evaluation, and try to execute' way. 
20091009 03:30:04< Crab_> so, when this phase does all possible moves and returns true from evaluate, it will return >0 in evaluation during next step, and, if selected for execution, do nothing (thus, returning false). this is ok - that candidate action will be blacklisted by rca loop till end of turn. 
20091009 03:31:09< Crab_> so, the end of ai turn will often look like this: 'move to targets selected as best, move to target moves units, move to targets selected as best, move to targets does nothing, move to targets blacklisted, rca loop ends due to no good candidate actions remaining'
20091009 03:31:24 * Arc just finished verifying and repairing the wesnoth-lib/en@shaw
20091009 03:32:59< Crab_> so, if you debug, don't be surprised if it's executed at least twice (last time - doing no actions) 
20091009 03:33:59< Crab_> this can be 'fixed' if we separate the 'can I do at least something?' and 'I do all I can' functions, but this is OK as is, since it's last by priority CA.
20091009 03:34:15< Arc> sorry if im unresponsible ive been plowing through hundreds of lines of script verifying that the transliterations were correct
20091009 03:34:22< Crab_> night
20091009 03:34:25< alink> Crab_: no actions, but no too much thinking too ?
20091009 03:34:45< alink> gn Crab_
20091009 03:34:55< shadowmaster> okay, so it shouldn't look like this, right? http://shadowm.nfshost.com/font_2.jpg
20091009 03:35:02< Crab_> alink: well, it will try to find an action. in most cases, it will stop very fast because there's no units remaining to move.
20091009 03:35:35< alink> good
20091009 03:36:05< Crab_> alink: btw, if it's sure that it has nothing to do, it can blacklist itself. but, it's not really necessary to complicate it with such checks :)
20091009 03:36:26< Crab_> since it will do that check by itself,  next time it's called.
20091009 03:36:42< shadowmaster> Arc: ^^^^^
20091009 03:36:58< Crab_> also, there are some problems which might arise if that phase won't be lowest-priority. 
20091009 03:37:24< Arc> shadowmaster: no.
20091009 03:37:26< Crab_> but, it's all ok now.
20091009 03:37:45< Arc> shadowmaster: the font you see in the upper right of that screenshot is correct.  the one in the dialog is not.
20091009 03:38:17< Crab_> since it's lowest, and it will stay so (almost by definition - because it moves units which haven't found a good thing to do during the turn)
20091009 03:38:25-!- Crab_ [i=crab@wesnoth/developer/crab] has quit ["Leaving."]
20091009 03:38:49< Arc> is this a font issue (Andagii.ttf is something i put together last night, compacted to be only 7k, with fontforge) or a code issue?
20091009 03:38:55< Arc> or a translation file issue
20091009 03:39:48< shadowmaster> not sure. I'm looking at the po file with a text editor
20091009 03:40:05< Arc> i just finished wesnoth-lib it didnt have any serious bugs in it though
20091009 03:40:09< shadowmaster> all msgstr's look like strings of blocks, and I'm using Andagii.ttf
20091009 03:40:30< Arc> monospaced?
20091009 03:40:46< Arc> im thinking the problem may be that wesnoth wants a monospaced font in the dialog
20091009 03:40:54< Arc> Andagii.ttf has no monospace glyphs
20091009 03:40:57< shadowmaster> no, it's not monospaced.
20091009 03:41:05< shadowmaster> why would wesnoth want a monospaced font?
20091009 03:41:10< Arc> no idea
20091009 03:41:21< shadowmaster> deja vu sans is certainly not a monospaced font
20091009 03:41:29< Arc> if Andagii.ttf didnt work you wouldn't see the font in the upper right
20091009 03:41:55< Arc> it reads "Summer Decidinous"
20091009 03:46:40< CIA-62> alink * r39182 /trunk/src/ (ai/default/ai.cpp menu_events.cpp unit.cpp unit.hpp): 
20091009 03:46:40< CIA-62> Continue to clean unit constructor:
20091009 03:46:40< CIA-62> Better control when we do random generation of traits, name and gender (to avoid OOS)
20091009 03:46:40< CIA-62> Activate them for debug-created units.
20091009 03:47:00< shadowmaster> maybe data/hardwired/fonts.cfg 's definition for that font doesn't include the codepoints sed in those strings
20091009 03:47:04< shadowmaster> *used
20091009 03:47:36< shadowmaster> can't really check the .po since my system doesn't want to use the font properly :/
20091009 03:48:40< alink> shadowmaster: the gui2::create-unit  'generate name' option seems useless now (but still work with a hack), maybe I should remove it?
20091009 03:49:01< shadowmaster> sure.
20091009 03:49:31< alink> ok, i will do it then
20091009 03:50:04< alink> it will also a first easy gui2 exercise for me :)
20091009 03:50:17< shadowmaster> Arc: maybe wait till tomorrow when Ivanovic reappears
20091009 03:50:40< shadowmaster> I don't think mordante can help since this does not seem to be a Pango issue; that dialog still uses SDL_ttf
20091009 03:51:13< Arc> http://media.apogean.org/img/wesnoth-1.png
20091009 03:51:19< Arc> this is the squares i was talking about
20091009 03:51:32< Arc> so is it a "SDL doesnt have proper ttf support" issue?
20091009 03:52:15< shadowmaster> nah, I don't think so.. if there's a bug it is probably in Wesnoth
20091009 03:52:46< Arc> maybe lack of support for compact TTF files?
20091009 03:53:15< Arc> yea and all the terrain hover-overs in the editor are squares too.  this seems to be the dialogs
20091009 03:53:47-!- noy [n=Noy@wesnoth/developer/noy] has quit [Connection timed out]
20091009 03:54:12< shadowmaster> no idea if the editor ui is using pango.
20091009 03:54:26< Arc> its whenever there's a popup or hover-over dialog
20091009 03:54:36< Arc> i get squares, you get.. weird characters
20091009 03:54:58< Arc> maybe its a unicode page1 issue.  GNU Screen has that
20091009 03:55:16< Arc> works fine with page0 (u0000 to uffff) but page1 is broken
20091009 03:55:33< Arc> oh, thats maybe why your editor doesnt work.
20091009 03:55:54< shadowmaster> the editor preferences dialog (which uses pango) seems to display the correct glyphs
20091009 03:56:11< Arc> so maybe its sdl-ttf and/or the way wesnoth is using it?
20091009 03:56:23< shadowmaster> yeah, that could be it.
20091009 03:56:28-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has quit []
20091009 03:56:38< Arc> oh well i'll keep plugging away
20091009 03:57:14 * shadowmaster leaving
20091009 03:57:14< Arc> thanks for looking into this shadowmaster 
20091009 03:57:38< shadowmaster> alink: do you know anything about SDL_ttf and GUI1 and anything about them related to internationalization?
20091009 03:57:56< shadowmaster> especially if it involves using a non-DejaVu Sans font specified in /data/hardwired/fonts.cfg
20091009 03:58:34< alink> no, just worked a bit on word-wrapping for this stuff
20091009 03:59:07< shadowmaster> ah, okay.
20091009 04:02:47-!- Chusslove [n=Chusslov@brsg-d9bee09b.pool.mediaWays.net] has quit [Read error: 110 (Connection timed out)]
20091009 04:02:54-!- Blueblaze [n=nick@adsl-99-171-161-30.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20091009 04:03:12< alink> After more thinking, I let the "generate name" option lives for the moment. The code is there and is not so small so maybe later reuse its code for something else or even just for gui2 debugging
20091009 04:03:51< alink> and i heard that some people want to mark debug unit somehow ;-p
20091009 04:04:57< alink> a possible idea is a "give loyal trait" option
20091009 04:05:48< alink> or "put it in recall list"
20091009 04:06:51< alink> or "use current playing side instead of 1"
20091009 04:10:10-!- Chusslove [n=Chusslov@brsg-d9beeab2.pool.mediaWays.net] has joined #wesnoth-dev
20091009 04:12:14< meowreka> are you luxurious spenders
20091009 04:36:27< Arc> nope not an issue with being compact
20091009 04:43:43-!- ilor_ [n=user@wesnoth/developer/ilor] has joined #wesnoth-dev
20091009 04:47:33-!- meowreka [n=Administ@c-69-242-154-116.hsd1.mo.comcast.net] has left #wesnoth-dev []
20091009 04:54:19-!- Ivanovic_ [n=ivanovic@dtmd-4db22a3a.pool.mediaWays.net] has joined #wesnoth-dev
20091009 05:01:35-!- ilor [n=user@wesnoth/developer/ilor] has quit [Read error: 110 (Connection timed out)]
20091009 05:05:48-!- alink [n=alink@wesnoth/developer/alink] has quit [Remote closed the connection]
20091009 05:11:28-!- Ivanovic [n=ivanovic@wesnoth/developer/ivanovic] has quit [Read error: 110 (Connection timed out)]
20091009 05:12:17-!- Ivanovic_ is now known as Ivanovic
20091009 05:37:10-!- DDR_ [n=chatzill@66.183.125.196] has joined #wesnoth-dev
20091009 05:43:53-!- ilor [n=user@wesnoth/developer/ilor] has joined #wesnoth-dev
20091009 05:47:53-!- DDR [n=chatzill@66.183.125.196] has quit [Read error: 110 (Connection timed out)]
20091009 05:50:58-!- ilor_ [n=user@wesnoth/developer/ilor] has quit [Read error: 60 (Operation timed out)]
20091009 06:24:43-!- Appleman1234 [n=Appleman@131.181.102.110] has joined #wesnoth-dev
20091009 06:35:29-!- ancestral [n=ancestra@97-116-105-75.mpls.qwest.net] has quit ["And that’s the end of THAT chapter."]
20091009 06:41:35-!- Aethaeryn [n=Michael@wesnoth/umc-dev/developer/aethaeryn] has quit ["'night"]
20091009 06:50:08-!- ancestral [n=ancestra@97.116.105.75] has joined #wesnoth-dev
20091009 06:51:47-!- Blueblaze [n=nick@adsl-99-171-161-30.dsl.hstntx.sbcglobal.net] has quit [Remote closed the connection]
20091009 06:52:04-!- Blueblaze [n=nick@adsl-99-171-161-30.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20091009 07:29:18< ilor> shadowmaster: you're welcome :)
20091009 07:31:43-!- noy [n=Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20091009 07:35:03-!- ardesh [n=ardesh@port-92-195-81-56.dynamic.qsc.de] has joined #wesnoth-dev
20091009 07:44:52-!- ilor_ [n=user@wesnoth/developer/ilor] has joined #wesnoth-dev
20091009 07:59:47-!- Sirp [n=user@wesnoth/developer/dave] has quit [Read error: 113 (No route to host)]
20091009 08:01:05-!- ilor [n=user@wesnoth/developer/ilor] has quit [Read error: 110 (Connection timed out)]
20091009 08:01:15-!- ancestral [n=ancestra@97.116.105.75] has quit ["And that’s the end of THAT chapter."]
20091009 09:47:09-!- Blueblaze [n=nick@adsl-99-171-161-30.dsl.hstntx.sbcglobal.net] has quit [Remote closed the connection]
20091009 10:18:21-!- ugasoft [n=ugasoft@89-96-141-83.ip12.fastwebnet.it] has joined #wesnoth-dev
20091009 10:19:01-!- ugasoft [n=ugasoft@89-96-141-83.ip12.fastwebnet.it] has left #wesnoth-dev []
20091009 10:19:18-!- ugasoft [n=ugasoft@89-96-141-83.ip12.fastwebnet.it] has joined #wesnoth-dev
20091009 10:22:23-!- SonIcco [n=SonIcco@pD9510AA9.dip0.t-ipconnect.de] has joined #wesnoth-dev
20091009 10:28:56-!- ilor_ [n=user@wesnoth/developer/ilor] has quit [Read error: 60 (Operation timed out)]
20091009 10:33:10-!- loonybot [n=loonybot@79.139.139.50] has joined #wesnoth-dev
20091009 10:33:57-!- loonycyborg [n=sergey@79.139.139.50] has joined #wesnoth-dev
20091009 10:45:43-!- valgor [i=5d7d107f@gateway/web/freenode/x-hqcbcxkusokxsmjb] has joined #wesnoth-dev
20091009 11:24:28< Ivanovic> moin
20091009 11:25:21< Ivanovic> Arc: one this you should definitely try is put the font into first position in the wesnoth textdomain
20091009 11:25:28< Ivanovic> this might already help
20091009 11:27:36< Ivanovic> Arc: beside this i have no idea what the issues might be though *maybe* (not sure at all since the problems might be related to oldgui stuff only, not sure) mordante is able to help you
20091009 11:28:08-!- ugasoft [n=ugasoft@89-96-141-83.ip12.fastwebnet.it] has left #wesnoth-dev []
20091009 11:36:21< loonycyborg> Ivanovic: I tried the Shavian translation with dummy locales and it doesn't work. In fact messages are in the same language as before switching to it.
20091009 11:36:38< Ivanovic> loonycyborg: yes, this suprises me, too
20091009 11:36:58< Ivanovic> loonycyborg: using --loginfo=all it shows that setlocale seems to be borked
20091009 11:42:21< loonycyborg> With the LANGUAGE patch it uses English, probably because gettext thinks that en@shaw should use english messages.
20091009 11:43:21< Ivanovic> hmm, i think en_GB works nicely, so do we have to switch things to use en_SW@shaw or soemthing like this?
20091009 11:43:50< Ivanovic> that maybe setlocale/gettext expects the country code and really relies on it?
20091009 11:44:39< loonycyborg> I said *probably*. That means that I really need to poke around more :P
20091009 11:46:48< Ivanovic> :)
20091009 11:54:33< loonycyborg> Hm.. en@shaw seems to work if I call gettext from python, but wesnoth actually sets LANGUAGE to something bogus.
20091009 11:54:58< loonycyborg> Namely to string 'en@shaw, en_AG@shaw, en_AU@shaw, en_CA@shaw, en_BW@shaw, en_DK@shaw, en_GB@shaw, en_HK@shaw, en_IE@shaw, en_IN@shaw, en_NG@shaw, en_NZ@shaw, en_PH@shaw, en_SG@shaw, en_US@shaw, en_ZA@shaw, en_ZW@shaw'
20091009 12:01:03< loonycyborg> ^^ somehow that string gets passed as slocale argument to wesnoth_setlocale()
20091009 12:03:32-!- Appleman1234 [n=Appleman@131.181.102.110] has quit [Read error: 104 (Connection reset by peer)]
20091009 12:05:32-!- noy [n=Noy@wesnoth/developer/noy] has quit []
20091009 12:06:14< Ivanovic> wow
20091009 12:06:30< Ivanovic> in theory this one should work, it was meant to do so since this is a list of *alternatives* to try
20091009 12:06:57< Ivanovic> (as in "if locale en@shaw is not available, try en_AG@shaw, ...")
20091009 12:07:15< Ivanovic> i think this stuff was once implemented by elias to also support de_AT
20091009 12:07:40< Ivanovic> (so that there was no need for him to install the locale de_DE to make the german translation work)
20091009 12:08:04 * Rhonda listens up to Ivanovic
20091009 12:08:55< Ivanovic> that is IIRC it in fact does work at least for data/languages/de_DE.cfg
20091009 12:09:31< Rhonda> There is no need to install locale de_DE anyway. de_AT falls back to *de*, not de_DE. Such a fallback is just stupid.
20091009 12:09:43< Soliton> well, en@shaw just put all those alternates in the locale key...
20091009 12:10:06< Rhonda> .. and totally contrary to how it is meant to work and actually *does* work in most other applications out there.
20091009 12:10:46< Ivanovic> argh, then it explains everything
20091009 12:10:49< Rhonda> what is en@shaw meant to be anyway?
20091009 12:11:02< Rhonda> Is that actually a proper locale?
20091009 12:11:52< Ivanovic> no, it is not
20091009 12:11:57< CIA-62> ivanovic * r39183 /trunk/data/languages/en@shaw.cfg: fix locale string for en@shaw, now it should be selectable...
20091009 12:12:02< Ivanovic> but shavian does not have a valid glibc locale so far
20091009 12:12:55< Rhonda> it's an english dialect?
20091009 12:12:59< Rhonda> Or what?
20091009 12:14:02< Rhonda> And what's sr@latin, btw.?
20091009 12:14:12< Chusslove> That's same as en@shaw :)
20091009 12:14:29< Chusslove> A language written in its non-default alphabet.
20091009 12:16:43< Chusslove> So Shavian is just plain English, only in a proper phonetic alphabet.
20091009 12:17:21< Chusslove> Meaning it has... how much letters? Given that English has *around* 43 phonemes...
20091009 12:18:22< Ivanovic> 43 i'd guess
20091009 12:18:24< Ivanovic> ;)
20091009 12:18:50 * Chusslove is counting...
20091009 12:19:10< Chusslove> 48. Yay.
20091009 12:38:00-!- zookeeper [n=l@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20091009 12:44:16-!- ardesh_ [n=ardesh@port-92-195-52-181.dynamic.qsc.de] has joined #wesnoth-dev
20091009 12:52:31-!- ardesh__ [n=ardesh@port-92-195-5-37.dynamic.qsc.de] has joined #wesnoth-dev
20091009 13:03:26-!- ardesh [n=ardesh@port-92-195-81-56.dynamic.qsc.de] has quit [Read error: 110 (Connection timed out)]
20091009 13:03:54-!- ardesh__ [n=ardesh@port-92-195-5-37.dynamic.qsc.de] has quit [Read error: 60 (Operation timed out)]
20091009 13:12:09-!- DDR [n=chatzill@66.183.125.196] has joined #wesnoth-dev
20091009 13:13:45-!- ardesh_ [n=ardesh@port-92-195-52-181.dynamic.qsc.de] has quit [Read error: 113 (No route to host)]
20091009 13:15:11< Ivanovic> strangely after the fix of the locale string setlocale *still* fails for en@shaw
20091009 13:18:27-!- DDR_ [n=chatzill@66.183.125.196] has quit [Read error: 60 (Operation timed out)]
20091009 13:28:51< loonycyborg> It seems to work for me with the LANGUAGE patch :P
20091009 13:31:59< Ivanovic> loonycyborg: i'd say that it should be cleaned up and included to replace the good old dummy locales
20091009 13:32:28< Ivanovic> (cleaned up as in "mark those without 'full featured' locales")
20091009 13:33:14< loonycyborg> Yes. I'll need to ask Mordante about how to do that..
20091009 13:33:29 * Ivanovic tries to summon mordante
20091009 14:00:32-!- isaac [n=isaac@debian/developer/isaac] has quit [Read error: 54 (Connection reset by peer)]
20091009 14:00:40-!- isaac [n=isaac@debian/developer/isaac] has joined #wesnoth-dev
20091009 14:04:07-!- valgor [i=5d7d107f@gateway/web/freenode/x-hqcbcxkusokxsmjb] has quit ["Page closed"]
20091009 14:22:39-!- ardesh [n=ardesh@port-92-206-92-167.dynamic.qsc.de] has joined #wesnoth-dev
20091009 14:36:54-!- Aethaeryn [n=Michael@wesnoth/umc-dev/developer/aethaeryn] has joined #Wesnoth-dev
20091009 14:39:25-!- SonIcco [n=SonIcco@pD9510AA9.dip0.t-ipconnect.de] has quit [Remote closed the connection]
20091009 14:39:41-!- SonIcco [n=SonIcco@pD9510AA9.dip0.t-ipconnect.de] has joined #wesnoth-dev
20091009 14:40:17-!- alink [n=alink@wesnoth/developer/alink] has joined #wesnoth-dev
20091009 14:54:58-!- SonIcco_ [n=SonIcco@pD9510AA9.dip0.t-ipconnect.de] has joined #wesnoth-dev
20091009 14:55:12-!- SonIcco [n=SonIcco@pD9510AA9.dip0.t-ipconnect.de] has quit [Remote closed the connection]
20091009 14:55:30-!- boucman [n=rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20091009 14:56:14-!- Crab_ [i=crab@wesnoth/developer/crab] has joined #wesnoth-dev
20091009 15:14:16-!- ardesh is now known as purple_shirted_e
20091009 15:14:26-!- purple_shirted_e is now known as ardesh
20091009 15:16:54-!- SonIcco__ [n=SonIcco@pD9510AA9.dip0.t-ipconnect.de] has joined #wesnoth-dev
20091009 15:18:25-!- stikonas [n=and@bcm-131-111-216-119.girton.cam.ac.uk] has joined #wesnoth-dev
20091009 15:21:50-!- Aethaeryn [n=Michael@wesnoth/umc-dev/developer/aethaeryn] has quit [Read error: 110 (Connection timed out)]
20091009 15:29:48-!- Netsplit farmer.freenode.net <-> irc.freenode.net quits: Chusslove, ardesh
20091009 15:31:39-!- SonIcco_ [n=SonIcco@pD9510AA9.dip0.t-ipconnect.de] has quit [Read error: 113 (No route to host)]
20091009 15:38:24-!- stikonas [n=and@wesnoth/translator/stikonas] has quit [Remote closed the connection]
20091009 15:41:29-!- stikonas [n=and@wesnoth/translator/stikonas] has joined #wesnoth-dev
20091009 15:44:57-!- ardesh [n=ardesh@port-92-206-39-186.dynamic.qsc.de] has joined #wesnoth-dev
20091009 15:47:58-!- Crab_ [i=crab@wesnoth/developer/crab] has quit [Read error: 110 (Connection timed out)]
20091009 15:50:44-!- Chusslove [n=Chusslov@brsg-d9beeab2.pool.mediaWays.net] has joined #wesnoth-dev
20091009 15:57:08< Arc> Ivanovic: you need to have the glib locale.  I use en_US@shaw and have the file for this if you'd like
20091009 15:57:54< Arc> Chusslove: yes George Bernard Shaw *hated* how the latin alphabet was shoehorned onto english making it so cumbersome, this is the result, and it's been in unicode for some time
20091009 15:59:09< Arc> I know a lot of people who keep all their personal notes in it, this translation effort is more or less to attract those people
20091009 15:59:28< Arc> we're transliterating the entire Ubuntu desktop; http://ubuntu.shavian.co.uk/
20091009 16:00:46< Arc> en_US@shaw en_GB@shaw etc differ in non-language ways such as date format/etc
20091009 16:01:02< Arc> so how should we specify this in wesnoth?
20091009 16:03:33-!- wesbot changed the topic of #wesnoth-dev to: string/feature freeze active! | 102 bugs, 237 feature requests, 10 patches | logs: http://irclogs.wesnoth.org/ | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com | http://imagebin.org
20091009 16:05:03< Ivanovic> Arc: those differences are handled by gettext
20091009 16:05:22< Ivanovic> that is the stuff is taken from the glibc locale (if there is one available!)
20091009 16:06:49< Arc> Ivanovic: I know.  I wrote a en_US@shaw locale.
20091009 16:07:14< Arc> so what should wesnoth set itself to in order to fallback to en_US@shaw en_GB@shaw etc
20091009 16:07:24< Ivanovic> ???
20091009 16:07:49< Ivanovic> Arc: you are switching to the system locale using the formatting it gives
20091009 16:07:59< Ivanovic> when you rely on dummylocales it uses basically the C locale
20091009 16:08:10< Arc> I use en_US@shaw on my system locally.  Thomas uses en_GB@shaw.  what locale string can wesnoth use which works for both of us
20091009 16:08:38< Arc> Ivanovic: we're adding these to glibc in the not so distant future
20091009 16:08:40< Ivanovic> it *should* all work for both of use due to me finally fixing the data/languages/en@shaw.cfg file
20091009 16:09:14< Ivanovic> if there is only en_GB@shaw available in the system, wesnoth will use that locale from the alternate list
20091009 16:09:27< Arc> AH
20091009 16:09:32< Arc> so its alternatives=
20091009 16:09:47< Arc> got it, i should have run a svn up before asking
20091009 16:10:10< Arc> I am 108 words from completing wesnoth-units translation
20091009 16:11:04< ardesh> Arc, *congratz*
20091009 16:11:36< Ivanovic> that is the first one in that list you got in your system will be used
20091009 16:11:41-!- zookeeper [n=l@wesnoth/developer/zookeeper] has quit [Read error: 145 (Connection timed out)]
20091009 16:11:48< Arc> we have a 120k+ word dictionary, most of the effort was a script to apply this properly for wesnoth and adding custom words like ogres and pikemen
20091009 16:12:06< ardesh> O_o
20091009 16:12:37< Arc> ardesh: Thomas setup a "word wiki" to collect translated words from the larger community, it grew pretty quickly
20091009 16:12:51< Arc> it's kinda like the Amiga community - small but the people using it are rabid about it ;-)
20091009 16:12:59< ardesh> hrr hrr hrr :D
20091009 16:13:08< ardesh> sure... but u still have to do most of it manually
20091009 16:13:27-!- zookeeper [n=l@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20091009 16:13:34< Arc> actually no, all I do manually is read over the translations and delete the fuzzy lines
20091009 16:13:55< Arc> now the 2700 "random" names in wesnoth's .po - THAT is a manual effort being shared by everyone
20091009 16:14:20< Arc> with 20 vowels we can put a bit of artistic pronounciation based on the race
20091009 16:15:17< ardesh> lol
20091009 16:16:03< Arc> Ivanovic: your fix was intended to make the dialogs render the font properly?  is there something special that needs to be done to enable that?
20091009 16:16:16< Arc> or make wesnoth rebuild that part?
20091009 16:17:15< Arc> I still get http://media.apogean.org/img/wesnoth-1.png in both the game and editor
20091009 16:17:42< Ivanovic> no, my fix was intended to make the game be able to switch to the locale it all
20091009 16:18:17< loonycyborg> Arc: Are buttons in language dialog affected too?
20091009 16:19:01< loonycyborg> AFAICT only GUI 2 can display that font properly.
20091009 16:20:59< loonycyborg> That is buttons in main menu-->language
20091009 16:21:25< loonycyborg> It's definitely using GUI2 already,
20091009 16:25:29< CIA-62> ivanovic * r39184 /trunk/po/ (3 files in 3 dirs): updated Shavian translation
20091009 16:26:58< loonycyborg> Actually, I see some shavian text rendered properly in that screenshot.
20091009 16:27:55< Arc> loonycyborg: yes, in the upper right.  no the menu buttons in-game are not rendered properly nor the drop-down menus from them
20091009 16:29:10-!- allefant [n=elias@allegro/developer/allefant] has joined #wesnoth-dev
20091009 16:29:21< loonycyborg> Main menu itself is using old GUI still but language selection dialog is already upgraded.
20091009 16:30:03< Arc> and the shaw text shows up fine in the language selection dialog
20091009 16:30:30< Arc> http://media.apogean.org/img/wesnoth-2.png
20091009 16:31:02< loonycyborg> It does for me too.
20091009 16:31:11< Arc> notice again that it shows up properly in the top-middle but the menus are all boxes.  it is also boxes for the hover-over on the right land-types
20091009 16:31:33< loonycyborg> That's what I get when I select shavian: http://imagebin.org/66891
20091009 16:31:53< alink> remote add-on 'Age_of_Guilds' has an icon which cannot be found: 'portraits\monsters\transparent\gryphon.png'   <-- does it work on windows? and if yes maybe print a warning about \ for campaign creators
20091009 16:32:04< Arc> ah yea im using the completed wesnoth-editor .po
20091009 16:32:15< Arc> that looks similar to http://shadowm.rewound.net/font_2.jpg
20091009 16:32:20< alink> shadowmaster: ^maybe your area
20091009 16:32:32< Arc> i wonder why I get boxes and you and shadowmaster get these weird glyphs
20091009 16:33:44< Soliton> alink: \ works on windows, yes. it might be nice to just convert it to / on other systems.
20091009 16:34:24< loonycyborg> Arc: Dunno. Maybe you don't have those fonts installed.
20091009 16:34:58< alink> Soliton: ah yes convert it is probably better. maybe check what we do elsewhere (in WML include for example)
20091009 16:34:58< Arc> you want to hear something great? for python 3.1 to handle UTF-8 strings properly from either script or interactive shell, you need to append them with .encode('utf-16').decode('utf-16')  - and Python 3.1 has the best UTF-8 support
20091009 16:35:11< Arc> loonycyborg: the question is which glyphs is it actually displaying?
20091009 16:35:45< loonycyborg> It looks like a mix of greek, russian and arabic :P
20091009 16:36:31< Arc> it seems to be doubling up the number of glyphs, so given that this is page1 unicode perhaps it only supports 16-bit unicode internally?
20091009 16:37:00< Arc> except the first byte would be an illegal character to render on its own
20091009 16:37:31< Arc> the "no" is 𐑯𐑴
20091009 16:38:51< CIA-62> alink * r39185 /trunk/src/construct_dialog.cpp: 
20091009 16:38:51< CIA-62> Fix an incorrect update of unit-preview when using filter and some sorting.
20091009 16:38:51< CIA-62> (to reproduce the bug, type "arc" in english create-unit dialog)
20091009 16:39:36< Arc> 𐑯𐑴 is '\ud801\udc6f\ud801\udc74'
20091009 16:41:53< Arc> or more appropriately '\U0001046f\U00010474'
20091009 16:42:49< Arc> or \xff\xfe\x01\xd8o \xdc\x01\xd8t\xdc
20091009 16:43:23< boucman> hey all
20091009 16:45:43< Arc> that was UTF-16, UTF-8 is \xf0\x90\x91\xaf \xf0\x90\x91\xb4
20091009 16:49:00-!- alink [n=alink@wesnoth/developer/alink] has quit [Remote closed the connection]
20091009 16:51:08-!- alink [n=alink@wesnoth/developer/alink] has joined #wesnoth-dev
20091009 17:06:53-!- Aethaeryn [n=Michael@wesnoth/umc-dev/developer/aethaeryn] has joined #Wesnoth-dev
20091009 17:16:22-!- stikonas [n=and@wesnoth/translator/stikonas] has quit [Remote closed the connection]
20091009 17:17:58-!- stikonas [n=and@wesnoth/translator/stikonas] has joined #wesnoth-dev
20091009 17:19:41-!- Arc is now known as Arc_
20091009 17:20:04-!- Arc [n=arc@pysoy/developer/ArcRiley] has joined #wesnoth-dev
20091009 17:21:21-!- Arc_ [n=arc@pysoy/developer/ArcRiley] has quit ["switching to TMUX"]
20091009 17:24:51-!- icco [n=SonIcco@pD9510AB6.dip0.t-ipconnect.de] has joined #wesnoth-dev
20091009 17:35:07< Soliton> Ivanovic: is there any use in 'key = _"$variable"'? i.e. a translatable string that consist only of a variable.
20091009 17:35:26< Ivanovic> i don't think so
20091009 17:35:50< Ivanovic> at least i can't think of a usecase directly out of my head
20091009 17:37:24< Soliton> shouldn't confuse gettext though like empty strings, right?
20091009 17:37:35< Ivanovic> it should not since there is an actual string
20091009 17:40:56< Soliton> i suppose it's possible that a translation would want to add something around the variable..
20091009 17:41:38-!- SonIcco__ [n=SonIcco@pD9510AA9.dip0.t-ipconnect.de] has quit [Read error: 110 (Connection timed out)]
20091009 17:43:24< Ivanovic> that is possible, but the *content* of the variable either is something that is translated itself or just some number or the likes
20091009 17:43:36< Ivanovic> hard to imagine that someone wants to add something there
20091009 17:44:27< Soliton> yeah.
20091009 17:45:38-!- blarumyrran [n=minaise@81-20-159-197.levira.ee] has joined #wesnoth-dev
20091009 17:50:18< Arc> Ivanovic: wesnoth-units text, "agrssiveness to mAtch" << I'm assuming that's not intentional
20091009 17:51:32< Ivanovic> nope, sounds like a clear case of a badass typo
20091009 17:52:11< Ivanovic> Arc: please add all typos you find at http://www.wesnoth.org/wiki/SpellingMistakes
20091009 17:52:24< Ivanovic> they will then fixed in a batch using pofix.py
20091009 17:53:32< Arc> ok
20091009 17:58:16< Ivanovic> and please make a page / list yourself as maintainer of the translation in the wiki: http://www.wesnoth.org/wiki/WesnothTranslations
20091009 17:58:27< Ivanovic> and: how do you want to be listed in the wiki?
20091009 17:58:31< Ivanovic> s/wiki/credits
20091009 17:58:39 * boucman tests google wave
20091009 18:11:14-!- Nayela [n=Nayela@cpc3-lich6-0-0-cust673.brhm.cable.ntl.com] has joined #wesnoth-dev
20091009 18:16:42 * Arc is Arc Riley <arcriley@ubuntu.com>
20091009 18:17:18< Arc> all three of us are working on the translation, if all team members belong in the credits
20091009 18:18:24< Arc> Thomas wrote most of the 120k word dictionary and is helping with batches of proper names, Jon is plowing through the random names list, im mostly doing script, cleanup, and adding new words to the dictionary
20091009 18:19:29-!- grzywacz [n=grzywacz@wesnoth/developer/grzywacz] has joined #wesnoth-dev
20091009 18:20:31< Ivanovic> Arc: have a look at data/core/about.cfg
20091009 18:20:45< Ivanovic> and just send me a diff with the ones to be added (make sure that it is in utf8)
20091009 18:21:28< Arc> ok
20091009 18:21:38< Arc> we have a month right?
20091009 18:21:56< Ivanovic> to complete everything?
20091009 18:21:59< Ivanovic> basically yeah
20091009 18:22:08< Ivanovic> for the next release: no
20091009 18:22:17< Ivanovic> i might get it out on sunday, no idea so far...
20091009 18:22:27< Arc> ah ok, i'll get you the diff now then
20091009 18:22:39< Ivanovic> just send it with the next updated files
20091009 18:23:15< Arc> ok
20091009 18:29:20< Arc> oh its generally good to list as a variant of english rather than just "Shavian" as the alphabet is used for other languages, notable esperonto
20091009 18:29:38< Arc> I usually write it as Shavian English or English (Shaw)
20091009 18:32:25< blarumyrran> Wow
20091009 18:32:39< Arc> in fact I know more people who primarily use the shaw alphabet for esperonto than with english
20091009 18:35:03< Arc> though its commonly confused with La Ŝava Alfabeto which is different
20091009 18:36:19< Arc> or quickscript, or, or, or - none of which are in unicode and are not really in use at all
20091009 18:39:13-!- Blueblaze [n=nick@adsl-99-171-161-30.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20091009 18:39:33< Arc> unless I'm mistaken Shaw is the only phonetic alphabet for english which has reached wide enough use to be included in unicode
20091009 19:06:18-!- mordante [n=mordante@wesnoth/developer/mordante] has joined #wesnoth-dev
20091009 19:06:31< mordante> servus
20091009 19:08:33< mordante> loonycyborg, Ivanovic if you can use translations without the locale installed it makes no longer sense to gray out items in the listbox, so the patch could be simplified
20091009 19:09:34< mordante> alink, no strong opinion about MP randomness for the names as long as the random generator is called the same amount of times for all users
20091009 19:09:48< mordante> is the delay due to server randomness too high?
20091009 19:10:57< alink> mordante: no idea, I just saw that these randoms where a not small part of WML data in replay
20091009 19:11:50< mordante> ah ok, but I'm not too familiar with that code as well, only fixed one bug there
20091009 19:12:00< alink> since complex randomness of name is not really important...
20091009 19:12:31< Arc> hey mordante 
20091009 19:12:45< mordante> true, we can use one random value and a simple algo to generate more numbers
20091009 19:12:47< alink> mordante: ok, and looked like a tricky bug, must have been fun to hunt ;)
20091009 19:12:52< mordante> hi Arc 
20091009 19:12:53< Arc> could use some help figuring out why page1 unicode isnt being rendered properly in dialogs and menus
20091009 19:13:14< alink> mordante: kind of one random seed for the name
20091009 19:13:20< Arc> http://media.apogean.org/img/wesnoth-2.png
20091009 19:13:29< Arc> note it renders correctly at the top center, but not the menu
20091009 19:13:39< mordante> alink, IIRC the biggest problem was to find the cause of the OOM... after which it was rather easy to fix
20091009 19:14:04< alink> OOM=OOS ?
20091009 19:14:25< mordante> yes was musing about the OOM server problem... ;-)
20091009 19:14:59< mordante> alink, well the random needs to stay in sync so players using the same language need to get the same names
20091009 19:15:02< Arc> this /may/ be a font support issue as it uses fonts/Andagii.ttf which I compacted, though using the traditional (and 5x the size) version resulted in the same
20091009 19:15:41< alink> i am currenlty trying to better control all these RNG calls, at least in the unit ctor, I hope it will help that
20091009 19:15:55< mordante> the unit ctor is a mess :-(
20091009 19:16:07< mordante> Arc, which language and which OS?
20091009 19:16:27< Arc> linux and en@shaw
20091009 19:16:41< Arc> which you should find at the bottom of the languages list in a non-roman script
20091009 19:16:45< alink> mordante: yes that's why i started to clean it (at least all these bool parameters having complex interactions)
20091009 19:17:47< Arc> i can send you the glibc locale file if you want to test on your own
20091009 19:18:44< alink> reduced 4 flags to 1 (real unit) which control all the rng calls (instead of having each flags interacting and triggering rng calls)
20091009 19:19:35< alink> but there still some details to polish
20091009 19:20:03< mordante> Arc, I assume Debian also ships those locales, are you the translator for this language?
20091009 19:20:56< Arc> mordante: no it does not.  I am a member of the Ubuntu translation team for the locale, we haven't gotten glibc to add the locale yet.  and yes we're contributing the wesnoth translations upstream so everyone can use them
20091009 19:21:56< mordante> I noticed the translation is in gettext, but haven't followed Wesnoth the last few days, still catching up ;-)
20091009 19:23:02< mordante> very funny since last weekend we talked about using some older English dialects ;-)
20091009 19:23:09< Arc> we're plugging away at it, have a 120k dictionary we've compiled since the translation is word-for-word, still a lot of custom names
20091009 19:23:40< Arc> you may also consider en@shaw to be a better source if you ever want to do text-to-speech, since the 20 vowels carry accents much better
20091009 19:24:06< Arc> there are 5 a's: 𐑩𐑱𐑨𐑭𐑷
20091009 19:24:36< Arc> age ash ado ah awe
20091009 19:25:20< mordante> I don't have the proper fonts installed to read it 
20091009 19:25:35< Arc> check fonts/Andagii.ttf - just 7k
20091009 19:26:10-!- Blueblaze [n=nick@adsl-99-171-161-30.dsl.hstntx.sbcglobal.net] has quit [Remote closed the connection]
20091009 19:27:22< mordante> do you know whether there's a debian package which contains that font?
20091009 19:28:12< Arc> no, but the glyphs are also available in the debian package ttf-mph-2b-damase
20091009 19:29:37< Arc> not a very good font, IMHO, the style changes radically between code blocks because it's the author's all-in-one
20091009 19:29:55< Arc> but it'll get you most of the native american fonts, shavian block, and a few others
20091009 19:30:12< Arc> s/american fonts/american glyphs/
20091009 19:30:49< mordante> ok I'll install them see whether it helps to display them
20091009 19:31:49< Arc> could the rendering bug be caused by a lack of full unicode support in sdl-ttf?
20091009 19:32:08< Arc> ie, is the part of the gui which is rendered properly using pango, while the menus and dialogs sdl-ttf?
20091009 19:32:12< freim> played through almost all of Delfador's Memoirs these last few days on trunk. No crashes, no major bugs expect "noise ticks" on music playback after playing a long time. Otherwise very enjoyable :)
20091009 19:32:50< mordante> no both that menu and the label should use the SDL-TTF
20091009 19:33:19< Arc> the label at the top of the screen?
20091009 19:33:34< mordante> yep
20091009 19:33:41< Arc> hmm, then what's different?
20091009 19:37:10< mordante> btw the name of your language in en@shaw.cfg is bad, between brackets there should be a name readable in English
20091009 19:37:35< freim> mordante: btw, wouldn't mind having mouse wheel scrolling in widgets
20091009 19:38:12< mordante> freim, recently broken https://gna.org/bugs/index.php?14459
20091009 19:38:19< freim> ah
20091009 19:38:22< mordante> will look at it tonight
20091009 19:39:04< mordante> Arc, did you translate the strings in the title screen?
20091009 19:39:08< freim> widgets in the text area also look a bit strange, but I guess that was something I was supposed to look at :)
20091009 19:39:33< mordante> yes 
20091009 19:39:49< mordante> I hope it's ugly enough to convince you to improve them :-P
20091009 19:40:08< freim> should probably make something completely new, that fits in the transparent area
20091009 19:40:22< mordante> would love that even better :-)
20091009 19:40:44< Arc> mordante: not yet
20091009 19:40:55< freim> mordante: do they have their own ID's, so I can make new widgets that will not overwrite the default?
20091009 19:41:22< mordante> freim, yes
20091009 19:42:48< Arc> mordante: see sr_RS@latin.cfg
20091009 19:43:13< Arc> should it instead be English (·𐑖𐑭𐑝𐑾𐑯) ?
20091009 19:43:43< freim> DM-showdown in the northern swamp: error config: Illegal character in map: (Qea) 'Qea'
20091009 19:46:58< mordante> Arc, no have a look at ar_AR.cfg, the goal is that somebody can at least have a clue what kind of language they look at
20091009 19:50:03< Arc> mordante: ok, i'll push that in my next update to Ivanovic 
20091009 19:50:15< Arc>        name="·𐑖𐑷𐑝𐑾𐑯 ·𐑰𐑙𐑜𐑤𐑦𐑖 (Shavian English)"
20091009 19:50:19< Arc>         sort_name = "English [Shaw]"
20091009 19:50:20< Arc> does that look right?
20091009 19:51:51< Aethaeryn> so many ?s
20091009 19:51:53< mordante> from the other files it looks like short name should be Shavian English
20091009 19:52:30< Arc> i assumed sort_name was for grouping in the menu
20091009 19:52:32< Aethaeryn> Arc: are you fluent in Shavian?
20091009 19:52:59< Arc> i dont know if the word fluent applies to writing, but as much as anybody
20091009 19:53:29< loonycyborg> mordante: Still the language dialog should probably provide some indication if locale is unavailable. locale isn't only the language but also date/time conventions etc.
20091009 19:53:33< mordante> we use if for sorting but use the normal English name for that, but otherwise ask Ivanovic 
20091009 19:53:34< Arc> Aethaeryn: you may need to install a font, such as the one included in the wesnoth svn under fonts/Andagii.ttf (just 7k)
20091009 19:54:29< mordante> loonycyborg, maybe, but it shouldn't use the disabled state for that, I rather see an icon
20091009 19:54:36< Arc> mordante: could just call it "English Shavian" then to get the order right
20091009 19:54:47< Arc> there's also Shavian Esperonto
20091009 19:55:19< mordante> also good, don't have a strong opinion, as long as I can read the language in English
20091009 19:57:51< Aethaeryn> Arc: Wouldn't it be better to call it "English (Shavian)" and "Esperonto (Shavian)"?
20091009 19:58:48< mordante> would look bad since it's already between parenthesis, so would become (English (Shavian))
20091009 19:58:50< Arc> thats how it was, but Ivanovic wants the name in the locale and mordante wants the name in english in ()
20091009 19:59:03-!- mjs-de [n=mjs-de@vpw.wh.uni-dortmund.de] has quit [Remote closed the connection]
20091009 19:59:25-!- mjs-de [n=mjs-de@vpw.wh.uni-dortmund.de] has joined #wesnoth-dev
20091009 19:59:34< loonycyborg> mordante: icon or no how easy it will be to extend that list widget to provide indication of availability of locales?
20091009 19:59:49-!- giusef [n=giusef@unaffiliated/giusef] has joined #wesnoth-dev
20091009 19:59:52< mordante> maybe we should in this case just make a difference between the sort_name and what's between brackets
20091009 20:00:51< mordante> loonycyborg, well it can use the same code as the current to detect availability 
20091009 20:01:12< freim> mordante: are the widgets drawn behind the transparent area?
20091009 20:01:30< Arc> mordante: *nod*
20091009 20:01:35< mordante> freim, behind which transparent area?
20091009 20:02:08< loonycyborg> mordante: That is obvious to me, though I don't know much about that widget itself.
20091009 20:02:12< Arc> it sounds weird to put it inverse order since Shavian is the adjective
20091009 20:02:46< freim> mordante: the text area itself
20091009 20:03:18< mordante> loonycyborg, well I can implement it if we decide to do it, not too much trouble
20091009 20:03:51< mordante> freim, it's drawn upon the transparent area
20091009 20:10:10< loonycyborg> mordante: If it's not too much trouble then I see no reason not to implement it :)
20091009 20:10:49< mordante> loonycyborg, well the problem I see is the stringfreeze since we can't explain things properly
20091009 20:12:31< loonycyborg> AFAIK icons don't need translating :)
20091009 20:13:21< mordante> well if you manage to make an icon that explains properly what the status of the translation is, I see no problem
20091009 20:13:42< mordante> but I've trouble to imagine how that icon should possibly look
20091009 20:14:21< mordante> off for diner
20091009 20:20:46< alink> an icon to show which locale are installed or which one is not installed ?
20091009 20:22:04< alink> if it only affect date/time maybe a clock icon, grayed-out when not installed ?
20091009 20:22:22< alink> or any other disabled variant
20091009 20:22:36< Aethaeryn> Arc: Actually, if there's English [US] and English [GB] why not English [Shavian]?
20091009 20:23:20< Arc> Aethaeryn: as a sort order, thats what I did, except i used the unicode for [Shavian] so it would be under [US]
20091009 20:23:34< Arc>         name="·𐑖𐑭𐑝𐑾𐑯 ·𐑰𐑙𐑜𐑤𐑦𐑖 (Shavian English)"
20091009 20:23:37< Arc>         sort_name = "English [·𐑖𐑭𐑝𐑾𐑯]"
20091009 20:24:33< Arc> the sort order is just that, a sorting, and because the namer dot · is greater than any latin letter it will always appear last
20091009 20:24:52< Aethaeryn> How do I do dummy locales?
20091009 20:24:54< Aethaeryn> what's the command?
20091009 20:29:33< Aethaeryn> nevermind
20091009 20:30:22< Arc> edit .scons-option-cache and drop in dummy_locales = True
20091009 20:30:28< Arc> rebuild, then use the flag
20091009 20:30:55< Arc> LANG=en_US.UTF-8@shaw
20091009 20:31:05< Arc> you need the glibc locale file for that installed first though
20091009 20:31:11< Arc> i can send if you'd like
20091009 20:31:52< loonycyborg> Also, it requires passing --dummy-locales to wesnoth
20091009 20:32:48-!- stikonas [n=and@wesnoth/translator/stikonas] has quit [Remote closed the connection]
20091009 20:33:36< Aethaeryn> Personally, m main motive is to see how complete the Latin translation is :P
20091009 20:33:46< Aethaeryn> Seeing Shavian is just a sceondary thing
20091009 20:33:54< loonycyborg> Theoretically, glibc locale file should make dummy locales unnecessary.
20091009 20:36:58-!- stikonas [n=and@bcm-131-111-216-119.girton.cam.ac.uk] has joined #wesnoth-dev
20091009 20:42:12< Arc> loonycyborg: "theoretically"
20091009 20:42:25< Arc> oh hey if anyone wants to take a crack at it, http://www.wesnoth.org/wiki/SpellingMistakes
20091009 20:42:49< Arc> one of the benefits of translation-by-dictionary is it highlights mispellings
20091009 20:43:33< loonycyborg> I didn't actually tried it. I used my experimental patch that sets the language without using dummy locales.
20091009 20:47:47< Arc> so whats the process for handling misspellings?
20091009 20:48:53< Aethaeryn> I suppose someone with SVN access patches it.
20091009 20:49:54< Arc> i know Ivanovic said he would need to run a command to update all the .po files
20091009 20:49:54< Arc> but does that merge existing translations?
20091009 20:52:22< Arc> im trying to figure out what *I* should do
20091009 20:53:11< Arc> i have wesnoth-units translated except for those words, if he runs that update script will my translations for those strings remain the same?
20091009 20:53:53-!- Bellerophon_true [n=chatzill@BSN-142-66-83.dial-up.dsl.siol.net] has joined #wesnoth-dev
20091009 20:55:38< Ivanovic> yes, your translations will remain the same
20091009 20:55:51< Ivanovic> sorry, currently afk, so don't expect any fast answers
20091009 20:58:53-!- faryshta [n=faryshta@168.255.251.50] has joined #wesnoth-dev
20091009 20:59:14< faryshta> Can I get a little question from c++? The c++ irc is dead.
20091009 20:59:31< faryshta> How do I declare a function which I expect to return me an array?
20091009 21:01:11< Rhonda> hmmmmm
20091009 21:01:32< Rhonda> optipng on the images within wesnoth decreases the file size by about 8.8%
20091009 21:08:30< loonycyborg> faryshta: You'd better not return arrays with functions :P
20091009 21:08:54-!- Netsplit farmer.freenode.net <-> irc.freenode.net quits: shikadibot, kimrhh, Smar, dfranke_, rtfb, yann, allefant, DDR
20091009 21:09:32-!- noy [n=Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20091009 21:09:40< loonycyborg> But if you really need to, it will be something like T* func() where T is the type of array elements.
20091009 21:09:43< faryshta> loonycyborg, why not?
20091009 21:10:11< loonycyborg> Because you'll need to manage memory.
20091009 21:10:45-!- Netsplit over, joins: allefant, DDR, rtfb, shikadibot, yann, kimrhh, Smar, dfranke_
20091009 21:11:21< loonycyborg> E.g. if you return an array that is a local variable in func(), you'll get undefined behavior since it no longer exists after func returns.
20091009 21:11:55< loonycyborg> If you use new/malloc then you'll have to remember to call delete/free in the caller.
20091009 21:12:07-!- Netsplit farmer.freenode.net <-> irc.freenode.net quits: allefant, DDR
20091009 21:12:07< loonycyborg> Better use std::vector or something.
20091009 21:12:25-!- Netsplit over, joins: allefant
20091009 21:13:01-!- Netsplit over, joins: DDR
20091009 21:13:33< mordante> alink, the problem I see is to have a good icon, I'm sure most users don't think about locales when seeing a clock
20091009 21:17:01-!- [Relic] [n=[Relic]@99-58-54-211.lightspeed.milwwi.sbcglobal.net] has joined #wesnoth-dev
20091009 21:18:05-!- Blueblaze [n=nick@adsl-99-171-161-30.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20091009 21:18:31< [Relic]> Hello :)
20091009 21:19:08< mordante> hi [Relic] 
20091009 21:19:30< [Relic]> :)
20091009 21:19:35-!- YogiHH [n=chatzill@d096134.adsl.hansenet.de] has joined #wesnoth-dev
20091009 21:19:46-!- nono9 [n=nono@92.96.179.226] has joined #wesnoth-dev
20091009 21:19:51< YogiHH> hello
20091009 21:19:55< nono9> Enter text here...hi
20091009 21:19:56< mordante> hi YogiHH 
20091009 21:20:17< mordante> hi nono9 
20091009 21:20:31< YogiHH> anyone: What's the proper way to show a map on the forum?
20091009 21:21:30< nono9> hello evry one
20091009 21:21:56< blarumyrran> To repair the damn images used by [map], duh
20091009 21:22:36< YogiHH> blarumyrran: so for now better use a screenshot?
20091009 21:22:40< nono9> hi mordante
20091009 21:22:55< YogiHH> hi nono
20091009 21:23:09< nono9> hi yogihh
20091009 21:23:16< nono9> how very one 
20091009 21:23:19< blarumyrran> If you use terrains newer than 1.6? then yes
20091009 21:23:34< YogiHH> yes, it's for 1.7
20091009 21:23:49< blarumyrran> or the new ones will show as question marks: http://www.wesnoth.org/forum/includes/WesnothMap.php
20091009 21:24:22< blarumyrran> bleh bad link
20091009 21:26:01-!- nono9 [n=nono@92.96.179.226] has quit ["It is often the case that the man who can´t tell a lie thinks he is the best judge of one.   -- Mark Twain, Pudd´nhead Wilson]
20091009 21:30:06< Rhonda> Ivanovic: I suggest you to give optipng a try instead of using pngcrush. :)
20091009 21:41:10< mordante> Arc, I can't find the problem directly and have some other bugs I need to work on, so best file a bug report
20091009 21:41:36< Arc> ok
20091009 21:42:02< mordante> alink, if you're bored you might want to look at https://gna.org/bugs/index.php?14486
20091009 21:47:48-!- Aethaeryn [n=Michael@wesnoth/umc-dev/developer/aethaeryn] has quit ["bah"]
20091009 21:47:57< noy> anybody speak english and italian?
20091009 21:48:51< boucman> noy: I speak french, is that close enough ?
20091009 21:49:25< faryshta> Ivanovic, hi.
20091009 21:49:26< noy> can you tell some italian players on the server why I keep kicking them because they are having extended conversations on the server?
20091009 21:49:39< noy> that
20091009 21:49:41< noy> type
20091009 21:49:43< noy> like 
20091009 21:49:45< noy> this
20091009 21:49:47< noy> all
20091009 21:49:49< Arc> Ivanovic: i just sent you up units and updated data files for the credits and language strings
20091009 21:49:49< noy> the 
20091009 21:49:51< noy> time?
20091009 21:50:34< noy> Blueblaze: around?
20091009 21:50:38< Blueblaze> yes
20091009 21:50:39< Arc> I edited the language config in accordance with mordante's suggestion to include the latin text so non-shavian readers know what it is
20091009 21:50:50< Arc> also set it to order below English [US]
20091009 21:50:58< noy> can you help me out on the server, you speak italian, right?
20091009 21:51:53< boucman> noy: nope, that too advanced for me. sorry
20091009 21:51:59< Blueblaze> I speak a little, but it is a little rusty. If I had time though, I'd address it but I have to be at a conference in a few hours for which I'm not at all prepared for
20091009 21:52:23< noy> do you know anybody off the top of your head who can speak?
20091009 21:52:43< Blueblaze> Google translate?
20091009 21:53:10-!- silene [n=plouf@wesnoth/developer/silene] has joined #wesnoth-dev
20091009 21:54:03< Blueblaze> It sound be effective enough to get the message across
20091009 21:54:09< Blueblaze> should be*
20091009 21:54:27< silene> hi
20091009 21:54:31< noy> ... not really
20091009 21:54:40< noy> I can't really reply
20091009 21:54:47< noy> if they are speaking italian back
20091009 21:55:05< loonycyborg> You can translate your replies too :P
20091009 21:55:08< Blueblaze> ^
20091009 21:55:25< loonycyborg> (I bet this will be a fun conversation :P )
20091009 21:55:33< Blueblaze> heh
20091009 22:00:34-!- thespaceinvader [n=chatzill@wesnoth/artist/thespaceinvader] has joined #wesnoth-dev
20091009 22:11:06< freim> mordante: still here?
20091009 22:12:05< mordante> freim, yes
20091009 22:12:11< boucman> hey thespaceinvader
20091009 22:12:33< thespaceinvader> hye boucman 
20091009 22:12:37< thespaceinvader> how's things?
20091009 22:13:02< boucman> i comitted a fix for your defense problem... it was a bit more tricky than I thought, could you test that everything is ok now ?
20091009 22:15:16-!- faryshta [n=faryshta@168.255.251.50] has quit ["Saliendo"]
20091009 22:15:19< blarumyrran> YogiHH, you do know that imagebin regularly (10 days?) clears all its images?
20091009 22:16:48< YogiHH> blarumyrran: no, i didn't know that
20091009 22:17:02< freim> mordante: http://feathers.os-tr.net/~hogne/buttons/
20091009 22:17:35< freim> mordante: tried out something. Not sure what kind of style you really want?
20091009 22:17:38< mordante> shadowmaster, I'm not sure whether bug #14458 part two is really an important bug and best file a report per issue, part one is fixed (to be committed soon)
20091009 22:17:43< freim> I think they should be rather simple
20091009 22:18:29< mordante> freim, do you have screenshots of them in the dialog?
20091009 22:18:47< mordante> also I prefer to leave the artistic look to the artists ;-)
20091009 22:20:51< freim> a bit hard to catch the feeling of mouse-over and click with a screenshot, but I can try
20091009 22:21:26< Ivanovic> Rhonda: have you ever had a look at the script used to run the image compression stuff?
20091009 22:21:40< CIA-62> mordante * r39186 /trunk/src/ (3 files in 2 dirs): 
20091009 22:21:40< CIA-62> Escape no longer escapes the gui2 dialog.
20091009 22:21:40< CIA-62> This caused the quit message to show when using escape in the message
20091009 22:21:40< CIA-62> dialog. (Fixes part 1 of bug #14458.)
20091009 22:21:44< Ivanovic> Rhonda: it is not using pngcrush anymore (already for ages!)
20091009 22:23:55< thespaceinvader> boucman: i can't get trunk
20091009 22:23:59-!- Blueblaze [n=nick@adsl-99-171-161-30.dsl.hstntx.sbcglobal.net] has quit [Remote closed the connection]
20091009 22:24:00< boucman> ok
20091009 22:24:05< Arc> mordante: done; https://gna.org/bugs/index.php?14487
20091009 22:24:18< boucman> just remember to give me feedback once it's tested
20091009 22:24:21< thespaceinvader> or at least, can't compile trunk
20091009 22:24:26< thespaceinvader> yep, will do
20091009 22:25:06< YogiHH> boucman: is it a known bug that idle animations are not cleared if they reach out into the next hex? I saw this when looking at gryphons in 1.6.
20091009 22:25:18< boucman> hmm
20091009 22:25:52< boucman> at some point we had hardwired that idle anims should fit in a hex for perf reasons, but I thought that limitation had been removed
20091009 22:26:02< boucman> (we're talking of idle, not standing anims, are we ?
20091009 22:26:27< YogiHH> what's again the difference?
20091009 22:26:45< thespaceinvader> YogiHH: idles play once in a while when the unit's not doign anything else
20091009 22:26:47< boucman> standing anim is the animation when the unit doesn't move (gryphon flying)
20091009 22:26:51< thespaceinvader> standing play all the time
20091009 22:27:00< boucman> idle anim is when the unit has nothing to do and does something funny from time to time
20091009 22:27:14< YogiHH> yeah, i am talking about idle animation
20091009 22:27:15< CIA-62> ivanovic * r39187 /trunk/ (3 files in 3 dirs): 
20091009 22:27:15< CIA-62> updated Shavian translation
20091009 22:27:15< CIA-62> added about entry for the translation
20091009 22:27:15< CIA-62> add a sort entry for the lang selection menu
20091009 22:27:31< boucman> ok
20091009 22:27:40< mordante> Arc, thanks
20091009 22:27:42< boucman> (gryphon has no idle anim afiak, are you adding one ? )
20091009 22:28:08< YogiHH> errm, i think it does, it flaps its wings once
20091009 22:28:13< freim> mordante: http://feathers.os-tr.net/~hogne/buttons/button_states.png
20091009 22:28:21< freim> mordante: normal, mouse-over, clicked
20091009 22:28:21< boucman> hmm
20091009 22:28:22< Ivanovic> mordante: is it known that the mouse wheel is problematic in the lang selection dialog?
20091009 22:28:31< Ivanovic> mordante: that is: it does not seem to work for me anymore
20091009 22:29:57< YogiHH> boucman: interestingly enough, it only seems one hex that is affected (directly north of the unit). I am pretty sure, the animation reaches into other neighbouring hexes as well
20091009 22:30:29< thespaceinvader> http://svn.gna.org/viewcvs/wesnoth/trunk/data/core/units/gryphons/Gryphon.cfg?rev=34139&view=markup no idle animation there, you must mean standing
20091009 22:30:38< boucman> YogiHH: ok, those type of bugs are very hard to debug... if you could describe a precise case where it happen, it would help me greatly
20091009 22:31:11< YogiHH> boucman: just recruit a gryphon and wait for the animation. You can't miss it.
20091009 22:31:20< boucman> ok
20091009 22:31:38< boucman> is this the gryphon (animal) or gryphon rider ?
20091009 22:32:12< YogiHH> gryphon rider. It is best seen if you put it into the water
20091009 22:32:12-!- meowreka [n=Administ@c-69-242-154-116.hsd1.mo.comcast.net] has joined #wesnoth-dev
20091009 22:32:28-!- meowreka [n=Administ@c-69-242-154-116.hsd1.mo.comcast.net] has left #wesnoth-dev []
20091009 22:33:22 * mordante hits Ivanovic with bug #14459
20091009 22:33:49< Ivanovic> do you think i read the bug tracker?!?
20091009 22:34:11< mordante> well it has been discussed on irc a few times, including today
20091009 22:34:23< mordante> but it's next on my list
20091009 22:34:25< Ivanovic> yeah, and i was away until just some mins ago
20091009 22:34:27< Ivanovic> ;)
20091009 22:34:30< shadowmaster> wesbot: bug #14458
20091009 22:34:31< wesbot> Bug #14458  Assigned to: Mark de Wever <mordante>  Status: In Progress  Priority: 7 - High
20091009 22:34:34< wesbot> Summary: GUI2 and GUI1 event handling do not cooperate with each other in 1.7.6+svn
20091009 22:34:37< wesbot> Original submission: Since mordante's various changes in the event handling me
20091009 22:34:40< wesbot> chanism for GUI2 after 1.7.6 was released, both user interface systems' event ha
20091009 22:34:43< wesbot> URL: https://gna.org/bugs/?14458
20091009 22:34:45< YogiHH> Ivanovic: As if you never read the logs ;-P
20091009 22:35:00< shadowmaster> mordante: it *is* a n issue
20091009 22:35:14< Ivanovic> YogiHH: i just grep the logs for my nick
20091009 22:35:30< Ivanovic> YogiHH: unless it is really short which is normally not the case on a friday evening...
20091009 22:35:44< YogiHH> ok :)
20091009 22:36:18< shadowmaster> mordante: and TBH I don't understand your last two comments on #14458 in the tracker
20091009 22:36:20< mordante> shadowmaster, can't read odd utf-8 chaters
20091009 22:36:27< mordante> characters
20091009 22:36:52< shadowmaster> mordante: ehm?
20091009 22:37:04< shadowmaster> what do you mean? I'm sure I'm talking in ASCII :p
20091009 22:37:52< mordante> freim, looks nice think the mouse-over is a bit too opaque, but feel free to commit it
20091009 22:38:25< freim> can adjust that
20091009 22:38:26< mordante> 007F is no ascii
20091009 22:38:42< shadowmaster> mordante: is this related to Arc's probles?
20091009 22:38:44< freim> can commit and see what feedback it gets
20091009 22:38:45< shadowmaster> *problems
20091009 22:38:50< freim> then adjust later on
20091009 22:39:18< mordante> freim, agreed
20091009 22:39:59< mordante> shadowmaster, why should it have anything to do with Arc's problem?
20091009 22:40:05< shadowmaster> mordante: ahhh, sorry, I now understand. Ignore that stray 0x7F.
20091009 22:40:13< shadowmaster> it is a SSH-lag garbage
20091009 22:40:46< mordante> and what exactly is unclear about those message, read them again and still think they make sense
20091009 22:40:49< shadowmaster> it didn't display at all in irsis
20091009 22:40:52-!- ilor [n=user@wesnoth/developer/ilor] has joined #wesnoth-dev
20091009 22:40:54< shadowmaster> *irssi, here
20091009 22:41:51< shadowmaster> mordante: part 1 was a duplicate of the bug you earlier said was a duplicate of bug ##14461 then?
20091009 22:41:51< mordante> ok
20091009 22:42:23< mordante> yes part 1 is a duplicate and part 2 is a different issue
20091009 22:42:58< shadowmaster> note part 2 is also related to your changes in event handling, since it wasn't present in 1.7.6
20091009 22:43:00< mordante> and part 2 might be a bit annoying to fix
20091009 22:43:08< shadowmaster> hm.
20091009 22:43:09< CIA-62> hogne * r39188 /trunk/images/buttons/ (4 files): New transparent buttons
20091009 22:43:16< mordante> I know it's new
20091009 22:43:41< shadowmaster> so you are suggesting that I make a different report for part 2? that's a bit annoying to fix :P
20091009 22:44:10< mordante> no just encourage you to do so next time when you report an issue
20091009 22:44:31< mordante> tracking half finished reports is very annoying
20091009 22:44:45< shadowmaster> I had assumed they were symptoms of a single bug
20091009 22:44:50< boucman> YogiHH: can'tr reproduce... you're seeing it with trunk ?
20091009 22:45:10< YogiHH> boucman: no, 1.6
20091009 22:45:14< shadowmaster> since both worked like "GUI1 catching an event before GUI2 processes it"
20091009 22:45:15< boucman> oh
20091009 22:45:26< boucman> well, there might be that sort of stuff in 1.6, yes
20091009 22:45:47< boucman> knowing how painfull these are to fix, I won't attempt them in 1.6
20091009 22:46:03< boucman> esp since most of the invalidation had to be rewritten to catch the corner cases
20091009 22:46:10< boucman> but thx for the report 
20091009 22:46:52< CIA-62> boucman * r39189 /trunk/ (4 files in 2 dirs): 
20091009 22:46:52< CIA-62> fix the anim engine sometime choosing the default animation when a valid user
20091009 22:46:52< CIA-62> animation was available. Detected with drake on-water anim, but could be
20091009 22:46:52< CIA-62> triggered by anim animation. Also add a security against force redraws between
20091009 22:46:52< CIA-62> the end of a strike and the next strike
20091009 22:47:07< shadowmaster> alink: no iddea if that portraits\blah\blah.png path should work in Windows.
20091009 22:47:14< boucman> thespaceinvader: the commit I was mentionning... and that I actually forgot to commit ;)
20091009 22:47:36< thespaceinvader> boucman: yeah?
20091009 22:47:58< shadowmaster> alink: I guess it does, but woul it be too expensive to make the WML preprocessor and the image cache system complain about such usage instead?
20091009 22:48:04< boucman> thespaceinvader: the next time you get a trunk build (or release) that's the commit that should fix the drake problem
20091009 22:48:44< thespaceinvader> boucman: i'll let you know when i do =)
20091009 22:48:49< boucman> thx
20091009 22:48:51< thespaceinvader> probably the next dev version
20091009 22:49:28< shadowmaster> alink: e.g. hacking the addons management dialogs to warn about a windows-style icon path doesn't seem to me like the definitive solution.
20091009 22:49:48< Rhonda> Ivanovic: No, haven't. I just found three images in the 1.7.6 release though that were reduced by 8.8% by simply running optipng on them. :)
20091009 22:50:38-!- blarumyrran [n=minaise@81-20-159-197.levira.ee] has quit []
20091009 22:50:40< Arc> is that lossless?
20091009 22:50:53< shadowmaster> yes it is
20091009 22:51:29< shadowmaster> well, at least the wesnoth-optipng command line, although IIRC optipng's defaults also produce lossless output.
20091009 22:51:53-!- allefant [n=elias@allegro/developer/allefant] has quit ["Leaving"]
20091009 22:55:22-!- icco [n=SonIcco@pD9510AB6.dip0.t-ipconnect.de] has quit [Remote closed the connection]
20091009 22:58:59< alink> yes, indeed. I don't feel like touching such OS-specific url stuff for the whole engine, but maybe someone will do.
20091009 22:59:06< alink> shadowmaster: ^
20091009 22:59:18< shadowmaster> wow, even the keyboard is hot
20091009 23:01:22< alink> also i have a feeling that considering url/like/this equal to url\like\this will make some STL strings stuff more complex. but maybe already solved in the current code somewhere
20091009 23:02:06< shadowmaster> alink: windows knows how to handle paths/like/this
20091009 23:03:29< shadowmaster> (hidden system feature)
20091009 23:03:37< alink> yes but if we allows \url\ in wesnoth WML etc, we must translate it each time
20091009 23:04:07< shadowmaster> we should just disallow it.
20091009 23:04:41< alink> well it's already disallowed on a linux system
20091009 23:04:46< shadowmaster> it's currently possible to use such notation only because the various components do not check whether it is a Unixly correct path.
20091009 23:05:27< mordante> s/unixly/the C++ way/ ;-)
20091009 23:05:28< shadowmaster> so we are basically able to throw any kind of garbage to an ifstream constructor AFAIK
20091009 23:06:01< alink> a warning or forbidding  will avoid UMC creator writing addon only working on windows
20091009 23:06:35< mordante> I'd say forbid otherwise the odds are high it won't be noticed
20091009 23:06:43< alink> at least  it seems to be the case for the addon icon case
20091009 23:07:31< shadowmaster> sure I could parse the .pbl and print some warnings that nobody will see because windows users don't read stderr.txt
20091009 23:07:54< shadowmaster> or throw a dialog box at them with a new translatable string..
20091009 23:08:02< alink> well just don't display addon icon if they have a "\"
20091009 23:08:18< alink> as it happens on linux
20091009 23:08:59< shadowmaster> since most people do not care whether their paths even have the .png extension, I doubt that'll help a lot, really :P
20091009 23:09:05< boucman> night all
20091009 23:09:22-!- boucman [n=rosen@wesnoth/developer/boucman] has quit ["Leaving."]
20091009 23:09:28< shadowmaster> but sure, that could be done... although I'd prefer get_image() to complain instead.
20091009 23:09:50< shadowmaster> e.g. to avoid [unit_type] image="sucky\path\like\this.png" as well
20091009 23:12:17< alink> i suppose that a small string operation is ok compared to the cost of finding+reading+uncompress_png+SDL stuff
20091009 23:12:20-!- Espreon [n=espreon@wesnoth/developer/espreon] has joined #wesnoth-dev
20091009 23:13:17< alink> I wonder why such problems don't show up more
20091009 23:13:41< shadowmaster> dunno, I don't play MP :P
20091009 23:13:58< shadowmaster> I believe that MP add-on authors tend to be lazier than SP authors
20091009 23:14:16< shadowmaster> [citation needed]
20091009 23:15:23< shadowmaster> OTOH not even linux users look at the stderr, so nobody will notice that add-on XYZ has a broken icon path in the add-ons list and they'll just assume that no icon is provided unless they are using debug mode and the Working Peasant of Terror™ appears 
20091009 23:15:34< shadowmaster> *most linux users
20091009 23:17:36< alink> we seems to already search for "../" and ".." (yes both, there probably is a possible optimization there)
20091009 23:18:38-!- giusef [n=giusef@unaffiliated/giusef] has quit ["exit (-1);"]
20091009 23:20:16< alink> mordante: btw, I suspect that most users don't know about locale and how to install them, so make an icon explaining that seems difficult ;-)
20091009 23:20:39< alink> or perhaps a big animated icons <:o)
20091009 23:22:08< alink> but indeed a clock is certainly not helpful for them
20091009 23:22:49< mordante> alink, exactly, that's why I think it needs documentation
20091009 23:23:25< shadowmaster> mordante: did your fix to part 1 of bug #IForgotTheNumberButItWasMine also fix any other keyboard input, or just the ESC?
20091009 23:23:55< shadowmaster> okay, it seems like it did.
20091009 23:25:17< shadowmaster> at least I'm not able to bring the Preferences dialog while the characters are talking :P
20091009 23:25:35-!- Bellerophon_true [n=chatzill@BSN-142-66-83.dial-up.dsl.siol.net] has quit ["ChatZilla 0.9.85 [Firefox 3.5.3/20090824101458]"]
20091009 23:26:01< mordante> keyboard and mouse since it also fixes #IForgotTheNumberButItWasNotYours :-P
20091009 23:30:17< shadowmaster> gui2::show_transient_message() is defined where?
20091009 23:30:38< shadowmaster> ah, found it.
20091009 23:36:31< CIA-62> alink * r39190 /trunk/src/ (unit.cpp unit.hpp): 
20091009 23:36:31< CIA-62> Clean a little generate_gender:
20091009 23:36:31< CIA-62> Unify with other unit::generate_* function, also to better show that it has side-effect.
20091009 23:36:31< CIA-62> Spare one useless RNG call when unit_type has only one gender.
20091009 23:38:51< alink> YogiHH ToD areas seems not saved. Is it a known bug ?
20091009 23:39:31< YogiHH> alink: not until now :). Is it for replays or normal saves?
20091009 23:40:01< alink> normal save. check south caves in test-scenario, they stop being cave after a reload
20091009 23:40:09< shadowmaster> probably for replays too I guess.
20091009 23:41:32< shadowmaster> um, well, apparently not, although I guess that's only because the [time_area] is defined by the start of scenario.
20091009 23:41:40< Ivanovic> Rhonda: i just have not run the update script for a while so some images will be (really) outdated
20091009 23:42:14< Ivanovic> Rhonda: though we have to at least keep some things that might otherwise be stripped away to keep the images editable
20091009 23:42:28< Ivanovic> (you know, that is why we have utils/wesnoth-optipng)
20091009 23:43:02< Rhonda> Alright. :)
20091009 23:43:13< shadowmaster> s/start of scenario/WML event or start of scenario/
20091009 23:43:29< Ivanovic> running the script now though it will probably take an hour to finish
20091009 23:43:41< Ivanovic> (yes, running it with 4 threads and it still takes ages!)
20091009 23:44:08< alink> YogiHH: not that I also see a [illuminated_time] in the savegame, but that's never supported by the engine, so you can ignore it
20091009 23:44:24< shadowmaster> Ivanovic: do you want some distributed wesnoth-optipng now? you know, like distcc?
20091009 23:44:29< shadowmaster> :P
20091009 23:44:42< Ivanovic> shadowmaster: for this i would need some more hardwarepower at hand first
20091009 23:44:55< Ivanovic> (no, i won't get a checkout on one of the machines at my university that has 16 cores)
20091009 23:45:22< mordante> why not sounds like a good use of those cores ;-)
20091009 23:45:35< Ivanovic> mordante: space issues
20091009 23:45:50< Ivanovic> a wesnoth checkout is simply too large for the (very limited) amount of quota we got
20091009 23:45:52< Ivanovic> ;)
20091009 23:46:21< mordante> :-(
20091009 23:46:51< mordante> finally we found something important for those cores to do, is there that silly quota thingy
20091009 23:47:00< mordante> ;-)
20091009 23:47:13< Ivanovic> ah, there is enough other important stuff for those machines to handle
20091009 23:47:26< Ivanovic> some strange algorithm stuff...
20091009 23:47:28< Ivanovic> ;)
20091009 23:48:15< mordante> nah if it's not Wesnoth it can't be important ;-P
20091009 23:48:58< Ivanovic> hmm, somehow you are right...
20091009 23:48:59< Ivanovic> ;)
20091009 23:49:43< mordante> I know :-)
20091009 23:51:04-!- stikonas [n=and@wesnoth/translator/stikonas] has quit [Remote closed the connection]
20091009 23:52:06< alink> mordante: i just finished to understand how the xp sorting work, and now i see that somone posted a comment exlpaining it 5 min ago :-/
20091009 23:52:23< alink> i say, usefull but not very intuitive
20091009 23:52:59< mordante> it was a gsoc patch and I'm also not too happy with it
20091009 23:53:25< mordante> but you're usually good at turning something crappy in the gui to something nice
20091009 23:53:36< alink> the funny thing is that i initially wanted to implement something like that, but even me, when I had it in front of my eyes, i didn't recognize the logic
20091009 23:54:24< alink> thanks, but in this case, at least without testing, i had the same bad idea :-)
20091009 23:54:35-!- stikonas [n=and@bcm-131-111-216-119.girton.cam.ac.uk] has joined #wesnoth-dev
20091009 23:54:46< alink> but testing filter a lot of my crazy UI ideas
20091009 23:55:54-!- stikonas [n=and@wesnoth/translator/stikonas] has quit [Remote closed the connection]
20091009 23:56:27-!- YogiHH [n=chatzill@d096134.adsl.hansenet.de] has left #wesnoth-dev []
20091009 23:57:23< mordante> well I hope you can turn this into something nice
20091009 23:57:40< mordante> I'm not really fond of this implementation
20091009 23:58:20< Ivanovic> mordante, freim: is there any reason to have the gui related images not in data/core/images ?
20091009 23:58:55< mordante> Ivanovic, not for me, but the old gui had them in there so didn't move them
20091009 23:58:58< shadowmaster> wmllscope?
20091009 23:59:06-!- stikonas [n=and@wesnoth/translator/stikonas] has joined #wesnoth-dev
20091009 23:59:23< mordante> might be
20091009 23:59:44< shadowmaster> it doesn't like images that aren't referenced from WML, no idea if it checks data/gui for WML though
--- Log closed Sat Oct 10 00:00:02 2009